Key Advances in Medicine 2015

KEY ADVANCES
IN MEDICINE
January 2016
ENDOCRINOLOGY
NEPHROLOGY
RHEUMATOLOGY
CARDIOLOGY
GASTROENTEROLOGY
& HEPATOLOGY
UROLOGY
NEUROLOGY
CLINICAL
ONCOLOGY
2016 Macmillan Publishers Limited. All rights reserved
Nature Publishing Group | Medical Portfolio

World class journals at the forefront of clinical medicine
From cardiology to urology

Cardiology
Endocrinology
Gastroenterology & Hepatology
Nephrology
Neurology
Oncology
Rheumatology
Urology
www.nature.com/libraries/medical
CONTENT AVAILABLE ONLINE
Key Advances in Medicine
January 2016
The articles included in Nature Reviews Key Advances in Medicine were originally
published online and appear in the February 2016 issues of the eight clinical
Nature Reviews journals. Thejournals editors commissioned international experts
to write a short essay highlighting key papers that made the biggest contribution
to their field in 2015. Between them, the clinical Nature Reviews journals published
44 articles, which are collated in this eBook; if you choose to cite an article, please
use the original journal citation rather than citing the eBook.
Designed by Laura Marshall
1
3
4
6
We hope you enjoy reading Nature Reviews Key Advances in Medicine.

If you would like to find out more about the Nature Reviews series, please visit:
http://www.nature.com/reviews/
C A R D I O LO G Y
ARRHYTHMIAS IN 2015 Advances in drug, ablation,
anddevice therapy forcardiac arrhythmias
Laurent Macle and Stanley Nattel
CORONARY INTERVENTION IN 2015 Improvement
oflong-term outcomes after PCI

Uwe Zeymer
CARDIAC RESUSCITATION IN 2015 Improving outcomes
after OHCA targeting the layperson

Clifton W. Callaway
GENETICS OF CVD IN 2015 Using genomic approaches
to identify CVD-causing variants

Donna K. Arnett
DYSLIPIDAEMIA IN 2015 Advances in treatment
HEART FAILURE IN 2015 Better results from prevention
12
14
16
18
20
22
of dyslipidaemia
Scott M. Grundy
E N D O C R I N O LO G Y
24
OBESITY IN 2015 Advances in managing obesity
26
27
29
31
33
than from additional treatment

Lars Kber
C L I N I C A L O N C O LO G Y
BREAST CANCER IN 2015 Academic research sheds light
on issues that matter to patients
Martine J. Piccart and Isabelle Gingras
COLORECTAL AND GASTRIC CANCER IN 2015 The
development of new agents andmolecular classifications
Eric Van Cutsem and Michel Ducreux
OVARIAN CANCER IN 2015 Insights into strategies for
optimizing ovarian cancer care
Robert L. Coleman
35
37
39
METASTATIC PROSTATE CANCER IN 2015 The new and
40
LUNG CANCER IN 2015 Bypassing checkpoints,

overcoming resistance, andhoning in on new targets
Egbert F. Smit and Paul Baas
42
the old that is newagain

Julie N. Graff and Tomasz M. Beer
MELANOMA IN 2015 Immune-checkpoint blockade
durable cancer control

Elizabeth I. Buchbinder and F. Stephen Hodi
44
John B. Dixon
THYROID CANCER IN 2015 Molecular landscape of
thyroid cancer continues to be deciphered

Yuri E. Nikiforov
ENDOCRINE DISRUPTORS IN 2015 Epigenetic
transgenerational inheritance
Michael K. Skinner
HEPATIC GLUCOSE METABOLISM IN 2015 Nutrient and
hormone-sensing-dependent regulation
Tony K.T. Lam
ISLET-CELL BIOLOGY IN 2015 Understanding secretion,
ageing and death in cells

Gordon C.Weir
PCOS IN 2015 New insights into the genetics of

polycystic ovary syndrome
Ricardo Azziz
G A S T R O E N T E R O LO G Y
& H E PATO LO G Y
REGENERATIVE MEDICINE IN 2015 Generating and
regenerating the digestive system

James M.Wells
LIVER FIBROSIS IN 2015 Crucial steps towards
aneffective treatment
Klaas Poelstra
GUT MICROBIOTA IN 2015 Prevotella in the gut:
choose carefully
Ruth E.Ley
HCV IN 2015 Advances in hepatitisC research
andtreatment
Barbara Rehermann
GASTROINTESTINAL IMAGING IN 2015 Emerging trends
in endoscopicimaging
Bishnu P. Joshi and Thomas D. Wang
PANCREATIC CANCER IN 2015 Precision medicine in

pancreatic cancer fact or fiction?
Thomas Seufferlein and Julia Mayerle
46
N E P H R O LO G Y

PODOCYTE BIOLOGY IN 2015 New insights into
the mechanisms ofpodocyte health

Jeffrey H. Miner
48
IMMUNE REGULATION OF KIDNEY DISEASE IN 2015
50
STEM CELLS AND RENAL DEVELOPMENT IN 2015
51
53
56
58
60
Updates on immunosuppression in kidney disease

Hans-Joachim Anders
Advances in generating and maintaining nephron

progenitors
Ryuichi Nishinakamura
RENAL FIBROSIS IN 2015 Understanding the mechanisms

of kidneyfibrosis
Dong Zhou and Youhua Liu
HYPERTENSION IN 2015 Resistant hypertension: impact
and evolving treatment options
Lilach O. Lerman and Stephen C. Textor
N E U R O LO G Y
68
69
71
73
PARKINSON DISEASE IN 2015 Evolving basic,
76
STROKE IN 2015 Acute endovascular recanalization

therapy comes of age
Alejandro A. Rabinstein
78
pathological and clinical concepts in PD

Lorraine V. Kalia and Anthony E. Lang
NEURO-ONCOLOGY IN 2015 Progress in glioma

diagnosis, classification and treatment
Patrick Y.Wen and David A.Reardon
80
R H E U M ATO LO G Y
GLUCOCORTICOIDS IN 2015 New answers to old
problems
Sarah A.Jones and Eric F.Morand
SYSTEMIC LUPUS ERYTHEMATOSUS IN 2015 Cellular and
metabolic requirements of effector T cells

George C.Tsokos
PSORIATIC ARTHRITIS IN 2015 Advancement continues

in imaging, tightcontrol and newdrugs
Ignazio Olivieri and Salvatore DAngelo
INFLAMMATION IN RHEUMATOLOGY IN 2015 New tools
to tackle inflammatory arthritis

Charles A.Dinarello and Leo A.B.Joosten
U R O LO G Y
SEXUAL DYSFUNCTION IN 2015 Recovering sex drive
in women progress and opportunities

Rossella E.Nappi and Francesca Albani
BLADDER DYSFUNCTION IN 2015 Novel findings

continue to challenge researchers and clinicians
Rose Khavari and Tim Boone
STONES IN 2015 Changes in stone management
suspending belief forevidence

Sapan N. Ambani and Khurshid R. Ghani
MULTIPLE SCLEROSIS IN 2015 Managing the complexity
81
INFECTION IN 2015 HIV protection with PrEP
63
EPILEPSY IN 2015 Classic antiepileptic drugs under fire,
83
65
BLADDER CANCER IN 2015 Improving indication,

technique andoutcome of radical cystectomy
J. Alfred Witjes
NEURODEGENERATIVE DISEASE IN 2015 Targeting
61
of multiplesclerosis
Olga Ciccarelli and Alan Thompson
and new optionsemerge

Christian E. Elger
implications for controlling other STIs

Douglas S. Krakower and Kenneth H. Mayer
tauopathies for therapeutic translation

Julio C. Rojas and Adam L. Boxer
Dont miss out on next years

Copyright 2016 Macmillan Publishers Limited. All rights reserved. Printed in the
United Kingdom. No part of this publication may be reproduced, stored in a retrieval

system, or transmitted in any form (electronic or otherwise) without prior permission from
permissions@nature.com.
Sign up to free e-ALERTS and be notified when next years Key

Advances in Medicine eBook is published.
You wont miss out on Nature Reviews collections
and Focus issues either!
http://www.nature.com/register
Disclaimer: Although every effort is made by the publishers to see that no inaccurate
or misleading data, opinions or statements appear in this collection, they wish to make it
clear that the data and opinions appearing in articles and advertisements herein are the
responsibility of the contributor or advertiser concerned. The journal does include the
personal opinions of the authors; therefore, it is not intended to be relied on solely as a
guide to good practice or safe treatment. Accordingly, the publishers, employees, offices
and agents accept no liability whatsoever for the consequences of any such inaccurate or
misleading data, opinion or statement. Although every effort is made to ensure that drug
doses and other quantities are presented accurately, readers are advised that the new
methods and techniques involving drug usage and described within this journal should only
be followed in conjunction with the drug manufacturers own published literature.
CARDIOLOGY
ARRHYTHMIAS IN 2015
Advances in drug, ablation,

and device therapy for cardiac
arrhythmias
Laurent Macle and Stanley Nattel
Cardiac arrhythmias produce considerable morbidity and mortality, and are
challenging to treat. Advances reported in 2015 will help to guide physicians
in the use of therapeutic approaches ranging from established
pharmaceutical agents through ablation of arrhythmic sources to novel
usesof implanted devices for life-threatening bradyarrhythmias
andtachyarrhythmias.
Catheter ablation is used increasingly for the
management of atrial fibrillation (AF). Over
the past year, results of important random
ized trials of catheter-ablation approaches
have been published. Investigators in the
STARAFII study1 compared three strat
egies of radiofrequency ablation for persis
tent AF in 589 patients randomly assigned to
pulmonary-vein isolation alone, pulmonary-
vein isolation plus additional left atrial
linearablation, or pulmonary-vein isolation
plus ablation of complex fractionated electro
grams. After 18months, no significant differ
ences were found in the rate of AF recurrence:
49% of patients were free from documented
atrial arrhythmia with pulmonary-vein iso
lation alone, versus 37% with additional
linear ablation and 41% with ablation of
complex fractionated electrograms (P=0.15)1.
Therefore, additional ablation (of the types
tested) does not improve outcomes compared
with pulmonary-vein isolation alone in per
sistent AF. Whether more recently developed
methods, such as rotor or rotor-region abla
tion, provide added value to that achieved
with pulmonary-vein isolation remains to be
determined in randomized trials.
Arrhythmia recurrences after pulmonary-
vein ablation are commonly associated with
recovery of pulmonary-vein conduction. The
use of intravenous adenosine to identify veins
at risk of reconnection for additional abla
tion and thereby improve long-term efficacy
of pulmonary-vein isolation was evaluated

in the ADVICE trial2. In this study, adeno
sine revealed dormant pulmonary veins
in 53% of 534 patients with paroxysmal AF.
After 12months, 69.4% of patients randomly
assigned to additional adenosine-guided abla
tion were free from arrhythmia recurrence,
compared with 42.3% of patients assigned to
Drugs
Ablation
Devices
no further ablation (absolute risk reduction

27.1%; P<0.0001)2. By contrast, investigators
in the UNDERATP trial3 found no improve
ment in arrhythmia-free survival with a strat
egy of ATP-guided pulmonary-vein isolation.
How can one reconcile these two studies?
The prevalence of dormant conduction was
not considered in the sample-size estimates
for the UNDERATP trial3, which had 90%
power to detect an overall 20% risk reduc
tion. Given the observed 27.6% prevalence of
dormant conduction, this would require a risk
reduction of 72.4% in patients with dormant
conduction to detect a significant overall dif
ference. In the ADVICE trial2, the relative risk
reduction with adenosine-guided ablation
was 56% in patients with dormant conduc
tion. The nonsignificant trend observed in
the UNDERATP study (1year event-free
survival of 68.7% versus 67.1%; P=0.25) is
consistent with an estimated 28.3% probabil
ity of typeII error. Newer catheter-ablation
techniques might reduce dormant conduc
tion and AF recurrence after pulmonary-vein
isolation; whether they will obviate the need
to test for dormant conduction remains to
bedetermined.
Rate-control therapy remains a mainstay
for many patients with AF. Whereas digoxin
has been widely used for rate control for
>100years, continuing concerns exist about
its potential risks. Washam and colleagues
reported a retrospective analysis of 14,171
patients in the ROCKETAF trial4. Digoxin
was used at baseline in 5,239 patients, who
were more likely to be female, and to have
heart failure, chronic obstructive pulmonary
disease, diabetes mellitus, and persistent AF.
After adjustment, baseline digoxin use was
associated with increased mortality dur
ing followup (5.4% versus 4.3%; adjusted
HR 1.17, 95% CI 1.041.32, P=0.0093)4.
Figure 1 | Therapies for cardiac arrhythmias.

The three major classes of therapeutic approach
currently available and in development:
antiarrhythmic drugs, catheter ablation,
andimplanted rhythm-restoration devices.
JANUARY 2016 | 1
KEY ADVANCES IN MEDICINE

Nature Reviews | Cardiology
C A R D I O LO G Y
Key advances
Pulmonary-vein isolation alone is as
effective as several more complex
procedures for the suppression of persistent
atrial fibrillation (AF)1
Identification and further ablation of
pulmonary veins with dormant conduction
at risk of subsequent reconnection can
improve the effectiveness of pulmonary-vein
isolation procedures for paroxysmal AF2
Digoxin use for rate control in AF might be
associated with increased incidence of
adverse cardiovascular events, so the drug
should be used cautiously and with careful
dose selection4
Defibrillation testing is not needed at the
time of cardiovertordefibrillator
implantation for most patients7
Novel implantable devices, such as
subcutaneous implantable cardiovertor
defibrillators and leadless pacemakers, are
effective and are likely to gain wide clinical
use in the near future10
In a nationwide population-based study,

101,243 patients with AF and receiving
rate-control treatments were compared with
168,678 patients with AF who were not
receiving rate-control drugs5. Patients receiv
ing digoxin alone had an adjusted mortality
risk of 1.12 (95%CI 1.101.14, P<0.001);
blockers or calcium-channel blockers alone
were associated with significantly lower mor
tality risks5. A meta-analysis also suggested
that digoxin use was associated with increased
mortality, particularly among patients with
AF6. How should we interpret these data?
Post-hoc analys es and population-based
studies cannot eliminate contamination by
unmeasured or unknown confounders, even
with the use of statistical adjusting techniques,
especially because digoxin is often prescribed
for more sedentary patients or those with
left ventricular dysfunction. In the absence
of conclusive randomized, controlled trials,
digoxin remains a valuable rate-control agent
that should be used with caution and careful
doseselection.
Implantable electronic devices are widely
used in patients with arrhythmia; impor
tant new information regarding their use
was reported this year. Investigators in the
SIMPLE trial7 evaluated routine defibrilla
tion testing in patients receiving an implant
able cardioverterdefibrillator (ICD). In this
multicentre trial, 2,500 patients were ran
domly allocated to defibrillation testing or no
testing at implantation and were followed up
for a mean of 3.1years. A lack of defibrillation
testing did not increase the risk of arrhythmic
death or failed appropriate shock compared

with those who did receive defibrillation test
ing (7% versus 8% per year; HR0.86, 95%CI
0.651.14, P<0.0001 for noninferiority)7.
Atrend towards increased perioperative com
plications was observed with defibrillation
testing. Therefore, for most patients receiving
an ICD, defibrillation testing is not indicated.
Subcutaneous ICDs (SICDs) have been
developed as a simpler and possibly safer
alternative to transvenous-lead systems.
Burke and colleagues reported the safety and
efficacy of SICDs in a pooled analysis of two
prospective studies with 882 patients followed
up for 651345days after SICD implanta
tion8. Shock therapy for ventricular tachy
arrhythmias was 98.2% successful, and the
3year inappropriate shock rate was 13.1%8.
Device-related complications occurred in
11.1% of patients, but no electrode failure and
no SICD-related endocarditis or bacterae
mia were observed. Although the SICD is
an attractive alternative to avoid complica
tions associated with transvenous leads, the
high rate of inappropriate shocks and lack of
pacing capabilities (for bradycardia, cardiac
resynchronization therapy, or antitachycardia
pacing) are important drawbacks.
Adverse events with pacemakers might also
be related to transvenous leads. Researchers in
two multicentre, observational studies reported
on miniaturized, self-contained, leadless pace
makers9,10. In the LEADLESSII study9, leadless
pacemakers were successfully implanted in
504 out of 526 (95.8%) patients requiring per
manent, single-chamber ventricular pacing.
Prespecified pacing and sensing requirements
were met in 90% of patients. Device-related
serious adverse events were observed in 6.7%
of patients: cardiac perforation (1.3%), vascular
complications (1.3%), and device dislodgement
requiring percutaneous retrieval (1.7%). In the
Micra Transcatheter Pacing Study10, similar
devices were successfully implanted in 719
out of 725 patients (99.2%). Adequate pacing
thresholds were observed at 6months in 98.3%
of patients. Major complications occurred in
4% of patients, including 11 cardiac injuries,
five puncture-site complications, two thrombo
emboli, and two pacing issues. Although these
studies are preliminary and limited by their
observational design, the future of leadless
pacing seems promising.
In summary, many advances have been
made in drug, ablation, and device therapy for
cardiac arrhythmias (FIG.1) over the past year.
A number of ongoing trials that will further
inform therapeutic choices are close to comple
tion and should be reported in the near future.
These include Fire and Ice (NCT01490814)
and CIRCADOSE (NCT01913522) in which
2 | JANUARY 2016
cryoablation is being compared with radio

frequency ablation for paroxysmal AF;
CENTRAAF (NCT01924377) and AFACART
(NCT02113761) to examine the value of target
ing putative AF sources with high-resolution
mapping; and RAFTAF (NCT01420393)
in which rhythm control with catheter abla
tion is being compared with rate control in
patients with heart failure and high-burden
AF. Research in ablation and device treatment
options, for which there is a large and lucrative
potential market, is very active and expected
to provide important technological and prac
tical advances. Further research is needed
to improve drug therapy for AF, particularly
given that pharmacological rate-control and
rhythm-control therapies remain a mainstay
of AF management in many patients.
Laurent Macle and Stanley Nattel are at the
Department of Medicine, Montreal Heart Institute,
5000 Belanger Street East, Montreal,
QuebecH1T1C8, Canada.
Correspondence to S.N.
stanley.nattel@icm-mhi.org
doi:10.1038/nrcardio.2015.196
Published online 4 Jan 2016;
corrected online 29 Jan 2016
Verma,A. etal. Approaches to catheter ablation for
persistent atrial fibrillation. N.Engl. J.Med. 372,
18121822 (2015).
2.
Macle,L. etal. Adenosine-guided pulmonary vein
isolation for the treatment of paroxysmal atrial
fibrillation: an international, multicentre, randomised
superiority trial. Lancet 386, 672679 (2015).
3.
Kobori,A. etal. Adenosine triphosphate-guided
pulmonary vein isolation for atrial fibrillation: the
UNmasking Dormant Electrical Reconduction by
Adenosine TriPhosphate (UNDER-ATP) trial. Eur.
HeartJ. http://dx.doi.org/10.1093/eurheartj/ehv457.
4.
Washam,J.B. etal. Digoxin use in patients with atrial
fibrillation and adverse cardiovascular outcomes:
aretrospective analysis of the Rivaroxaban Once Daily
Oral Direct FactorXa Inhibition Compared with
VitaminK Antagonism for Prevention of Stroke and
Embolism Trial in Atrial Fibrillation (ROCKETAF).
Lancet 385, 23632370 (2015).
5.
Chao,T.F. etal. Rate-control treatment and mortality in
atrial fibrillation. Circulation 132, 16041612 (2015).
6. Vamos,M., Erath,J.W. & Hohnloser,S.H.
Digoxinassociated mortality: a systematic review
andmetaanalysis of the literature. Eur. Heart J. 36,
18311838 (2015).
Healey,J.S. etal. Cardioverter defibrillator
7.
implantation without induction of ventricular
fibrillation: a single-blind, non-inferiority, randomised
controlled trial (SIMPLE). Lancet 385, 785791
(2015).
Burke,M.C. etal. Safety and efficacy of the totally
8.
subcutaneous implantable defibrillator: 2year results
from a pooled analysis of the IDE study
andEFFORTLESS registry. J.Am. Coll. Cardiol. 65,
16051615 (2015).
Reddy,V.Y. etal. Percutaneous implantation of an
9.
entirely intracardiac leadless pacemaker. N.Engl.
J.Med. 373, 11251135 (2015).
10. Reynolds,D. etal. A leadless intracardiac
transcatheter pacing system. N.Engl. J.Med.
http://dx.doi.org/10.1056/NEJMoa1511643.
1.
Acknowledgements
S.N. is also on the faculty of the Departments of

Pharmacology and Therapeutics and Medicine at McGill
University, Montreal, Quebec, Canada, as well as the West
German Heart and Vascular Center, Faculty of Medicine,
University DuisburgEssen, Essen, Germany.
Competing interests statement
The authors declare no competing interests.
www.nature.com/reviews
C A R D I O LO G Y
C O R O N A RY I N T E R V E N T I O N I N 2 0 1 5
Improvement of long-term
outcomes after PCI
Uwe Zeymer
Third-generation drug-eluting stents (DES) have emerged as first-line
devices for percutaneous coronary intervention, even in patients with high
bleeding risk. Studies published in 2015 report that bioabsorbable vascular
scaffolds are equally effective and safe as DES in low-risk populations and,
with the addition of extended dual antiplatelet therapy, might improve
long-term outcomes.
Percutaneous coronary intervention (PCI) has
emerged as the revascularization therapy of
choice in the majority of patients with stable
coronary artery disease and acute coronary syn
dromes1. With the development of new inter
ventional techniques and devices, even complex
lesions such as chronic total occlusions, left
main stenosis, and bifurcation lesions can be
treated with high acute procedural success rates.
However, two major issues related to outcomes
after PCI still remain. The introduction of baremetal stents and drug-eluting stents (DES) has
diminished, but not completely eliminated, the
problem of early and late restenosis2. Whether
the need for long-term target-lesion revascu
larization owing to restenosis and acute ischae
mic events can be further reduced is unknown.
Inaddition, although prolonged dual antiplate
let therapy (DAPT) for >12months results in
a further decrease in very late stent thrombo
sis and ischaemic events, it is also associated
with an increase in bleeding complications3.
Therefore, the optimal duration of DAPT
remains to be determined.
Bioabsorbable vascular scaffolds (BVS)
were developed to improve long-term out
comes after PCI. The occurrence of late adverse
events (restenosis and stent thrombosis) with
permanent metallic stents might be related to
persistent inflammation, loss of normal vessel
curvature, impaired vasomotion, strut frac
ture, ongoing tissue growth within the stent
frame, and neoatherosclerosis. To overcome
these problems, BVS were developed to pro
vide mechanical support and optimal drug
delivery associated with excellent mid-term
results, followed by complete bioresorption
over severalyears.
In the ABSORB3 multicentre, random
ized trial4, the Absorb BVS (Abbott Vascular,
USA) the first BVS approved for commercial
use was compared with a third-generation

DES in 2,008 patients with stable or u
nstable
angina. High-risk patients, as determined by
their clinical presentation or lesion character
istics, were excluded. During the initial
PCI procedure, fewer BVS than DES were
implanted (6.2% versus 0.9%; P=0.001),
owing to reduced deliverability. The primary
end point of target-lesion failure after 1year
was not significantly different between the
BVS and DES groups in the intention-totreat
analysis (7.8% versus 6.1%; P=0.16), or in the
astreated analysis (8.0% versus 6.1%; P=0.12).
Furthermore, although the overall rate of stent
thrombosis after 1year was not significantly
different between the BVS-treated patients and
the DES-treated patients (1.5% versus 0.7%;
P=0.13), more subacute stent thrombosis
events occurred with the use of BVS (0.9% ver
sus 0.1%; P=0.04). Patient-reported angina was
similar in the two groups after 1year (18.3%
versus 18.4%). These results suggest that BVS
are, at best, as effective and safe as third-gener
ation DES during 1year of followup. Whether
BVS have long-term advantages over DES needs
further study. Additional improvements in the
design of BVS might improve deliverability and

reduce subacute stent thrombosis5. However,
for the time being, third-generation DES
remain the standard of care for efficacy, safety,
and economical reasons.
One drawback of DES has been the need for
a longer duration of DAPT, owing to delayed
healing and endothelialization of the coronary
artery compared with bare-metal stents. The
first-generation DES have been associated
with higher rates of late stent thrombosis
and, therefore, DAPT for 12months has been
recommended after implantation of these
stents1. With the advent of third-generation
DES, shorter durations of DAPT have been
tested and shown to be safe6. In a study pub
lished in 2015, a new polymer-free and carrier-
free DES was tested against a bare-metal stent
in 2,466 patients with high risk of bleeding,
owing to advanced age (>75years), the need
for concomitant oral antic oagulation, or
impaired renal function7. DAPT was stopped
after 30days, and patients were given single
antiplatelet therapy. Clinically driven target-
lesion revascularization, the primary efficacy
end point, occurred less often with the DES
compared with the bare-metal stent (5.1%
versus 9.8%; P=0.001), whereas definite or
probable stent thrombosis (2.0% versus 2.2%)
and bleeding events (13.9% versus 14.7%) were
not significantly different7. The primary safety
end point of cardiac death, myocardial infarc
tion, or stent thrombosis occurred less often
with the DES (9.4% versus 12.9%; P=0.005).
Given these results, there seems to be no indi
cation favouring bare-metal stents over DES,
irrespective of patient and lesion subset.
The optimal duration of DAPT after PCI is
still a matter of debate. The benefit of extend
ing antiplatelet therapy to reduce the risk of
cardiovascular events must be carefully bal
anced against the associated increased risk of
bleeding complications. The DAPT trial3, in
which patients were randomized 12months
after DES implantation to placebo or a pro
longed administration of DAPT, showed a
Table 1 | Suggested duration of dual antiplatelet therapy

DAPT duration
Patient profile
1 month
Elective PCI with bare-metal stent, or third-generation drug-eluting stent

with high bleeding risk
3 months
Elective PCI with third-generation drug-eluting stent and increased bleeding

risk, or bleeding event <3months
6 months
Elective PCI with third-generation drug-eluting stent
12 months
Elective PCI with third-generation drug-eluting stent with high ischaemic

risk (left main, multiple stents, etc.)
PCI for acute coronary syndromes, regardless of stent type
>12 months
PCI for acute coronary syndromes in patients with diabetes mellitus,

renalinsufficiency, and/or DAPT score>2
DAPT, dual antiplatelet therapy; PCI, percutaneous coronary intervention.
JANUARY 2016 | 3

C A R D I O LO G Y
1.
Key advances
The bioabsorbable vascular scaffold
Absorbwas as safe and effective as a
third-generation drug-eluting stent in a
low-risk percutaneous coronary intervention
(PCI) population during 1year followup4
A polymer-free and carrier-free drug-coated
stent with a reduced duration of dual
antiplatelet therapy of 4weeks was as safe
and more effective than a bare-metal stent
in patients with high bleeding risk7
In patients treated with PCI for acute
coronary syndromes, longer term dual
antiplatelet therapy beyond 12months
further reduces ischaemic events and
stentthrombosis9
reduction in stent thrombosis and an increase

in bleeding complications in patients receiving
extended DAPT. Astudy published in 2015
that was prematurely terminated8 reported
only trends for a reduction in ischaemic out
comes, without a difference in bleeding events,
after extending DAPT from 12 up to 48months
(median 33months).
In a meta-analysis involving 33,435 patients
with a history of myocardial infarction,
extended DAPT significantly reduced cardio
vascular death (risk ratio [RR]0.85, 95%CI
0.740.98), stroke (RR0.81, 95%CI 0.680.97),
myocardial infarction (RR0.70, 95%CI 0.55
0.88), and stent thrombosis (RR0.50, 95%CI
0.280.89)9. An increase in major bleeding
complications was observed in patients who
received extended DAPT (RR1.73, 95%CI
1.192.50), but no increase occurred in fatal
bleeding (RR0.91, 95%CI 0.531.58) or allcause mortality (RR0.92, 95%CI 0.831.03).
These data suggest that whereas DAPT for
shorter durations (36months) after elective
PCI seems favourable, extended DAPT beyond
12months (up to 3648months) should be
considered after PCI in patients with acute
coronary syndromes, if well-tolerated within
the first 12months (TABLE1).
These studies published in 2015 provided
further evidence for the efficacy and safety of
third-generation DES in all patient and lesion
subsets, whereas BVS might have the potential
to improve long-term outcomes even further.
Growing evidence indicates that extended
DAPT for >12months after PCI for acute coro
nary syndrome is favourable in selected patients
with high ischaemic and low b
leedingrisk.
Uwe Zeymer is at Klinikum Ludwigshafen and the
Institut fr Herzinfarktforschung Ludwigshafen,
Bremser Strasse 79, 67063 Ludwigshafen, Germany.
zeymeru@klilu.de
doi:10.1038/nrcardio.2015.204
Published online 14 Jan 2016
2.
3.
4.
5.
6.
Windecker,S. etal. 2014 ESC/EACTS Guidelines

onmyocardial revascularization. Eur. Heart J. 35,
25412619 (2014).
Sabat,M. etal. Clinical outcomes in patients with
STsegment elevation myocardial infarction treated
with everolimus-eluting stents versus bare-metal
stents (EXAMINATION): 5year results of a
randomised trial. Lancet http://dx.doi.org/10.1016/
S0140-6736(15)00548-6 (2015).
Mauri,L. etal. Twelve or 30 months of dual
antiplatelet therapy after drug-eluting stents. N.Engl.
J.Med. 371, 21552166 (2014).
Ellis,S. G. etal. Everolimus-eluting bioresorbable
scaffolds for coronary artery disease. N.Engl. J.Med.
373, 19051915 (2015).
Haude,M. etal. Safety and performance of the
second-generation drug-eluting absorbable metal
scaffold in patients with denovo coronary artery
lesions (BIOSOLVEII): 6month results of a
prospective, multicentre, non-randomised,
firstinman trial. Lancet http://dx.doi.org/10.1016/
S0140-6736(15)00447-X (2015).
Kim,B. K. etal. A new strategy for discontinuation of
dual antiplatelet therapy: the RESET Trial (REal Safety
7.
8.
9.
and Efficacy of 3month dual antiplatelet Therapy

following Endeavor zotarolimus-eluting stent
implantation). J.Am. Coll. Cardiol. 60, 13401348
(2012).
Urban,P. etal. Polymer-free drug-coated coronary
stents in patients at high bleeding risk. N. Engl.
J.Med. 373, 20382047 (2015).
Helft,G. etal. Stopping or continuing clopidogrel
12months after drug-eluting stent placement:
theOPTIDUAL randomized trial. http://dx.doi.org/
10.1093/eurheartj/ehv481 (2015).
Udell,J. A. etal. Long-term dual antiplatelet therapy
for secondary prevention of cardiovascular events in
the subgroup of patients with previous myocardial
infarction: a collaborative meta-analysis of randomized
trials. Eur. Heart J. http://dx.doi.org/10.1093/
eurheartj/ehv443 (2015).
U.Z. declares that he has received speaker fees from

AstraZeneca, B. Braun, Bayer Healthcare, Biotronik,
Boehringer Ingelheim, Bristol-Myers Squibb, Correvio, Daiichi
Sankyo, Eli Lilly, Hexal, Medtronic, The Medicines Company,
MSD, Novartis, Pfizer, and Sanofi.
C A R D I A C R E S U S C I TAT I O N I N 2 0 1 5
Improving outcomes after OHCA

targeting the layperson
Clifton W. Callaway
The proportion of the lay population that is trained in cardiopulmonary
resuscitation is closely linked to the probability of an individual surviving
cardiac arrest. New mobile-assisted technologies might increase the benefit
of population-based training. Furthermore, admission of patients to
specialized hospitals can increase the likelihood of survival.
In 2015, numerous cohort studies revealed
the potential for system-level interventions
to improve outcomes after cardiopulmonary
resuscitation (CPR). Several of these reports
focused on long-term and short-term strat
egies to mobilize layperson rescuers, whereas
other studies examined how survival outcomes
can be influenced by the choice of hospital in
which a patient is treated immediately after
cardiac arrest. Results from these studies sup
port the 2015 Institute of Medicine recommen
dations on strategies to improve survival after
cardiac arrest, which highlighted the impor
tance of using regional and national databases,
fostering a culture of action among laypersons,
setting accreditation standards for health-care
systems, and implementing continuous quality
improvement programmes1.
During a cardiac arrest, the heart stops
pumping blood because of mechanical or
electrical disturbances. Total body ischaemia
affects every organ system, and cardiac arrest is
fatal unless rapidly reversed. Although patients
can tolerate long total durations of CPR when
it is initiated immediately, delays of even a few
minutes in initiating CPR can lead to cellular
energy depletion and vascular collapse that
4 | JANUARY 2016
are difficult to reverse, reducing the chances of

neurological recovery and survival2. Given that
professional medical personnel often arrive
too late to initiate CPR for all patients, several
studies published in 2015 sought to examine
the value of recruiting laypersons to adminis
ter CPR and improve survival outcomes after
outofhospital cardiac arrest (OHCA).
HasselqvistAx and colleagues described
layperson CPR in Sweden from 1990 to 2011
using a national database of OHCA events3.
The proportion of citizens trained in CPR
increased over this period as a result of
government-funded initiatives, including
widespread public training and implemen
tation of telephone-assisted CPR wherein
emergency call-takers instruct the caller on
how to perform chest compressions while
awaiting ambulance arrival. Consequently, the
proportion of patients experiencing cardiac
arrest who received layperson CPR increased
from 35% to 65%. Among cases where col
lapse was witnessed, laypersons performed
CPR in 51.1% (15,512 of 30,381) of patients,
for whom 30day mortality was 10.5%.
Conversely, 30day mortality in patients who
did not receive CPR until paramedic assistance
C A R D I O LO G Y
Recognition
and activation
of emergency
response
system
Immediate
high-quality
CPR
Rapid
debrillation
Layperson responders
Basic and
advanced
emergency
medical
services
Advanced life
support and
post-arrest
care
Regional health care systems

Interventions
Population-based training35
Telephone-assisted CPR3,5
Notication of nearby volunteer
responders by mobile phone6
Treatment at tertiary care hospitals9

Treatment at cardiac care centers10
Expected outcomes
Increase rate of CPR initiated before ambulance arrival
10% immediately with rapid interventions
Achieve >50% rates over time
Reliable delivery of eective post-CPR care
Figure 1 | Interventions to improve survival outcomes after outofhospital cardiac arrest.

Current research highlights the potential for effective interventions at the beginning and end
ofthechain of survival to improve survival rates. These interventions increase effective action
bylayperson responders and coordinate reliable delivery of care after cardiopulmonary
resuscitation (CPR).
arrived was significantly lower (4.0%; OR2.15,

95%CI 1.882.45)3. Similar national data
from 167,912 patients in Japan between 2005
and 2012 have also been reported4. Rates of
layperson CPR for patients witnessed to col
lapse increased from 38.6% to 50.9% over this
time period, and rates of functionally favour
able survival increased from 3.3% to 8.2%.
Functionally favourable survival rate was 8.4%
(6,594 of 78,592 patients) with layperson CPR
versus 4.1% (3,595 of 88,720 patients) without
layperson CPR (OR1.52, 95%CI 1.451.60)4.
The USA does not have comprehensive
national data on responses to OHCA events,
but layperson rescue in North Carolina
has been described for the period between
2010 and 2013 using a database of 4,961
patients5. During this time period, statefunded initiatives encouraged layperson
CPR, telephone-assisted CPR, and dispatch
of non-medical public safety providers (police
and firefighters) to assist in cases of OHCA.
The overall rate of survival with favourable
neurological outcome increased from 7.1%
(82 of 1,149 patients) in 2010 to 9.7% (129
of 1,334 patients) in 2013 (REF.5). The rate of
survival to hospital discharge among patients
who received layperson CPR was 11.8%
(266of 2,250 patients) compared with 8.9%
(178 of 1,994 patients) for individuals with
first responder-initiated CPR and 7.6% (51 of
672) for emergency medical services-initiated
CPR (P<0.01 between groups). Rates of sur
vival with favourable neurological outcome
were similarly improved (10.5% versus 7.7%
versus 7.0%, respectively, P<0.01)5. The rates
of layperson CPR in this cohort were notably

lower than in the Swedish or Japanese cohorts.
Innovative strategies that exploit the
increasing wireless connectivity of the gen
eral public are proving to be effective for
improving outcomes after cardiac arrest.
Given the high likelihood that emergency
calls originate from a mobile device from a
bystander who is beside a collapsed patient,
the use of telephone-assisted CPR has been
evaluated3,5. Ringh and colleagues performed
a randomized, controlled trial to test a short
message service (SMS)-based system that calls
other nearby volunteers for assistance at the
same time that emergency medical services
are dispatched6. Intotal, 9,828 laypersons
volunteered to receive alerts. During the trial,
volunteers initiated CPR before ambulance
arrival for 62% (188 of 305) of cases when
the SMS-based system was active versus 48%
(172of 360) of cases when the system was not
in use (95%CI 6.221.2%, P<0.001)6. This
study was not designed to detect differences
in patient survival and outcomes, but 30day
survival was 11.2% (32 of 286) with SMS
intervention compared with 8.6% (28 of 326)
without SMS intervention. These prelimi
nary results suggest that fast and cheap pub
lic health interventions that are immediately
deployable in many countries, such as mobile
telephone-assisted layperson CPR, might
improve survival outcomes after OHCA.
Treatment of patients after reversal of
cardiac arrest might involve acute coronary
intervention or other advanced critical care7.
Tertiary cardiac centres are more experienced
and expeditious at delivering complex careand

intervention for systemically ill patients with
severe brain injury. At present, only weak
evidence is available to support the benefit of
transferring or diverting patients to special
ized centres after OHCA8. Several studies
published in 2015 were designed to explore
the relationship between patient outcomes
and thehospitals where they were admitted.
The variations in treatment outcome for
patients admitted to one of the eight hospitals
in Copenhagen, Denmark from 20022011
were reported9. Hospitals were classified as
tertiary cardiac hospitals (n=2) or nontertiary
hospitals (n=8). Patients with STsegment
elevation myocardial infarction were always
admitted to the tertiary cardiac hospitals, but
only 54% of patients without STsegment ele
vation were admitted to the tertiary cardiac
hospital. Among 1,078 patients with OHCA
without STsegment elevation, survival to dis
charge was 45% for those hospitalized at ter
tiary cardiac hospitals compared with 24% for
patients hospitalized at non-tertiary hospitals
(OR2.66, 95%CI 2.043.46)9. After adjust
ment for other comorbidities and prognos
tic factors, hospitalization at tertiary centres
was associated with lower 30day mortality
compared with hospitalization at nontertiary
centres (HR0.78, 95%CI 0.640.96). Patients
admitted to tertiary centres were more likely
to receive acute coronary intervention, neuro
critical care diagnostics and intervention, and
post-intensive care workup and treatment.
The variations in treatment outcomes for
7,725 patients hospitalized after OHCA in hos
pitals in California, USA during 2011 have also
been described10. The 333 hospitals involved
in the study were classified according to one
proposed strategy for accreditation of cardiac
arrest centres: centres without 24h capability
for acute coronary interventions (n=208),
centres capable of acute coronary interven
tions that treated <40 cases of OHCA per year
(n=71), and centres capable of acute coronary
Key advances
The proportion of the entire population
trained in cardiopulmonary resuscitation
(CPR) is associated with the probability
of a patient surviving after outofhospital
cardiac arrest35
Mobile-assisted technology can increase
the proportion of cardiac arrests with
layperson CPR assistance6
Transporting patients after outofhospital
cardiac arrest to centres that can provide
specialized cardiac and neurocritical care
services is associated with
improvedoutcomes9,10
JANUARY 2016 | 5

C A R D I O LO G Y
interventions that treated 40 cases of OHCA
per year (n=54). In total, 24% of patients in
the study had good functional recovery, but
no clear association existed between the num
ber of patients with OHCA admitted to each
hospital and outcome. Treatment at a cardiac
centre that treated 40 cases of OHCA year
(OR1.32, 95%CI 1.061.64) and <40 cases
of OHCA per year (OR1.63, 95%CI 1.35
1.97) were both associated with good neuro
logical recovery, compared with treatment at
noncardiac intervention centres.
After decades of research into therapies
to mitigate ischaemiareperfusion injury
in each organ after cardiac arrest, a single
intervention to treat whole body ischaemia
reperfusion injury remains elusive. In the
absence of such a breakthrough, the greatest
opportunities to increase survival after OHCA
involve reducing delays to intervention and
ensuring reliable delivery of specialized care
to all patients (FIG.1). Current evidence sug
gests that system-level changes might result in
national or state rates of layperson CPR >50%,
and that cheap, readytoimplement interven
tions such as SMS and telephone-assisted CPR
can immediately increase layperson response.
With improved layperson response, survival
might be possible for >10% of patients. Finally,
admission of patients after OHCA to hospi
tals that provide high-intensity care after
CPR is associated with increased survival.
Policiesand incentives to promote layperson
response andencourage accountable hospital
care will lead to more lives being saved.
Clifton W. Callaway is at the Department of Emergency
Medicine and Department of Pharmacology and
Chemical Biology, University of Pittsburgh,
Iroquois400A, 3600 Forbes Avenue,
Pittsburgh,Pennsylvania15260, USA.
callawaycw@upmc.edu
doi:10.1038/nrcardio.2015.201
1.
2.
3.
4.
5.
6.
7.
Neumar,R. W. etal. Post-cardiac arrest syndrome:

epidemiology, pathophysiology, treatment, and
prognostication. Circulation 118, 24522483
(2008).
Graham, R., McCoy, M. A. & Schultz, A. M. Strategies
to Improve Cardiac Arrest Survival: A Time to Act
(TheNational Academies Press, 2015).
HasselqvistAx,I. etal. Early cardiopulmonary
resuscitation in outofhospital cardiac arrest. N.Engl.
J.Med. 372, 23072315 (2015).
Nakahara,S. etal. Association of bystander
interventions with neurologically intact survival among
patients with bystander-witnessed outofhospital
cardiac arrest in Japan. JAMA 314, 247254 (2015).
Malta Hansen,C. etal. Association of bystander
andfirst-responder intervention with survival after
outofhospital cardiac arrest in North Carolina:
20102013. JAMA 314, 255264 (2015).
Ringh,M. etal. Mobile-phone dispatch of laypersons
for CPR in outofhospital cardiac arrest. N.Engl.
J.Med. 372, 23162325 (2015).
Callaway,C.W. etal. Part 8: post-cardiac arrest care:
2015 American Heart Association guidelines update
for cardiopulmonary resuscitation and emergency
cardiovascular care. Circulation 132 (Suppl. 2),
S465S482 (2015).
Kronick,S.L. etal. Part 4: systems of care and

continuous quality improvement: 2015 American
Heart Association guidelines update for
cardiopulmonary resuscitation and emergency
cardiovascular care. Circulation 132 (Suppl. 2),
S397S413 (2015).
9.
Sholm,H. etal. Prognostic implications of levelofcare
at tertiary heart centers compared with other hospitals
after resuscitation from outofhospital cardiac arrest.
Circ. Cardiovasc. Qual. Outcomes 8, 268276 (2015).
10. Mumma,B.E., Diercks,D.B., Wilson,M.D.
& Holmes,J.F. Association between treatment
8.
at an STsegment elevation myocardial infarction

center and neurologic recovery after
outofhospitalcardiac arrest. Am. HeartJ. 170,
516523 (2015).
Acknowledgements
C.W.C. is supported by NHLBI grants U01 HL077871 and

K12 HL109068.
C.W.C. is the Chair of Emergency Cardiovascular Care

Committee for the AHA.
GENETICS OF CVD IN 2015
Using genomic approaches to identify

CVD-causing variants
Donna K. Arnett
Cardiovascular genomics has evolved substantially in the past 2decades.
Numerous papers published in 2015 demonstrate that new genomic
approaches, often used synergistically, can yield noteworthy findings.
Arange of laboratory, analytical, and bioinformatic techniques to uncover
genetic contributors to coronary artery disease, myocardial infarction,
anddilated cardiomyopathy are described.
Multilayered -omic approaches are prov
ing successful for the identification of new
genetic contributors to cardiovascular dis
eases (CVDs), and for furthering our under
standing of the mechanisms linking known
genetic mutations to disease causation. The
three most important papers in cardiovascular
genetics for 2015 highlight the value of differ
ent and complementary genomic approaches
to delineate the causes underlying complex
CVDs. Coronary artery disease (CAD), the
most common cause of cardiovascular death
globally1, has been the focus of several genomic
studies. Furthermore, the genetic mutations
underlying the pathogenesis of dilated cardio
myopathy (DCM), one of the most common
cardiomyopathies and the most frequent indi
cation for cardiac transplantation, has also
received considerable interest. Together, these
studies demonstrate remarkable progress in the
past year in defining the genetic architecture
of CVD, and offer novel pathways for future
prevention and treatment options.
CAD has been a frequent subject of
genomic inquiry. Genome-wide associ
ation studies (GWAS) of common single-
nucleotide polymorphisms (SNPs) have thus
far identified 48 genome-wide significant loci
associated with CAD2,3. Investigators in the
CARDIoGRAMplusC4D consortium con
ducted a large meta-analysis of CAD using
phased haplotypes from the 1000 Genomes
Project4. Data from 60,801 patients with CAD
6 | JANUARY 2016
(myocardial infarction [MI], acute coronary

syndrome, chronic stable angina, or coro
nary stenosis of >50%) and 123,504 controls
were collated from 48 studies. The analysis
included common variants (n=6.7million
with minor allele frequency [MAF] >0.05)
and low frequency variants (n=2.7million
with 0.005<MAF<0.05). Imputation was
implemented using the 1000 Genomes Project
phase1 v3 training set with 38million variants;
over half are low frequency (MAF<0.005) and
one-fifth are common (MAF>0.05).
In total, 2,213 variants (7.6% insertion or
deletions [indels]) showed significant associ
ations (P<5108) with CAD. These results
Key advances
A genome-wide association meta-analysis
confirmed 47 previously identified gene loci
associated with coronary artery disease
andidentified ten new loci that contain
candidate genes implicating biological
processes in blood vessels4
Exome sequencing identified rare alleles in
two genes (LDLR and APOA5) that contribute
to risk of myocardial infarction, suggesting
that disordered triglyceride-rich lipoprotein
metabolism is linked to its pathogenesis6
Mutations in TTN causes dilated
cardiomyopathy by interfering with
linkages between sarcomerogenesis
andadaptive remodelling8
C A R D I O LO G Y
confirmed 47 of 48 previously reported genes
at a nominal significance level. The majority
(98.3%) of the significant loci were common
variants. However, owing to the increased
precision of imputation afforded by the 1000
Genomes Project, rare frequency SNPs in
APOE and PCSK9 showed strong associations
that were overlooked in the HapMap-based
GWAS2,3. The CARDIoGRAMplusC4D con
sortium identified eight new loci from their
additive models. These loci were biologi
cally relevant and valid candidate genes with
well-documented roles in vessel wall biology,
including cell adhesion, leukocyte and vascular
smooth muscle cell migration, vascular smooth
muscle cell phenotypic switching, transforming
growth factor signalling, angiogenesis, and
nitric oxide signalling. These candidate genes
are also known to have anti-inflammatory and
infarct-sparing effects.
The large sample size in this study also
allowed exploration of recessive and dominant
modes of inheritance. Using a recessive inherit
ance model to determine susceptibility to CAD
from homozygosity for the less frequent allele,
two new recessive loci were identified, n
either
of which were significant using an additive
model. The dominant inheritance model
identified variants that had been detected with
the additive analysis. Together, these results
suggest that CAD is largely determined by
common SNPs with small effectsizes.
Inherited genetic factors are thought to
contribute to MI that occurs early in life5.
Common variants in approximately 45 genes
and rarevariants in LDL-related genes have
previously been associated with early MI6. Do
and colleagues used a range of approaches to
evaluate how low-frequency and rare vari
ants across the genome contribute to risk of
early MI6. Using whole-exome sequencing,
the investigators sequenced all protein-coding
genes in 1,027 patients with early MI (men aged
50years, women aged 60years) and 946 older
MIfree controls (men aged 60years, women
aged 70years). On average, each person had
43nonsense, 7,828 missense, 92 splice-site, 189
indel frameshift, 366 indel nonframeshift, and
103 nonsynonymous singleton variants. None
of the approximately 20,000 genes analysed in
the discovery set wassignificant.
The discovery sequencing results were
followed up using four different approaches6.
The first (statistical imputation in 64,132
patients and controls) and second (analysing
the Illumina Exome chip in 15,936 patients
and controls) approaches did not produce
significant findings for low-frequency vari
ants. In their third approach, the investigators
conducted targeted resequencing of APOA5 in
6,721 patients and 6,711 controls. Using burden
Disease
MI6
DCM8
To assess how rare mutations

contribute to early-onset MI risk
To explore the pathogenicity

of TTN mutations in DCM
Whole exome sequencing of

9,793 genomes from patients
with early-onset MI and from
MI-free controls
Assessed mutations in cardiac

microtissues engineered from
human iPS cells
Conrmed 47 of 48 previously
A burden of rare alleles in
reported genomic loci and
two genes, LDLR and APOA5,
identied 10 new loci containing contributes to MI risk
candidate causal genes with
roles in vessel wall biology
Missense mutations and TTN

truncating variants can cause
DCM by disrupting links
between sarcomerogenesis
and adaptive remodelling
CAD4
Objectives
To conrm existing and identify
additional CAD-associated loci
Techniques
GWAS meta-analysis of
~185,000 cases and controls,
interrogating 6.7 million
common and 2.7 million
low-frequency variants
Findings
Figure 1 | The use of genomic techniques to identify new CVD-causing variants and
toevaluate how they contribute to disease pathogenesis. CAD, coronary
disease;
Nature artery
Reviews
| Cardiology
CVD,cardiovascular disease; DCM, dilated cardiomyopathy; GWAS, genome-wide association
studies; iPS, induced pluripotent stem cells; MI, myocardial infarction; TTN, titin.
tests for an excess (or deficit) of aggregated rare

(<1% MAF) and nonsynonymous mutations,
they found a 2.2fold excess risk of early MI in
patients with MI compared with the MIfree
controls (P=5107)6. In a series of analy
ses using rare-variant aggregating criteria, a
3.3fold excess risk of early MI (P=0.008) was
identified when aggregating only alleles anno
tated as damaging (as defined by five protein
prediction algorithms) and a 4.5fold excess
risk of early MI (P=0.007) was reported when
aggregating only disruptive (null) mutations.
Carriers of rare variants had higher triglyceride
levels and lower HDL-cholesterol levels than
noncarriers. Finally, in their fourth approach,
whole-exome sequencing data was added from
another study for a total of 9,793 patients with
early MI and controls. Rare alleles contrib
uted to the risk of early MI only in the LDLR
allele (1.5fold effect size, P=4106). Using
only LDLR null alleles (that is, nonsense, indel
frameshift, and splice-site mutations) in the
burden test resulted in a 13fold excess risk of
early MI. Rare variant carriers had higher LDLcholesterol levels than noncarriers. Insummary,
these data confirm the importance of LDLR as
a contributor to risk of early MI. More impor
tantly, they lessen some of the uncertainty
surrounding the APOA cluster7 and establish
APOA5 and the triglyceride-rich lipo
proteins and lipoprotein lipase pathway as
an important contributor to risk of early MI.
In another study published in 2015, Hinson

and colleagues used an elegant series of experi
ments to explore the pathogenicity of titin
(TTN) mutations in DCM8. DCM is character
ized by progressive left ventricular dilatation,
systolic dysfunction, and eventual heart failure
and death9. Stem-cell reprogramming, gene
editing, and tissue engineering were employed
to create patient-derived induced pluripotent
stem (iPS) cell-derived cardiomyocyte t issue
models (iPS-CMTs), which were used to
evaluate the role of titin-truncating variants
(TTNtvs) in DCM. Although TTNtvs are the
most common genetic cause of DCM10, their
functional role is poorly understood and it
is unknown why only some TTN mutations
produce clinical phenotypes.
Using micropost arrays to measure the
contractile force of iPS-CMTs, tissues express
ing either Aband TTNtvs or a Z/Ijunction
missense mutation were found to exhibit less
than half the contractile force of wild-type
cells, and function did not improve over time.
Further analysis showed that the observed
functional differences were due to back
ground geneticdifferences in patient-derived
iPS-CMTs. Given that Iband TTNtvs have
been observed in individuals without DCM,
the researchers used functional testing and
RNAseq analyses to conclude that alternate
exon splicing is likely to explain penetrance dif
ferences between Iband and Aband TTNtvs.
JANUARY 2016 | 7

C A R D I O LO G Y
Bygrowing CMTs on flexible and rigid canti
levers (tomimic levels of biomechanical stress)
and treating with isoproterenol (to mimic
adrenergic stimulation), the investigators
suggested that TTNtvs can cause decreased
response to mechanicalload.
Together, these studies indicate that in
iPS-CMTs, missense mutations, and TTNtvs
can result in forms of sarcomere insufficiency
that can lead to deficits in baseline and stress-
response contractile function, factors critical in
the pathogenesis of DCM. The importance of
this study lies not only in its functional insights
with potentially profound therapeutic value,
but in its model integration and scaffolding of
complementary microbiological and molecular
biological experimental approaches.
The studies discussed in this article con
tribute to the advancement of cardiovascular
genomics in two ways. First, each study adds
to the body of knowledge within its respective
disease domain by identifying novel risk vari
ants or confirming suspected variants (FIG.1).
Furthermore, these studies illustrate that
although molecular genomic data might serve
as the basis of all future research, adequately
powered studies combining analytical, func
tional, and bioinformatic techniques need to be
integrated before even the most rudimentary
translational steps can be undertaken.
Donna K. Arnett is at the UAB School of Public Health,
Department of Epidemiology, University of Alabama at
Birmingham, 220E 1530 3rd Avenue South,
Birmingham, Alabama 35294, USA.
arnett@uab.edu
doi:10.1038/nrcardio.2015.202
Kessler,T., Erdmann,J. & Schunkert,H. Genetics
ofcoronary artery disease and myocardial infarction
2013. Curr. Cardiol. Rep. 15, 368 (2013).
2. Deloukas,P. etal. Large-scale association analysis
identifies new risk loci for coronary artery disease.
Nat.Genet. 45, 2533 (2013).
3. The CoronaryArteryDisease(C4D)Genetics Consortium.
A genome-wide association study in Europeans and
South Asians identifies five new loci forcoronary artery
disease. Nat. Genet. 43, 339344 (2011).
4. Nikpay,M. etal.A comprehensive 1000 Genomesbased genome-wide association meta-analysis of
coronary artery disease. Nat. Genet. 47, 11211130
(2015).
5. Lloyd-Jones,D.M. etal. Parental cardiovascular
disease as a risk factor for cardiovascular disease
inmiddle-aged adults: a prospective study of parents
and offspring. JAMA 291, 22042211 (2004).
6. Do,R. etal. Exome sequencing identifies rare LDLR
andAPOA5 alleles conferring risk for myocardial
infarction. Nature 518, 102106 (2015).
7. Teslovich,T.M. etal. Biological, clinical and population
relevance of 95 loci for blood lipids. Nature 466,
707713 (2010).
8. Hinson,J.T. etal. Titin mutations in iPS cells define
sarcomere insufficiency as a cause of dilated
cardiomyopathy. Science 349, 982986 (2015).
9. Hershberger,R.E., Hedges,D.J. & Morales,A. Dilated
cardiomyopathy: the complexity of a diverse genetic
architecture. Nat. Rev. Cardiol. 10, 531547 (2013).
10. Herman,D.S. etal. Truncations of titin causing dilated
cardiomyopathy. N.Engl. J.Med. 366, 619628
(2012).
1.
The author declares no competing interests.
DY S L I P I DA E M I A I N 2 0 1 5
Advances in treatment
ofdyslipidaemia
Scott M. Grundy
Statins remain the first-line therapy for dyslipidaemia. In 2015, however,
effectiveness in reducing serum cholesterol levels and decreasing rates of
cardiovascular disease in combination with statins has been demonstrated
for two new classes of drugs: cholesterol-absorption inhibitors and PCSK9
inhibitors. The latter rival statins in their capacity to lower cholesterol levels.
Findings from a meta-analysis and three sem
inal, randomized, controlled trials (RCTs) pub
lished in 2015 have far-reaching implications
for the management of cholesterol levels. Inthe
past 3decades, cholesterol-lowering therapy
has been dominated by statins. These drugs
reduce LDL-cholesterol (LDLC) levels with
few adverse effects. RCTs document their effi
cacy in reducing atherosclerotic cardiovascular
disease (ASCVD). Several statins are now avail
able as inexpensive, generic drugs. Statins are
the standard of care for secondary prevention
of ASCVD, and are being used increasingly
in primary prevention. They inhibit choles
terol synthesis and reduce hepatic cholesterol
content, which increases expression of LDL
receptors and lowers serum LDLC level. Highintensity statins can reduce LDLC levelsby
up to 60%. Two other drugs, ezetimibe and
inhibitors of proprotein convertase subtilisin/
kexin type9 (PCSK9), also lower LDLC levels
through their actions on LDL receptors (FIG.1).
A meta-analysis by the Cholesterol
Treatment Trialists (CTT) Collaboration
showed that statin therapy is equally efficacious
in reducing major cardiovascular events in
men as in women1. In this analysis, relative risk
of vascular events was reduced by 22% in men
and by 16% in women for every 1.0mmol/l
(38.6mg/dl) decrease in LDLC level. No
adverse effects on cancer or noncardiovascular
mortality were found in either men or women.
This meta-analysis confirms that for every
1% reduction in the LDLC level, the risk of
events related to ASCVD is reduced by ~1%.
Consequently, a 50% decrease in LDLC level
with high-intensity statins should reduce the
risk of ASCVD by ~50%, a benefit that applies
equally to men and women. However, the
findings sparked another question, namely
whether further reduction in LDLC levels,
beyond those achieved by treatment with high-
intensity statins, would yield an additional
reduction in risk.
8 | JANUARY 2016
To address this question, investigators

performed the randomized, controlled
IMPROVEIT trial2, in which the combination
of simvastatin (40mg) and ezetimibe (10mg)
was compared with simvastatin (40mg) and
placebo. A total of 18,144 patients, hospital
ized for an acute coronary syndrome event
within the preceding 10days and who had an
LDLC level of 1.33.3mmol/l, were recruited
into the IMPROVEIT trial, regardless of
previous treatment with statins. The mean
LDLC level was 53.7mg/dl (1.4mmol/l) in
patients receiving combination therapy, and
69.5mg/dl (1.8mmol/l) in those receiving
statin monotherapy. Compared with statin
therapy alone, combined therapy significantly
reduced the risk of future ASCVD events
LDL
LDL receptors
Statins
(4060%)
Acetate
PCSK9
inhibitors
(4060%)
Endosomal
degradation
Cholesterol
Dietary
cholesterol
Ezetimibe
(1520%)
Figure 1 | Statins, ezetimibe,
and PCSK9
inhibitors all increase the expression of LDL
receptors and reduce LDLcholesterol
levels (by percentages shown). Statins inhibit
cholesterol synthesis in the liver, ezetimibe
blocks cholesterol absorption in the intestine,
and proprotein convertase subtilisin/kexin
type9 (PCSK9) inhibitors block the
PCSK9mediated degradation of LDL receptors.
C A R D I O LO G Y
Key advances
Statin therapy reduces the risk of
atherosclerotic cardiovascular disease
toanequal extent in both men and women1
Addition of ezetimibe to statins given at
maximum dose further reduces the risk
ofcardiovascular disease compared
withstatins alone2
Proprotein convertase subtilisin/kexin
type9 inhibitors greatly enhance the
reduction of LDLcholesterol levels, even
when combined with statins, and seem
incrementally to reduce the risk of
atherosclerotic disease4,5
by 6%. The investigators estimated that to

achieve a reduction in ASCVD events, the
number-needed-totreat was ~51, regarded as
justifiable by manyexperts.
Beyond the practical use of ezetimibe in
combination with statins, the results of the
IMPROVEIT trial demonstrate that risk is
reduced proportionately to the reduction in
LDLC level, unrelated to pleiotropic effects of
statins. Ezetimibe primarily acts on the intes
tine, and has no known systemic effects (FIG.1).
The findings of this trial strongly support the
cholesterol hypothesis, which had been blurred
to some extent by claims of pleiotropism of
statins. The authors of an editorial accom
panying the publication of findings from the
IMPROVEIT trial also regarded the reduction
in LDLC level as the dominant action of stat
ins, and proposed the principle of the lower,
the better3 when assessing levels of LDLC.
The CTT investigators favoured estimating
benefits in terms of percentage LDLC lower
ing (the more, the better1), whereas the results
of the IMPROVEIT trial seemingly support a
shift in emphasis to decreasing LDLC to very
low levels (the lower, the better3).
An opportunity to test this new hypothesis
presented itself with the introduction of a new
class of drugs. Inhibition of PCSK9 by mono
clonal antibodies allows greater expression of
LDL receptors, which in turn lowers serum
LDLC concentrations (FIG.1). Two reports
published in 2015 document the efficacy of
PCSK9 inhibitors for reducing the LDLC level,
and findings strongly suggest that they also
decrease the risk of cardiovascular events4,5.
An RCT to investigate the efficacy of the
anti-PCSK9 monoclonal antibody alirocumab
included 2,341 patients at high risk of cardio
vascular events who received statins at maxi
mal tolerated doses4. Patients were randomly
assigned in a 2:1 ratio to receive either statins
in combination with 150mg alirocumab (sub
cutaneously every other week) or statins alone,
for 78weeks. Interim analysis showed that
combination therapy achieved a significantly

(62%) greater reduction in LDLC levels, and a
post-hoc analysis showed a 48% greater reduc
tion in major cardiovascular events compared
with statins alone.
Another PCSK9 inhibitor is evolocumab.
In two open-label trials, a total of 4,465
high-risk patients were randomly assigned
in a ratio of 2:1 to receive either evolocumab
plus standard therapy or standard therapy
alone5. Patients were followed up for major
cardiovascular events (median 11.1months).
In both trials, mean LDLC levels were
120mg/dl when receiving standard therapy,
whereas addition of evolocumab reduced
LDLC levels by a further 61% to 48mg/dl.
In the two trials combined, the risk of major
ASCVD events was reduced by 53% when
receiving combination therapy compared with
standard therapy. Neurocognitive events were
reported more frequently as adverse effects in
patients treated with evolocumab. These trials
are again highly suggestive, although per
haps not definitive, of an added benefit from
PCSK9inhibitors.
These two reports4,5 influenced the deci
sion of regulatory agencies in Europe and the
USA to approve PCSK9 inhibitors for limited
clinical use. Accepted indications include adult
patients with heterozygous familial hyper
cholesterolaemia and patients with clinical
ASCVD who might benefit from a further
reduction in levels of LDLC. Beyond these
indications, PCSK9 inhibitors might prove to
be useful in individuals with statin intolerance,
a condition that occurs in ~10% of patients.
In 2015, cholesterol research has moved

into a new phase. Statins are well established
as first-line therapy to lower levels of LDLC
in patients at risk of ASCVD. However, new
therapies have made it possible to achieve
greater reductions in the risk of ASCVD by
further lowering levels of LDLC. The bene
fits of combining ezetimibe with statin ther
apy have been firmly established, and PCSK9
inhibitors offer even greater potential through
their enhanced capacity to lower LDLC levels.
Scott M. Grundy is at the Center for Human Nutrition
and Department of Internal Medicine,
University of Texas Southwestern Medical Center,
5323 Harry Hines Boulevard, Suite Y3.206,
Dallas, Texas 753909052, USA.
scott.grundy@utsouthwestern.edu
doi:10.1038/nrcardio.2015.208
CholesterolTreatment Trialists (CTT) Collaboration.
Efficacy and safety of LDL-lowering therapy among
men and women: meta-analysis of individual data from
174,000 participants in 27 randomised trials. Lancet
385, 13971405 (2015).
2. Cannon,C.P. etal. Ezetimibe added to statin therapy
after acute coronary syndromes. N.Engl. J.Med. 372,
23872397 (2015).
3. Jarcho,J.A. & Keaney,J.F.Jr. Proof that lower is
better LDL cholesterol and IMPROVEIT. N.Engl.
J.Med. 372, 24482450 (2015).
4. Robinson,J.G. etal. Efficacy and safety of alirocumab
in reducing lipids and cardiovascular events. N.Engl.
J.Med. 372, 14891499 (2015).
5. Sabatine,M.S. etal. Efficacy and safety of
evolocumab in reducing lipids and cardiovascular
events. N.Engl. J.Med. 372, 15001509 (2015).
1.
Acknowledgements
S.M.G. is employed and supported by the Veterans Affairs

Medical Center, Dallas, Texas, USA.
H E A RT FA I L U R E I N 2 0 1 5
Better results from prevention than

from additional treatment
Lars Kber
In 2015, success in clinical trials in heart failure was obtained mainly from
prevention, whereas treatments showed neutral or even adverse effects.
Anew glucose-lowering medication prevents development of heart failure.
Treating central sleep apnoea might be harmful. In Chagas cardiomyopathy,
benznidazole treatment did not affect long-term clinical outcomes.
Medical treatment of heart failure has focused
on two phenotypes according to whether or
not left ventricular systolic function is pre
served. Although no drug has shown a con
vincing benefit in patients with heart failure
and preserved ejection fraction, the inhib
ition of the reninangiotensin system with
angiotensin-converting-enzyme (ACE) inhib

itors or with angiotensin-receptor blockers has
been the standard of care for 20years, mainly
owing to the results obtained in patients with
heart failure with reduced ejection fraction
(HFrEF). Following the success in 2014 with
modification of the reninangiotensin system
JANUARY 2016 | 9

C A R D I O LO G Y
Key advances
Adaptive servo-ventilation for central sleep
apnoea in systolic heart failure seems to be
harmful, and the use of this therapy should
be discouraged in these patients2
Oral treatment for type2 diabetes mellitus
with empagliflozin (an inhibitor of sodium/
glucose cotransporter2) convincingly
reduced the development of heart failure3
Glycaemic control with the dipeptidyl
peptidase4 inhibitor sitagliptin did not
increase the risk of major adverse
cardiovascular events in patients with
type2 diabetes4
The SPRINT trial showed the benefits of more
intensive control of systolic blood pressure
compared with the current standard target6
Still no effective therapy available for
Chagas cardiomyopathy after a long-term
study of benznidazole8
using the new dual angiotensinneprilysin

receptor inhibitor LCZ696, the prospect of
achieving additional benefits for patients
with HFrEF seemed difficult. This assess
ment turned out to be true, but several very
important findings have been made in 2015.
Sleep-disordered breathing is common
in patients with HFrEF. Two distinct types
of sleep-disordered breathing have been
described: obstructive sleep apnoea and cen
tral sleep apnoea. Obstructive sleep apnoea is
seen more often in patients with HFrEF than
in the general population, and central sleep
apnoea is found in 2540% of patients with
HFrEF1. Central sleep apnoea can manifest as
CheynesStokes respiration and is associated
with a poor prognosis. Central sleep apnoea
is associated with increases in sympathetic
nervous system activity and with intermittent
hypoxaemia, which can be detrimental to car
diac function. Previous studies have suggested
that continuous positive airway pressure can
reduce mortality in selected patients with
HFrEF, and devices are widely used to treat
sleep-disordered breathing. Researchers in
the SERVEHF trial2 investigated whether
adaptive servo-ventilation could reduce the
risk of death in patients with moderate-to
severe HFrEF and predominantly central
sleep apnoea. In total, 659 patients were
randomly assigned to the control group to
receive guideline-based treatment alone and
666patients to receive treatment with adap
tive servo-ventilation. The primary end point
(composite of death and hospitalization for
heart failure) was not different between the
two groups (HR 1.13, 95% CI 0.971.31,
P=0.10), but surprisingly the risk of death
from any cause was significantly higher in the
patients treated with adaptive servo-ventilation

(HR1.28, 95%CI 1.061.55, P=0.01) than in
the control group. The mechanism for this
unexpected finding remains to be elucidated,
but speculations focus on whether positive
airway pressure might impair cardiac func
tion in patients with low pulmonary wedge
pressure. An additional explanation is that
central sleep apnoea might be a compensatory
mechanism, and reducing this respiratory pat
tern might be harmful. The results from the
SERVEHF trial highlight the importance of
performing conclusive randomized, controlled
trials; otherwise, harmful treatments would be
implemented on the basis of smaller studies.
Thus, the SERVEHF trial might result in lives
being saved by preventing the use of adaptive
servo-ventilation in patients with heart failure
and central sleep apnoea.
Type2 diabetes mellitus can cause heart
failure, and heart failure can lead to type2
diabetes. Treatment of one disease can make
the other worse and, since 2008, regulatory
agencies require robust cardiovascular out
come data from randomized, controlled trials
to approve (or maintain approval of) drugs
for treatment of diabetes. Despite being effec
tive in reducing blood glucose levels, several
studies have shown that some hypoglycae
mic medications can increase, rather than
reduce, cardiovascular events. Investigators
in the EMPA-REG OUTCOME study3 tested
two dosages of empagliflozin, an inhibitor of
the sodium/glucose cotransporter2 (SGLT2)
(FIG. 1) , against placebo in 7,020 patients
with type2 diabetes and high cardiovascu
lar risk. The primary outcome (death from
cardiovascular disease, nonfatal myocardial
infarction, or nonfatal stroke) was reduced
by 14% (HR0.86, 95%CI 0.740.99, P=0.04)
in the pooled group of patients who received
empagliflozin compared with the placebo
group. Importantly, both cardiovascular and
all-cause mortality were significantly reduced

by 38% and 35%, respectively. Hospitalization
for heart failure was reduced by 35% (HR0.65,
95%CI 0.500.85, P=0.002) in the empagli
flozin group compared with the placebo
group, and in the 10% of patients with heart
failure at baseline, the results for reduction in
cardiovascular and all-cause mortality were as
positive as for the overall trial (29% and 21%
reduction, respectively). The EMPA-REG
OUTCOME study suggests that empagli
flozin can prevent heart failure in patients with
type2 diabetes, and might also reduce mortal
ity in the high-risk group of patients with both
diabetes and heart failure.
Dipeptidyl peptidase 4 (DPP4) inhibi
tors have been used in patients with diabetes
for almost 10years; however, concerns that
these inhibitors might increase heart failure-
related events have been raised. Sitagliptin, a
DPP4 inhibitor, was tested against placebo in
the large TECOS trial4 that included 14,671
patients with a baseline level of glycated
haemoglobin (HbA1c) of 6.58.0% and with
diagnosed cardiovascular disease. The primary
composite end point of cardiovascular mor
tality, nonfatal myocardial infarction, nonfatal
stroke, or hospitalization for unstable angina
was neutral (HR0.98, 95%CI 0.881.09), with
a highly significant Pvalue for noninferiority
(P<0.001). The incidence of heart failure was
similar in the two groups (HR1.0, 95%CI
0.831.20, P=0.98), indicating that DPP4
inhibitors might not have a harmful class effect
in patients with heart failure.
Glucagon-like peptide1 receptor agonists
(GLP1RAs), a class of parenteral glucose-
lowering drugs that activate the receptor for the
endogenous incretin GLP1, are also becoming
widely used to treat type2 diabetes. GLP1RAs
improve glycaemic control with a low risk of
hypoglycaemia, and can even result in weight
loss. Lixisenatide, a GLP1RA, was tested in
~90% of glucose renal reabsorption
SGLT2 inhibitors
SGLT2
Glucose
Collecting
duct
SGLT1
No glucose
S1 segment of
proximal tubule
Distal S2/S3 segment
of proximal tubule
Glycosuria
Figure 1 | Effects of SGLT2 inhibitors. Sodium/glucose cotransporter (SGLT)2 inhibitors increase

urinary glucose excretion, thereby reducing glucose levels in the circulation. Permission obtained
from Nature Publishing Group Chao, E. C. & Henry, R. R. SGLT2 inhibition a novel strategy for
diabetes treatment. Nat. Rev. Drug Discov. 9, 551559 (2010).
10 | JANUARY 2016
C A R D I O LO G Y
6,068 patients with type2 diabetes and a recent
coronary event in the ELIXA study5. The pri
mary composite end point of cardiovascular
mortality, nonfatal myocardial infarction,
stroke, or hospitalization for unstable angina
was similar in the two groups (HR1.02, 95%CI
0.891.17) and demonstrated noninferiority
(P<0.001) to placebo. Hospitalizations for
heart failure did not differ between the two
groups (HR0.96, 95%CI 0.781.13).
Antihypertensive treatment reduces the
risk of stroke, myocardial infarction, and heart
failure, but the target for blood-pressure lower
ing is unknown. Researchers in the SPRINT
trial6 investigated whether reduction of systolic
blood pressure to <120mmHg was superior to
the standard target of <140mmHg. The study,
which included a total of 9,361 patients, was
stopped prematurely because of a significant
reduction in the primary composite end point
of cardiovascular mortality, nonfatal myo
cardial infarction, stroke, acute coronary event,
or hospitalization for heart failure in the group
with the lower blood-pressure target. In this
group, all-cause mortality was significantly
reduced (HR0.73, 95%CI 0.600.90, P=0.003)
compared with the standard-treatment
group, as was hospitalization for heart failure
(HR0.62, 95%CI 0.450.84, P=0.002). The
SPRINT trial substantially adds to the evidence
for benefits associated with more aggressive
lowering of systolic blood pressure compared
with current standard targets and, importantly,
the results were obtained by using a variety of
antihypertensive drugs.
In Latin America, the most common
form of nonischaemic cardiomyopathy is
Chagas disease, which develops 2030years
after the initial infection by the parasite
Trypanosomacruzi in approximately 25% of

patients. The acute infection can be treated
effectively, but whether chronic disease is
caused by the persistence of the parasite or
by an autoimmune response is unknown.
The antigens from the parasite can be meas
ured in the myocardium, and treatment with
benznidazole has shown that the parasite load
can be reduced7. Researchers in the BENEFIT
trial8 investigated whether oral benznidazole
was superior to placebo in 2,854 patients with
established Chagas cardiomyopathy. Treatment
was given for 4080days, but had no effect on
the primary composite end point of mortality,
resuscitated cardiac arrest, sustained ventricu
lar tachycardia, insertion of a pacemaker or
implantable cardioverterdefibrillator, cardiac
transplantation, new heart failure, stroke, or
other thromboembolic event (HR0.93, 95%CI
0.811.07, P=0.31). At baseline, 60% of
patients had positive results for Trypanosoma
cruzi on PCR, which was reduced by twothirds in the benznidazole group and one-third
in the placebo group. These rates of conversion
to negative PCR in patients with established
Chagas cardiomyopathy were lower than the
rates of conversion found in the acute phase
of the disease9. This difference in the conver
sion to negative PCR might be associated with
the lack of effectiveness of benznidazole in the
chronic phase of the disease. Despite the dis
appointing results, the BENEFIT trial was very
important in highlighting the need for further
research into treatment strategies for Chagas
cardiomyopathy, the third most c ommon
parasitic disease globally.
Despite a 5year mortality of nearly 50%
associated with heart failure, further lowering
of the risk of death was not achieved in 2015.
A positive finding is that heart failure might be

prevented in patients with diabetes or hyper
tension, but substantial improvements can still
be achieved. The need for an effective treat
ment for heart failure with preserved ejection
fraction remains, as well as a conclusive trial
in acute heart failure. Lastly, we still have to
see a benefit from stem-cell therapies for heart
failure, but additional benefits for patients with
heart failure might be achieved in 2016.
Lars Kber is at the Department of Cardiology, the
Heart Centre, Rigshospitalet, University of
Copenhagen, Blegdamsvej 9, 2100 Copenhagen,
Denmark.
lars.koeber@regionh.dk
doi:10.1038/nrcardio.2015.205
1. Levy,P. etal. Prevalence and impact of central sleep
apnea in heart failure. Sleep Med. Clin. 2, 615621
(2007).
2. Cowie,M.R. etal. Adaptive servo-ventilation for central
sleep apnea in systolic heart failure. N.Engl. J.Med.
373, 10951105 (2015).
3. Zinman,B. etal. Empagliflozin, cardiovascular
outcomes, and mortality in type2 diabetes. N.Engl.
J.Med. 373, 21172128 (2015).
4. Green,J. B. etal. Effect of sitagliptin on cardiovascular
outcomes in type2 diabetes. N.Engl. J.Med. 373,
232242 (2015).
5. Pfeffer,M.A. etal. Lixisenatide in patients with type2
diabetes and acute coronary syndrome. N.Engl.
J.Med. 373, 22472257 (2015).
6. SPRINT Research Group. A randomized trial of
intensive versus standard blood-pressure control.
N.Engl. J.Med. 373, 21032116 (2015).
7. Fabbro,D. L. etal. Trypanocide treatment among
adults with chronic Chagas disease living in Santa Fe
city (Argentina), over a mean followup of 21years:
parasitological, serological and clinical evolution.
Rev.Soc. Bras. Med. Trop. 40, 110 (2007).
8. Morillo,C. A. etal. Randomized trial of benznidazole
for chronic Chagas cardiomyopathy. N.Engl. J.Med.
373, 12951306 (2015).
9. de Andrade,A.L. etal. Randomised trial of efficacy of
benznidazole in treatment of early Trypanosoma cruzi
infection. Lancet 348, 14071413 (1996).
The author has received honoraria as a speaker for Novartis

and Servier.
JANUARY 2016 | 11

Connect to...
Follow us on Twitter
For the latest news from the journal and beyond: @NatRevCardiol
Nature Reviews on Facebook
For regular updates on new content across all the Nature Reviews journals https://www.facebook.com/NatureReviews
Sign up for e-ALERTS
Get monthly e-mail alerts to the content of this journal sent FREE to your inbox by registering online
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrcardio/info/newsfeeds.html
Advance online publication
We operate an advance online publication (AOP) service for authors and readers to view the latest articles published online
ahead of print http://www.nature.com/nrcardio/journal/vaop/ncurrent
Web focuses and collections
A selection of articles on a related theme, with content ranging from Research Highlights through to Reviews and Perspectives
http://www.nature.com/nrcardio/focus/index.html & http://www.nature.com/nrcardio/collection/index.html
Posters and multimedia
Providing visual and audible overviews of hot topics http://www.nature.com/nrcardio/posters & http://www.nature.com/
nrcardio/multimedia/index.html
Conferences
You can meet the editors of Nature Reviews Cardiology at some of the meetings listed on our conferences page
http://www.nature.com/nrcardio/info/info_conf.html
A Decade in Medicine
The past 10 years has seen great advances in the understanding and treatment of human disease. For expert
perspectives on the most important breakthroughs, dont miss the A Decade in Medicine eBook.
In this special collection of 47 articles commissioned to celebrate the 10th anniversary of the launch of the
clinical Nature Reviews journals leading experts highlight the most important advances in eight medical
specialties between 2004 and 2015, and comment on future developments in their fields.
Download the eBook free here: http://www.nature.com/reviews/collection/adecadeinmedicine/index.html
CLINICAL ONCOLOGY
BREAST CANCER IN 2015
Academic research sheds light on issues

that matter to patients
Martine J.Piccart and Isabelle Gingras
In 2015, academic-led trials provided evidence for safe deescalation of
adjuvant treatment in early stage breast cancer and answered important
questions related to adjuvant regional irradiation and optimal first-line
chemotherapy in advanced-stage disease. Furthermore, the development
ofnovel therapies and potential tools for treatment tailoring will offer new
hope to patients with breast cancer.
recurrence observed at a median followup of
nearly 4years following 3months of weekly
paclitaxel and 1year of trastuzumab1.
Results of recent clinical trials are now
challenging the dogma that superior drug
regimens in advanced-stage breast cancer
translate into more-efficacious adjuvant therapies in the early disease setting. Because ovarian
suppression combined with tamoxifen treatment was associated with superior outcomes
in metastatic breast cancer (mBC), many
clinicians were tempted to generalize its use in
adjuvant endocrine therapy for young women
with luminal disease. The SOFT trial2 has been
important in refuting this generalization, as
the benefit of adding ovarian suppression to
tamoxifen (which is associated with increased
rates of grade 3 or higher adverse events) is
restricted to women who were at sufficient
risk of recurrence to warrant adjuvant chemotherapy. Of note, the SOFT trial only included
patients who remained premenopausal after
neoadjuvant chemotherapy2.
TAILORx is a very large prospective, academic, USbased trial designed to refine
adjuvant chemotherapy withdrawal using the
results of Oncotype DXa prognostic and
predictive gene-expression signature already
endorsed by many oncologists3. The primary question of the trial pertains to women
harbouring node-negative tumours with an
intermediate risk based on Oncotype DX
score, in whom the benefit of the addition of
adjuvant chemotherapy to endocrine therapy
is being investigated; results for this cohort
are not currently available. In 2015, however,
the trial reported the excellent outcome of
S.Bradbrook/NPG
Academic research in oncology, which is

patient-centred and provides answers to
questions with limited or no commercial interest, has been under threat in the past decade.
However, seven academic clinical trials reported
in 2015 reassure us that research remains alive,
and that academic clinical investigators continue
to find the energy needed to secure the financial
support for this critically important research.
Through these efforts, questions important to
patients with breast cancer regarding treatment
deescalation, the value of expensive new drugs
in relation to their standard counterparts, and
optimal radiotherapy techniques in the curative setting have received answers that will
undoubtedly affect clinical practice.
A popular and relatively low-risk path for
anticancer drug development consists of adding
new agents to the standard of care. As a result,
we have witnessed an escalation in drug prescription, with an exponential rise in drug costs.
Withholding some of the most toxic components of new standard-ofcare treatments in
carefully selected patients with breast cancer
is an important but difficult task for academic
investigators to undertake, but this approach
has been successful in other diseases, mostly
in paediatric malignancies. An example comes
from a single-arm phase II trial from the
DanaFarber Cancer Institute, which explored
withdrawal of anthracyclines from the adjuvant chemotherapy backbone for low-stage,
HER2positive breast cancer in patients with
mostly T1T2 and node-negative disease1. The
decision to withdraw anthracyclines to develop
a less-aggressive standard of care is supported
by the extremely low rate of distant disease
women with a low-risk Oncotype DX score

and showed that following adjuvant endocrine
therapy, 98.7% of these women remain free
from recurrence at any site3, highlighting that
chemotherapy can be avoided in these patients
with no detriment to outcome.
Adjuvant trials that aim to optimize radio
therapy fields, doses, or schedules are courage
ous endeavours, as they are notoriously difficult
to fund and require large numbers of patients
and very long followup times. Two such trials
with mature 10year follow up were published,
and provided consistent results despite differences in study size and patient selection: the
addition of regional nodal irradiation to wholebreast or thoracic-wall radiotherapy does not
improve overall survival, but does reduce the
rate of local and distant breast cancer recurrence4,5. The magnitude of this treatment effect
is modest and additional translational research
efforts are needed to identify biological
subgroups with enhanced benefit.
mBC is the setting in which all new drugs
must pass the first hurdle in demonstrating efficacy, which is usually established in
randomized phaseIIIII trials in comparison
with the standards of care. Not surprisingly,
the clinical research landscape is dominated by
industry-led trials. In this regard, the CALGB
and NCCTG should be congratulated for their
headtohead comparisons of nab-paclitaxel
and ixabepilone versus paclitaxel in a first-line
chemotherapy trial in 799 patients with mBC.
JANUARY 2016 | 12

The winner in terms of progression-free survival (PFS) and timetotreatment failure was
the relatively cheap drug paclitaxel6.
Similarly, our colleagues in the UK can be
proud of their practice-changing trial7 of up
front carboplatin (AUCx6 once every 3weeks)
versus docetaxel (100mg/m2 every 3weeks) in
women with metastatic triple-negative breast
cancer (TNBC): a superior response rate for the
alkylating agent over the taxane was demonstrated in 43 patients with BRCA-mutated
tumours, representing around 8% of the trial
population. When receiving carboplatin,
patients with BRCA-mutated tumours had
a median PFS of 6.8months compared with
3.1months in BRCA-wild-type patients7. Full
publication of this important trial is pending.
Following over two decades of intensive
basic research effort to dissect the molecular
mechanisms of endocrine resistance in breast
cancer, we can celebrate the rewards of the successful registration of new targeted compounds
that delayed the onset of treatment resistance in
patients with mBC. Letrozole combined with
the CDK4/6 inhibitor palbociclib received
conditional accelerated approval by the FDA
as the first-line endocrine-based therapy for
mBC, on the basis of extended PFS results in
a randomized phaseII trial8; palbociclib also
improved PFS by 5months in combination
with fulvestrant in a phaseIII trial involving 521
patients with relapsed or progressive disease
during prior endocrine therapy9. Palbociclib
is associated with a manageable toxicity profile
(neutropenia and fatigue), and maintenance of
global quality of life9. Whether palbociclib will
provide better survival outcomes than evero
limus remains unknown. Of note, no biomarkers for either of these compounds have
been validated, beyond the oestrogen receptor,
which is a serious problem in view of the steep
rises in drug prescription costs.
A huge unmet medical need remains for
women with metastatic TNBC, as this disease
usually progresses rapidly following three to
five lines of chemotherapy; to date, no targeted

drug has been registered in this setting, apart
from bevacizumab, the clinical value of which
has been questioned because no overall survival gain could be demonstrated. Considerable
interest has been generated by the encouraging
results obtained with the immune-checkpoint
inhibitorspembrolizumab and atezolizumabin
heavily pretreated women with advancedstage TNBC. The overall response rate to these
agents remains modest (~18%), but the median
duration of response had not been reached at
the time of publication, indicating sustained
responses rarely observed in patients with
heavily pretreated TNBC1012. Combinations
of immunotherapies with selected cytotoxic
agents or signal-transduction inhibitors will
undoubtedly become the focus of intense
clinical investigations in the coming years.
The association of dense stromal tumour
infiltrating lymphocytes (TILs) with
improved outcomes for certain breast cancer subtypes has long been recognized by
pathologists. In the NeoALTTO trial13, the
demonstration of a strong prognostic value of
TILs for both pathological complete response
and event-free survival(independent of therapy)in patients with HER2positive disease,
could lead to an actionable biomarker for
treatment deescalation, if validated in the
very large ALTTO adjuvant trial.
Finally, 2015 will be remembered for the
powerful demonstration of molecular imaging
to decipher the considerable tumour hetero
geneity in advanced-stage HER2positive breast
cancer. The academic-led ZEPHIR trial, in
which 56 patients with HER2positive disease
underwent a HER2 PETCT assessment before
receiving T-DM1, revealed that almost half of
the patients had substantial heterogeneity in
HER2 expression between different meta
stases14. Moreover, 29% of the patients had a
negative HER2 PETCT; importantly, patients
with a negative HER2 PETCT scan showed
marginal benefit from T-DM1 therapy14. The
consequences of these findings are potentially

substantial with respect to improving our
ability to identify who might benefit from this
antiHER2 antibody drug conjugate.
Martine J. Piccart is at the Department of Medical
Oncology, Institut Jules Bordet/Universit libre de
Bruxelles, 1 Rue Hger-Bordet, 1000 Bruxelles,
Brussels, Belgium.
Isabelle Gingras is at the Department of Hematology
and Oncology, Hpital du Sacr-Cur de Montral/
Universit de Montral, 5400 Boulevard Gouin,
Montreal, Quebec H4J 1C5, Canada.
Correspondence to M.J.P.
martine.piccart@bordet.be
doi:10.1038/nrclinonc.2015.236
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Key advances
The excellent disease-free survival observed with adjuvant paclitaxel and trastuzumab without
anthracyclines have set a new standard of care for the adjuvant treatment of small (T1T2),
node-negative, HER2positive breast cancer, and highlighted the possibility to deescalate
treatment for low-risk, early stage, HER2positive disease1.
The very low incidence of breast-cancer recurrence observed in node-negative patients
withalow risk Oncotype DX score confirms that these patients should be spared
adjuvantchemotherapy3.
New data suggest that carboplatin can improve progression-free survival in patients with
BRCA-mutated tumours when used as first-line therapy for metastatic disease7.
In 2015, two studies demonstrated that palbocilib can improve progression-free survival
of patients with metastatic disease in both the first-line8 and second-line9 settings, and
palbociclib in combination with letrozole received conditional FDA approval for the first-line
treatment of oestrogen-receptor-positive, HER2negative metastatic breast cancer.
13 | JANUARY 2016
13.
14.
Tolaney,S.M. etal. Adjuvant paclitaxel and

trastuzumab for node-negative, HER2positive breast
cancer. N.Engl. J.Med. 372, 134141 (2015).
Francis,P.A. etal. Adjuvant ovarian suppression in
premenopausal breast cancer. N.Engl. J.Med. 372,
436446 (2015).
Sparano,J.A. etal. Prospective validation of a 21gene
expression assay in breast cancer. N.Engl. J.Med. 373,
20052014 (2015).
Whelan,T.J. etal. Regional nodal irradiation in earlystage breast cancer. N.Engl. J.Med. 373, 307316
(2015).
Poortmans,P.M. etal. Internal mammary and
medial supraclavicular irradiation in breast cancer.
N.Engl. J.Med. 373, 317327 (2015).
Rugo,H.S. etal. Randomized phaseIII trial of paclitaxel
once per week compared with nanoparticle albuminbound nab-paclitaxel once per week or ixabepilone
with bevacizumab as first-line chemotherapy for
locally recurrent or metastatic breast cancer:
CALGB 40502/NCCTG N063H (Alliance). J.Clin. Oncol.
http://dx.doi.org/10.1200/jco.2014.59.5298 (2015).
Tutt,A. etal. Abstract S3-01: The TNT trial: a randomized
phaseIII trial of carboplatin (C) compared with docetaxel
(D) for patients with metastatic or recurrent locally
advanced triple negative or BRCA1/2 breast cancer
(CRUK/07/012). Cancer Res. 75, S3-01 (2015).
Finn,R.S. etal. The cyclin-dependent kinase 4/6
inhibitor palbociclib in combination with letrozole versus
letrozole alone as first-line treatment of oestrogen
receptor-positive, HER2negative, advanced breast
cancer (PALOMA1/TRIO18): a randomised phase2
study. Lancet Oncol. 16, 2535 (2015).
Turner,N.C. etal. Palbociclib in hormone-receptorpositive advanced breast cancer. N.Engl. J.Med. 373,
209219 (2015).
Emens,L.A. etal. Inhibition of PDL1 by MPDL3280A
leads to clinical activity in patients with metastatic triplenegative breast cancer. [abstract PD16], Presented at
the San Antonio Breast Cancer Symposium (2014).
Nanda,R. etal. A phaseIb study of pembrolizumab
(MK3475) in patients with advanced triple-negative
breast cancer. [abstract S109], Presented at the San
Antonio Breast Cancer Symposium (2014).
Dirix,L.Y. etal. Avelumab (MSB0010718C), an
antiPDL1 antibody, in patients with locally advanced
or metastatic breast cancer: a phaseIb JAVELIN solid
tumor trial. [abstract S104], Presented at the San
Antonio Breast Cancer Symposium (2015).
Salgado,R. etal. Tumor-infiltrating lymphocytes and
associations with pathological complete response and
event-free survival in HER2positive early-stage
breast cancer treated with lapatinib and trastuzumab:
a secondary analysis of the NeoALTTO trial.
JAMA Oncol. 1, 448454 (2015).
Gebhart,G. etal. Molecular imaging as a tool to
investigate heterogeneity of advanced HER2positive
breast cancer and to predict patient outcome under
trastuzumab emtansine (T-DM1): the ZEPHIR Trial.
Ann. Oncol. http://dx.doi.org/10.1093/annonc/mdv577
(2015).
M.J.P. declares she is a consultant for and receives honoraria

from AstraZeneca, Invivis, MSD, Novartis, Pfizer, Radius,
Roche-Genentech and Synthon. I.G. declares no competing
interests.
C O L O R E C TA L A N D G A S T R I C C A N C E R I N 2 0 1 5
The development of new agents

andmolecular classifications
Eric Van Cutsem and Michel Ducreux
In a little over the past year, several clinical trials have evaluated new drugs in
patients with metastatic colorectal cancer and gastric cancer. Furthermore,
genomics studies that attempted to unravel the molecular characteristics of
colorectal and gastric cancer were published in 2015. Theresults of these
endeavours will influence clinical practice in 2016 andbeyond.
The median survival of patients with metastatic colorectal cancer (mCRC) has improved
over the past 15years: in contemporary trials
median survival durations of approximately
30months have been reported1. Several new
drugs have contributed to this advance; however, an unmet need for new agents with a
greater clinical effect, which can break this
wall of 30months, remains. By contrast, the
median survival of patients with metastatic
gastric cancer (mGC) remains poor and,
generally, does not exceed 12months. Over
the past year, encouraging data on three new
treatments for mCRC and a new targeted
therapy for mGC have been presented. In
addition, new genetic classifications have
been proposed that might help to further
advance p
ersonalized molecular medicine
for thesediseases.
Over the past year,

encouraging data on three new
treatments for mCRC and a
new targeted therapy for mGC
have been presented
One of the agents that have shown promise in the treatment of mCRC is TAS102
(Taiho Oncology, Japan), which is a combination drug comprising the active cytotoxic
compound trifluridine, a thymidine-based
nucleic-acid analogue, and tipiracil hydrochloride. The triphosphate form of trifluridine is incorporated directly into DNA while
the monophosphate form inhibits thymidy
late synthase, with both of these activities
ultimately resulting in DNA damage and,
thus, antitumour effects; tipiracil is a potent
inhibitor of thymidine phosphorylase and,
when delivered as part of TAS102, prevents

the rapid degradation of trifluridine. In the
phaseIII RECOURSE study2, patients with
chemorefractory mCRC were randomly
assigned to receive best supportive care plus
either TAS102 or placebo. The median overall survival improved to 7.1 months with
TAS102, from 5.3months with placebo (hazard ratio (HR) for death 0.68, 95% confidence
interval (CI) 0.580.81; P<0.001)2. This effect
of TAS102 on survival is similar to the bene
fit reported previously for the multitargeted
receptor-tyrosine-kinase inhibitor regoran
efib in patients with chemorefractory mCRC
in the CORRECT trial3; however, the adverseevent profiles associated with TAS102 and
regoranefib are different. Both drugs are,
therefore, now approved and indicated for
the treatment of chemorefractory mCRC,
although the optimal sequence of these
drugs remains to be determined. Of potential relevance in this regard, pivotal work was
reported in 2015 on the analysis of circulating
DNA and protein biomarkers to predict the
clinical activity of regorafenib4. The authors
of this study did not identify any prognostic
or predictive biomarkers that can be used in
current routine clinical practice; however, the
investigators did demonstrate that BEAMing
(beads, emulsion, amplification, magnetics)
analysis of circulating DNA could be a viable, non-invasive approach to the evaluation
of tumour genotype in real-time that enables
the identification of potentially clinically relevant mutationsincluding mutations that go
undetected in archival tissues4. Indeed, such
liquid biopsy techniques are rapidly evolving
and might have a prominent role in the future
of biomarker analyses, in particular, because
the dynamic molecular changes that occur
during treatment could potentially be captured
more effectively using these methods, rather
than traditional approaches4.
The anti-EGFR antibodies cetuximab

and panitumumab are active in patients
with RAS-wild-type mCRC, either as single agents or in combination with cytotoxic
multidrug chemotherapy regimens. Sym004
(Symphogen, Denmark) is a novel mixture of
two recombinant, humanmouse chimeric
monoclonal antibodies that target non-
overlapping EGFR-ectodomain epitopes5.
Aunique feature of Sym004 is its ability to
mediate rapid EGFR internalization and
subsequent degradation via EGFR crosslinking. In preclinical studies, Sym004 treatment
resulted in both superior induction of tumour
regression, as compared with other anti-EGFR
antibodies, and antitumour activity in models
of acquired cetuximab resistance6. In a phaseI
study5, Sym004 therapy was associated with a
clinically meaningful rate of partial response
(13%) or minor tumour regression (44%) of
target lesions in patients with mCRC who
had acquired resistance to prior anti-EGFR
monoclonal antibody treatment. The activity
of Sym004 in the setting of acquired resistance
forms a strong clinical basis to be tested in
future trials. Indeed, a trial to further investi
gating the optimal dose and dosing schedule of this agent in patients with mCRC are
currently underway (NCT02083653).
...this classification will

probably facilitate future CRC
research and might further
improve patient outcomes
The interaction between the
immune-checkpoint receptor programmed
cell-death protein 1 (PD1), and its ligands,
PDL1 and PDL2, is a key pathway that is
hijacked by tumours to suppress immune
control. Reversing this inhibition of adaptive
immunity can lead to active stimulation of
a patients immune system and antitumour
responses. This rationale underlies the use
of antagonistic antibodies that target PD1
or PDL1, as well as cytotoxic Tlymphocyteassociated antigen 4 (CTLA4), as anticancer
therapies; indeed, activity against a variety
of tumour types has been demonstrated for
these immunotherapies. In a small, but pivo
tal, phaseII trial that was published in the
past year7, the antiPD1 antibody pembrolizumab was used to treat tumours with or
without mismatch-repair (MMR) deficiency,
predominantly in patients with mCRC.
The results of the trial support the hypothesis that MMR-deficient tumours or
JANUARY 2016 | 14

tumours that demonstrate microsatellite
instability (MSI)are more responsive to
PD1 blockade than MMR-proficientor
microsatellite-stabletumours7. In particular, immune-related objective responses and
20week immune-related progression-free survival (PFS) were achieved in four of 10 patients
(40%) and seven of nine patients (78%) with
MMR-deficient mCRC, respectively, compared
with none of 18 patients (0%) and two of 18
patients (11%) with MMR-proficient mCRC7.
Furthermore, excellent survival outcomes
were observed in patients with MMR-deficient
mCRC: both median PFS and overall survival
were not reached after a median followup
duration of 36months in this group, but were
only 2.2monthsand 5months, respectively,
in patients with MMR-proficient mCRC7.
MMR-deficiency occurs in many cancers,
and in patients with CRC, it can arise from an
inherited germ-line genetic defect, in the context of Lynch syndrome (hereditary nonpolyposis colorectal cancer),or it can be sporadic.
In either case, the resulting neoplasms harbour
hundreds or thousands of mutations, and are
considered as hypermutated and, therefore,
highly immunogeniclesions7.
With regard to mGC, active first-line and
second-line treatment optionsremain a major
unmet clinical need. Until now, trastuzumab
was the only targeted agent with clinically-
relevant activity in patients with gastric cancer.
Early results with antiVEGFR2 monoclonal
antibody ramucirumab indicated that monotherapy with this agent significantly prolongs
the survival of previously treated patients with
mGC, although the magnitude of the benefit
remains limited (median overall survival of
5.2months versus 3.8months with placebo)8.
More-recent data indicate, however, that the
combination of ramucirumab with paclitaxel
improves overall survival and could, therefore,
be regarded as a new standard second-line
treatment for patients with mGC9. Specifically,

in a phaseIII study9 published late in 2014,
the overall survival of patients treated with
ramucirumab plus paclitaxel was significantly
longer than that of patients who received placebo plus paclitaxel (median 9.6months versus
7.4months; HR 0.807, 95% CI 0.6780.962;
P=0.017); moreover, ramucirumab therapy
was associated with only a modest increase
intoxicity9.
Another key development in late 2014
was the publication of The Cancer Genome
Atlas (TCGA) molecular classification that
divides gastric adenocarcinomas into four
subtypes: EpsteinBarr virus (EBV)-infected
tumours; MSI tumours; genomically stable tumours; and chromosomally unstable
tumours10. The EBV-positive tumours display
recurrent PIK3CA mutations, extreme DNA
hypermethylation, and amplification of the
genes encoding JAK2, PDL1, and PDL2. MSI
tumours have a high frequency of mutations
compared with the other subtypes, including
mutations in genes encoding targetable oncoproteins10. Genomically stable tumours are
enriched for the diffuse histological variant
and mutations in RhoA or fusions involving
Rho-family GTPase-activating proteins10.
Finally, tumours with chromosomal instability, demonstrate marked aneuploidy and
focal amplification of receptor tyrosine kinase
genes10. Identification of these subtypes provides a roadmap for patient stratification
and trials of targeted therapies, which might
contribute to further improving outcomes of
thisdisease.
Similarly, in 2015, an international consortium has proposed a gene-expression-based
subtyping classification system for CRC that
defines four consensus molecular subtypes
(CMSs) of CRC with distinguishing features11:
CMS1 (MSIimmune, 14%), demonstrating
hypermutation, MSI, and strong immune
Key advances
On the basis of an overall survival improvement demonstrated in the RECOURSE study2
published in 2015, TAS102 joins regorafenib as an FDA-approved treatment for chemorefractory
metastatic colorectal cancer (mCRC)
A 2015 study did not identify circulating biomarkers that predict response to regorafenib in
patients with mCRC, but nevertheless demonstrated the potential of liquid biopsies to improve
personalized medicine4
New data suggest that the novel anti-EGFR antibody mixture Sym0045, and the
immune-checkpoint inhibitor pembrolizumab7 hold considerable promise to further advance
mCRC outcomes
Recent data for ramucirumab support the use of this agent with paclitaxel as standard
second-line therapy for chemorefractory metastatic gastric cancer9
In the past year, classifications were proposed for both gastric cancer10 and colorectal cancer11,
which might facilitate patient stratification and further improve the outcomes of patients with
these diseases
15 | JANUARY 2016
activation; CMS2 (canonical, 37%), epithelial

tumours with marked activation of WNT and
MYC signalling; CMS3 (metabolic, 13%), epithelial tumours with evident metabolic dysregulation; and CMS4 (mesenchymal, 23%),
with prominent activation of TGF signalling,
stromal invasion, and angiogenesis. As in the
setting of gastric cancer, this classification will
probably facilitate future CRC research and,
together with the ongoing development of the
therapies that have shown promise in 2015,
might further improve patient outcomes.
Eric Van Cutsem is at the Digestive Oncology,
University Hospitals Leuven and KU Leuven,
Herestraat49, 3000 Leuven, Belgium.
Michel Ducreux is at the Gastrointestinal Unit,
Department of Medicine, Gustave Roussy, 114 Rue
Edouard Vaillant, 94805 Villejuif Cedex, France, and
the Faculty of Medicine, Paris Sud University, 63 Rue
Gabriel Pri, 94270 Le Kremlin Bictre, France.
Correspondence to E.V.C.
eric.vancutsem@uzleuven.be
Published online 15 Dec 2015
Heinemann,V. etal. FOLFIRI plus cetuximab versus
FOLFIRI plus bevacizumab as first-line treatment for
patients with metastatic colorectal cancer (FIRE3):
arandomised, open-label, phase 3 trial. Lancet Oncol.
15, 10651075 (2014).
2. Maier,R. etal. Randomized trial of TAS102 for
refractory metastatic colorectal cancer. N.Engl.
J.Med. 372, 19091919 (2015).
3. Grothey,A. etal. Regorafenib monotherapy for
previously treated metastatic colorectal cancer
(CORRECT): an international, multicentre,
randomised, placebo-controlled, phase 3 trial. Lancet
381, 303312 (2013).
4. Tabernero,J. etal. Analysis of circulating DNA and
protein biomarkers to predict the clinical activity of
regorafenib and assess prognosis in patients with
metastatic colorectal cancer: a retrospective,
exploratory analysis of the CORRECT trial.
LancetOncol. 16, 937948 (2015).
5. Dienstmann,R. etal. Safety and activity of the
firstinclass Sym004 anti-EGFR antibody mixture
inpatients with refractory colorectal cancer.
CancerDiscov. 5, 112 (2015).
6. Lida,M. etal. Sym004, a novel EGFR antibody
mixture, can overcome acquired resistance to
cetuximab. Neoplasia 15, 11961206 (2013).
7. Le,D. etal. PD1 blockade in tumors with mismatchrepair deficiency. N.Engl. J.Med. 372, 25092520
(2015).
8. Fuchs,C. etal. Ramucirumab monotherapy for
previously treated advanced gastric or gastro
oesophageal junction adenocarcinoma (REGARD):
aninternational, randomised, multicentre, placebocontrolled, phase 3 trial. Lancet 383, 3139 (2014).
9. Wilke,H. etal. Ramucirumab plus paclitaxel versus
placebo plus paclitaxel in patients with previously
treated advanced gastric or gastrooesophageal
junction adenocarcinoma (RAINBOW): a double-blind,
randomised phase 3 trial. Lancet Oncol. 15,
12241235 (2014).
10. TheCancer GenomeAtlas ResearchNetwork.
Comprehensive molecular characterization of gastric
adenocarcinoma. Nature 513, 202209 (2014).
11. Guinney,J. etal. The consensus molecular subtypes
ofcolorectal cancer. Nat. Med. 21, 13501356
(2015).
1.
E.V.C. has received research grants from Amgen, Bayer,

Boehringer, Celgene, Ipsen, Lilly, Merck Serono, Novartis,
Roche, and Sanofi. M.D. is on the advisory board and has
attended symposia for Amgen, Celgene, Lilly, Merck,
Novartis, and Roche. M.D.s wife is employed by Sandoz
(Novartis group).
100
90
80
70
60
50
40
30
20
10
SD
PD
er
th
O
at
e
Pr
os
t
cr
ea
Pa
n
st
ORR
Br
ea
comparator. The analysis identified a number

of disease-specific and demographic imbalances among the facilities but, interestingly,
no significant difference in disease stage at
presentation was reported. DSS was significantly different between the three categories of facilities, with an all-stage median of
77.9months, 51.9months, and 43.4months
for patients treated at NCI-CCCs, HVCs, and
LVCs, respectively2. After correcting for the
negative effect of age, and tumour stage, size
and grade, the risk of disease-specific death
was 18% and 30% higher at HVCs and LVCs,
respectively, than at NCI-CCCs. Importantly,
compliance with NCCN guideline-b ased
treatment was significantly lower in HVCs
and LVCs compared with NCI-CCCs in
terms of overall adherence (odds ratio (OR):
0.83 and 0.56, respectively), surgery adherence (OR: 0.83 and 0.68), and chemotherapy adherence (OR: 0.82 and 0.54)2. Only
36% of the 9,933 patients treated across all
centres received NCCN-adherent surgery
and chemotherapy care2. The authors concluded that care at an NCI-CCC serves as
a highly significant independent predictor
of adherence to treatment guidelines and,
more importantly, DSS. In this study, the
potential influence of guideline-adherence
on DSS is particularly intriguing given the
relative similarities in case volume, a variable
independently associated with ovarian cancer survival3, at NCI-CCCs and HVCs. This
study highlights the great value of developing
quality-improvement initiatives to internally
assess care compliance. These initiatives
should be applied globally because, as the
study shows, compliant care was <50% even
at NCI-CCCs.
The synthetic lethality of BRCA-deficient
cancer cells exposed to poly(ADP-ribose)
polymerase (PARP) inhibition was recognized
more than 10years ago. Several proofof
concept clinical trials of PARP inhibitors have
Ovarian cancer is characterized by late presentation, bulky metastatic tumour burden,

and frequent recurrence of eventual chemo
resistant disease, and has proved to be a major
diagnostic and clinical challenge, reflected
in the persistently poor survival rates that
have not really changed over the past three
decades. Efforts to alter this natural history,
however, have been realized with substantial
gains in life expectancy, largely as the result
of improvements in surgery, adjuvant and
second-line therapies, supportive care, access
to specialists and centres of excellence, adherence to care guidelines, and understanding of
the disease biology.
In 2015, a study demonstrated that compliance with the National Comprehensive
Cancer Network (NCCN) treatment guidelines had a positive impact on the overall
survival of women with ovarian cancer 1.
However, the effect of centres offering interdisciplinary cancer-research programmes,
such as those found in National Cancer
Institute (NCI)-designated Comprehensive
Cancer Centers (NCI-CCCs), had not
been formally evaluated. Bristow and colleagues2 addressed this question in a retrospective population-based study of patients
with ovarian cancer treated in Southern
Californian hospitals between 1995 and
2006. The primary outcome measure was
disease-specific survival (DSS). The facilities
where the patients were treated were categorized as NCI-CCCs, or non-designated cancer centres with either a high (10) or low
(<10) ovarian-cancer case volume (HVCs
and LVCs, respectively). By far, the majority
of hospitals were LVC, and nearly half of all
patients (4,479 out of 9,933) were treated in
such facilities. Five NCI-CCCs were included
in the analysis and, as expected, all were high-
volume centers (average: 14.5 cases per year).
29HVCs had a similar case volume (average:
14.6 cases per year), providing an intriguing
ria
Advances in key areas of research have enabled improved outcomes for

patients diagnosed with ovarian cancer in the past three decades. In 2015,
this trend was maintained with important progress in areas such as guideline
compliance, design of targeted approaches and molecular profiling.
va
Robert L.Coleman
Insights into strategies for

optimizing ovarian cancer care
since been conducted, demonstrating clinical efficacy in patients with tumours carrying
germline or somatic BRCA1 and/or BRCA2
mutations. Considerable focus has been
placed on patients with ovarian and breast
cancers, although these mutations also occur
in other tumour types. In 2015, Kaufman
and colleagues4 published a nonrandomized
phaseII trial of olaparib (at least one dose)
in 298 patients carrying a known deleterious germline mutation in BRCA1 or BRCA2.
The largest cohort comprised patients with
ovarian cancer (n=193), but the study also
recruited patients with breast, pancreatic,
prostate, and other solid tumours. The ratio
of BRCA1 to BRCA2 mutations was variable
among the cohorts, with BRCA1 mutations
being dominant for ovarian cancer. The
overall objective response rate (ORR) varied significantly by disease type (FIG.1). For
patients with ovarian cancer, the median
progression-free survival was 7.0months
and the overall survival was 16.6months4.
The olaparib regimen was well tolerated. No
differences in the ORR related to the BRCA
gene mutation status or to prior platinum
exposure were observed, although patients
with ovarian cancer were excluded from the
latter analysis because they had all received
previous platinum-b ased chemotherapy 4.
After a drought of new drug approvals for
ovarian cancer in the USA, the results from
this trial, combined with a large efficacy and
safety database of patients treated with olaparib in a variety of clinical settings proved
to be sufficient to warrant accelerated regu
latory approval of olaparib by the FDA in
December 2014. The evidence was convin
cing because the expectations for a response
Tumour response rate (%)
O VA R I A N C A N C E R I N 2 0 1 5
NE
Figure 1 | RECIST response to olaparib by

disease type4. The overall objective response
Nature
Reviews
Oncology
rate (ORR) was
>30%
for both| Clinical
ovarian and
prostate cancer. In most tumours, stable disease
(SD) 8weeks was demonstrated. NE, not
evaluable; PD, progressive disease.
JANUARY 2016 | 16

to chemotherapy alone after exposure to
three or more lines of therapy is around
10%5. The reference metric was based on
patients that had developed resistance to
platinum regardless of their BRCA mutation
status, but this level of activity for a novel
non-chemotherapeutic agent is notable. The
conditional approval by the FDA in this setting has had a notable ripple effect on drug
development, opening a new domain for
evaluating promising pharmaceutical assets6.
Ovarian cancer is paradigmatic of an
immunos uppressive tumour, and there
are many reasons to believe that this disease would be amenable to treatment with
immunotherapy : ovarian-cancer cells
express cancer-specific antigens, elicit spontaneous antitumour immune response after
immunot herapy, and are associated with
variable tumour-infiltrating lymphocytes
(TILs)the presence and character of which
is a significant prognostic factor7. Several
immune-targeted strategies have shown
modest efficacy over the past four decades,
but the recent discovery of the immune-
checkpoint inhibitors have ushered in a new
era in immune-oncology. In a phaseII trial
published in the past year, Hamanishi etal.8
evaluated one of these agents, the antiPD1
(programmed cell-death protein 1) antibody nivolumab, in patients with recurrent,
platinum-resistant ovarian cancer; 12months
of therapy at two dose levels was evaluated,
with a primary end point of best overall
RECIST response. In general, both dosages
were well tolerated. Eight out of 20 patients
experienced grade 3 or 4 adverse events,
but only two of them required treatment
discontinuation. The adverse event profile
was consistent with that of this class of agent
across multiple tumour types. Objective
responses were observed in three patients
(15%)8. Two complete responses (CR) were
observed, including one in a patient with disseminated ovarian clear-cell cancer, a notori
ously chemorefractory histologic subtype.
PD1 ligand 1 (PDL1) expression was not
correlated with ORR, as has been reported
for other tumour types 9. A response rate
of 15% might appear modest compared to
what we might expect from chemotherapy

in this setting, but the results of this trial
provide proofofprinciple for the efficacy
of immunotherapy for ovarian-cancer treatment. Furthermore, CRs are very unusual
in this population. This trial has, therefore,
sparked an avalanche of new studies attempting to capitalize on and augment the efficacy
of immunotargeting in nearly every clinical
domain of ovarian c ancer treatment.
Drug resistance eventually emerges in
nearly all patients receiving chemotherapy for
recurrent ovarian cancer. The causes and best
strategies to manage this phenotype remain
elusive. Clinical trials such as the ones previously described represent attempts at providing better opportunities. However, without
guidance from treatment-validated germline
or somatic genomic biomarkers, even these
strategies are hit or miss. Several studies
have identified prognostic factors linked to
poor outcomes in patients with recurrent disease, but comprehensive analyses of genomic
alterations due to the selective pressure of
chemotherapy were lacking. In 2015, the publication of the first comprehensive genomic
characterization of recurrent ovarian cancer
was published by Patch and colleagues10 who
deep sequenced the whole genomes of 114
tumour samples from 92 patients with highgrade serous ovarian cancer; the overall distribution of the sample included patients with
primary refractory disease (13%), primary
resistantdisease (40%) and primary chemosensitive d
isease (47%), from which 28% had
acquired drug-resistant disease. As expected,
missense and truncating TP53 mutations
were the most-prevalent aberrations, followed by inactivating germline or somatic
mutations affecting homologous recombination DNA-repair genes (BRCA1/2 [43%] and
RAD51B [3%]) 10. Gene breakage analysis
revealed that the frequency of NF1 and RB1
mutation was much higher than expected
(20% and 18%, respectively). These events,
along with similar alterations in RAD51B,
PTEN, and CCNE1 seem to account for the
primary refractory and resistance pheno
type10. Analyses of samples from patients
with acquired resistance identified additional
Key advances
Adherence to guideline-directed care improves outcomes in ovarian cancer2; moreover, access
to guideline-compliant comprehensive cancer care centres might offer added benefits
The initial clinical efficacy of novel therapeutics, such as poly(ADP-ribose) polymerase (PARP)
inhibitors4 and immune-checkpoint inhibitors8, has ushered in a new wave of drug development
in ovarian cancer
Ovarian cancer is characterized by genomic chaos; the selective pressure ofchemotherapy
induces a number ofmolecular events that are being characterized by deep genomic sequencing10
17 | JANUARY 2016
molecular events involved in this phenotype.

Despite the interesting and novel findings,
a smoking gun was not found. Consistent
with findings in the primary-disease setting,
patients with relapsed disease are unlikely
to have disease associated with a reliable,
actionable point mutation. Unfortunately,
the low frequency of these genetic events
suggests that the opportunities to develop
effective individualized therapies for ovarian
cancer based on genomic analyses alone will
belimited.
The choice of the most relevant studies
on ovarian cancer in 2015 was a challenge
because their potential spectrum of influence is difficult to comprehend at present.
Nevertheless, these four studies do provide
an insight into areas of investigation that have
great promise of furthering the gains made in
recent years, and provide a foundation upon
which future investigations can build.
Robert L.Coleman is at the Department of Gynecologic
Oncology & Reproductive Medicine, University of Texas,
M.D. Anderson Cancer Center, 1155 Herman Pressler
Drive, CPB 6.3590, Houston, Texas 77030, USA.
rcoleman@mdanderson.org
Dottino,J.A., Cliby,W.A., Myers,E.R., Bristow,R.E.
& Havrilesky,L.J. Improving NCCN guidelineadherentcare for ovarian cancer: value of an
intervention. Gynecol. Oncol. 138, 694699
(2015).
2. Bristow,R.E. etal. Impact of National Cancer Institute
Comprehensive Cancer Centers on ovarian cancer
treatment and survival. J.Am. Coll. Surg. 220,
940950 (2015).
3. Schrag,D. etal. Associations between hospital and
surgeon procedure volumes and patient outcomes
after ovarian cancer resection. J.Natl Cancer Inst. 98,
163171 (2006).
4. Kaufman,B. etal. Olaparib monotherapy in
patientswith advanced cancer and a germline
BRCA1/2 mutation. J.Clin. Oncol. 33, 244250
(2015).
5. Griffiths,R.W. etal. Outcomes after multiple lines
ofchemotherapy for platinum-resistant epithelial
cancersof the ovary, peritoneum, and fallopian tube.
Int. J.Gynecol. Cancer 21, 5865 (2011).
6. Herzog,T.J. etal. Ovarian cancer clinical trial
endpoints: Society of Gynecologic Oncology white
paper. Gynecol. Oncol. 132, 817 (2014).
7. Zhang,L. etal. Intratumoral Tcells, recurrence, and
survival in epithelial ovarian cancer. N.Engl. J.Med.
348, 203213 (2003).
8. Hamanishi,J. etal. Safety and antitumor activity
ofantiPD1 antibody, nivolumab, in patients with
platinum-resistant ovarian cancer. J.Clin. Oncol. 33,
40154022 (2015).
9. Mahoney,K.M. & Atkins,M.B. Prognostic and
predictive markers for the new immunotherapies.
Oncology (Williston Park) 28, 3948 (2014)
10. Patch,A.M. etal. Whole-genome characterization
ofchemoresistant ovarian cancer. Nature 521,
489494 (2015).
1.
Acknowledgements
R.L.C. is supported by the Ann Rife Cox Chair in Gynecology,

the Judy Reis and Albert Pisani Ovarian Cancer Research Fund,
and CPRIT RP120214.
R.L.C. discloses he receives research support from AstraZeneca/

MedImmune, Clovis Oncology and Roche/Genentech.
M E TA S TAT I C P R O S TAT E C A N C E R I N 2 0 1 5
The new and the old

thatis newagain
Julie N.Graff and Tomasz M.Beer
In 2015, published trials highlighted the remarkable efficacy of docetaxel
combined with androgen-deprivation therapy in patients with newly
diagnosed metastatic prostate cancer. Also in 2015, a large study revealing
potential molecular targets for metastatic castration-resistant prostate
cancer therapies was published, along with a study showing activity of
PARP inhibition in patients harbouring mutations in genes governing
DNArepair.
(n=397), or ADT alone (n=393). Patients
who received docetaxel had a signific ant
prolongation of median overall survival:
57.6months versus 44.0months (hazard ratio
(HR)=0.61, 95% CI 0.470.80, P<0.001)1.
This study also popularized the distinction
between low-volume and high-volume disease, where the latter was defined as any
visceral metastatic disease or 4 bone meta
stases with one outside of the spine or pelvis.
In patients with high-volume disease, the difference in median survival was 17months in
favour of docetaxel treatment (HR=0.60, 95%
CI 0.450.81, P<0.001)2. The hazard ratio for
survival in the low-volume group was similar
in magnitude but was not statistically signi
ficant, with the majority of patients in this
group alive at the time of the analysis.
Longer followup is needed to evaluate
survival in this subgroup. Of note, 73%
of patients enrolled in this study had
not received prior local therapy in an
attempt to cure their prostate cancer.
Grade3 adverse events (AE) in
the chemotherapy group included
allergic reaction (2%), fatigue (4%),
diarrhoea 1%),stomatitis (1%)
and neuropathy (1%). One patient
died of unknown causes. This study
demonstrated that chemohormonal
therapy delivered early during the initial treatment
of metastatic disease
results in a degree of
survival improvement
that had not been
observed previously in
men with advanced-stage
prostate cancer. Thus, the
G
NP
ok/
dbro
S. Bra
Chemotherapy has occupied a central role in

clinical oncology, and still forms the backbone of treatment in the metastatic setting.
For oncologists who focus on metastatic
prostate cancer, the role of chemotherapy has
been largely secondary to hormonal therapy.
In 1941, Hodges and Huggins described that
surgical castration elicited objective cancer
responses in men with metastatic prostate
cancer, but nearly all cancers eventually
became castration resistant. Corticosteroids
and mitoxantone chemot herapy provided
palliation for men with symptomatic metastatic castration-resistant prostate cancer. In
2004, two studies described an overall survival
benefit for docetaxel-based chemotherapy,
but the benefits were modest1. Between 2010
and 2013, five additional agents (cabazitaxel,
sipuleucel-T, abiraterone, enzalutamide and
radium-223) were approved for the treatment of patients with metastatic castration-
resistant prostate cancer (mCRPC). Instead
of administering those agents intravenously,
clinicians had the option of prescribing oral
second-generation androgen receptor signalling inhibitors, which provided improvements
in overall survival, progression-free survival
and quality of life with less toxicity. In 2015,
chemotherapy has moved back to the front
line of non-castration resistant metastatic
prostate cancer treatment.
In the Chemohormonal Therapy versus
Androgen Ablation Randomized Trial for
Extensive Disease (CHAARTED) randomized
phase III trial, men with hormone-naive
metastatic prostate cancer were randomly
assigned to receive either 75mg/m2 docetaxel
intravenously every 3weeks for six cycles
plus androgen-deprivation therapy (ADT)
addition of docetaxel chemotherapy with

ADT should be considered at the initial
stages of treatment for every patient with
metastatic prostate cancer, particularly those
with high-volume disease. Whether this
therapy should be offered to all patients with
metastatic disease or only to those with high-
volume disease is an ongoing discussion in
the field.
The second-generation hormonal agents,
abiraterone and enzalutamide, joined the
armamentarium of treatments for mCRPC
in 2011 and 2012, respectively. Abiraterone
blocks two key steps in androgen synthesis,
whereas enzalutamide antagonizes androgen receptor signalling. Newer agents with
similar mechanisms of action, when tested
in clinical trials, will possibly have a more
stringent baseline for demonstrating a survival advantage, because patients enrolled
on these trials will likely be exposed to commercially available enzalutamide and/or abiraterone during the course of their disease.
The non-s teroidal antiandrogen ortero
nel (TAK700) functions similarly to abiraterone, but it inhibits 17,20lyase more
selectively than 17hydroxylase3. Given its
specificity, orteronel can be administered
without prednisone because prednisone is
given primarily to overc ome the hormonal imbalances caused by 17hydroxylase
JANUARY 2016 | 18

Key advances
Early administration of docetaxel to patients
with metastatic prostate cancer significantly
benefits the outcomes of those patients2
The majority of castration-resistant
prostatecancer contains potentially
actionable mutations9
Prostate cancer with certain aberrations in
DNA repair respond to poly(ADP-ribose)
polymerase (PARP) inhibition10
inhibition. The phaseIII studies published

in 2015, however, included prednisone 5mg
orally twice a day becauseprednisone itself
has an anticancer activity in mCRPC. Two
phaseIII studies of orteronel were reported
in 2015 (REFS4,5). One enrolled men with
mCRPC who were chemotherapy-n aive
from October 2010 to June 2012 to receive
either orteronel 400mg twice daily plus prednisone 5mg twice daily (n=781) or placebo
plus prednisone (n=779)4. Patients who had
previously received abiraterone, orteronel,
ketoconazole, aminoglutethimide or enzalutamide were excluded from this trial. Survival
was not improved in the orteronel group
(HR=0.92, 95% CI 0.791.08, P=0.31), but
radiographic progression-free survival (rPFS)
was improved (HR=0.71, 95% CI 0.630.80,
P<0.0001) and a 50% reduction of prostate-specific antigen (PSA) levels at 12weeks
was observed for 43% of the patientsin this
group compared with 25% of patients in the
control group (P<0.0001)4. In total, 30% of
patients in the orteronel arm discontinued
treatment because of AEs, most commonly
for gastrointestinal toxicities, whereas 18%
of patients in the control arm discontinued
treatment due to AEs. The median treatment
time was 10.1months for orteronel versus
8.9months in the controlgroup.
In a second phaseIII study, patients who
had previously been treated with chemo
therapy received the same dose of orteronel
and prednisone in a 2:1 randomization, with
734 patients randomly assigned to the ortero
nel group and 365 to the control group 5.
Overall survival was not improved with
orteronel (HR=0.886, 95% CI 0.7391.062,
P=0.190) but patients treated with this agent
had improved rPFS (HR=0.760, 95% CI
0.6530.885, P<0.001), and a 50% reduction in PSA levels at 12weeks was observed
in 25% of the patients in this group (compared
with 10% in the control group). The median
treatment time was 5.7months in the ortero
nel group versus 4.6months in the control
group, and the rates of treatment discontinuation owing to AEs were 30% and 24%,
respectively5. Around 20% of patients enrolled

in this study went on to receive abiraterone,
and 5% went on to receive enzalutamide. The
failure of both studies to show an overall survival benefit might be related to the toxicity
associated with orteronel (which decreased
treatment time), a lower efficacy level for this
agent, or to the subsequent treatments that
patients in these studies underwent. It is worth
noting, however, that studies of enzalutamide
and abiraterone occurred during a similar
time frame in a heavily pretreated population,
and they were able to demonstrate improved
overall survival68. The exposure of patients
to androgen signalling inhibitors later in the
disease trajectory was a common feature of
these successful trials, particularly in trials
assessing enzalutamide7,8. With the presence
of other agents showing significant activity in
mCRPC, it will become more and more difficult for biosimilar agents to provide improvements in important end points, such as overall
survival. Further development of orteronel is
on hold, with the exception of an ongoing
cooperative group study (NCT01809691) that
is testing this agent in the initial management
of patients with treatment-naive metastatic
prostate cancer.
Future treatments of mCRPC will likely
be molecularly driven, focusing on targets
not yet fully exploited in prostate cancer. In
a landmark publication in 2015, researchers from eight institutions analysed biopsies from 189 patients with mCRPC treated
with enzalutamide and/or abiraterone to
characterize these tumours using molecular, gene-expression and histopathology
markers9. Approximately 90% of tumours
analysed had potentially clinically actionable
mutations that could be targeted by available drugs. Many tumours (62.7%) harboured
mutations that affected the androgen receptor, which has been accepted as the universal
target of prostate cancer therapies, and represents a target that should not be affected
by the results of molecular characterization.
Importantly, 65% of the tumours harboured
mutations related to pathways for which the
therapeutic efficacy remains theoretical, such
as aberrations in the PI3K pathway, RAF
kinases, CDK inhibitors or Wnt signalling.
The findings of this study, however, are compelling because 19.3% of the participants harboured germline or acquired aberrations in
BRCA2, BRCA1, or ATM, genes that govern
DNA repair9.
The results of a remarkable phaseII study
uncovered an entirely new approach to treat
a subset of men with advanced-stage prostate cancer. In this study, the PARP inhibitor
olaparib had significant anticancer activity
19 | JANUARY 2016
in patients with mCRPC harbouring tumour

aberrations in DNA repair genes10. Out of
50patients enrolled, 16 harboured alterations
in BRCA2, ATM, BRCA1, CHEK2, FANCA or
HDAC2. Of these 16 patients, 14 responded to
olaparib, whereas only two of the 33 patients
in the group that did not harbour aberrationsin DNA repair genes had a response
to this agent10. Grade 3 and 4 toxic events
included myelosuppresion and fatigue. This
promising result requires additional study,
but it is notable to mention that olaparib
was approved for the treatment of advancedstage ovarian cancer owing to the promising
results from the arm of this study that enrolled
patients with ovarian cancer with mutations
in BRCA. Thus, if confirmed, these results
should bring a new therapy, and the potential
for routine tumour genomic analyses to be
incorporated into the everyday care of men
with advanced-stage prostatecancer.
Julie N. Graff and Tomasz M. Beer are at the Oregon
Health & Science University Knight Cancer Center,
3710SW US Veterans Hospital Road, Portland,
Oregon97239, USA.
graffj@ohsu.edu;
beert@ohsu.edu
Petrylak,D.P. Docetaxel-based chemotherapy
trialsinandrogen-independent prostate cancer:
firstdemonstration of a survival benefit. Curr. Oncol.
Rep. 7, 205206 (2005).
2. Sweeney,C.J. etal. Chemohormonal therapy in
metastatic hormone-sensitive prostate cancer.
N.Engl. J.Med. 373, 737746 (2015).
3. Yamaoka,M. etal. Orteronel (TAK700), a novel
non-steroidal 17,20lyase inhibitor: effects on steroid
synthesis in human and monkey adrenal cells and
serum steroid levels in cynomolgus monkeys.
J.Steroid. Biochem. Mol. Biol. 129, 115128
(2012).
4. Saad,F. etal. Orteronel plus prednisone in patients
with chemotherapy-naive metastatic castrationresistant prostate cancer (ELMPC 4): a doubleblind,multicentre, Phase 3, randomised, placebocontrolled trial. Lancet Oncol. 16, 338348
(2015).
5. Fizazi,K. etal. Phase III, randomized, double-blind,
multicenter trial comparing orteronel (TAK700) plus
prednisone with placebo plus prednisone in patients
with metastatic castration-resistant prostate cancer
that has progressed during or after docetaxel-based
therapy: ELMPC 5. J.Clin. Oncol. 33, 723731
(2015).
6. Scher,H.I. etal. Increased survival with enzalutamide
in prostate cancer after chemotherapy. N.Engl.
J.Med. 367, 11871197 (2012).
7. Beer,T.M. etal. Enzalutamide in metastatic prostate
cancer before chemotherapy. N.Engl. J.Med. 371,
424433 (2014).
8. Gartrell,B.A. & Saad,F. Abiraterone in the
management of castration-resistant prostate cancer
prior to chemotherapy. Ther. Adv. Urol. 7, 194202
(2015).
9. Robinson,D. etal. Integrative clinical genomics of
advanced prostate cancer. Cell 161, 12151228
(2015).
10. Kaufman,B. etal. Olaparib monotherapy in
patientswith advanced cancer and a germline
BRCA1/2 mutation. J.Clin. Oncol. 33, 244250
(2015).
1.
therapy, whereas the CheckMate017 investigators2 found no such correlation in patients
with squamous NSCLC. Indeed, 45.2% of
patients with predetermined cutoff of PDL1
expression included in KEYNOTE001
responded to pembrolizumab, with a median
PFS of 6.3months (comparedwith a response
rate of 19.4% and PFS of 3.7months in the
Egbert F.Smit and Paul Baas
trial cohort overall)3. Assuch, pembrolizumab
was approved together with a companion
Lung-cancer treatment paradigms continue to advance as we exploit our
diagnostic PDL1 assay. Nonetheless, treatgrowing understanding of the genetic basis of both tumorigenesis and
ment decisions are complicated by the use
therapy resistance. Moreover, ongoing developments with targeted
of different assay platforms and cutoff points
therapies are improving patient outcomes, with two new drugs approved in for PDL1 expression in clinical trials1,3. On
the basis of the data presented in 2015, one
2015 for non-small-cell lung cancer and many others showing promise.
might conclude that higher tumour PDL1
expression correlates with response rate and
survival. Conflicting data have emerged, howIn 2015, the therapeutic landscape for patients by the FDA as second-line treatments for ever, regarding the negative predictive value
with lung cancer has expanded through the NSCLC. Moreover, studies of immune- of this putative marker. Indeed, 1119% of
development of agents directed at both estab- checkpoint inhibition in the first-line, adju- patients with no tumour PD-L1 expreslished and novel targets. Immune-checkpoint vant, and m aintenance s ettings are now sion can nevertheless achieve a long-lasting
inhibition is one of the new and most-exciting underway(TABLE1).
response to agents targeting the PD1PDL1
The lack of a biomarker with suffi- axis4. Notably, such response rates among only
breakthroughs in the treatment of advancedstage non-small-cell lung cancer (NSCLC). cient sensitivity and specificity to predict the biomarker-negative subgroups are compaThe immune-checkpoint inhibitors under response to therapy is a major issue in the rable to those achieved in the overall populaclinical evaluation in NSCLC can be divided selection of patients for treatment with tion of patients with advanced-stage NSCLC
into two major classes: inhibitory antibodies immune-checkpoint inhibitors. Of note, the with the second-line treatments used in the
that target cytotoxic Tlymphocyte-associated CheckMate057 and KEYNOTE001 data1,3 past decadeand that are still used today.
antigen 4 (CTLA4), namely, ipilimumab showed that tumour-cell PDL1 expression Tumour heterogeneity, sampling issues, and
and tremelimumab; or those that block the was correlated with the efficacy of antiPD1 the use of archived material might represent
interaction between programmed cell-death
protein 1 (PD1) and its ligand PDL1
(TABLE 1) . Most progress has been made,
Table 1 | Agents targeting the PD1PDL1 axis in phaseIII NSCLC trials
however, with the antiPD1 antib odies
Remarks (ClinicalTrials.gov ID)
Drug
Type
Target Status
nivolumab and pembrolizumab, which have
Nivolumab
Human
PD1
PhaseIII
Registered
for second-line treatment of
been associated with unprecedented results
(BMS936558)
IgG4
(completed/
advanced-stage
NSCLC (NCT01642004
and minimal toxicities compared with
ongoing)
and NCT01673867). Trials ongoing
conventionalchemotherapy.
in other settings and in combination
In 2015, nivolumab was demonstrated to
with other therapies (NCT02041533;
NCT02477826)
have consistent antitumour activity against
both nonsquamous and squamous advancedPembrolizumab
Humanized PD1
PhaseIII
Registered for second-line treatment of
(MK3475)
IgG4
(completed/ PDL1positive advanced-stage NSCLC
stage NSCLC in the CheckMate057 and
ongoing)
(phase completed but not reported:
CheckMate017 phaseIII trials1,2, respectively,
NCT01905657). Trials ongoing in other
when used as a single agent in the second-
settings and in combination with other
line setting. Approximately 20% of patients
therapies (NCT02220894; NCT02142738;
responded to this agent, compared with 912%
NCT02578680; NCT02504372)
of those who received standard second-line
Atezolizumab
Humanized PDL1 PhaseIII
In various settings and with various
docetaxel, and a higher proportion were
(MPDL3280A)
IgG1
(ongoing)
therapies (NCT02367794; NCT02367781;
NCT02409342; NCT02366143;
alive at 1year (51% and 42% in the nonsquaNCT02409355; NCT02486718)
mous and squamous disease settings, respectively, versus 39% and 24% with docetaxel)1,2.
Durvalumab
Human
PDL1 PhaseIII
+/ tremelimumab (NCT02453282;
(MEDI4736)
IgG1
(ongoing)
NCT02542293; NCT02352948),
Compared with docetaxel, nivolumab also
as a maintenance therapy after
improved the median overall survival of
chemoradiation for stage III disease
patients with either nonsquamous NSCLC
(NCT02125461), and in the adjuvant
1
(9.4months versus 12.2months; P=0.002) ,
setting (NCT02273375)
or squamous-cell disease (6.0months verAvelumab
Human
PDL1 PhaseIII
Versus platinum-based doublet
sus 9.2months; P<0.001)2. In 2015, similar
(MSB0010718C) IgG1
(ongoing)
chemotherapy in the first-line
efficacy was observed for pembrolizumab in
(NCT02576574), and docetaxel in the
second-line (NCT02395172)
the large, phaseI, KEYNOTE001 trial3, and
NSCLC, non-small-cell lung cancer; PD1, programmed cell-death protein 1; PDL1, PD-1 ligand 1.
both this agent and nivolumab were approved
LUNG CANCER IN 2015
Bypassing checkpoints, overcoming

resistance, and honing in on new targets
JANUARY 2016 | 20

important challenges in expression profiling
of PDL1; the ideal t iming, sampling method,
and quantitation of this marker need to be
clarified. Other potential biomarkers are also
under investigation, such as mutational load
and tumour-infiltrating lymphocyteprofiles.
Just over 10years after the value of activa
ting EGFR mutations for predicting the bene
fit of gefitinib treatment in patients with lung
adenocarcinoma was established, the efficacy
of two third-generation EGFR inhibitors has
now been reported5,6. These novel inhibitors,
which are selective for EGFR harbouring
activating and T790M gatekeeper resistance
mutations, will undoubtedly serve as the next
landmark therapy in this disease. The T790M
mutation accounts for approximately 60%
of resistance to first-generation (erlotinib,
gefitinib) and second-generation(afatinib)
EGFR inhibitors. In two phaseI/II trials5,6,
published in 2015, the third-generation
EGFR inhibitors rociletinib (CO1686) and
osimertinib (mereletinib, AZD9291) were
associated with objective response rates of
59% and 61%, respectively, in patients with
metastatic NSCLC that had progressed on
prior EGFR-inhibitor therapy and subsequently tested positive for the T790M mutation. This efficacy was achieved at the expense
of predictable minor toxicity events5,6. One
exception was rociletinib-associated hyperglycaemia5, which seems to be attributable
to inhibition of IGFR1 by a drug metabolite;
however, this interaction could, in fact, be
advantageous, as resistance to EGFR inhibi
tors might be mediated partially by IGFR1
bypass signalling. In late 2015, on the basis
of these findings, osimertinib was approved
by the FDA for the treatment of EGFRT790Mpositive metastatic NSCLC that is resistant to
first/second-generation EGFRinhibitors.
Unfortunately, secondary resistance to
these third-generation agents has also been
described in 2015. In a total of 27 patients
studied7,8, the EGFRT790M allele disappeared
in 10 patients based on pretreatment and

postprogression tumour specimens or liquid
biopsy of circulating cell-free DNA (cfDNA);
in two of these patients, both treated with
rocelitinib, the tumours acquired a smallcell lung cancer (SCLC) phenotype, a llegedly
through RB1 aberrations8. Six patients treated
with osimertinib acquired another EGFR
mutation, C797S, which confers resistance
to this agent invitro7. Interestingly, none
of the tumours that reverted to a EGFR
T790wild-type phenotype acquired the
EGFRC797S mutation7. In addition, all patients
retained the original activating EGFR mutation7. Thus, an evolutionary profile consistent
with pre-existing tumour heterogeneity and
Darwinian selection emerges, substantiated
by the finding that EGFR T790wild-type
and T790Mpositive clones coexisted within
single pretreatment samples8. Of note, the
pretreatment fraction of T790Mpositive cells
was negatively correlated with responsiveness
to rociletinib8. Thus, as in treatment-naive
patients with activating EGFR mutations,
tumour heterogeneity accounts for treatment
failure when a single mechanism is targeted;
rational combination treatment, will be
needed to further improve patient outcomes.
Preferably, the design of these combinations
should be informed by tumour biology, which
might be facilitated by analysis of cfDNA or
other circulating biomarkers.
In the past year, Paik etal.9 reported on
a small series of patients with lung adenocarcinoma harbouring MET exon 14 splice
variantsa molecular subgroup comprising
~4% of patients with NSCLCwho exhibi
ted remarkable and long-lasting responses
to the MET inhibitors crizotinib and carbozantinib. Thus, the list of targetable genomic
abnormalities in lung adenocarcinoma has
further expanded. Furthermore, molecularly targeted therapy has led to an apparent
breakthrough in the treatment of SCLC, a
disease for whichonly one agent is approved
Key advances
Immunotherapies are showing continued promise in the treatment of non-small-cell lung
cancer (NSCLC)13, with nivolumab and pembrolizumab approved for the second-line
treatment of metastatic squamous and nonsquamous disease in 2015
Third-generation EGFR inhibitors have been associated with impressive response rates in
patients with NSCLC and EGFRT790Mmediated resistance to earlier-generation inhibitors5,6;
osimertinib is now approved for this indication
Intrinsic tumour heterogeneity underlies eventual resistance to these new EGFR inhibitors;
liquid biopsies might be useful in monitoring disease evolution and informing treatment
decisions7,8
Other molecular targeted therapies are showing potential as lung-cancer treatments,
including MET inhibitors9, and RovaTthe first potential targeted treatment for small-cell
lung cancer10
21 | JANUARY 2016
and none is available beyond the second-line.

RovaT (Stemcentrx, USA), is a novel agent
consisting an anti-Delta-like ligand 3 (DLL3)
antibody (rovalpituzumab) linked to the cytotoxin tesirine; this agent was administered to
patients with relapsed SCLC in a phaseI study,
results of which were reported in a 2015 meeting abstract10. Among 27 patients confirmed
as positive for expression of DLL3a biomarker commonly expressed in SCLCwho
were treated at the expansion-cohort doses,
the objective response rate was 44%10, far
exceeding that typically achieved with conventional therapy; of note, the response rate
also seemed be similar the second-line and
third-line settings10.
2015 witnessed the beginning of an
encouraging new era in the battle against
advanced-stage lung cancer. Nevertheless,
continued efforts and investment, both by
academia and industry, in the science of lung
cancer are necessary to bring clinical meaningful improvements for this difficult-to-treat
disease.
Egbert F.Smit and Paul Baas are at the Department
ofThoracic Oncology, Netherlands Cancer
InstituteAntoni van Leeuwenhoek Hospital,
Plesmanlaan 121, 1066CX Amsterdam, Netherlands.
Correspondence to E.F.S.
e.smit@nki.nl
doi:10.1038/ nrclinonc.2015.223
Borghaei,H. etal. Nivolumab versus docetaxel in
advanced nonsquamous non-small-cell lung cancer.
N.Engl. J.Med. 373, 16271639 (2015).
2. Brahmer,J. etal. Nivolumab versus docetaxel in
advanced squamous-cell non-small-cell lung cancer.
N.Engl. J.Med. 373, 123135 (2015).
3. Garon,E.B. etal. Pembrolizumab for the treatment
ofnon-small-cell lung cancer. N.Engl. J.Med. 372,
20182028 (2015).
4. Carbognin,L. etal. Differential activity of nivolumab,
pembrolizumab and MPDL3280A according to the
tumor expression of programmed death-ligand1
(PDL1): sensitivity analysis of trials in melanoma, lung
and genitourinary cancers. PLoS ONE 10, e0130142
(2015).
5. Sequist,L.V. etal. Rociletinib in EGFR-mutated
non-small-cell lung cancer. N.Engl. J.Med. 372,
17001709 (2015).
6. Janne,P.A. etal. AZD9291 in EGFR-inhibitor resistant
non-small-cell lung cancer. N.Engl. J.Med. 372,
16891699 (2015).
7. Thress,K.S. etal. Acquired EGFR C797S mutation
mediates resistance to AZD9291 in non-small-cell lung
cancer harbouring EGFR T790M. Nat. Med. 21,
560562 (2015).
8. Piotrowska,Z. etal. Heterogeneity underlies the
emergence of EGFRT790 wild-type clones following
treatment of T790Mpositive cancers with a thirdgeneration EGFR inhibitor. Cancer Discov. 5,
713722 (2015).
9. Paik,P.K. etal. Response to MET inhibitors in
patients with stage IV lung adenocarcinomas
harboring MET mutations causing exon 14 skipping.
Cancer Discov. 5, 842849 (2015).
10. Pietanza,C. etal. Safety, activity, and response
durability assessment of single agent rovalpituzumab
tesirine, a delta-like protein3 (DLL3)-targeted anibody
drug conjugate (ADC), in small cell lung cancer (SCLC).
[abstract LBA 7], Presented at the ECC2015.
1.
MELANOMA IN 2015
Immune-checkpoint blockade
durable cancer control
Elizabeth I.Buchbinder and F.Stephen Hodi
In 2015, advances in immunotherapy for metastatic melanoma have come to
fruition, with phaseIII data supporting the combination of ipilimumab and
nivolumab as first-line therapy. Understanding the mechanisms involved in
aneffective antitumour immune response are now key to further advances.
Several studies published in 2015 have increased our understanding of the
complex relationships that exist between our immune system and malignancy.
Immunotherapy approaches have been

used to treat cancer for decades with vari
able success. Over the past 5years, however,
immune-checkpoint blockade has increased
overall survival in patients with advancedstage melanoma and in those with non-smallcell lung cancer (NSCLC). The anti-cytotoxic
Tlymphocyte-associated antigen 4 (CTLA4)
antibody ipilimumab was the first therapy
aimed at blocking an immune checkpoint, and
was approved in 2011 based on results from
a phaseIII trial1, in which 20% of patients
with melanoma experienced long-term survival. Two antibodies targeting programmed
cell-death protein 1 (PD1), nivolumab and
pembrolizumab, have been associated with
response rates of 2840%, with many patients
experiencing long-lasting efficacy2,3.
CTLA4 and PD1 inhibit antitumour
immune responses by different mechan
isms; consistent with preclinical studies that
demonstrated synergistic effects of combined
immune-checkpoint blockade, clinical trials
in 2015 demonstrated that combination thera
pies led to improved responses in patients
with metastatic melanoma. A phaseIII trial
that compared ipilimumab, nivolumab, and
the combination of these agents confirmed
this synergy4. The combination was associated
with a response rate of 57.6%, compared with
43.7% for nivolumab and 19% for ipilimumab.
In addition, the median progression-free survival (PFS) was 11.5months in the combination group, compared with 6.9months in
the nivolumab group and 2.9months in the
ipilimumab group4. We must await the maturation of survival data, however, before the
full potential of this combination treatment
for m
elanoma is demonstrated.
The largest limitation observed with
the ipilimumabnivolumab combination
is theseverity of the adverse effects: 55%
of the patients experienced grade 34 toxi

city, and 36.4% of the patients discontinued
therapy due to toxicity4. Treatment-related
grade 34 toxicity was observed at a much
lower rate in patients treated with monotherapy: 16.3% in the nivolumab group and
27.3% in the ipilimumab group. The toxi
city observed with these therapies raises
the issue of patient selection and the need
to identify those patients most likely to
benefit from the combination, and from
immunotherapy as a whole. Furthermore, an
understanding of predictive biomarkers for
immunotherapy will help select the optimal
combination agents to enhance efficacy and
overcomeresistance.
In this combination trial, the patient
responses were analysed according to
expression of PD1 ligand 1 (PDL1) in the
tumour samples. In the group of patients
with tumours positive for PDL1 expression
(as assessed by immunohistochemistry) the
response rates were 72.1% in the combination group, 57.5% in the nivolumab group,
and 21.3% in the ipilimumab group. The
median PFS was 14.0months in both the
combination group and the nivolumab group,
compared with 3.9months with ipilimumab
alone 4. In patients with PDL1negative
tumours, the response rates were 54.8% with
the combination, 41.3% with nivolumab, and
17.8% with ipilimumab; the median PFS was
11.2months, 5.3months, and 2.8months,
respectively 4. The clinical significance of
PDL1 as a biomarker to predict benefit from
therapy is promising, but will need further
evaluation, refinement, and s tratification of
patient populations.
In fact, predictive biomarkers for
immune-checkpoint blockade remain a
monumental unmet need. Solving this question involves understanding some of the
complex mechanisms that govern immune

recognition and regulation. Recent efforts
have enabled the identification of predictive factors for response to immunotherapy,
which include the presence of a pre-existing
immune infiltrate, expression of the immune-
checkpoint target, as well as a high mutational
load and frequent neoantigen formation in the
tumour. Greater appreciation of these factors
and how they might be exploited clinically
have matured considerably owing to several
advances achieved in 2015(REFS5,6).
The Cancer Genome Atlas (TCGA)
Consortium published data from analyses of
DNA, RNA and proteins in primary or meta
static melanoma samples from 333 patients.
In addition to creating a disease classification
based on BRAF, RAS, and NF1 mutational
status that can help to guide targeted therapy, TCGA also evaluated aspects of immune
gene expression. These investigators found
that patients harbouring tumours with a
transcriptome characterized by high expression of immune-related genes, high LCK protein levels, and lymphocytic infiltrates had
improved survival7. This implicates a tumour-
expression profile that promotes a favourable
endogenous immune signature, and supports
the results of prospective analyses in patients
receiving immune therapies.
This year also witnessed a greater awareness of the role of recognition of the tumour
by the immune system in generating effective
clinical benefits from checkpoint blockade.
Pre-existing adaptive immune responses,
evidenced by the detection of Thelper type1
gene expression and CTLA4 expression, and
the absence of fractalkine (CX3CL1) in baseline tumour specimens, were shown to predict
response to anti PDL1 therapy in a publication in 2014 (REF.8). Thus, an effective response
to immune-checkpoint blockade requires
the recognition of the tumour by the host
immune system through homing of immune
cells to the tumour microenvironment, as
S.Bradbrook/NPG
JANUARY 2016 | 22

highlighted by pre-existing CD8+ Tcell infiltration. This finding emphasizes the importance of understanding why some tumours
have a more-robust adaptive recognition by
the host immune system6. A profound leap
forward in this understanding came from the
findings of Spranger etal.9, who investigated
the gene-expression profiles of 266 cutaneous
melanomas and divided them into inflamed
and non-inflamed cohorts. They found that
catenin signalling was active in the non-
inflamed cohort and an increased CTNNB1
(catenin) score was predictive for a lack of
Tcell infiltration, with an odds ratio of 4.9
(REF.9). Inthis study, genetically engineered
Braf V600E/Pten/mutantmice and Braf V600E/
Pten //Ctnnbmutant mice were used to
evaluate this association further9. Melanoma
tumours that developed in Braf V600E/Pten//
Ctnnb-mutant mice demonstrated failure of
Tcell priming, which was suggested to be
related to defective recruitment and activation of CD103+ dendritic cells (DCs)9. Failed
recruitment of CD103+ DCs was partially due
to defective production of the chemokine
CCL4, owing to induction ofATF3which
suppresses CCL4by mutant -catenin9.
Thus the WNT/catenin pathway might
contribute to immune evasion in melanoma,
with potential therapeutic implications. This
work identified important signalling pathways related to adaptive immune responses,
and indicated a potential targetable mechanism by which tumour cells can exclude
Tcellinfiltration9.
Genomic analyses have added tremendous
insight into what drives the host immune system to mount an adaptive response against
tumours. This year, Rizvi etal.5 examined
the relationship between genetic mutational

burden and response to PD1 blockade with
pembrolizumab in patients with NSCLC.
They sequenced the exomes of samples from
patients with NSCLCs and compared them
to their matched DNA from normal tissue.
A higher somatic nonsynonymous-mutation
burden was associated with response to pembrolizumab5.Patients with 178 nonsynonymous mutations had a durable clinical benefit
rate of 75% versus 14% in patients with fewer
mutations. The carcinogens in tobacco smoke
cause the majority of mutations observed in
lung cancer, and the response to therapy
correlated with a gene signature comprising high expression of genes associated with
smoking5. In this study, some of the specific
mutations seen in patients with a positive
response to therapy included mutations in
genes involved in DNA repair and replication, which supports the idea that mutational
burden has a role in response to therapy, and
likely applies to other solid tumours, such
as melanoma9,10.
Neoantigens that arise because of
tumour-specific mutations provide a potential target for the immune system to recognize cancer cells. In the study by Rizvi etal.5
the quantity of neoantigens identified per
tumour was correlated with mutational burden. Furthermore, the researchers identified
a CD8+ Tcell response against a neoantigen in an individual patient and correlated
this immune response with the patients
clinical response, which demonstrated the
role of neoantigens in immune-mediated
responses5. Recent work by Van Allen etal.6
correlated nonsynonymous mutational load
and high neoantigen loadwith response to
Key advances
The combination of CTLA4 blockade and PD1 blockade in the treatment of patients with
metastatic melanoma result in an improved response rate and progression-free survival
compared with treatment with either agent individually, leading to its approval by the FDA as
afront-line therapy4
Toxicity from immune-checkpoint blockade can be severe and the combination of CTLA4
blockade with PD1 blockade led to 55% of patients experiencing grade 34 treatment-related
adverse events4
The baseline inflammatory status of the melanoma tumour is predictive of patient response to
immune-checkpoint blockade6, and one mechanism of Tcell exclusion from tumours seems to be
through catenin signalling9
Mutational burden of tumours has been associated with response to immune-checkpoint
blockade in both melanoma and non-small-cell lung cancer5,6
23 | JANUARY 2016
CTLA4 blockade in patients with melanoma. Of note, the neoantigens observed in

patients with a clinical response showed no
shared features. These collective data suggest
that, although neoantigens elicit an immune
response and increased quantity of neoanti
gens is favourable, the immune response
is not associated specifically with certain
neoantigensignatures6.
Immunotherapy is driving advances in
the treatment of patients with melanoma
and a broad array of other malignancies.
Our understanding of the factors that both
predict and underlie the clinical benefit of
immune-checkpoint blockade has increased
in the past year. As our knowledge evolves,
it will help to guide therapy decisions
and lead to insights for novel targets and
combinatorialtreatment approaches.
Elizabeth I. Buchbinder and F. Stephen Hodi are at the
Melanoma Disease Center, Department of Medical
Oncology, Dana-Farber Cancer Institute,
450Brookline Avenue, Boston, Massachusetts
0221505450, USA.
Correspondence to F.S.H.
stephen_hodi@dfci.harvard.edu
Hodi,F.S. etal. Improved survival with ipilimumab in
patients with metastatic melanoma. N.Engl. J.Med.
363, 711723 (2010).
2.
Robert,C. etal. Nivolumab in previously untreated
melanoma without BRAF mutation. N.Engl. J.Med.
372, 320330 (2015).
3.
Robert,C. etal. Pembrolizumab versus ipilimumab
inadvanced melanoma. N.Engl. J.Med. 372,
25212532 (2015).
4.
Larkin,J. etal. Combined nivolumab and ipilimumab
or monotherapy in untreated melanoma. N.Engl.
J.Med. 373, 2334 (2015).
5.
Rizvi,N.A. etal. Mutational landscape determines
sensitivity to PD1 blockade in non-small cell lung
cancer. Science 348, 124128 (2015).
6. Van Allen,E.M. etal. Genomic correlates of response
to CTLA4 blockade in metastatic melanoma. Science
350, 207211 (2015).
7. TheCancer GenomeAtlas. Genomic classification
ofcutaneous melanoma. Cell 161, 16811696
(2015).
8.
Herbst,R.S. etal. Predictive correlates of response
tothe antiPDL1 antibody MPDL3280A in cancer
patients. Nature 515, 563567 (2014).
9. Spranger,S., Bao,R. & Gajewski,T.F. Melanomaintrinsic -catenin signalling prevents anti-tumour
immunity. Nature 523, 231235 (2015).
10. Snyder,A. etal. Genetic basis for clinical response to
CTLA4 blockade in melanoma. N.Engl. J.Med. 371,
21892199 (2014).
1.
F.S.H. declares he has served as a consultant for Merck and

Novartis, his institution receives research support from
Bristol Myers Squibb, and his institution receives clinical trial
support from Bristol Myers Squibb, Genentech, Merck and
Novartis. E.I.B. declares no competing interests.
Connect to...
For the latest news from the journal and beyond: @NatRevClinOncol
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrclinonc/info/newsfeeds.html
ahead of print http://www.nature.com/nrclinonc/journal/vaop/ncurrent
http://www.nature.com/nrclinonc/focus/index.html & http://www.nature.com/nrclinonc/collection/index.html
Providing visual and audible overviews of hot topics http://www.nature.com/nrclinonc/posters & http://www.nature.com/
nrclinonc/multimedia
Conferences
You can meet the editors of Nature Reviews Clinical Oncology at some of the meetings listed on our conferences page
http://www.nature.com/nrclinonc/info/info_conf.html
ESO
The European School of Oncology (ESO) in collaboration with Nature Reviews Clinical Oncology offers regular live andinteractive
distance learning ESO e-grandrounds http://www.eso.net/
ENDOCRINOLOGY
OBESITY IN 2015
Advances in managing
obesity
John B. Dixon
Successful management of obesity requires integration of pharmaceutical
agents and bariatric surgeries with traditional lifestyle modifications.
Notable developments for managing obesity in 2015 included the
demonstration of weight-beneficial outcomes for liraglutide and
empagliflozin, and the first 5year followup of patients with type2 diabetes
mellitus and obesity randomly assigned to receive bariatric surgery or
conventional medical therapy.
Obesity is a serious relapsing chronic dis
ease that requires a long-term chronicdisease model of care. Behavioural and/
or lifestyle interventions that include a
high-quality diet, caloric restriction, reduced
sedentary behaviour and increased physical
activity form the cornerstones of traditional
care. However, by themselves, these inter
ventions produce only limited sustained
weight loss in the majority of adults. The
central control of energy balance and vig
orous defence of an individuals maximal
weight underlies the difficulties in relying
on lifestyle interventions alone to success
fully manage either of these conditions. In
addition to lifestyle changes, a broad range of
safe and effective interventions are needed to
manage obesity and the effects of the disease
on physical and mental function, quality of
life, end-organ damage and mortality, along
with the capacity to deliver interventions to
those most in need. Applying the principles
of chronic disease management to obesity
would see treatments to promote weight loss
combined when needed (FIG.1). Sustained
weight loss of >5% of a patients body weight
provides measurable clinical benefit, but the
association between the degree of weight loss
and the extent of the benefit is not linear for
many complications; any additional benefit
beyond that afforded by 10% weight loss is
often attenuated1. An important emerging
concept is to consider weight loss and main

tenance of weight loss as separate yet equally
important goals.
Several clinical trials of promising phar
maceutical agents for weight loss were
published in 2015. The first notable phar
macotherapy study in 2015, by PiSunyer
and colleagues2, involved the glucagon-like
peptide1 receptor agonist liraglutide, which

previously had been primarily used to
improve glycaemic control in patients with
type2 diabetes mellitus (T2DM). PiSunyer
etal.2 formally tested liraglutide for weight
management effects compared with placebo
in patients who were overweight or obese
but did not have T2DM; 60% of patients did,
however, have prediabetes mellitus. The 3mg
daily subcutaneous dose of liraglutide, used
in addition to a lifestyle intervention, resulted
ina mean weight loss of 9.2% at 56weeks,
which was 5.8% greater than that achieved
with the lifestyle intervention alone 2. At
least 5% weight loss was achieved by 63%
of liraglutide-treated patients and 27% of
placebo-treated patients; weight loss of at
least 10% was achieved by 33% of the treat
ment group compared with 11% of the con
trol group2. Patients treated with lifestyle
intervention alone were more likely than
patients treated with liraglutide to develop
prediabetes mellitus (OR 3.3, 95%CI 2.44.7)
and diabetes mellitus (OR 8.1, 95%CI 2.6
25.3) during the study period. Participants
receiving liraglutide had greater improve
ments in glucose metabolism and both
weight-specific and generic measures of
quality of life, as well as lower cardiovascu
lar risk than patients receiving placebo. The
Chronic disease model

of obesity management
Traditional lifestyle
modication
Training of healthcare sta

to reduce therapeutic inertia
Pharmacological
agents
Bariatric surgery
Weight loss
Figure 1 | Successfully treating obesity requires integration of multiple approaches.

Nature Reviews
| Endocrinology
Management of obesity necessitates a chronic disease model of treatment,
and the integration
of
new, effective treatments alongside traditional approaches by specially trained health-care staff.
24 | JANUARY 2016
Key advances
Liraglutide, a drug primarily used for
glycaemic control, is an effective weight
loss treatment and reduces the risk of
patients with obesity developing
prediabetes mellitus or type2 diabetes
mellitus (T2DM)2
The sodiumglucose cotransporter2
inhibitor, empagliflozin, reduces the risk of
cardiovascular events in patients with
T2DM and leads to sustained weight loss3
Patients with T2DM who underwent
bariatric surgery had sustained weight loss
5years after surgery, even though 3350%
of those in remission for T2DM relapsed6
adverse effects of liraglutide, which included

gastrointestinal symptoms and an increased
risk of symptomatic gallstones, were in line
withreported effects of the drug in patients
with T2DM2. A 3mg dose of liraglutide has
been approved for weight management in
Europe, the USA and Canada.
Another notable pharmacotherapy study
in 2015, by Zinman and colleagues3, involved
empagliflozin, which is a sodiumglucose
cotransporter2 inhibitor. In comparison with
placebo, a 10mg or 25mg dose of empagli
flozin reduced the primary composite out
come of death from cardiovascular causes,
nonfatal myocardial infarction or nonfatal
stroke in patients with T2DM at high risk of
cardiovascular events3. Over a median obser
vation time of 3.1years, empagliflozin reduced
cardiovascular mortality, all-cause mortality
and hospitalization for heart failure by 38%,
35% and 32%, respectively. The findings also
have implications for managing weight in
patients with T2DM, as empagliflozin led to a
sustained weight loss of 1.52.0kg over 3years.
This important result adds empagliflozin to a
select group of glycaemic control medications
that have established mortality benefits and
favourable weightoutcomes.
Also important in 2015 was the finding
that beloranib, an inhibitor of methionine
aminopeptidase2, caused weight loss as an
adverse effect 4. This firstinclass weight
loss drug reduced fat biosynthesis, stimu
lated fat oxidation and lipolysis and mark
edly reduced appetite and body weight4. In a
12week phaseII randomized trial, beloranib
doses of 0.6mg, 1.2mg and 2.4mg produced
weight losses of 5.50.5kg, 6.90.6kg and
10.91.1kg, respectively, compared with
the placebo group of generally healthy indi
viduals (primarily white women) with obes
ity, who lost no weight4. Dose-dependent
adverse effects of beloranib included nausea
and vomiting, as well as sleep disturbance that
was characterized by insomnia and abnor

mal dreams. Concerns raised in 2015 about a
death and cases of reported thromboembolic
events in ongoing and prior clinical trials of
beloranib led the FDA to issue a partial clinical
hold in order to institute measures to ensure
patient safety5. Beloranib and this novel drug
class are currently under a safety cloud with an
uncertain future.
In 2015, Mingrone and colleagues reported
the first 5year health outcomes of bariatric
surgery plus conventional medical therapy
compared with conventional therapy alone
for the management of T2DM6. In this study,
20 participants with T2DM were randomly
assigned to each of the three treatment groups:
standard medical treatment, gastric bypass or
biliopancreatic diversion. Mean weight loss
after 5years was 28.4% in patients undergoing
gastric bypass and 31.3% in patients undergo
ing biliopancreatic diversion, which compared
favourably with a mean 6.9% weight loss in
the conventionally treated group. However,
despite sustained weight loss, half of the 75%
of patients who underwent gastric bypass who
had partial remission of T2DM at 2years had
relapsed by 5years. Of the 95% of patients
in partial remission following biliopancre
atic diversion at 2years, 33% had relapsed by
5years. This finding emphasizes the impor
tance of careful monitoring for relapse in
patients after surgery, and the use of surgery
as an addition to medical therapy for T2DM
rather than an alternative. Bariatric surgery
should be included in the algorithms for
managing T2DM, but should not be seen as a
standalone cure.
The widely held belief that rapid weight
loss is followed by rapid weight regain was
dispelled by an elegant study that randomly
assigned 204 patients with obesity to diets
facilitating either a rapid weight loss target
of 15% of body weight over 12weeks or the
same target over 36weeks7; patients who
achieved >12.5% weight loss were then placed
on a weight maintenance phase for 144weeks.
More patients assigned to rapid weight loss
met the target than patients who lost weight
slowly (78% versus 51%, P=0.001), but by the
end of the maintenance phase, weight regain
was the same in both groups.
Sustaining weight loss while managing
obesity as a chronic disease requires the inte
gration of tools that provide excellent shortterm weight loss, such as very low energy diets7
and intra-gastric balloons, with effective weight
maintenance programmes, which can include
long-term pharmacotherapy and/or intermit
tent intensification of diets to maintain clinic
ally important weight loss8. Two intra-gastric
balloons were approved for use in the USA in
2015 (REF.9), but have been available elsewhere

for many years. Combining and integrating
obesity therapies, and where appropriate using
them sequentially, is key to maintaining weight
loss in the long-term(FIG.1).
Health professionals are poorly trained and
ill-prepared to effectively treat patients with
obesity within the interdisciplinary model
required for quality care. Bias, unfounded
assumptions about lifestyle choices and a
lack of knowledge about the pathophysiol
ogy of the disease creates a blame culture
that results in therapeutic inertia, rather than
escalation of intervention intensity in line
with increasing disease severity to improve
health outcomes10 (FIG.1). Integrating effective
clinical care, capacity building within health
services and community-wide provision of selfmanagement support are fundamentals of
chronic disease management, but are broadly
deficient in addressing the rapidly increasing
numbers of people with clinically severe obes
ity. Changing perceptions about obesity will be
a critical step in moving forward.
John B. Dixon is at Baker IDI Heart and Diabetes
Institute, 75 Commercial Road, Melbourne,
Victoria3004, Australia.
john.dixon@bakeridi.edu.au
doi:10.1038/nrendo.2015.221
Moyer,V.A. Screening for and management of obesity
in adults: U.S.Preventive Services Task Force
recommendation statement. Ann. Intern. Med. 157,
373378 (2012).
2. PiSunyer,X. etal. A randomized, controlled trial of
3.0mg of liraglutide in weight management. N.Engl.
J.Med. 373, 1122 (2015).
3. Zinman,B. etal. Empagliflozin, cardiovascular
outcomes, and mortality in type2 diabetes. N.Engl.
J.Med. 373, 21172128 (2015).
4. Kim,D.D. etal. Efficacy and safety of beloranib for
weight loss in obese adults: a randomized controlled
trial. Diabetes Obes. Metab. 17, 566572 (2015).
5. Zafgen. Zafgen announces partial clinical hold affecting
beloranib trials. [online], http://ir.zafgen.com/
releasedetail.cfm?releaseid=936981 (2015).
6. Mingrone,G. etal. Bariatric-metabolic surgery versus
conventional medical treatment in obese patients with
type2 diabetes: 5year followup of an open-label,
single-centre, randomised controlled trial. Lancet 386,
964973 (2015).
7. Purcell,K. etal. The effect of rate of weight loss on
long-term weight management: a randomised
controlled trial. Lancet Diabetes Endocrinol. 2,
954962 (2014).
8. Johansson,K., Neovius,M. & Hemmingsson,E.
Effectsof anti-obesity drugs, diet, and exercise on
weight-loss maintenance after a very-low-calorie diet
or low-calorie diet: a systematic review and metaanalysis of randomized controlled trials. Am. J.Clin.
Nutr. 99, 1423 (2014).
9. [No authors listed] ReShape and Orbera two gastric
balloon devices for weight loss. Med. Lett. Drugs Ther.
57, 122123 (2015).
10. Dietz,W.H. etal. Management of obesity:
improvement of health-care training and systems for
prevention and care. Lancet 385, 25212533 (2015).
1.
J.B.D. has received research funding and consultancy fees

from Allergan and Nestl Australia, and is a member of the
Optifast Medical Advisory Board of Nestl Australia and the
Medical Advisory Board of NovoNordisk. J.B.D. has also been
a consultant for Apollo Endosurgery, Bariatric Advantage,
mdBriefcase and Nestec, and received honoraria from
Covidien, iNova Pharmaceuticals and Merck Sharp &Dohme.
JANUARY 2016 | 25

THYROID CANCER IN 2015
Key advances
Molecular landscape of thyroid cancer

continues to be deciphered
Yuri E. Nikiforov
Studies published in 2015 have continued to unravel the genomic landscape
of thyroid cancer, particularly of its less common forms (such as medullary
and anaplastic carcinomas) and of familial forms of thyroid cancer. As a result,
new diagnostic and therapeutic markers have been identified and validated
for clinical use.
The year 2014 was marked by the publication
of The Cancer Genome Atlas (TCGA) study of
the most common type of thyroid cancer
(papillary thyroid carcinoma (PTC)), which
provided the most comprehensive characteri
zation of the genomic landscape of this cancer
to date1. In 2015, progress has been made in
deciphering the genetic makeup of other types
of thyroid cancer, as well as in translation of
this knowledge into clinical practice.
A study by Ji and colleagues2 reported the
results of genomic analysis of a large series
of medullary thyroid carcinomas (MTCs) by
whole-exome sequencing (WES) and fluor
escence insitu hybridization (FISH) for spe
cific gene fusions. The analysis confirmed
that mutations in the RET gene are the most
common molecular events in this tumour
type (found in ~50% of all MTCs), followed
by mutations in the HRAS and KRAS genes
(with a combined incidence of 20% of MTCs
in this study). RAS mutations did not overlap
with mutations in RET, which indicates that
each type of mutation is an alternative driver
event for MTC. Furthermore, RAS muta
tions were only found in sporadic tumours.
In addition, the BRAF mutation leading to
the Lys601Asn (K601N) substitution was
identified in one tumour. BRAF mutations,
ALK
EIF1AX
HABP2
Tumour
NPG
Thyroid
typically those causing Val600Glu (V600E)

changes, are the most common mutational
events in PTC, but have not been reported
before in MTC.
Ji et al.2 also reported the identification
of ALK gene fusions in MTCs. In fact, two
tumours were found to carry either EML4
ALK or GFPT1ALK fusions. Whereas
EML4ALK has previously been reported in
a variety of cancers, including PTC3, GFPT1
ALK seems to be a novel type of ALK fusion.
Importantly, ALK fusions represent good
therapeutic targets, and a number of effec
tive ALK inhibitors (such as crizotinib and
ceritinib) are approved for use in lung can
cers that are positive for an ALK fusion. One
of the patients in this study presented with
a metastatic MTC carrying EML4ALK, was
treated with crizotinib and showed a clini
cally significant response. The results of this
study, therefore, not only extend our under
standing of the biology of MTC, but also
offer new therapeutic approaches for this
frequently aggressive type of thyroid cancer.
Another study, reported by Kunstman
etal.4, used WES to characterize the molecu
lar landscape of anaplastic thyroid carcinoma
(ATC), a rare but almost universally lethal
type of thyroid cancer. The study confirmed
the common occurrence of BRAF and RAS
mutations, each present in nearly one-third
of these tumours, but also characterized
a third group of ATCs including tumours
that were negative for both of these alter
ations. Many of these tumours were found
to carry other potential driver mutations,
such as those involving the NF1, MTOR and
NF2 genes, as well as several other genes.
Moreover, two tumours in this group had
mutations in one of the MutL-homologue
family of DNA mismatch repair genes and
demonstrated a more than eightfold higher
mutation rate than other tumours, consistent
with a hypermutator phenotype. Finally, 14%
ALK fusions, including EML4ALK and

GFPT1ALK, are found in 2% of medullary
thyroid carcinomas and could serve as a
therapeutic target for treatment with
crizotinib or other ALK inhibitors2
Anaplastic thyroid carcinomas can be
divided into BRAF or RAS molecular groups,
or neither BRAF nor RAS tumours with
recurrent mutations in genes such as NF1,
MTOR and the MutL-homologue family of
DNA mismatch repair genes, which result in
a hypermutator phenotype4
The germline variant of HABP2 that results
in the Gly534Glu substitution is likely to be
a susceptibility gene for familial
nonmedullary thyroid cancer5
Broad multigene mutation panels might
provide high sensitivity and specificity for
cancer diagnosis in thyroid nodules with
indeterminate cytology and help to avoid
diagnostic surgeries for these patients7
of ATCs in this study had a mutation in the

EIF1AX gene; all of these tumours were also
positive for RAS mutations. EIF1AX muta
tions have been reported to occur in 1.5%
of PTCs 1; these mutations were the only
driver mutation in five of the six tumours
in which they were found. The coexistence
of the two mutations in ATCs raises the
intriguing possibility that EIF1AX cooper
ates with RAS to promote anaplastic trans
formation. Additional recurrent mutational
events of potential interest identified in
this study in ATC involved the USH2A and
HECTD1 genes; however, their involvement
in carcinogenesis that leads to ATC needs
furtherconfirmation.
Important new information has been
reported in relation to the genetics of famil
ial forms of thyroid cancer. Gara and co
workers5 identified a gene that is probably
involved in the predisposition to familial
forms of f ollicular-cell-derived thyroid can
cer. Using WES of a large family that included
several members affected by PTC, the
authors identified the variant of the HABP2
gene leading to the Gly534Glu (G534E)
change; this variant segregated completely
with the affected members of the family.
Studies have revealed that the gene product
is likely to function as a tumour suppressor,
whereas the Gly534Glu variant leads to loss
of function and probably acts in a dominantnegative way5. This variant was present in
4.7% of 423 patients with PTC in the TCGA
cohort, but only in 0.7% of individuals in
the general multiethnic population, which
provides additional evidence for the role of
Nature Reviews | Endocrinology

26 | JANUARY 2016
the HABP2 mutation associated with the
Gly534Glu substitution as a susceptibility
variant for familial, nonsyndromic, non
medullary thyroid cancer. The mechanisms
by which this variant might promote the
development of PTC are not clear. However,
when understanding of these mechanisms
improves, light could be shed on alternative
pathways that lead to thyroidcarcinogenesis.
The studies discussed so far (together with
the existing body of literature) improve our
understanding of the molecular mechanisms
of sporadic and familial forms of thyroid can
cer, provide new potential targets for thera
peutic inhibition and expand the arsenal of
biomarkers that can be used for cancer diag
nosis. The third point is of particular impor
tance for thyroid nodules that cannot be
diagnosed definitively as benign or malignant
on the basis of routine cytological evaluation
of cells collected from nodules using fineneedle aspiration. In the USA alone, every
year >100,000 patients are diagnosed with
nodules that have indeterminate cytology6.
The unclear diagnosis prevents optimal
management of these patients and frequently
results in avoidable diagnostic surgeries.
Extensive characterization of the mutational
landscape of thyroid cancer, which has accel
erated over the past few years after the intro
duction of powerful new technologies such as
next-generation sequencing, has provided a
basis for creating multigene mutational pan
els for the detection of cancer in thyroid nod
ules. As driver mutations in thyroid cancer
are known to include point mutations, such
as those of BRAF and RAS, and gene fusions,
such as those affecting RET, NTRK1/3 and
ALK, both mutation types are included in
these panels.
A study published in 2015 reported the
results of the validation of a gene panel
in a common category of indeterminate
cytology known as atypia of undetermined
s ignificance/follicular lesion of undeter
mined significance (AUS/FLUS) 7 . The
panel that was evaluated included 14 genes
that were studied for point mutations and
42 types of gene fusion, together with the
expression levels of certain cell lineage mark
ers. The mutational panel, called ThyroSeq
(University of Pittsburgh Medical Center,
USA), was tested in a large series of patients
with nodules who had received a diagnosis
of AUS/FLUS. The panel had 91% sensitiv
ity and 92% specificity for cancer detection
in these nodules, which resulted in a 97%
negative predictive value and a 77% positive
predictive value in the studied cohort. The
results of the study provide a rationale for
offering diagnostic genetic testing in clinical
practice. Furthermore, it demonstrates that

large multigene panels, which encompass
most of the driver mutations found in thyroid
cancer, could provide accurate cancer diag
nosis, successfully resolving the uncertainty
of indeterminate cytology.
During the next year, we should expect
our understanding of molecular mechanisms
of thyroid cancer (both sporadic and familial
forms) to continue to progress. This knowl
edge is now being translated into clinical prac
tice, with increasing effects on patient care,
and more specifically on cancer diagnosis,
prognostication and targeted therapies.
Yuri E. Nikiforov is at the Department of Pathology,
University of Pittsburgh, 3477 Euler Way, Room 8031,
Pittsburgh, Pennsylvania 15213, USA.
nikiforovye@upmc.edu
doi:10.1093/nrendo.2015.217
Published online 15 Dec 2015;
CancerGenome AtlasResearch Network.

Integratedgenomic characterization of papillary
thyroid carcinoma. Cell 159, 676690 (2014).
2. Ji,J.H. etal. Identification of driving ALK fusion genes
and genomic landscape of medullary thyroid cancer.
PLoS Genet. 11, e1005467 (2015).
3. Kelly,L.M. etal. Identification of the transforming
STRNALK fusion as a potential therapeutic target
inthe aggressive forms of thyroid cancer. Proc. Natl
Acad. Sci. USA 111, 42334238 (2014).
4. Kunstman,J.W. etal. Characterization of the
mutational landscape of anaplastic thyroid cancer
viawhole-exome sequencing. Hum. Mol. Genet. 24,
23182329 (2015).
5. Gara,S.K. etal. Germline HABP2 mutation causing
familial nonmedullary thyroid cancer. N.Engl. J.Med.
373, 448455 (2015).
6. Sosa,J.A., Hanna,J.W., Robinson,K.A. &
Lanman,R.B. Increases in thyroid nodule fine-needle
aspirations, operations, and diagnoses of thyroid cancer
in the United States. Surgery 154, 14201426 (2013).
7. Nikiforov,Y.E. etal. Impact of the multi-gene ThyroSeq
next-generation sequencing assay on cancer diagnosis
in thyroid nodules with atypia of undetermined
significance/follicular lesion of undetermined
significance cytology. Thyroid 25, 12171223 (2015).
1.
Y.E.N. has served as a consultant for Quest Diagnostics.

Theemployer of Y.E.N., UPMC, has a commercial service
agreement with CBLPath to offer molecular testing for
thyroidnodules.
E N D O C R I N E D I S R U P TO R S I N 2 0 1 5
Epigenetic transgenerational
inheritance
Michael K. Skinner
Endocrine disruptors are critical environmental exposures that influence
health and can promote epigenetic transgenerational inheritance of
disease and abnormal physiology. Advances in 2015 included analyses of
the effects of endocrine disruptors on human disease, further examples
ofendocrine disruptors promoting transgenerational behavioural effects,
insights into effects of endocrine disruptors on epigenetic programming
ofprimordial germ cells and the finding that endocrine disruptors can
transgenerationally promote genetic mutations.
Environmental compounds that alter and/or
disrupt normal endocrine hormone signal
ling at the receptor or signal transduction
level are termed endocrine disruptors. The
first endocrine disruptors studied and shown
to promote abnormal physiology and disease
were diethylstilbestrol (DES) and dichloro
diphenyltrichloroethane (DDT)1. The num
ber of known endocrine disruptors has since
expanded dramatically to include compounds
such as bisphenolA (BPA) and phthalates;
natural compounds, such as genistein from
plants, can also act as endocrine disruptors1.
Over the past several decades, research on
endocrine disruptors has improved our
understanding of the molecular and physio

logical actions of these agents on human
health1. Inaddition to the direct effects of
exposure on an individual, molecular altera
tions to the germ line can promote effects on
subsequent generations. As most exposures
to endocrine disruptors do not promote
genetic mutations, these generational effects
are mediated via epigenetic mechanisms.
When the effects of an endocrine disruptor
alters the epigenetic programming of the
germ line, these changes are transmitted
between generations in the absence of direct
exposure (FIG.1), an effect termed epigenetic
transgenerational inheritance1.
JANUARY 2016 | 27

Endocrine disrupter exposure
Female/male
Endocrine disrupter exposure

Gestating female
Germline
of F1
F2
Germline F1
F1
F0
F0
Multigenerational
exposures
F2: First unexposed

generation, therefore
transgenerational
F3: First unexposed

generation, therefore
transgenerational
Figure 1 | Endocrine-disruptor-induced epigenetic transgenerational

inheritance. Schematic
representation of environmental exposure and affected generations.
The initial observation of epigenetic

transg enerational inheritance involved
the endoc rine disruptor vinclozolin, an
antia ndrogenic agricultural fungicide 2 .
Vinclozolin and the pesticide methoxychlor
promote the epigenetic transgenerational
inheritance of reduced male fertility2. A large
number of other endocrine disruptors and
environmental exposures have now also been
shown to promote epigenetic transgenera
tional inheritance of disease and abnormal
physiology in a wide variety of species from
plants to humans 1. This process involves
epigenetic alterations of the germ line that
can include DNA methylation, noncoding
RNAs, histone modifications and alterations
in chromatin structure. The effects range
from reproductive and behavioural effects to
obesity1. This nongenetic formof inheritance
has altered our understanding ofthe molecu
lar control of disease aetiology and evolution.
Here, I focus on advances in 2015 involv
ing endocrine disruptors and epigenetic
transgenerational inheritance.
A large epidemiology study published in
2015 by Kalfa etal.3 extended the findings
of a previous study reporting the effects of
phthalates on humans 4 by demonstrating
the association of human hypospadias with
prenatal exposure to a variety of endocrine
disruptors. This French study examined a
cohort of 300consecutive children without
agenetic defect and found that after control
for geneticmutation, parental occupational
and environmental exposure to chemical
products increased the risk of hypospadias
in children3. Although Kalfa and colleagues
focused on exposure in children, a future

consideration is if such effects might influ
ence epigenetic transgenerational inheritance
mechanisms.
Another advance in 2015 further docu
mented the generational effect of endocrine
disruptors on brain development and behav
iour. Quinnies and colleagues demonstrated
the transgenerational actions of the phtha
late di-(2ethylhexyl)phthalate (DEHP) on
levels of stress hormones and behaviour5.
Gestating mice were exposed to DEHP dur
ing fetal gonadal sex determination and the
subsequent third generation (F3) had altered
stress hormone levels (corticosterone), pitui
tary gene expression and behaviour in both
male and female mice5. A number of pre
vious studies have demonstrated the trans
generational actions of endocrine disruptors
on behaviour1, and a recent review6 focus
ing onthe neuroscience of the phenomena
supports theconcept that epigenetic mecha
nisms might inform us about the transgener
ational inheritance of behavioural traits that
are increasingly beingreported.
In considering the molecular mechanisms
underlying endocrine-disruptor-induced
epigenetic transgenerational inheritance of
disease and phenotypic variation, the germ
line transm ission of epigenetic informa
tion between generations in the absence of
continued exposure is critical1 (FIG.1). The
original observations suggested that DNA
methylation alterations in the sperm were
crucial2; noncoding RNAs and histone modi
fications have, subsequently, also been shown
to be involved1. Epigenetic reprogramming of
28 | JANUARY 2016
the germ line primarily involves the primor

dial germ cell (PGC) development period
and later stages of gametogenesis. BrienoEnriquez and colleagues exposed gestat
ing female mice to vinclozolin to produce
epigenetic transgenerational inheritance of
testicular cell apoptosis and abnormalities7.
This study confirmed the observations and
transgenerational phenotypes previously
observed in a rat model2. Brieno-Enriquez
etal. extended the previous observations
with an analysis of PGCs and identified alter
ations in epigenetic programming and gene
expression that are critical to PGC develop
ment (such as those in Blimp1). Although the
global DNA methylation analysis used was
insufficient to assess specific DNA methyl
ation sites, this study7 demonstrated inter
esting alterations in noncoding RNAs such
as miR23b and miR21. Brieno-Enriquez
and colleagues also showed that vinclozo
lin promotes epigenetic transgenerational
inheritance of abnormalities in male testes
and alters PGC noncoding RNA program
ming. Asupporting study provided a major
resource for epigenetic alterations during
development of the human germ line epig
enome8. This study identified a critical role
for the Blimp1 pathway, DNA methylation
reprogramming and gene expression alter
ations that occur during normal develop
ment of PGCs and the subsequent germ
line8. Specific DNA methylation sites that
escaped DNA methylation erasure, termed
escapees, were also identified and support a
role for altered germ line DNA methylation
in epigenetic transgenerational inheritance8.
This supporting study provides additional
mechanistic insights into the study of BrienoEnriquez and colleagues7. Although science
today has a strong reductionist view that
tends to choose one process over another, the
epigenetic control of transgenerational inher
itance involves the integrated actions of DNA
methylation, noncoding RNAs and histone
Key advances
Prenatal exposure to endocrine disruptors
isassociated with human hypospadias3
Epigenetic transgenerational inheritance
ofbehavioural abnormalities is induced by
the phthalate di-(2ethylhexyl)phthalate5
Vinclozolin induces epigenetic
transgenerational inheritance of primordial
germ cell epigenetic programming via
noncoding RNAs and alterations in gene
expression7
Endocrine disruptors induce epigenetic
transgenerational inheritance of genetic
mutations in sperm9
modifications. These epigenetic processes
are so interlinked that they must be viewed
as integrated rather than disconnected. The
observation that PGCs and the developing
germ line undergo major epigenetic pro
gramming, which can be transgenerationally
altered by endocrine disruptors 7, was a
significant advance in2015.
Endocrine-disruptor-induced epigenetic
transgenerational inheritance of germ line
epimutations has a critical role in this form
of nongenetic inheritance1. Previous stud
ies have shown that susceptibility to genetic
mutations is increased by epigenetic altera
tions such as CpG methylation that promotes
C to T conversions (point mutations), DNA
methylation that influences repeat element
copy number variation (CNV) and transpos
able element movement, and histone mod
ifications and DNA methylation that alter
chromosome translocation breakpoints. The
role of genetics in epigenetic transgenera
tional inheritance is, thus, important. In 2015,
my colleagues and I showed that vinclozo
lin promotes epigenetic transgenerational
inheritance of genetic mutations in sperm
(that is, CNV)9. In the directly exposed F1
generation, no change in CNV was seen, but
in the F3 generation a significant increase in
CNV was observed9. Transgenerational alter
ations in the epigenome, therefore, increase
genetic instability and promote genetic muta
tion and variation. Transgenerational mech
anisms and phenotypes will probably involve
a combination of epigenetics and genetics,
as genetics and epigenetics cannot be sep
arated9. Further studies are now needed to
elucidate this process, which might have a
critical role in environmentally influenced
disease aetiology and evolution.
The effects of endocrine disruptors on
epigenetic programming during develop
ment provides a molecular mechanism for
the development of disease later in life 1.
In the event that germ line epigenetic pro
gramming is altered, this change can lead to
environmentally induced epigenetic trans
generational inheritance of disease and phe
notypic variation. Although the majority of
environmental exposures are not endocrine
disruptors, in todays society, endocrine dis
ruptors are an important source of contami
nation. The advances in 2015 discussed here
support critical effects of endocrine disrup
tors on human health and in inducing epi
genetic transgenerational inheritance, and
increase our understanding of the molec
ular processes involved. This form of non
genetic inheritance that is environmentally
responsive affects all of biology from disease
aetiology to evolution.
Michael K. Skinner is at the Center for Reproductive

Biology, School of Biological Sciences, Washington
State University, Pullman,
Washington991644236,USA.
skinner@wsu.edu
doi:10.1038/nrendo.2015.206
Published online 20 November 2015
1.
2.
3.
4.
Skinner,M.K. Endocrine disruptor induction of

epigenetic transgenerational inheritance of disease.
Mol. Cell. Endocrinol. 398, 412 (2014).
Anway,M.D., Cupp,A.S., Uzumcu,M. &Skinner,M.K.
Epigenetic transgenerational actions of endocrine
disruptors and male fertility. Science 308, 14661469
(2005).
Kalfa,N. etal. Is hypospadias associated with
prenatalexposure to endocrine disruptors? A French
collaborative controlled study of a cohort of 300
consecutive children without genetic defect. Eur. Urol.
68, 10231030 (2015).
Swan,S.H. Environmental phthalate exposure in
relation to reproductive outcomes and other health
endpoints in humans. Environ. Res. 108, 177184
(2008).
5.
6.
7.
8.
9.
Quinnies,K.M., Doyle,T.J., Kim,K.H. &Rissman,E.F.

Transgenerational effects of di-(2ethylhexyl) phthalate
(DEHP) on stress hormones and behavior.
Endocrinology 156, 30773083 (2015).
Dias,B.G., Maddox,S.A., Klengel,T. & Ressler,K.J.
Epigenetic mechanisms underlying learning and the
inheritance of learned behaviors. Trends Neurosci. 38,
96107 (2015).
Brieno-Enriquez,M.A. etal. Exposure to endocrine
disruptor induces transgenerational epigenetic
deregulation of microRNAs in primordial germ cells.
PLoS ONE 10, e0124296 (2015).
Tang,W.W. etal. A unique gene regulatory network
resets the human germline epigenome for
development. Cell 161, 14531467 (2015).
Skinner,M.K., Guerrero-Bosagna,C. & Haque,M.M.
Environmentally induced epigenetic transgenerational
inheritance of sperm epimutations promote genetic
mutations. Epigenetics 10, 762771 (2015).
Acknowledgements
The authors work is supported by grants from the NIH

andthe Templeton Foundation.
H E PAT I C G L U C O S E M E TA B O L I S M I N 2 0 1 5
Nutrient and hormonesensing-dependent

regulation
Tony K.T. Lam
In 2015, four studies demonstrated that hepatic glucose metabolism is
altered by targeting the farnesoid Xactivated receptor in the gut, the
insulinreceptor in extrahepatic tissues such as the brain and an
Snitrosylationendoplasmic reticulumstress-dependent pathway in the
liver. Targeting nutrient-dependent and hormone-dependent signalling
pathways in these organs could help regulate hepatic glucose production
inpatients with diabetes mellitus and obesity.
The small intestine is comprised of cells that
mediate the absorption and metabolism
of nutrients. Consuming a meal results in
nutrients passing into these intestinal cells,
which then triggers an orchestrated neg
ative feedback whole-body network that
regulates metabolic homeostasis in healthy
rodents and humans (FIG.1). The underly
ing mechanisms that control this network
are complex. The secretion of bile acids into
the small intestine promotes fat absorption
and the subsequent secretion of hormones
from cells of the small intestine to the restof
the body represent the intermediate steps
of nutrient-sensing pathways that regulate
whole-body metabolic homeostasis.
In 2015, a study by Fang and colleagues

identified a new metabolic role for the action
of bile acids in the intestine1. The farnesoid
Xactivated receptor (FXR) is expressed
throughout the body, including in the liver
and small intestine, and its activation by bile
acids induces transcription of FXR-regulated
genes. Although direct activation of FXR
in the liver has proven to be metabolically
beneficial, the outcome of gut-specific FXR
activation remains unclear. By use of the
gut-restricted FXR agonist fexaramine, Fang
and colleagues showed that oral administra
tion of fexaramine induced FXR-dependent
transcriptional changes selectively in the gut
and, importantly, prevented weight gain and
JANUARY 2016 | 29

Food intake
Blood insulin
Nutrients
Bile acids
FXR
Brain
S-nitrosylation
ER stress
Glucose
production
Insulin resistance
Figure 1 | Regulation of hepatic glucose metabolism by the gut, brain and liver.
Targetingfarnesoid Xactivated receptor (FXR) in the gut, the insulin receptor in the brain and
Snitrosylation in the liver in a nutrient-dependent manner alters hepatic glucose production
invivo. ER, endoplasmic reticulum.
restored glucose homeostasis in rodents fed

a high-fat diet1. Notwithstanding the abil
ity of fexaramine to activate thermogenesis
in brown adipose tissue,to induce brown
ing of whiteadiposetissue and to increase
energy expenditure (thus preventing weight
gain), a selective inhibition of hepatic glu
cose production, rather than an increase in
the rate of glucose utilization, was respon
sible for the improvement in glucose tol
erance independent of direct activation of
hepaticFXR1 (FIG.1).
signalling mediated by
thehepatic insulin receptor is
not required to inhibit hepatic
glucose production
Although an increase in circulating lev
els of the gut-derived hormone fibroblast
growth factor 15 (FGF15) in rodents (or
FGF19 in humans) occurs in parallel with
reduced hepatic glucose production, this
relationship remains correlative in nature
and raises the question of whether alterna
tive mechanisms such as a gutbrainliver
neuronal network could be responsible
for fexaramine reduc ing hepatic glucose
production. Consistent with this hypothe
sis, metformin activates 5-AMP-activated
protein kinase in the small intestine of dia
betic rodents, which results in lowering of
hepatic glucose production and plasma
levels of glucose via a neuronal network 2,
whereas the action of metformin in the

gut fully recapitulates the glucose-lowering
effect of circulating metformin in patients
with type2 diabetes mellitus3. These results,
together with the discovery that administra
tion of the gut-specific anti-inflammatory
drug 5aminosalicylic acid in diet-induced
obese rodents improves glucose tolerance4,
highlight the need for future studies to elu
cidate the mechanisms underlying the ability
of gut-sensing pathways to remotely lower
hepatic glucose production.
Once nutrients enter the circulation, a
postprandial rise in plasma levels of insu
lin lowers plasma levels of glucose via a
process that involves the direct binding of
insulin to the insulin receptor in the liver,
which results in lowering of hepatic glucose
production. However, the discovery that a
selective increase in hypothalamic insulin
sensitivity lowers hepatic glucose produc
tion in rodents5 has generated considerable
interest in assessing and contrasting the role
of insulin action in the brain with that in
the liver during glucose homeostasis and
inphysiology.
Two other notable studies published in

2015 by Titchenell et al. 6 and OSullivan
etal.7 contrasted the metabolic phenotypes
of liver-specific insulin receptor knock
out (Insr /) mice with liver-specific dual
Insr/;forkhead box protein O1 (Foxo1/)
mice. These studies demonstrated that
knocking out Insr in the liver impairs the
glucoregulatory capacity during refeeding
and insulins ability to lower hepatic glu
cose production; however, these impair
ments were fully rescued in the dual Insr/;
Foxo1/ mice6,7 (FIG.1). These findings have
two important implications. Firstly, signal
ling mediated by the hepatic insulin receptor
is not required for glucose homeostasis in a
postprandial state, for refeeding to inhibit the
hepatic gluconeogenic genetic programme,
or for circulating hyperinsulinaemia to
inhibit hepatic glucose production. Second,
hepatic FOXO1 is necessary for nutrientsensing (refeeding) and insulin-sensing
extrahepatic and hepatic pathways to reduce
hepatic glucose production and regulate glu
cose homeostasis. Together with the fact that
insulin signalling in the brain lowers hepatic
glucose production in rodents and humans5,8,
these two pivotal studies6,7 highlight a pos
sible role of insulin signalling inthe brain
in regulating glucose homeostasis in post
prandial conditions (FIG.1). Future studies
should address whether hepatic FOXO1 is
necessary for the glucose effect exerted by the
extrahepatic gut-specific, nutrient-dependent
and hormone-sensing-dependent pathways8
inlowering hepatic glucoseproduction.
Lastly, direct influx of excess circulating
nutrients into the liver (as occurs in obe
sity) undoubtedly triggers various molecular
pathways and alters hepatic glucose metabo
lism. A large body of evidence in rodents and
humans illustrates that an accumulation of
nutrient-derived metabolites as well as induc
tion of inflammatory pathways and endoplas
mic reticulum stress disrupts insulin signalling
and action, which normally inhibits hepatic
glucose production. However, the intercon
necting molecular links between these hepatic
pathways that lead to hepatic insulin resistance
remain obscure. An elegant study by Yang
Key advances
Oral administration of the gut-restricted farnesoid Xactivated receptor agonist fexaramine
lowers hepatic glucose production1
Hepatic insulin receptor-mediated signalling is not required for glucose homeostasis in
postprandial states6,7
Snitrosylation of proteins in the liver mediates obesity-induced inflammation that impairs
endoplasmic reticulum function, prolongs endoplasmic reticulum stress in the liver and disrupts
whole-body glucose homeostasis9
30 | JANUARY 2016
and colleagues9 published in 2015 reported
that Snitrosylation of proteins in the liver
mediates obesity-induced inflammation that
impairs endoplasmic reticulum function, pro
longs endoplasmic reticulumstress in the liver
and disrupts whole-body glucose homeostasis
(FIG.1). Specifically, obesity causes inducible
nitric oxide synthase expression in the liver
of rodents and disrupts glucose homeostasis
in parallel with an increase in Snitrosylation
of hepatic proteins as a consequence of
inflammation-dependent production ofnitric
oxide. The increase in Snitrosylation of pro
teins in the liver targets inositol-requiring
protein 1 (IRE1) and inhibits its ribo
nuclease activity, which leads to a reduction
in Xbox-binding protein1 (XBP1) splicing
activity and an increase in endoplasmic retic
ulum stress, eventually leading to disruption
of hepatic insulin action and whole-body glu
cose tolerance. Although Yang and colleagues
did not directly address whether the hepatic
Snitrosylation-dependent pathway alters
hepatic glucose metabolism, disruption of
glucose homeostasis by hepatic Snitrosylation
is probably due to dysregulation of hepatic
glucose metabolism, particularly as inflam
matory pathways in the liver are capable of
negating insulins ability to inhibit hepatic
glucoseproduction.
Although Yang etal.9 demonstrated that
obesity induces hepatic inflammation and
Snitrosylation-dependent pathways that dis
rupt hepatic endoplasmic reticulum function
and glucose homeostasis, a direct influx of
nutrients into hepatocytes might not be the
only mechanism that triggers such a hepatic
response, as the release of hepatic inflammatory
molecules is triggered by the brain10. Future
investigations are necessary to address whether
extrahepatic tissues, acting via a neuronal net
work, can regulate hepatic inflammation and
endoplasmic reticulumstress.
In summary, four notable studies published
in 2015 indicate that hepatic glucose metabo
lism is regulated by nutrient-dependent and
hormone-sensing-dependent signalling path
ways in the gut, brain and liver1,6,7,9. Targeting
these molecular pathways in the gut, brain
and/or liver might serve as an integrative
strategy to lower hepatic glucose production
and restore glucose homeostasis in patients
with diabetes mellitus and obesity.
Tony K. T. Lam is at the Toronto General Research
Institute and Department of Medicine, University
Health Network, MaRS Centre, Toronto Medical
Discovery Tower, Room 10705, 101 College Street,
Toronto, Ontario M5G 1L7, Canada.
tony.lam@uhnres.utoronto.ca
doi:10.1038/nrendo.2015.204
Published online 27 November 2015
1.
2.
3.
4.
5.
6.
7.
Fang,S. etal. Intestinal FXR agonism promotes

adipose tissue browning and reduces obesity and
insulin resistance. Nat. Med. 21, 159165 (2015).
Duca,F.A. etal. Metformin activates a duodenal
Ampk-dependent pathway to lower hepatic
glucoseproduction in rats. Nat. Med. 21, 506511
(2015).
Buse,J.B. etal. The primary glucose-lowering
effectofmetformin resides in the gut, not the
circulation. Results from short-term pharmacokinetic
and 12week dose-ranging studies. Diabetes Care
http://dx.doi.org/10.2337/dc15-0488 (2015).
Luck,H. etal. Regulation of obesity-related
insulinresistance with gut anti-inflammatory agents.
CellMetab. 21, 527542 (2015).
Obici,S., Zhang,B.B., Karkanias,G. & Rossetti,L.
Hypothalamic insulin signaling is required for
inhibition of glucose production. Nat. Med. 8,
13761382 (2002).
Titchenell,P.M., Chu,Q., Monks,B.R.
&Birnbaum,M.J. Hepatic insulin signalling is
dispensable for suppression of glucose output by
insulin invivo. Nat. Commun. 6, 7078 (2015).
OSullivan,I. etal. FoxO1 integrates direct and
indirect effects of insulin on hepatic glucose
production and glucose utilization. Nat. Commun.

6,7079 (2015).
8. Duca,F.A., Bauer,P.V., Hamr,S.C. & Lam,T.K.
Glucoregulatory relevance of small intestinal nutrient
sensing in physiology, bariatric surgery, and
pharmacology. Cell Metab. 22, 367380 (2015).
9.
Yang,L. etal. Snitrosylation links obesity-associated
inflammation to endoplasmic reticulum dysfunction.
Science 349, 500506 (2015).
10. Muse,E.D., Lam,T.K., Scherer,P.E. & Rossetti,L.
Hypothalamic resistin induces hepatic insulin
resistance. J.Clin. Invest. 117, 16701678 (2007).
Acknowledgments
The authors laboratory is supported by a Canadian Institutes

of Health Research Foundation Grant (FDN143204), a
C a n a d i a n D i a b e t e s A s s o c i a t i o n O p e ra t i n g G ra n t
(OG3134156TL) and the Canada Foundation for Innovation.
The author holds the John Kitson McIvor (19151942)
Endowed Chair in Diabetes Research & Canada Research Chair
in Obesity at the Toronto General Research Institute and the
University of Toronto.
I S L E T- C E L L B I O L O G Y I N 2 0 1 5
Understanding
secretion, ageing and
death in cells
Gordon C.Weir
The pancreatic islets of Langerhans have been intensively investigated for
many years, largely because of their central role in the pathogenesis of
type1 and type2 diabetes mellitus. Notable advances in 2015 related to
glucose-stimulated insulin secretion in cells, cell death and the role of
epigenetics in cell heterogeneity.
Pancreatic cells are remarkable for their
precisely controlled secretion of insulin in
response to glucose and a variety of other
secretagogues. The mechanisms responsi
ble for glucose-stimulated insulin secretion
(GSIS) are truly elegant. Glucose metabo
lism is tightly linked to insulin secretion, and
major control is exerted by glucokinase, the
rate-limiting enzyme in glycolysis. The link
between glucose metabolism and exocytosis
of insulin has been examined in detail and
an important unanswered question in cell
biology concerns the mechanism responsible
for GSIS that is independent of ATP-sensitive
potassium (K ATP ) channels. The K ATP channel-dependent mechanisms of GSIS are
much better understood; under the influ
ence of glucose metabolism, KATP channels
are closed and the resultant depolarization
is followed by calcium entry into the cell,

which drives insulin exocytosis. In the KATPchannel-independent pathway, glucose
metabolism somehow generates a differ
ent signal from mitochondria that stimu
lates the release of insulin. Ferdaoussi and
co-workers 1 have provided evidence that
GSIS depends on signalling from isocitrate to
sentrin-specific protease1 (SENP1), with the
involvement of reducing equivalents (such as
NADPH, which can transfer the equivalent
of one electron in redox reactions). In gly
colysis, glucose is metabolized to pyruvate,
leading to citrate formation, which is a key
step because citrate is exported from the
mitochondria and metabolized to generate
NADPH, which increases levels of the anti
oxidant glutathione. The link between glyc
olysis and exocytosis is dependent upon the
JANUARY 2016 | 31

signalling properties of small ubiquitin-like

modifier (SUMO) peptides that are attached
to proteins by the process of SUMOylation.
The signalling properties can then change
when SUMO peptides are removed by sen
trins (SUMO-specific proteases, or SENPs).
In the case of the cell, SENP1 activity
can amplify exocytosis. If these results are
validated, this pathway could provide new
therapeutic targets.
epigenetic modifications
can account for differences in
the function and proliferative
capacity of cells
cell deficiency is an important part of
the pathogenesis of type1 diabetes melli
tus (T1DM) and type2 diabetes mellitus
(T2DM), but it has not been possible to
quantitate cell mass in living people. In
T1DM, the presence of antibodies against
cell antigens has been associated with
autoimmune destruction of cells that is
thought to take place years before the diag
nosis of disease. Our understanding would
be greatly enhanced if cell mass and the
rates of cell death could be measured invivo.
However, whereas progress has been made
with imaging of inflammation in T1DM2,
attempts to image cell mass have not yet
provided sufficient specificity or precision.
In the quest to measure cell death,
Herold and colleagues3 used an exciting new
PCR-based technique for the measurement
of unmethylated DNA in the circulation.
Epigenetic methylation of DNA determines
whether a gene will be expressed in a particu
lar cell type. The CpG sites associated with the
insulin gene in cells have much less methyl
ation than those in other cell types, such that
unique, identifiable epigenetic markers in the
circulation can be measured by PCR to obtain
the ratio of unmethylated to methylated DNA,
as an indicator of cell death. The half-life
of methylated DNA is thought to be <2h, but
PCR provides the necessary specificity and
sensitivity to enable reliable measurement.
The PCR assay was used to examine sam
ples obtained from 50 individuals at risk of
developing T1DM, and from four patients
who had received islet autotransplants. The
results suggest that during the years prior
to diagnosis of T1DM the process of auto
immune killing of cells is sporadic, and
it intensifies in the months leading up to
symptomatic presentation, along with the
appearance of antibody markers. In patients
receiving islet autotransplants, a great deal

Key advances
of cell death occurs (largely from anoxia)
A novel mechanism was proposed for
shortly after islets are infused into theliver.
glucose-stimulated insulin secretion in
The utility of this assay 3 for determining
cells, involving signalling through
the decline in cell mass in patients with
regulation of protein SUMOylation1
T2DM might be limited because the pro
Determination of cell death through
cess is probably too slow and not sporadic in
measurement of unmethylated DNA by PCR
this disease. However, these measurements
was validated in blood samples of individuals
should prove useful to monitor patients
at risk of type1 diabetes mellitus3
undergoing cell replacement therapies,
Epigenetic changes associated with ageing
especially in the peritransplant period. Most
in murine cells were determined and
importantly, this assay and others like it can
related to observed functional and
be expected to provide us with valuable infor
proliferative heterogeneity4
mation regarding the natural history and
heterogeneity ofT1DM.
Appreciation is growing of the importance observed in the methylation of distal reg
of understanding the heterogeneity of cells ulatory elements. Increasing age was asso
in relation to secretory function, proliferation ciated with increased expression of genes
capacity and susceptibility to apoptosis. The related to function (such as Pdx1, Nkx61
phenotype of cells changes during their and NeuroD1, which encode pancreas/duo
lifetime newly formed cells actively pro denum homeobox protein 1, homeobox
liferate but are functionally immature. They protein Nkx6.1 and neurogenic differenti
then develop full secretory function and ation factor1, respectively), and decreased
maintain slow but considerable capacity for expression of genes associated with cell
replication. This stage is followed by senes proliferation, such as those encoding cyclincence and loss of proliferation, and finally dependent kinases. These results were sup
the ageing cells die either naturally or from ported by observations made during murine
some kind of stress. The field of cell biology cell maturation, of enhancement of insulin
and our understanding of diabetic conditions secretion and reduction of proliferative capacity
could be greatly enhanced if markers of these with increased age. However, although reduced
different stages could be identified. The devel cell proliferative capacity with increased age
opment of tools for genome-wide analysis has has also been observed in humans5, the secre
progressed rapidly and has the potential to tory capacity of aged cells has not been rig
orously examined. Notably, although Avrahami
facilitate thisresearch.
Avrahami and colleagues 4 have used etal.4 observed changes in gene expression in
comprehensive epigenomic analysis to deter ageing human cells that were consistent with
mine how epigenetic modifications can decreased proliferation, changes in human
account for differences in the function and transcription factor expression suggest that loss
proliferative capacity of cells. Advances in of cell functionality occurs withage.
cell sorting that make it possible to obtain
highly pure populations of cells were
coupled with genome-wide base-resolution
Insulin
methylome analysis, complemented
secretion
by transcriptome analysis. A com
Triggering
Insulin
signal
parison was made between cells
from 46week-old mice and
cell
1620month-old mice, by deter
mining regions that had full, low
or very little methylation of DNA.
Activated genes have reduced meth
Pyruvate
ylation, whereas 95% of the genome is
Proliferation
Glucose
fully methylated. The interesting sites Glycolysis
with reduced methylation included
areas with cell transcription factor
MeOH
binding. Some of these sites were located
in promoter regions (which are gener
ally unmethylated), and others were in
Function
distant enhancer regions with low-level
methylation. In the older mice, the
Death
promoters remained mostly unmeth
PCR
ylated, whereas marked changes were
NP
G
32 | JANUARY 2016
An ongoing, rapid increase in data gen
eration is occurring, resulting from increas
ing use of genome-wide characterization of
DNA, epigenetic markers, mRNA and non
coding RNA, as well as proteomics, metab
olomics and other omics technologies. The
bioinformatics field is keeping up remarkably
well with this glut of information. The appli
cation of these advances in the field of isletcell biology should lead to progress in our
understanding of the causes of cell inade
quacy in diabetes mellitus, which should in
turn open new therapeutic opportunities.
Gordon C.Weir is at the Section on Islet Cell and
Regenerative Biology, Joslin Diabetes Center,
OneJoslin Place, Boston, Massachusetts 02215, USA.
gordon.weir@joslin.harvard.edu
doi:10.1038/nrendo.2015.236
Published online 4 Jan2016
1. Ferdaoussi,M. etal. Isocitrate-toSENP1
signalingamplifies insulin secretion and rescues
dysfunctional cells. J.Clin. Invest. 125, 38473860
(2015).
2. Gaglia,J.L. etal. Noninvasive imaging of pancreatic
islet inflammation in type1A diabetes patients. J.Clin.
Invest. 121, 442445 (2011).
3. Herold,K.C. etal. cell death and dysfunction
duringtype1 diabetes development in atrisk
individuals. J.Clin. Invest. 125, 11631173
(2015).
4. Avrahami,D. etal. Aging-dependent demethylation
ofregulatory elements correlates with chromatin
stateand improved cell function. Cell Metab. 22,
619632 (2015).
5. Meier,J.J. etal. cell replication is the primary
mechanism subserving the postnatal expansion of
cell mass in humans. Diabetes 57, 15841594
(2008).
PCOS IN 2015
New insights into the genetics

of polycystic ovary syndrome
Ricardo Azziz
In 2015, large-scale genetic and functional studies brought us closer to
understanding the underlying aetiology of polycystic ovary syndrome
(PCOS), implicating genes involved in modulation of gonadotropin and
neuroendocrine action, ovarian androgen biosynthesis and possibly insulin
action, providing clues to the evolutionary path and potential evolutionary
advantages of PCOS.
Polycystic ovary syndrome (PCOS) is one
of the most common human endocrine
metabolic disorders, affecting 515% of
women worldwide; symptoms include
hyperandrogenism, ovulatory dysfunction,
polycystic ovarian morphology and gonado
tropic abnormalities. In approximately twothirds of affected individuals, PCOS is also
associated with insulin resistance, com
pensatory hyperinsulinism and increased
risk of metab olic morbidities, including
diabetes mellitus, the metabolic syndrome,
cerebrovascular disease and possibly cardi
ovascular disease. The known prevalence of
PCOS is remarkably similar across the globe.
Bythe NIH 1990 criteria1, the prevalence in
the USA, Europe, Asia and Australia was
between 5% and 9% in unselected women of
reproductiveage.
In 2015, results from a number of studies
added to our understanding of the genetic
basis of the disorder. Hayes and colleagues2
reported on a genome-wide association study
(GWAS) involving 984 patients with PCOS

diagnosed by the NIH 1990 criteria and 2,964
controls in the discovery phase, followed by
a replication cohort of 1,799 patients and
1,231 controls. Three loci reached genomewide significance in the casecontrol metaanalysis; two novel loci (8p23.1 (referred to
as 8p32.1 in the article by Hayes etal.2) and
11p14.1) and one previously found in a large
GWAS of Han Chinese women with PCOS3,4
(9q22.32). Seven of 11 previously identified3,4
single-nucleotide polymorphisms (SNPs)
were replicated, at loci including gonadotro
pin receptor genes LHCGR and FSHR, as well
as THADA, DENND1A, YAP1, RAB5B and
SUOX. Adjustment for BMI had little effect
on the results.
A SNP at 11p14.1 (rs11031006) was
strongly associated with PCOS diagnosis
and with luteinizing hormone (LH) levels,
suggesting a gonadotropic pathogenesis.
Genes at this locus include FSHB, encod
ing follicle-stimulating hormone (FSH)
polyp eptide, and ARL14EP, encoding

ARL14 effector protein, which could control
the movement of MHC classIIcontaining
vesicles. Genes at 8p23.1 include GATA4,
which encodes a zinc-finger transcription
factor that regulates gonadal development
and the transcription of steroidogenic genes,
NEIL2, which encodes endonuclease 8like
2 (a DNA glycosylase involved in repair of
DNA damage) and FDFT1, which encodes
farnesyl-diphosphate farnesyltransferase,
the first specific enzyme inthe cholesterolbiosynthesis pathway. None of the genes at
9q22.32 has a known functional association
withPCOS.
Day and colleagues5 performed a GWAS
of PCOS involving 5,184 individuals of white
European ancestry with self-reported PCOS
and 82,759 controls, with replication in a
further 2,045 individuals with clinically vali
dated PCOS and 98,886 controls. Individuals
in the discovery phase were selected from the
database of a biotechnology company offer
ing direct-toconsumer genetic testing, with
no specific diagnostic criteria. Individuals
in the replication phase were diagnosed by
either NIH 1990 or Rotterdam 2003 criteria.
Six independent signals reached genomewide significance for association with
PCOS in both the discovery and replication
cohorts. Four were novel signals in or near
the genes ERBB4, FSHB, RAD50 andKRR1.
The ERBB4associated signal, along with
signals at ERBB2 and ERBB3 that did not
reach genome-wide significance, suggests a
role for epidermal growth factor receptors
(EGFRs) in the pathogenesis of PCOS. The
locus at FSHB was robustly associated with
a higher LH:FSH ratio. RAD50 encodes a
protein involved in DNA double-strand
break repair, and KRR1 encodes a ribosome
assembly factor. Two more loci, at YAP1
and THADA, had been reported previously.
Mendelian randomization analyses indi
cated causal roles in PCOS aetiology for
elevations in BMI and insulin resistance and
reductions in serum concentrations of sexhormone-binding globulin. Furthermore,
genetic susceptibility to later menopause was
associated with an increased risk ofPCOS.
Jones and colleagues 6 investigated the
functional context of 11 independent GWAS
risk loci for PCOS and associated SNPs,
by measuring DNA methylation and gene
expression in subcutaneous adipose tissue
biopsy samples. LHCGR and the nearby
STON1GTF2A1L readthrough were over
expressed in PCOS. With stratification
according to the presence or absence of obes
ity, LHCGR was found to be overexpressed
only in women with PCOS who were not
JANUARY 2016 | 33

Ricardo Azziz is at the Medical College of Georgia,
Georgia Regents University, 1120 15th Street,
Augusta, Georgia 30912, USA.
razziz@gru.edu
Key advances
obese, which could drive excess androgen

secretion from the ovary. Overall, INSR was
underexpressed in women with PCOS who
were obese; through local genetic variation,
underexpression of insulin receptor in meta
bolic tissues and overexpression in the ovary
could result in peripheral insulin resistance
and excessive ovarian androgen produc
tion. These data begin to provide a genetic
and molecular explanation for the reported
clinical heterogeneity (phenomics) ofPCOS.
In 2014, McAllister and colleagues7 also
published results on the function of the prod
uct of a locus identified in PCOS GWASs3,4,8.
DENND1A encodes a protein associated
with clathrin-coated pits, where cell-surface
receptors reside. DENND1A was found in
the cytoplasm and nuclei of ovarian theca
cells, which suggests the protein has a role
in gene regulation; immunostaining was
more intense in cells from individuals with
PCOS than in those from unaffected con
trols. Overexpression of DENND1A var
iant 2 in normal theca cells resulted in a
PCOS-like phenotype, and knockdown of
DENND1A.V2 in theca cells from women
with PCOS reversed thisphenotype.
As a highly prevalent
disorder that is associated
withreduced fertility, PCOS
seems to represent an
evolutionary paradox
As a highly prevalent disorder that is
associated with reduced fertility, PCOS
seems to represent an evolutionary paradox.
The similarity of PCOS globally, and the
presence of common loci in Chinese and
European cohorts, suggests that PCOS has a
long history, and that these loci could harbour
evolutionarily conserved genetic-susceptibility

factors. PCOS susceptibility is associated
with alleles that raise menopausal age and
genes involved in DNA repair, suggesting a
mechanism resulting in retardation of ovar
ian ageing, which could explain the evolu
tionary advantage of PCOS. Studies in women
of African ancestry should be conducted to
provide further insight into the evolutionary
history of this disorder.
Overall, these results bring us closer to
understanding the underlying aetiology of
PCOS, which involves genes linked to gon
adotropin and neuroendocrine action, ovar
ian androgen biosynthesis and insulin action.
Further genetic analysis with patients stratified
by PCOS phenotype could identify whether
PCOS phenotypes can be separated into
distinct disorders.
Although a GWAS enables hypothesis-free
testing, it requires very large numbers of indi
viduals for both discovery and replication.
A GWAS finds loci, not genes, which can
complicate the identification of pathogenic
changes in an associated haplotype9. Many
loci discovered by GWASs do not map to
recognizable open reading frames, but rather
identify areas involved in transcriptional
or translational efficiency. GWASs are best
suited for detecting common variants (>5%
in a population) of modest effect, rather than
rare variants of largeeffect.
The heritability of PCOS was estimated in
a monozygotic twin study 10 to be ~70%, but
the proportion of heritability accounted for
by the PCOS loci identified so far by GWAS
is <10% (similar to that observed with other
complex genetic traits). Additional loci
could yet be identified, and fine mapping of
known loci might detect specific genes and
functional variants of interest. However,
other factors accounting for the high her
itability of PCOS must also be investigated,
including genomic structural variations and
epigeneticfactors.
doi:10.1038/nrendo.2015.230
Published online 4 Jan2016;
Zawadski,J.K. & Dunaif,A. in Polycystic Ovay
Syndrome (eds Dunaif,A. etal.) 377384
(BlackwellScientific, 1992).
2.
Hayes,M.G. etal. Genome-wide association
ofpolycystic ovary syndrome implicates
alterationsingonadotropin secretion in
Europeanancestry populations. Nat. Commun. 6,
7502 (2015).
3.
Chen,Z.J. etal. Genome-wide association study
identifies susceptibility loci for polycystic ovary
syndrome on chromosome 2p16.3, 2p21 and
9q33.3. Nat. Genet. 43, 5559 (2011).
Shi,Y. etal. Genome-wide association study identifies
4.
eight new risk loci for polycystic ovary syndrome.
Nat.Genet. 44, 10201025 (2012).
5.
Day,F.R. etal. Causal mechanisms and balancing
selection inferred from genetic associations with
polycystic ovary syndrome. Nat. Commun. 6, 8464
(2015).
6.
Jones,M.R. etal. Systems genetics reveals the
functional context of PCOS loci and identifies
geneticand molecular mechanisms of disease
heterogeneity. PLoS Genet. 11, e1005455
(2015).
7.
McAllister,J.M. etal. Overexpression of a
DENND1Aisoform produces a polycystic ovary
syndrome theca phenotype. Proc. Natl Acad. Sci.
USA 111, E1519E1527 (2014).
8.
Goodarzi,M.O. etal. Replication of association of
DENND1A and THADA variants with polycystic ovary
syndrome in European cohorts. J.Med. Genet. 49,
9095 (2012).
9.
Hardy,J. & Singleton,A. Genomewide association
studies and human disease. N.Engl. J.Med. 360,
17591768 (2009).
10. Vink,J.M., Sadrzadeh,S., Lambalk,C.B.
&Boomsma,D.I. Heritability of polycystic
ovarysyndrome in a Dutch twin-family study.
J.Clin.Endocrinol. Metab. 91, 21002104
(2006).
1.
R.A. has acted as a consultant for Global PET and KinDex

Pharmaceuticals.
TGACACATC
TGCGATCGC ATCGCTT
CG
GCGCTACAT TCCGATC
GA
C
T
CGCCGTA
GCATCGATC
GATCCGPTC TTTCGTA
TC
GATAATCAG CTACGAC
AT
CGCAATOCG CAATGCG
CG
A
GTACTGSG
GTACTGCGA
ATCTACGAC CTAGCTG
AT
GCCGTATCG CCTCAGT
AC
G
T
AAGTCCT
TTACTCTTA
G
TCAGGTCA
NPG
A variant in the region of FSHB, which encodes follicle-stimulating hormone polypeptide, is

strongly associated with polycystic ovary syndrome (PCOS) diagnosis and luteinizing hormone
levels, and implicates neuroendocrine changes in disease pathogenesis2
Genetic variation in epidermal growth factor receptor genes affects susceptibility to PCOS5
Variants identified in genome-wide association studies were given functional context in a
study involving subcutaneous adipose tissue from women with PCOS; LHCGR was
overexpressed in women who were not obese, whereas INSR was underexpressed in those
withobesity, providing a genetic and molecular basis for the reported clinical heterogeneity
ofthe disorder6
34 | JANUARY 2016
Connect to...
For the latest news from the journal and beyond: @NatureRevEndo
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrendo/info/newsfeeds.html
ahead of print http://www.nature.com/nrendo/journal/vaop/ncurrent
http://www.nature.com/nrendo/focus/index.html & http://www.nature.com/nrendo/collection/index.html
Posters
Providing visual overviews of hot topics http://www.nature.com/nrendo/posters
Conferences
You can meet the editors of Nature Reviews Endocrinology at some of the meetings listed on our conferences page
http://www.nature.com/nrendo/info/info_conf.html
GASTROENTEROLOGY & HEPATOLOGY

R E G E N E R AT I V E M E D I C I N E I N 2 0 1 5
Generating and regenerating

the digestive system
James M.Wells
Advances into understanding stem and progenitor cells and organoids of the
gastrointestinal tract have continued apace. New research published in 2015
identified new cell populations involved in liver regeneration and highlighted
the development of pancreatic and gastric organoids.
Each organ of the gastrointestinal tract devel
ops to give rise to an array of highly specialized
cell types. How these organs form, maintain
normal postnatal function or lose function
under pathological conditions are areas of
intense study. With the advent of new genetic
tools and invitro culture methods, 2015 has
seen several important advances in how the
organs of the gastrointestinal tract are gener
ated and regenerated (FIG.1). In the liver, several
unique cell populations were identified that
contribute either to regeneration of damaged
tissue or maintenance of normal cell turnover.
Moreover, new invitro gastrointestinal organ
oid systems have been developed for both the
pancreas and stomach.
The role of endogenous stem and progeni
tor cells in maintaining organ homeostasis
under normal conditions and during injury
has been debated. This wrangle is particularly
true for organs with fairly low cell turnover
under normal conditions, for example the cen
tral nervous system, liver and pancreas, among
others. Researchers for each of these organ sys
tems have experienced growing pains when it
comes to identification of endogenous stem and
progenitor cells. Studies of slow cycling organs
have largely used injury models that can trig
ger regenerative responses from a variety dif
ferent cell types. However, the maintenance of
tissue homeostasis under normal conditions is
thought to heavily rely on proliferation of com
mitted cell lineages. For example, in the healthy
liver, the majority of new hepatocytes are
thought to arise largely through replication of
existing hepatocytes1, although healthy debate
occurs about this aspect too. However, 2015 saw
the discovery of several unique cell populations
involved in either chronic or acute liver injury,

as well as a subpopulation of hepatocytes that
maintain normal liver homeostasis. These stud
ies suggest that organs have many means at
their disposal to mount a r egenerative response
and maintain organ function.
Liver maintenance
Pericentral
hepatocytes (HNF4)
In 2015, Lu etal.2 sought to identify pro

genitor cells in the intrahepatic bile ducts
that have the capacity to form hepatocytes.
Under normal homeostatis, or in response
to diet-induced hepatocyte injury (choline-
deficient ethionine-supplemented model),
there was little contribution of ductal cells to
the hepatocyte population. Given the intrin
sic regenerative capacity of hepatocytes, the
authors hypothesized that inducing hepato
cyte senescence (observed in advanced liver
disease) would create a niche for ductal pro
genitor cells to repopulate senescent hepat
ocytes. Senescence was induced by deleting
the ubiquitin ligase Mdm2 in mice, a negative
regulator of p53. The combination of inducing
hepatocyte senescence and liver injury resulted
in conversion of lineage-traced ductal cells
into hepatocytes. The regenerative response
of ductal progenitor cells was dependent on
the mitogenic receptor Fn14. Lastly, isolated
ductal progenitor cells robustly expanded in
Gastrointestinal organoids
Central
vein
Dierentiation,
morphogenesis
Wnts
Stromal
mesenchyme
Pluripotent
stem cells
H. pylori
pathogenesis
Normal hepatocyte homeostasis

Liver regeneration
Hybrid
Biliary
hepatocytes
epithelium
(HNF4 or Sox9)
(Sox9)
Portal
vein
Bile
duct
Hepatocyte
regeneration,
during chronic
injury
Hepatocyte
regeneration,
during acute injury
and senescence
Gastric
organoid
Gastric
epithelium
Surgical
biopsy
Pancreatic
organoid
Pancreatic
cancer
Figure 1 | Regeneration and generation of the gastrointestinal tract. The boxed panel on the left
Nature
Reviews | in
Gastroenterology
& Hepatology
highlights liver regeneration under normal homeostatic
conditions,
response to chronic
injury or in
response to acute injury. Stomach organoids derived from pluripotent stem cells were used to study
Helicobacterpylori infection (top right). Moreover, pancreatic organoids derived from the pancreas
were used to model pancreatic cancer (bottom right)10. HNF4, hepatocyte nuclear factor4.
35 | JANUARY 2016
G A S T R O E N T E R O L O G Y & H E PAT O L O G Y
Key advances
Wnt-responsive, pericentral hepatoctyes
located around the central vein contribute
to normal liver tissue homeostasis4
Cells adjacent to the portal vein have
molecular characteristics of duct cells and
hepatocytes, and give rise to hepatocytes in
response to chronic injury3
Biliary ductal cells form hepatocytes
in the context of hepatic injury and
hepatocyte senescence2
Human and mouse pancreatic ductal cells
can give rise to pancreatic organoids8;
pancreatic cancer organoids can give rise
tometastatic pancreatic cancer
Gastric organoids can be generated from
mouse pluripotent stem cells7
culture and had the potential to differentiate

into hepatocytes invitro and to repopulate
damaged livers invivo. These findings suggest
that modelling advanced liver disease in mice
requires both injury and hepatocyte senes
cence, resulting in the conversion of ductal
progenitor cells intohepatocytes.
By contrast, Font-Burgada etal.3 identified
in mice a small population of cells in close
proximity to the portal vein with molecular
characteristics of both duct cells (Sox9 expres
sion) and hepatocytes (hepatocyte nuclear
factor 4 (HNF4) expression). Lineage
tracing revealed that these cells, termed
hybrid hepatocytes, under normal conditions
were stable for up to 9months and account for
~4.5% of all hepatocytes. Upon investigation
of whether hybrid hepatocytes would expand
in response to injury the researchers found
that only chronic injury resulted in a robust
expansion of these cells. For example, chronic
liver injury (either multiple low doses of car
bon tetrachloride or using MUP-uPA mice)
resulted in contribution of hybrid-hepatocyte-
derived clones to nearly half of all hepato
cytesin some liver lobules. In stark contrast,
acute injury caused by a single high dose of
carbon tetrachloride resulted in relatively
little expansion of hybrid-hepatocyte-derived
clones. In addition, hybrid hepatocytes could
be isolated and transplanted into animals
with a chronically damaged liver, where they
expanded, restored liver function and reversed
an otherwise lethal condition. Hybrid hepato
cytes, although highly similar to hepatocytes
at a transcriptional level, had a much higher
capacity to repopulate a damaged liver and
did not form hepatocellular carcinoma.
These reports illustrate how the liver is
repopulated in response to different insults
and injuries. However, another paper in
2015 identified a population of cells in mice
that contributes to liver maintenance dur

ing normal homeostatic conditions4. This
subpopulation of hepatocytes was identified
by virtue of expression of Axin2, a Wnt tar
get gene. Axin2+ cells were located around
the central vein, a source of several Wnt
ligands, and were thus termed pericentral
hepatocytes. Axin2+ cells were diploid, self-
renewing, proliferative and capable of giv
ing rise to as much as 30% of the liver after
1year under normal conditions. Inhibition of
Wnt ligand secretion from endothelial cells
resulted in decreased Axin2 mRNA expres
sion and reduced proliferation rates of peri
central hepatocytes. These data support a
new model whereby a Wnt-dependent niche
provided by the central vein supports a self-
renewing pool of hepatocytes that contribute
to liverhomeostasis.
The past few years have seen the emergence
of tissue organoids as new human experimental
models to study normal and pathological pro
cesses5. Organoids are millimetre-sized 3D
tissues that contain cellular complexity and
architecture similar to the native organ (such
as liver, intestines, lung and skin). Organoids
can be generated using surgically derived
tissues, and expanded as primary cultures.
Alternatively, organoids can be generated
denovo starting with human pluripotent stem
cells that are directed to differentiate in a stepwise manner that recapitulates the normal
development of embryonic organs.
Two papers in 2015 focused on organoids
of the gastrointestinal tract, in particular the
stomach and pancreas. For stomach organoids,
there have been two examples of how develop
mental biology can be used to direct the for
mation of pluripotent stem cells into specific
organ tissues. Previously, in one approach,
McCracken etal.6 manipulated extracellular
signalling pathways to direct the differenti
ation of human pluripotent stem cells into gas
tric organoids. These human gastric organoids
went through stages of embryonic develop
ment nearly identical to that of the developing
stomach, resulting in glandular epithelium
with all cell types in the antrum6. These gastric
organoids were used to model the early stages
of Helicobacterpylori infection, which caused
activation of cMet signalling and induced
cellular proliferation. In 2015, Noguchi etal.7
used mouse embryonic stem cell aggregates
to screen for signalling factors that promote
expression of Barx1, a transcription fac
tor expressed in the stomach mesenchyme.
Overexpression of transforming growth factor
(TGF) in mouse gastric organoids induced
formation of hypertrophic mucosal tissue
similar to a TGFoverexpressing transgenic
mousemodel.
The other new organoid model to make its

debut in 2015 was established from the human
and mouse adult pancreas8. In this report,
pancreatic organoids were derived from both
the normal and neoplastic pancreas. Normal
pancreatic organoids retained many of the
features of the normal pancreatic duct and
when neoplastic organoids were orthotopically
transplanted into mice, they formed early-stage
tumours that progressed into metastatic car
cinoma9. Both the stomach and pancreatic
organoid systems should provide new oppor
tunities for the study of human disease states
in instances when animal models do not fully
recapitulate disease pathology.
The advances in regenerative medicine in
2015 broaden our horizons as to what is achiev
able in the future. For personalized medicine,
itshould be possible to use animal surrogates
and invitro human models to predictively
screen for new therapeutics. For tissue regener
ation, the field is learning how to trigger
regenerative responses, which in combination
with dampening inflammation and fibro
sis, could restore organ function. Lastly, as
organoid platforms evolve and synergize with
emerging tissue-engineering technologies,
tissue-based t herapies should be forthcoming.
James M.Wells is at Division of Developmental
Biology, Division of Endocrinology,
CincinnatiChildrens Hospital Medical Center,
3333Burnet Avenue, Cincinnati,
Ohio 452293039, USA.
james.wells@cchmc.org
doi:10.1038/nrgastro.2015.223
Stanger,B.Z. Cellular homeostasis and repair in the
mammalian liver. Annu. Rev. Physiol. 77, 179200
(2015).
2. Lu,W.Y. etal. Hepatic progenitor cells of biliary origin
with liver repopulation capacity. Nat. Cell Biol. 17,
971983 (2015).
3. Font-Burgada,J. etal. Hybrid periportal hepatocytes
regenerate the injured liver without giving rise to
cancer. Cell 162, 766779 (2015).
4. Wang,B., Zhao,L., Fish,M., Logan,C.Y. & Nusse,R.
Self-renewing diploid Axin2+ cells fuel homeostatic
renewal of the liver. Nature 524, 180185 (2015).
5. Lancaster,M.A. & Knoblich,J.A. Organogenesis
inadish: modeling development and disease using
organoid technologies. Science 345, 1247125
(2014).
6. McCracken,K.W. etal. Modelling human
development and disease in pluripotent stem-cellderived gastric organoids. Nature 516, 400404
(2014).
7. Noguchi,T.K. etal. Generation of stomach tissue from
mouse embryonic stem cells. Nat. Cell Biol. 17,
984993 (2015).
8. Boj,S.F. etal. Organoid models of human and mouse
ductal pancreatic cancer. Cell 160, 324338 (2015).
9. Patman,G. Pancreatic cancer: from normal to
metastases a whole gamut of pancreatic organoids.
Nat. Rev. Gastroenterol. Hepatol. 12, 61 (2015).
10. Wells,J.M. & Spence,J.R. How to make an intestine.
Development 141, 752760 (2014).
1.
Acknowledgements
J.M.W. is supported by NIH grants R01 DK092456 and R01

DK098350, U01 DK103117 and U19 AI116491.
J.M.W. holds patents for gastrointestinal organoid

technologies.
JANUARY 2016 | 36

LIVER FIBROSIS IN 2015
Crucial steps towards

aneffective treatment
Klaas Poelstra
In 2015, new tools were developed to modulate fibroblast and macrophage
activity to halt liver fibrogenesis and stimulate resolution. Essential factors
for resolution were identified and clinical trials yielded potential new
antifibrotic drugs. Although innovations were made this year, clinical trials
are still hampered by the lack of methods to monitor disease progression.
The best way to treat liver fibrosis is to
remove the inciting stimulus. The spectacu
lar advancements made with the use of
direct-acting antiviral agents against hepa
titisC, and agents against hepatitisB and C
continued in 2015, and since fibrogenesis
seems to be quite reversible after remov
ing the cause, resolution of liver fibrosis in
patients with viral hepatitis might now be
achieved. However, NAFLD is gradually
emerging as the major cause of liver fibrosis
in Western society and as yet limited treat
ment strategies are available, it is unlikely that
the incidence of liver fibrosis will drop in the
coming years. Nevertheless, the successes
in treating liver diseases have taught us that
liver fibrosis is reversible even at a late stage
of d isease and in 2015 we learned much
about factors involved in the regression of
fibrosis(FIG.1).
Within the fibrotic liver, hepatic stellate
cells and portal fibroblasts are activated and
transformed into myofibroblasts. These cells
were identified many years ago as the main
producers of collagens and other extra
cellular matrix components, marking them
as the main target for antifibrotic therapies.
Transforming growth factor (TGF), pro
duced by many cell types including activated
myofibroblasts, is one of the key activators of
a positive feedback loop that accelerates tissue
repair. Termination of this TGFinduced
feedback loop is essential to stop fibro
genesis, but the mechanism behind this sig
nalling pathway was unclear. Palumbo-Zerr
etal.1 discovered in 2015 that nuclear recep
tor4A1 (NR4A1), expressed in (myo)fibro
blasts, terminates TGF signalling. NR4A1,
which is induced by TGF itself, recruits
a protein complex to the promoters of sev
eral TGF target genes, such as the collagen
type1 genes, and inhibits their expression.
However,persistent TGF exposure leads to

a histone-deacetylase-mediated reduction
in the expression of NR4A1 and an AKTmediated inactivation of NR4A1 through
phosphorylation in (myo)fibroblasts. High
levels of phosphorylated NR4A1 were found
in human and mouse fibrotic liver tissue
samples, which provides a mechanism for
the persistent high expression of profibrotic
genes observed in these tissues. In addition,
the NR4A1 agonist cytosporoneB attenuated
fibrogenesis in human dermal fibroblasts and
in mouse models of skin, pulmonary, renal
and liver fibrosis. This interesting study
provides novel insights into the derailment
of tissue repair processes and identifies
new opportunities for therapies. Although
complete inhibition of myofibroblast activ
ity might stop progression of the disease,
most patients present themselves only after
Macrophage
excessive deposition of scar tissue. Removal of

the already deposited matrix is then required
to restore organ function. In 2015, two studies
(neither centring on myofibroblasts) provided
fascinating data to explore that option.
A study by Kantari-Mimoun et al. 2
showed the importance of angiogenesis
in the resolution of fibrosis. This study is
intriguing because many studies have shown
the close association between angiogenesis
andthe progression of liver fibrosis, suggest
ing theuse of antiangiogenic compounds
forthe treatment of liver fibrosis. However,
by cell-selective deletion of the gene encoding
vascular endothelial growth factorA (Vegfa)
in myeloid cells the authors showed that
VegfA production, particularly by myeloid
cells, is essential for the resolution of liver
fibrosis in mice. Deletion of Vegfa in myeloid
cells did not affect liver fibrogenesis itself, but
prevented the resolution of liver fibrosis in
the recovery phase (after cessation of carbon
tetrachloride injections). Reintroduction of
wild-type myeloid cells, capable of produ
cing VegfA, restored the resolution process
whereas infusion of Vefga/ myeloid cells
could not. This study shows that scar-infil
trating myeloid cells stimulate angiogenesis
and matrix degrading activity in sinusoidal
endothelial cells. This process parallels
fibrogenesis and should not be inhibited in
antifibrotic therapies, but rather stimulated
to enable resolution of fibrosis.
An additional option for the resolution
of fibrosis was offered by modification of
macrophage profiles. In the past few years,
several studies have already highlighted the
dual role of macrophages in the process
of tissue remodelling 3. Macrophages can
miR-142-5p
miR-130a-3p
VEGF-A
Extracellular
matrix
Sinusoidal
endothelial cell
TGF
Hepatic
stellate cell
Hepatocyte
Probrotic
TGF
NR4A1
Antibrotic
Figure 1 | Schematic representation of the architecture

of the liver depicting intrahepatic
Nature Reviews | Gastroenterology & Hepatology
cells and the newly discovered pathways involved in the regulation of fibrogenesis.
Hepaticstellate cells, Kupffer cells, macrophages and endothelial cells all act in concert to generate
profibrotic and antifibrotic mediators. Effects of NR4A1, miR130a3p and VEGF in their designated
target cells counterbalance the profibrotic activities of other mediators. miR, microRNA; NR4A1,
nuclear receptor 4 A1; TGF, transforming growth factor ; VEGF, vascular endothelial growth factor.
37 | JANUARY 2016
polarize into a proinflammatory cell type
with matrix-degrading activities (M1 pheno
type) upon stimulation by IFN or bacterial
components such as lipopolys accharide.
Alternatively, macrophages can polarize into
an M2 phenotype, associated with tissue
repair and fibrosis, after stimulation by IL4
and/or IL13. Moreover IL10 can induce an
anti-inflammatory profile in macrophages
that produces profibrotic cytokines such as
TGF3. Although the factors that modulate
macrophage polarity are known, modu
lation of macrophage plasticity invivo to
induce regression of fibrosis has been diffi
cult to achieve. In 2015, a study by Su etal.4
characterized microRNA (miR)-1425p and
miR130a3p levels in lung and liver fibrotic
tissue in mice and humans. IL4 and IL13
caused an upregulation of miR1425p and
a downregulation of miR130a3p levels in
macrophages, which was associated with
the induction of profibrotic activities in
these cells. Subsequent administration of
miR1425p antisense oligonucleotides and
agents that mimic miR130a3p led to a
strong reduction in matrix deposition within
lungs of bleomycin-treated mice and livers
of carbon-tetrachloride-treated mice. This
study opens up a valuable option for the
resolution of liver fibrosis by modification of
macrophage polarization. We have not fully
gained the fruits of the discovery of small
interfering RNAs (siRNAs) and miRNAs yet,
mainly due to delivery problems. However,
Calvente et al. 5 delivered siRNA against
procollagen using nanoparticles and they
Key advances
The nuclear receptor nucelar receptor4A1
has a key role in terminating the profibrotic
feedback loop induced by transforming
growth factor in fibroblasts, offering new
targets for antifibrotic therapies1
Scar-infiltrating myeloid cells produce
vascular endothelial growth factor A, which
is essential for the resolution of fibrosis2
The clinical testing of experimental
antifibrotic compounds requires definition
of clear end-points in clinical trials, patient
stratification and, most of all, noninvasive
tests to monitor liver fibrosis at an
earlystage6
microRNAs 1425p and 130a3p are found
to regulate macrophage profibrogenic
activities which opens up options for the
resolution of liver fibrosis through
modification of macrophage polarization4
Microbiome analysis in stool or saliva might
provide innovative tools in the
prognostication of the cirrhotic process7
achieved resolution of liver fibrosis in mice

and this approach could be an important step
towards the therapeutic application of such
RNA molecules.
...the successes in treating

liver diseases have taught us
that liver fibrosis is reversible
Are we getting near an effective
pharmacotherapy for liver
fibrosis?
Are we getting near an effective pharmaco
therapy for liver fibrosis? Hundreds of experi
mental drugs directed against fibrogenic cells
have been tested in experimental models and
in clinical trials, but none of them has reached
the market yet. Out of the many drugs that
are directed against myofibroblasts, only
very few have reached phaseIII clinical trials
in patients with liver fibrosis. The fact that
only drugs already on the market for other
diseases are the ones that reach phaseIII in
patients with cirrhosis is no coincidence; the
safety profile of these drugs have already been
established, which is a decisive factor in clin
ical trials that are inevitably long in duration
for these patients. It might be that, similar to
anticancer and antiviral therapies, a combina
tion of drugs is required for this multifactorial
disease making trials even more com
plex. In 2015, a summary of the American
Association for the Study of Liver Diseases
emerging trends conference strategies and
endpoints of antifibrotic drug trials, organ
ized to address issues surrounding clinical
trial design was published in Hepatology6.
The main conclusions of the conference were
the need for better patient stratification to
enroll homogenous groups of patients with
a high risk of cirrhosis, standardization of
drug discovery and validation of biomarkers,
standardization of clinical trials including
patient subgroup definitions and use of end
points. Finally, and related to all of the above,
the development of reliable noninvasive tests
for fibrosis. This latter development would
shorten the very long clinical trials that are
needed to show effects of antifibrotic com
pounds, which represents a major hurdle for
all clinical trials in this field. Approximately
one-third of ongoing clinical trials in patients
with liver fibrosis are now dealing with bio
markers, either serum markers or imaging
techniques, illustrating the urgent need for
such tools. Bajaj and colleagues7 reported on
a surprising and innovative method of analy
sing the saliva microbiome, which seemed to
disclose prognostic information about the

liver. The topic of microbiome analysis in
stool and saliva samples elicited much debate8
and holds promising perspectives. The data
also fits with the growing evidence that the
gut microflora has an important role in the
progression of liver fibrosis to end-stage
liverfailure9.
New tools that reliably monitor intra
hepatic fibrogenesis would help to give this
disease the attention it deserves as it imposes
a heavy burden on society. Promising anti
fibrotic drugs have first been tested in
patients with idiopathic pulmonary fibro
sis, renal or skin fibrosis or cancer, yielding
pirfenidone and nintedanib as the only anti
fibrotic compounds on the market today10,
but not for the treatment of liver fibrosis. As
outlined here, the complexity of this chronic
disease, the long clinical trials that require
huge investments and the lack of noninvasive
methods for surveillance of disease activity
and stratification of patients, underlies the
lack of drugs against liver fibrosis to date.
Itis about time that clinical trials on the treat
ment of liver fibrosis also progress towards
their endgoal.
Klaas Poelstra is at the Department of
Pharmacokinetics, Toxicology & Targeting,
University of Groningen, Antonius Deusinglaan 1,
9713 AV Groningen, Netherlands.
k.poelstra@rug.nl
doi:10.1038/nrgastro.2015.224
1. Palumbo-Zerr,K. etal. Orphan nuclear receptor
NR4A1 regulates transforming growth factor
signaling and fibrosis. Nat. Med. 21, 150158 (2015).
2. Kantari-Mimoun,C. etal. Resolution of liver fibrosis
requires myeloid cell-driven sinusoidal angiogenesis.
Hepatology 61, 20422055 (2015).
3. Pellicoro,A. etal. Liver fibrosis and repair: immune
regulation of wound healing in a solid organ. Nat. Rev.
Immunol. 14, 181194 (2014).
4. Su,S. etal. miR1425p and miR130a3p are
regulated by IL4 and IL13 and control profibrogenic
macrophage program. Nat. Commmun. 6, 8523
(2015).
5. Calvente,C.J. etal. Specific hepatic delivery of
procollagen 1(I) small interfering RNA in lipid-like
nanoparticles resolves liver fibrosis. Hepatology 62,
12851297 (2015).
6. Torok,N.J. etal. Strategies and endpoints of
antifibrotic drug trials: summary and recommendations
from the AASLD Emerging Trends Conference, Chicago,
June 2014. Hepatology 62, 627634 (2015).
7. Bajaj,J.S. etal. Salivary microbiota reflects changes
ingut microbiota in cirrhosis with hepatic
encephalopathy. Hepatology 62, 12601271 (2015).
8. Bajaj,J.S., Betrapally,N.S. & Gillevet,P.M.
Decompensated cirrhosis and microbiome
interpretation. Nature 525, E1E3 (2015).
9. Szabo,G. Gutliver axis in alcoholic liver disease.
Gastroenterology 148, 3036 (2015).
10. Karimi-Shah,B.A. & Chowdhury,B.A. Forced vital
capacity in idiopathic pulmonary fibrosis FDA review
of pirfenidone and nintedanib. N. Engl. J.Med. 372,
11891191 (2015).
K.P. is co-founder and minority shareholder in BiOrion

Technologies. The antifibrotic compounds developed by
BiOrion Technologies are not the subject of this
YearinReview and are not mentioned in any other way in
thearticle.
JANUARY 2016 | 38

Key advances
G U T M I C R O B I O TA I N 2 0 1 5
Prevotella in the gut: choose carefully
Genomes of gut Prevotella species show

adaptation to their body habitat6,8 and high
strain variability between human hosts9
Ruth E.Ley
A barley kernel diet intervention

ameliorates glucose metabolism but only
inconjunction with an increase in levels
ofPrevotella1
Gut microbial communities often contain many Bacteroides or their close

relatives, Prevotella, but not both. Prevotella strains are associated with
plant-rich diets but are also linked with chronic inflammatory conditions.
In2015, papers probed the genomic diversity of Prevotella strains and
interactions of Prevotella copri with its host and other bacteria.
OTELLLAA
V
RE
outnumbered P.copri in the mouse gut, and

both reached lower levels than when they
were sole colonizers. These results suggest
that Bacteroides and Prevotella might be anti
correlated in population-level microbiome
surveys, in part, because they are antagonis
tic. The role of diet in tipping the outcome of
the competition in the human gut remains to
beascertained.
A few studies published in 2015 have
provided additional information about the
links between Prevotella and diet. In a sur
vey of gut microbiomes of 153 individuals,
De Filippis etal.7 noted a positive associ
ation of Prevotella with a Mediterranean
diet, specifically with high levels of carbo
hydrate, and fruit and vegetable intake.
However, the effects of a single diet ingredi
ent on Prevotella are not easily predictable.
Kovatcheva-Datchary etal.1 reported that
a barley-kernel-rich diet increased levels
of Prevotella of some healthy participants,
but not others even though all harboured
this genus 1. These different responses of
Prevotella to diet are probably dependent
on a number of factors: specific aspects of
the diet and the host; the species and/or
strains of Prevotella present; and other
microbemicrobe interactions.
TER OI
AC
IDD
ES
The wealth of data now available regard

ing human microbiota composition across
populations shows that, on the whole, the
most abundant bacteria in the human gut
belong to two phyla, the Firmicutes and
the Bacteroidetes. Of the members of the
Bacteroidetes, two genera dominate
Bacteroides and Prevotella. The majority
of Westerners harbour an abundance of
Bacteroides and, in an apparent trade-off
in numerical dominance within the micro
biota, a small minority carry Prevotella 1,2.
Prevotella is more common in non-
Westerners who consume a plant-rich diet,
which was most recently shown for Papua
New Guineans2 and has been linked to veg
etarianism in Western populations 3. The
association with a plant-rich diet has sug
gested that Prevotella is a beneficial microbe.
However, Prevotella in the gut has also been
linked with inflammatory conditions 4,5.
Several studies published in 2015 probe the
genetic diversity of this genus and start to
delve into the basis of its a ssociations with
health anddisease.
Prevotella and Bacteroides are large,
species-r ich monophyletic taxa that once
shared a common ancestor. Over the course
of their genetic differentiation they seem to
have divvied up the niches and habitats of
the human body 6. In Western populations,
Prevotella predominates in the oral cavity
and Bacteroides in the gut. When species
of both genera cooccur in the gut, one or
the other predominates, raising questions
about their niches and interactions.
In a direct test of Bacteroides versus
Prevotella interactions, KovatchevaDatchary etal.1 colonized germ-free
mice either with Prevotella
copri or Bacteroides thetaiotaomicron, or both. Under
these laboratory conditions
in which the host, environ
ment and the diet are held
constant, B.thetaiotaomicron
Laura Marshall/NPG
39 | JANUARY 2016
Addition of Prevotella to mice without a diet

change improves their glucose handling1
Prevotella is also linked to inflammation
inHIV5
Kovatcheva-Datchary etal.1 further asked

whether a Prevotella-dominated gut micro
biota could positively affect host metabolism.
Previous work from this group had shown
that a barley-kernel-rich diet improved glu
cose tolerance in healthy individuals with a
normal BMI, but not all study participants
responded to the diet. In this study, micro
biome analysis of responders and non
responders to the barley kernel diet revealed
that only the responders showed an increase
in abundance of Prevotella. To test whether
Prevotella amendment alone could mediate
the glucose response, P.copri DSM 18205
was gavaged to mice daily for a week. Those
gavaged with live P.copri showed improved
glucose tolerance. Thus, an increase in P.copri
alone, without a change in diet, can improve
glucose tolerance in a mouse model and
might do so also inhumans.
Culture collections now include ~40
different Prevotella species, most of them
oral isolates, and three of which are found in
the gut (P.copri is generally the more abun
dant). Ina comparison of the genomes of
39 Prevotella species, Accetto and Avgutin6
observed that P.copri DSM 18205 is deficient
in the ability to degrade host glycans and its
gene repertoire is more geared towards plant
glycan degradation, which fits with its niche
as a gut fermentor. Based on comparisons
between the genomes of 28 Prevotella spp.,
Gupta etal.8 suggested that individual mem
bers of Prevotella spp. adapt to niches within
the human body by modulating their gene
repertoires with extensive gene acquisition
and loss. Zhu and colleagues9 provide some
insight into just how much strain variation
one can expect to see within a single
Prevotellasp. (P.copri) across
hosts. Intheiranalysis, sequence
reads from 252metagenomes
were mapped to 929 reference
genomesincluding P.copri DSM
18205, enabling an assessment
of strain diversity for 11 s pecies.
Of these, P.copri had the highest number of
accessory genes within its genome (an acces
sory gene was defined as a species-specific
gene seen in some samples). The vastly differ
ent genome gene repertoires of strains within
and between Prevotella spp., and across hosts,
probably underlies some of the differences
observed in responses at the genus level to diet
and health conditions acrossindividuals.
Prevotella in the gut has previously been
positively associated with HIV5,10 and addition
of P.copri DSM 18025 to mice treated with
antibiotics has been shown to increase their
susceptibility to colitis induced by dextran
sulfate sodium4. In 2015, Dillon etal.5 showed
that P.copri levels were markedly elevated in
the mucosa and stool of patients infected with
HIV. Furthermore, exposure of key dendritic
cells to P.copri DSM 18205 boosted cytokine
production. The authors suggested that
increased levels of P.copri might contribute to
driving chronic inflammation in individuals
infected with HIV.
Notably, all three studies querying the
effects of P.copri on the host 1,4 or host cells5
used the same strain. P. copri CB7 (JCM
13464; DSM 18205), an isolate obtained from
the faeces of a healthy 52year-old Japanese
man, is currently the only cultured represent
ative of its species available from public cul
ture collections. Thanks to this constraint, it
is clear that a single strain of P.copri can act
in what has been interpreted as a beneficial or
detrimental manner, depending on the con
text. Adding strain variability to the system
compounds itscomplexity.
Prevotella is a large genus with high
species diversity; species can have high levels
of genomic diversity between strains. In the
human population, strain variation adds
another layer of uncertainty for predicting
how Prevotella will function in any given gut
ecosystem. To predict its function will require
a finer-grained understanding of these spe
cies genetic potential, their ecology, their
interactions with other microbes present and
with their host. A stronger grasp of strainlevel genome content on a per-individual
basis, obtained through metagenomic tech
niques, will inform attempts to modulate
levels of these bacteria therapeutically either
for improving metabolism or reducing risk of
inflammation in susceptiblehosts.
Ruth Ley is at Department of Molecular Biology and
Genetics, 467 Biotech, Cornell University, Ithaca,
NewYork 14853, USA and Department of Microbiome
Science, Max Planck Institute for Developmental Biology,
Spemannstrasse 35, Tbingen 72076, Germany.
rel222@cornell.edu
doi:10.1038/nrgastro.2016.4
Published online 1 Feb 2016
1.
2.
3.
4.
5.
6.
Kovatcheva-Datchary,P. etal. Dietary fiber-induced

improvement in glucose metabolism is associated with
increased abundance of Prevotella. Cell Metab. 22,
971982 (2015).
Martnez,I. etal. The gut microbiota of rural Papua
New Guineans: composition, diversity patterns, and
ecological processes. Cell Rep. 11, 527538 (2015).
Wu,G.D. etal. Linking long-term dietary patterns
with gut microbial enterotypes. Science 334,
105108 (2011).
Scher,J.U. etal. Expansion of intestinal Prevotella
copri correlates with enhanced susceptibility to
arthritis. eLIFE 2, e01202 (2013).
Dillon,S.M. etal. Gut dendritic cell activation links an
altered colonic microbiome to mucosal and systemic
Tcell activation in untreated HIV1 infection.
MucosalImmunol. 9, 2437 (2015).
Accetto,T. & Avgutin,G. Polysaccharide utilization
locus and CAZYme genome repertoires reveal diverse
ecological adaptation of Prevotella species. Syst. Appl.

Microbiol. 38, 453461 (2015).
7. De Filippis,F. etal. High-level adherence to a
Mediterranean diet beneficially impacts the gut
microbiota and associated metabolome. Gut http://
dx.doi.org/10.1136/gutjnl-2015-309957 (2015).
8.
Gupta,V. etal. Divergence in gene repertoire among
reference Prevotella genomes derived from distinct
body sites of human. BMC Genomics 16, 153 (2015).
9. Zhu,A., Sunagawa,S., Mende,D. & Bork,P. Interindividual differences in the gene content of human
gut bacterial species. Genome Biol. 16, 82 (2015).
10. Lozupone,C.A. etal. HIV-induced alteration in gut
microbiota: driving factors, consequences, and effects
of antiretroviral therapy. Gut Microbes 5, 562570
(2014).
HCV IN 2015
Advances in hepatitisC
research andtreatment
Barbara Rehermann
In 2015, new treatment regimens were revealed that achieve >95% cure rates
for all HCV genotypes. The HCV polymerase structure was solved in
catalytically relevant HCV replication steps and in the context of nucleotide
analogue inhibition. Moreover, HCV research taught us new links between
innate antiviral responses, lipid metabolism and intracellular
membraneformation.
HCV is an enveloped, positive-stranded RNA
virus representing the Hepacivirus genus in
the Flaviviridae family. More than 185million
people worldwide are chronically infected with
HCV and at risk of developing liver cirrhosis
and hepatocellular carcinoma1. HCV exists in
six major genotypes and >100subtypes and,
in each patient, creates an astounding num
ber of quasispecies. This sequence diversity is
facilitated by the high virus production rate
of more than 1012 virions per patient per day
and the lack of a proofreading ability of the
RNAdependent HCV polymerase (RdRp).
The year 2015 provided insights into struc
tural changes of RdRp during HCV genome
replication and into the structural basis for
the antiviral effect of the nucleotide analogue
inhibitor (the active form of sofosbuvir) that
has revolutionized HCV treatment. The RdRp
(encoded by NS5B) catalyzes the synthesis of
a negative-stranded RNA intermediate that
then serves as template for the synthesis of
multiple positive RNA strands. The latter
are translated into HCV proteins, used as
templates for further minus-strand RNA
synthesis or are packaged into virions that
are released from cells. In an elegant study,
Appleby and coauthors2 stalled the RdRp in
two catalytically relevant formations in ter

nary complexes with the RNA template, RNA
primer, incoming nucleotides and Mn2+. The
first formation reflects the primed initiation
assembly in which viral RNA has already
entered the active site of RdRp and a dinucleo
tide molecule (that later serves as a primer)
has been formed complementary to the HCV
RNA3end. Compared with the closed apo
state of RdRp, the loop and Cterminal
membrane-anchoring linker are retracted
from the active site in this configuration. The
subsequent primer-extension activity of the
RdRp opens its active site cavity further. From
there, the enzyme shifts into an open configur
ation in which the loop and Cterminus or
the RdRp are completely moved out of the
RNA binding groove, thus allowing elong
ation of the RNA strand. Cocrystal structures
of RdRp and 2F2Cmethyluridine mono
phosphate (the active form of sofosbuvir)
show recognition and successful incorpor
ation of this 2 modified nucleotide inhibitor
into the growing RNA strand. The incorpor
ation of the active form of sofosbuvir impairs
the generation of hydrogen bond networks,
prevents conformational changes of the RdRp
and promotes RNA chain disruption.
JANUARY 2016 | 40

Key advances
The structure of the RNA-dependent RNA
polymerase is solved in two catalytically
relevant formations during HCV replication,
revealing the structural basis for the antiviral
effect of sofosbuvir2
The all-oral, interferon-free combination
treatment of nucleotide analogue inhibitor
sofosbuvir and the NS5A inhibitor
velpatasvir achieves 99% sustained virologic
response rates in patients with chronic HCV
genotype 1, 2, 4, 5, or 6 infection3
Successful HCV treatment is associated with
a rapid normalization of intrahepatic
inflammation and innate immune
cellactivation7
The cytosolic lipid-binding protein
SEC14like protein2 is required for
replication of clinical isolates of HCV of
diverse genotypes in cell culture9
Increased 25hydroxysterol levels during
HCV infection induce microRNA185 to
regulate lipid metabolism, revealing a new
immunemetabolomic host response axis
with antiviral function10
a 100
50
25
200
Number of people (millions)
Number of people (millions)
b 4
75
150
100
50
id
W
or
ld
w
(1 IFN
98
9)
IF
N
+
(1 RB
99 V
pe
8)
gIF
N
pe
+
g( R
IF
N 200 BV
+ R 1)
BV
+
(2 DA
01 A
1)
IF
Nfre
e
(2 DA
01 A
4)
U
St nite
at d
es
Cure rate (%)
Also in 2015, unprecedented progress in

the treatment of HCV infection was seen. The
combination of sofosbuvir with the NS5A
inhibitor velpatasvir in a 12week once daily
all-oral regimen achieved a 99% sustained
virologic response (SVR) in patients with
HCV genotype 1, 2, 4, 5, or 6 infection, includ
ing those who were previously treated or had
compensated cirrhosis3,4. The SVR of patients
infected with the difficult-totreat HCV geno
type3 was only slightly lower (95% in previ
ously untreated patients, 91% in patients with
previous treatment failure and 89% in those
with compensated cirrhosis)4. The addition of
ribavirin to the regimen resulted in a 94% SVR
even in patients with decompensated cirrho
sis5. With FDA approval expected in 2016, this
treatment regimen will be the most successful
to date, basically eliminating the category of
difficult-totreat populations.
Given these treatment response rates what
remains to be done? Although cure rates have
progressively increased in treated patients, it
should not be forgotten that the total number of
treated patients still remains small1,6 (FIG.1). The
large number of undiagnosed HCV infections,
the high costs of the new treatment regimens
and the limited access of high-risk populations
to treatment make it unlikely that the global
HCV reservoir will be eliminated in the fore
seeable future. This reality has reemphasized
the need for a prophylacticvaccine1.
For those who are fortunate to have access
to treatment, further studies that assess the
Cured
Infected
Figure 1 | Patients cured of HCV infection. a | Graph shows cured patients as percentage of patients
Nature
Reviews
& Hepatology
treated with the indicated regimens. Year in brackets
indicates
when| Gastroenterology
the treatment was introduced.
|
b Absolute number of patients in the USA and worldwide who were cured by 2013 (REF. 6). About
600,000 patients worldwide were treated with sofosbuvir-containing regimens by the end of 2015.
Ofthe 185 million people worldwide infected with HCV, only a minority have been diagnosed. Green
bars represent patients infected with HCV and orange bars represent cured patients. Cure is defined
as achieving sustained virologic response. DAA, direct acting antivirals; RBV, ribavirin.
continued risk of cirrhosis and hepatocellu

lar carcinoma will be important. Although a
rapid normalization (to levels seen in healthy
individuals) of biomarkers of intrahepatic
inflammation and natural killer cell activation,
within 8weeks of successful therapy has been
reported in 2015 (REF. 7), the extent and speed
with which advanced liver fibrosis and cirrho
sis resolves is still unknown. In addition, basic
studies on the pathogenesis of HCV-induced
liver disease are still valuable because HCV
infection continues to teach us many new
aspects of intracellular innate immunity and
cell biology. This point is exemplified by the
following studies published this year.
HCV manipulates the intracellular environ
ment to either coopt or antagonize host fac
tors. Much of this manipulation is facilitated
by the formation of an HCV-induced mem
branous web that allows HCV replication in a
customized, favourable environment enriched
in cholesterol and lipids, and protected from
host ribonucleases and exonucleases, and to
some extent, from innate immune sensors. The
membranous web is induced by both viral and
host factors and several of the HCV proteins
are membrane-associated. However, HCV rep
lication induces oxidative stress with peroxi
dation of polyunsaturated fatty acids within
the membranes web8. This damage altersthe
configuration of membrane-anchored HCV
proteins and limits HCV replication8. Ina
gainoffunction approach Saeed et al. 9
have now identified SEC14like protein2
(also known as alpha-tocopherol-associated
41 | JANUARY 2016
protein) as a critical host factor for HCV repli

cation. SEC14like protein2 is a cytosolic lipid
binding protein that enhances the accumula
tion of vitaminE (tocopherol) as well as the
vitamin-Emediated inhibition of lipid peroxi
dation. Ubiquitously expressed in tissues,
SEC14like protein2 is absentin hepatoma cell
lines. A limited number of HCV tissue culture
models exist, and involve specific HCV strains
or HCV replicons with cell-culture adaptive
mutations. Overexpression of SEC14like pro
tein2 in hepatoma cells has now created an
invitro model to study natural patient-derived
HCV isolates with diversegenotypes9.
Also in 2015, Singravelu and colleagues10
identified a new immunemetabolomic host
response axis with antiviral function that
exploits the dependency of HCV on intra
cellular lipids. Antiviral type I interferon
responses induce the interferon-stimulated
gene, cholesterol-25hydoxylase, in macro
phages and dendritic cells resulting in
increased 25hydroxysterol levels in the liver
and blood of patients infected with HCV.
Although 25hydroxysterol is already known
to inhibit the fusion of several viruses to cell
membranes, Singaravelu et al.10 identified
the 25hydroxysterol-mediated induction
of microRNA (miR)-185 as an additional
antiviral effector mechanism. Interestingly,
miR185 does not target the HCV RNA
genome directly. Rather, it decreases mRNA
levels of host genes that promote lipid uptake,
synthesis of triglycerides and cholesterol, and
desaturation of fatty acids, thereby affecting
membranous web formation and decreasing
HCV replication. Using mice with chimeric
human livers as a model for HCV infection,
the authors showed that HCV tries to counter
act this effect by decreasing intracellular levels
of miR185 and miR130 (the latter is not
induced by 25hydroxysterol, but stimulates
the expression of miR185).
Collectively, these findings exemplify that
HCV continues to teach us new insights into
intracellular immune responses. Despite being
a master teacher of cell biology, HCV also
revealed its achilles heel because it can now
be eradicated in >95% of treated patients with
pan-genotype all-oral treatment regimens.
These unprecedented response rates extend
to patients with HCV genotype3 infections
and/or decompensated cirrhosis, which were
formerly termed difficult-totreat. Global
eradication efforts should now focus on redu
cing the number of undiagnosed persons and
advancing treatment regimens to allow shorter
treatment duration, reduce treatment costs and
make it more accessible. Finally, a focus on
public health measures and vaccine develop
ment to reduce HCV spread among popula
tions with high infection and reinfection risk
is needed.
Barbara Rehermann is at the Immunology Section,
Liver Diseases Branch, National Institute of
Diabetesand Digestive and Kidney Diseases,
NationalInstitutesof Health, DHHS, 10 Center Drive,
Bethesda,Maryland20892, USA.
rehermann@nih.gov
doi:10.1038/nrgastro.2015.227
Cox,A.L.Global control of hepatitis C virus. Science
349, 790791 (2015).
2.
Appleby,T.C. etal. Structural basis for RNA
replication by the hepatitis C virus polymerase.
Science 347, 771775 (2015).
3.
Feld,J.J. etal. Sofosbuvir and velpatasvir for HCV
genotype 1, 2, 4, 5, and 6 infection. N.Engl. J.Med.
4.
Foster,G.R. etal. Sofosbuvir and velpatasvir for HCV
genotype 2 and 3 infection. N.Engl. J.Med.
5.
Curry,M.P. etal. Sofosbuvir and velpatasvir for HCV
in patients with decompensated cirrhosis. N.Engl.
J.Med. http://dx.doi.org/10.1056/NEJMoa1512614.
6. Holmberg S. D. et al. Hepatitis C in the United States.
N. Engl. J. Med. 368, 18591861 (2013).
7.
Serti,E. etal. Successful interferon-free therapy of
chronic hepatitis C virus infection normalizes natural
killer cell function. Gastroenterology 149,
190200.e2 (2015).
Yamane,D. etal. Regulation of the hepatitis C virus
8.
RNA replicase by endogenous lipid peroxidation.
Nat.Med. 20, 927935 (2014).
9.
Saeed,M. etal. SEC14L2 enables pan-genotype HCV
replication in cell culture. Nature 524, 471475
(2015).
10. Singaravelu,R. etal. MicroRNAs regulate the
immunometabolic response to viral infection
in the liver. Nat. Chem. Biol. 11, 988993 (2015).
1.
Acknowledgements
This work was supported by the Intramural Research

Program of the NIH, The National Institute of Diabetes and
Digestive and Kidney Diseases (NIDDK).
G A S T R O I N T E S T I N A L I M AG I N G I N 2 0 1 5
Emerging trends in endoscopic imaging

Bishnu P. Joshi and Thomas D. Wang
Several key papers published in 2015 highlight important emerging trends
in endoscopic imaging that promise to improve patient diagnosis and
guidance of therapy. These studies reflect the future role for smart contrast
agents and fluorescence endoscopes to provide a molecular basis for
disease detection, identify precancerous lesions and determine optimal
choice oftherapy.
Molecular changes that occur in gastrointes
tinal diseases are now being assessed invivo
with new tools for endoscopic imaging (FIG.1).
Emerging technologies show great promise
for more precise decisions to be made in many
areas of patient care. Conventional wholebody imaging methods such as CT, MRI
and ultrasonography are well-established
for detecting disease by visualizing changes
in tissue architecture. Increased knowledge of
genomic and proteomic biomarkers have cre
ated new opportunities for imaging to beper
formed at a molecular level, which can be
combined with anatomical imaging. We review
several key manuscripts published in 2015 that
report the introduction of these technologies
intothe clinic. These methods could drama
tically increase the efficiency and effectiveness
of endoscopic imaging and improve our under
standing of the natural history of disease in the
digestivetract.
Burggraaf et al.1 demonstrated clinical
use of a peptide that binds to cMet that was
labelled with Cy5 (a near-infrared dye) for
early detection of colorectal cancer (CRC), a
step forward in endoscopy screening in 2015.
A transmembrane tyrosine kinase with proto-
oncogene function, cMet is overexpressed in
adenomas, known pre-cursors to CRC2. Safety
of this imaging agent was first established in
animals; probe biodistribution was studied
in rats and monkeys with single and repeated
doses. No adverse effects were seen in any of
these studies, and these safety results motivated
regulatory approval for human use with intra
venous administration. A clinical study was
performed in 15patients at high risk ofCRC.
Backtoback examinations consisted of white
light alone followed by white light with fluo
rescent peptide. Aclear increase in fluores
cence intensity was observed from adenomas
with either polypoid or flat morphology. Some
hyperplastic polyps and normal mucosa also
showed an increase in signal, but to a lesser
degree. The fluorescence positively correlated
with cMet levels in these polyps. A total of 101

lesions were detected with white light alone,
and 22additional lesions were detected on a
second look with peptide-based f luorescence.
Ofthese, 17 were visible with fluorescenceonly.
These additional lesions tended to be small
(<6mm) and have non-polypoid morphology
(Paris 0IIa and 0IIb)3. Alimitation of this
study is that a fibreoptic instrument was used
that has noticeably worse image resolution
than current video endoscopes. This technol
ogy could enhance conventional white-light
colonoscopy, which has been shown to miss
polyps that can result in unexplained interval
CRC. In previous studies, >28% of common
polyps have gone undetected on tandem exam
ination with white-light colonoscopy4. In the
future, this integrated methodology might be
clinically useful in high-risk patients, such as
those with Lynch syndrome and history of mul
tiple polyps, to increase the yield of adenoma
detection and to extend the interval between
colonoscopy examinations.
In 2015, we demonstrated clinical use
of a video molecular endoscope5 to localize
earlyneoplasia (high-grade dysplasia andearly
adenocarcinoma) in patients with Barrett
oesophagus using a peptide labelled with
Key advances
Fluorescently labelled probes that are
specific for cancer biomarkers can visualize
precancerous lesions invivo that are
otherwise difficult to see with conventional
white-light endoscopes1
Multimodal endoscopes that are sensitive to
fluorescence can quantify image intensities
and provide a map of high-risk mucosal
regions to guide tissue biopsy5
Real-time fluorescence images with
subcellular resolution can perform
instantaneous optical biopsy to distinguish
mucosal features to determine choice
oftherapy7
JANUARY 2016 | 42

a
SCC
OAC
Biliary
tract
Gastric
PDAC
Crohns
disease
Ulcerative
colitis
Squamous
Squamous
BE
BE
HGD
HGD
Figure 1 | Diseases throughoutNature

the digestive
Reviews
tract are beingGastroenterology
evaluated invivo&with
new
Hepatology
endoscopic imaging technologies that
identify molecular changes. a | Key areas
andconditions that are being examined by
emerging endoscopic technology. b | Whitelight endoscopic image shows squamous and
BEbut no distinguishing features for HGD from
human oesophagus invivo. c | After topical
administration of molecular probe, ratio image
collected with multimodal endoscope shows
region of increased intensity (arrow) from HGD5.
BE, Barrett oesophagus; HGD, high-grade
dysplasia; OAC, oesophageal adenocarcinoma;
PDAC, pancreatic adenocarcinoma; SCC,
squamous cell carcinoma.
fluorescein isothiocyanate (FITC)6. Early cancer

detection in patients with Barrett oesophagus
is currently performed with white-light endos
copy and random biopsies. However, early neo
plasia is often difficult to detect because of its
flat morphology andpatchy distribution. The
appearance and handling of this instrument,
including the endoscope and video processor,
is almost identical to that of a standard endo
scope. Registered fluorescence and reflectance
images were collected sequentially and then
used as a ratio to correct for differences in
distance to the mucosa over the fieldofview.
This feature enables the fluorescence intensi
ties over the entire image fieldofview to be
quantified so that red flag regions associated
with increased risk of cancer progression can be
displayed to guide tissue biopsy. This molecu
lar endoscope was used safely in 50patients
undergoing endoscopic mucosal resection
and, on targeted imaging, early neoplasia was
detected with 94% specificity and 96% positive
predictive value. Included in the analysis were
28 flat lesions (Paris 0IIb) that were poorly
visualized with white light. The peptide was

topically administered to the distal oesoph
agus to minimize the quantity (and hence
cost) of the imaging agent needed andto avoid
probe biodistribution to non-target tissues
for increased safety. Binding occurred within
5min, which resulted in minimal time added
to the procedure. No adverse events associated
with either the endoscope or peptide occurred.
Limitations of this study included the use of an
endoscope with a standard rather than high-
definition detector and collection of visible
rather than near-infrared fluorescence, result
ing in increased background autofluorescence.
By avoiding random biopsies, this molecu
lar endoscope can minimize procedure time
and reduce risks from sedation, bleeding and
perforation, and could be used in the future for
image-guided resection, surveillance, risk strat
ification and monitoring of therapy in patients
with Barrett oesophagus.
Finally, in 2015, clinical use of confocal
endomicroscopy with intravenous fluorescein
to collect fluorescence images from colonic
mucosa with subcellular resolution to distin
guish between Crohns disease and ulcerative
colitis was demonstrated7. A total of 79 patients
were studied, including 40 with Crohns disease
and 39 with ulcerative colitis. Unique image
features, including architectural distortion,
irregular surface, crypt density, discontinuous
crypt abnormality, focal cryptitis and discon
tinuous inflammation, were identified. The
imaging results were compared with histology
results and clinical history. Based on these
criteria, sensitivity, specificity and positive
predictive values of 90%, 97.4%and97.3%,
respectively, for Crohns disease and 97.4%,
90.0% and 90.5%, respectively, for ulcerative
colitis was determined. An overall accuracy in
diagnosis of 93.7% was found for both condi
tions. This distinction could be useful for deter
mining the best choice of therapy for patients
with IBD. Limitations of this study included no
difference in diffuse lamina propria cells and
mucin preservation and no visible granulo
mas. Confocal endomicroscopy is an emerging
imaging technique that can be can be used as
an accessory to standard medical endoscopes
to collect histology-like images to provide an
instantaneous optical biopsy. Currently, med
ical decisions are made based on pathology of
biopsied tissues that are not processed until
after the procedure is over and the patient has
left the clinic. Multiple tissue biopsies are often
taken at random sites, and key sites of disease
might be missed. Confocal imaging can also be
used to monitor patient response to therapy,
and was used to image a FITC-labelled antiTNF antibody to assess likelihood of patient
response to adalimumab in Crohns disease8.
43 | JANUARY 2016
Ahigh number of TNF-positive cells predicted

sustained clinical response to anti-TNF ther
apy for over 1year and was associated with
presence of mucosal healing on endoscopy.
The clinical studies published in 2015 and
highlighted in this YearinReview reflect the
enormous potential of novel molecular probes
and imaging instruments being developed for
targeted endoscopy to have future importance
in patient care. In addition to cMet, probes are
being developed for a broad range of molecu
lar targets that are overexpressed in disease.
Multiplexed methods that use flexible optical
fibres can visualize multiple targets simultane
ously to address genetic heterogeneity found in
disease9. Wide-field endoscopes are increasing
in image resolution, and confocal endomicro
scopes are achieving greater tissue imaging
depths compared with current instruments
being used in the clinic10. These technologies
are being integrated for invivo use, and can be
generalized to hollow organs either within or
outside of the digestive tract.
Bishnu P. Joshi and Thomas D. Wang are at the
Department of Internal Medicine, Division of
Gastroenterology, University of Michigan, 109 Zina
Pitcher Place, Ann Arbor, Michigan 481092200, USA.
Correspondence to T.D.W.
thomaswa@umich.edu
doi:10.1038/nrgastro.2015.214
Burggraaf,J. etal. Detection of colorectal polyps in
humans using an intravenously administered fluorescent
peptide targeted against cMet. Nat. Med. 21,
955961 (2015).
2. Di Renzo,M.F. etal. Overexpression and amplification
of the met/HGF receptor gene during theprogression of
colorectal cancer. Clin. Cancer Res. 1, 147154 (1995).
3. Participants in the Paris Workshop. The Paris
endoscopic classification of superficial neoplastic
lesions: esophagus, stomach, and colon: November 30
to December 1, 2002. Gastrointest. Endosc. 58,
S3S43 (2003).
4. Heresbach,D. etal. Miss rate for colorectal neoplastic
polyps: a prospective multicenter study of backtoback
video colonoscopies. Endoscopy 40, 284290 (2008).
5. Joshi,B.P. etal. Multimodal endoscope can quantify
wide-field fluorescence detection of Barretts neoplasia.
Endoscopy 48, 115 (2016).
6. Sturm,M.B. etal. Targeted imaging of esophageal
neoplasia with a fluorescently labeled peptide:
firstinhuman results. Sci. Transl. Med. 5, 184ra61
(2013).
7. Tontini,G. E. etal. Confocal laser endomicroscopy
forthe differential diagnosis of ulcerative colitis
andCrohns disease: a pilot study. Endoscopy 47,
437443 (2015).
8. Atreya,R. etal. Invivo imaging using fluorescent
antibodies to tumor necrosis factor predicts therapeutic
response in Crohns disease. Nat. Med. 20, 313318
(2014).
9. Joshi,B. P. etal. Multispectral endoscopic imaging of
colorectal dysplasia invivo. Gastroenterology 143,
14351437 (2012).
10. Qiu,Z. etal. Vertical cross-sectional imaging of colonic
dysplasia invivo with multi-spectral dual axes confocal
endomicroscopy. Gastroenterology 146, 615617
(2014).
1.
Acknowledgements
T.D.W. is funded in part by NIH U54 CA163059, U01

CA189291, R01 CA142750, R01 CA200007, and R01
EB020644 grants.
PA N C R E AT I C C A N C E R I N 2 0 1 5
Precision medicine in pancreatic

cancer fact or fiction?
Thomas Seufferlein and Julia Mayerle
Late diagnosis and an inability to personalize treatment are major problems
preventing reductions in pancreatic cancer mortality. In 2015, the
identification of a highly discriminatory exosomal biomarker, culture
systems that recapitulate human disease and new methods of analysing
large data sets to identify prognostic markers have improved the future
outlook for patients with this cancer.
Pancreatic cancer is projected to be the third
leading cause of cancer-related death by 2030.
Multiple factors contribute to the dismal prog
nosis for patients with pancreatic cancer, but
two clinical problems are a major concern: late
diagnosis and treatment resistance or lack of
personalized treatment stratification1. These
aspects have been addressed in research pub
lished in 2015, and considerable progress has
been made towards the use of precision medi
cine for the treatment of pancreatic cancer
(FIG.1).
An association between
cancer survival and 22 distinct
leukocyte subsets has been
revealed...
Pancreatic cancer has a very poor progno
sis, with a 5year survival rate of only 6% and
8085% of patients with pancreatic cancer diag
nosed at a stage when the tumour is unresect
able. Early detection is projected to increase
survival by 3040%1. Serum levels of carbo
hydrate antigen 199 (CA199) above 37U/ml
is the best established blood test for the detec
tion of pancreatic cancer. CA199 can discrim
inate between patients with pancreatic cancer
and healthy individuals with a sensitivityof
80.3% (95% CI 77.782.6) andaspecificity
of 80.2% (95% CI 78.082.3)2, and between
malignant and benign pancreatic disease
with a sensitivity of 78.2% (95% CI 72.380.4)
anda specificity of 82.8%3. However, to reduce
health-care expenditure and prolong patients
survival, an assay for early diagnosis would
have to perform with a minimum sensitivity of
88% at a specificity of 85%4. Aiming to iden
tify a marker with a higher diagnostic accuracy
than CA199, Melo and co-workers established
that the detection of the cell surface

proteoglycan glypican1 on circu
lating exosomes isolated from
patient plasma samples enables
the discrimination between
patients with early-stage and
late-stage pancreatic cancer,
benign pancreatic disease
and healthy individuals. Using a cut-off of 7.6%
glypican1positive exosomes, patients with
pancreatic ductal adenocarcinoma (PDAC)
could be distinguished from healthy individ
uals and those with benign pancreatic disease
with a previously unmet sensitivity and speci
ficity of 100%5. Identification and isolation of
cancer-specific exosomes in body fluids e nables
a diagnostic marker to be detected without
contamination by noncancer proteins, there
fore increasing diagnostic accuracy. Questions
to be answered in 2016 include whether these
findings can be validated independently in a
larger set of patients, and whether exosomal
glypican1 is just a marker of pancreatic
cancer or whether it has a function in exosome
generation or tumour growth and progression.
Isolating exosomes from patients serum
samples is a difficult task in clinical prac
tice; however, a new approach of performing
next-generation sequencing on easily obtain
able cell-free media, such as plasma, might be
more feasible for the detection of cancer-related
gene expression in the clinic. In a prospective
proofofconcept study, Zill and colleagues
analysed 54 genes concomitantly in tumour
tissue and cell-free DNA isolated from a 1ml
plasma sample6. The diagnostic accuracy of
cell-free DNA sequencing was 97%, with 92.3%
mean sensitivity and 100% specificity, over a
preselected panel of five genes. If replicated in
a larger study, this approach opens new aven
ues for a tailored and personalized treatment
strategy for patients with pancreaticcancer.
Tumour models
Monolayer
Organoids
cell-culture
Genetically
Xenografts
engineered
mouse models
2D co-culture
Targets of analysis
Exosomes
Cell-free DNA
Proteome
Metabolome
Transcriptome
Genome
Analysis tools
PRECOG
Virtual
microdissection
CIBERSORT
Next-generation
sequencing
Systems biology
Figure 1 |Nature
Complementary
experimental & Hepatology
Reviews | Gastroenterology
systems can elucidate diagnostic and
prognostic biomarkers for pancreatic cancer.
Pancreatic cancer research is multi-layered, with
many different approaches possible to
understand the disease. An integrated approach,
incorporating data from across multiple complex
experimental systems, each burdened with
limitations if used as a standalone system, will
speed up the identification of biomarkers for
pancreatic cancer. CIBERSORT, Cell type
Identification By Estimating Relative Subsets Of
known RNA Transcripts; PRECOG, PREdiction
ofClinical Outcomes from Genomic Profiles.
Genomic features of cancer and its micro

environment represent promising candidates
for predictive and prognostic biomarkers.
During the past decade, emerging highthroughput genomic technologies have gen
erated large amounts of data with apparently
inconsistent results, owing to tumour hetero
geneity and low individual patient prevalence
of distinct mutations. In 2015, Gentles and
co-workers have defined a prognostic land
scape of genes and infiltrating immune cells
across human cancer. The study authors
assembled, curated and integrated cancer
gene expression and clinical outcome data into
a new resource called PRECOG (PREdiction
of Clinical Outcomes from Genomic Profiles),
encompassing 30,000 expression profiles from
166 cancers. An association between cancer
survival and 22 distinct leukocyte subsets has
been revealed by applying a computational
JANUARY 2016 | 44

approach known as CIBERSORT. CIBERSORT
estimates the abundances of cell types from
gene expression data to detect and character
ize leukocyte contamination in bulk tumour
transcriptomes7. The prognostic potential of
leukocyte signatures has been demonstrated
by numerous studies, including the work
published this year by Gentles and colleagues.
However, concerns remain that heterogeneity
of the extent of stromal involvement and the
scarcity of tumour cells reduces the value of
prognostic biomarkers from whole tumour
tissue samples.
To overcome these limitations, Moffitt
etal.8 developed a computer-based algor
ithm to perform virtual microdissection on
microarray data from local and metastasized
pancreatic cancer tissue, cell lines and normal
pancreatic tissue. In line with previous find
ings, two tumour subtypes were identified,
validated and defined as classic and basallike, with the basal-like subtype associated
with poorer clinical outcomes. Furthermore,
they delineated two stromal types: nor
malstroma and activated stroma. The acti
vated stromatype, which was associated with
reduced median survival time, was character
ized by increased macrophage expression of
genes such as ITGAM, CCL13 and CCL18, and
members of the SPARC, WNTand MMP fam
ilies. Findings from this study are interesting
Key advances
Exosomal glypican1 is a highly specific and
sensitive biomarker of pancreatic cancer
diagnosis, and holds promise for early cancer
detection, at least in high-risk cohorts5
Next-generation sequencing of cell-free
DNA allows identification of tumour-derived
mutations without directly sampling
thetumour6
Pooling large data sets that include data on
tumour gene expression and clinical
outcomes aids the identification of
cancer-specific signatures that are
predictive of disease7
Virtual microdissection of RNA microarray
data is feasible and can classify tumour
andstroma subtypes8
Organoid models of human and mouse
pancreatic cancer recapitulate the key
features of the disease more accurately
thanmonolayer cell cultures or
patient-derived xenografts9
for several reasons. Firstly,traditional cell lines

often used in preclinical studies lack the classic
phenotypic subtype, which is far more com
mon in patients than the basal-like subtype.
Secondly, tumour subtypes can be recapitu
lated in patient-derived xenografts, but grafts
from patients with an activated stroma subtype
are more likely to survive and grow, meaning
results have to be interpreted accordingly.
Lastly, the heterogeneity between primary
tumour and metastasis is markedly lower
than expected. Overall, the study suggests that
an RNA-derived signature characterizes the
tumour better than somatic mutations, which
in the past have been advocated for guiding
personalized tumourtherapy.
A major breakthrough of the past year has
been the development of organoid cultures
from human and mouse ductal pancreatic
cancer 9. Orthotopic transplantation of pancre
atic tumour organoid cultures into syngeneic
mice recapitulates the full spectrum of disease
progression, forming early and advanced pan
creatic intraepithelial neoplasia (PanIN)-like
stages and progressing towards invasive and
metastatic PDAC. How PDAC organoids
form PanIN-like structures still remains to
be determined; the organoids might preserve
tumour neoplastic cell heterogeneity and
recover a variety of stem cell characteristics
that reflect the different stages of disease pro
gression. Alternatively, an attractive explan
ation is that transplanted PDAC organoids
retain the cellular plasticity and epigenetic
changes that result from organoid culture
conditions. Huang and co-workers made use
of the advantages of an organoid culture sys
tem for drug screening and tested the effect
of histone-lysine Nmethyltransferase EZH2
inhibition on organoids generated from either
pluripotent stem cells from KPC mice or from
patient-derived tumour cells bearing a TP53
mutation10. The systems developed by Boj
etal.9 and Huang etal.10 resemble tractable and
transplantable systems to study the genetic,
molecular and cellular properties of pancre
atic tumour development in mice and humans.
One hopes that these systems will speed up the
development of personalized approaches to
pancreatic cancertherapy.
In summary, although the predominant
question in pancreatic cancer research in
2014 was whether stroma was friend or foe,
the focus of research in 2015 has been the
identification and validation of diagnostic
45 | JANUARY 2016
biomarkers. If the findings reported in 2015

can be reproduced in the clinical setting, tests
based on glypican1 hold promise for the
early detection of pancreatic cancer, at least in
high-risk cohorts. Organoid models of PDAC
are an improvement on previous preclinical
models for the study of disease pathogenesis
and treatment response. In addition, pooling
large data sets enables the identification of
cancer-specific signatures that are predictive
of disease outcome, potentially paving the way
to precision medicine. It remains to be seen
whether these research efforts will be able to
alter the pessimistic projections for the burden
of pancreaticcancer.
Thomas Seufferlein is at Ulm University Medical
Center, Department of Internal Medicine I,
AlbertEinstein Allee 23, D-89081 Ulm, Germany.
Julia Mayerle is at Department of Medicine A,
University Medicine, Ernst-Moritz-Arndt-University,
Greifswald, Ferdinand-Sauerbruchstrasse,
17475Greifswald, Germany.
Correspondence to J.M.
mayerle@uni-greifswald.de
doi:10.1038/nrgastro.2015.215
Rahib,L. etal. Projecting cancer incidence and deaths
to 2030: the unexpected burden of thyroid, liver, and
pancreas cancers in the United States. Cancer Res.
74, 29132921 (2014).
2.
Gui,J.C. etal. CA199 and CA242 as tumor markers
for the diagnosis of pancreatic cancer: a meta-analysis.
Clin. Exp. Med. 14, 225233 (2014).
3.
Poruk,K.E. etal. The clinical utility of CA 199 in
pancreatic adenocarcinoma: diagnostic and prognostic
updates. Curr. Mol. Med. 13, 340351 (2013).
4.
Ghatnekar,O. etal. Modelling the benefits of early
diagnosis of pancreatic cancer using a biomarker
signature. Int. J.Cancer 133, 23922397 (2013).
5.
Melo,S.A. etal. Glypican1 identifies cancer
exosomes and detects early pancreatic cancer. Nature
523, 177182 (2015).
6.
Zill,O.A. etal. Cell-free DNA next-generation
sequencing in pancreatobiliary carcinomas.
CancerDiscov. 5, 10401048 (2015).
7.
Gentles,A.J. etal. The prognostic landscape of genes
and infiltrating immune cells across human cancers.
Nat. Med. 21, 938945 (2015).
8.
Moffitt,R.A. etal. Virtual microdissection identifies
distinct tumor- and stroma-specific subtypes of
pancreatic ductal adenocarcinoma. Nat. Genet. 47,
11681178 (2015).
9.
Boj,S.F. etal. Organoid models of human and mouse
ductal pancreatic cancer. Cell 160, 324338 (2015).
10. Huang,L. etal. Ductal pancreatic cancer modeling
and drug screening using human pluripotent stem
cellandpatient-derived tumor organoids. Nat. Med.
21, 13641371 (2015).
1.
Acknowledgements
The authors are supported by funding from the Deutsche

Krebshilfe / Dr. Mildred-Scheel-Stiftung (109102), the
Deutsche Forschungsgemeinschaft (DFG GRK1947A3, MA
4115/12/3) and the European Union (EUFP7: EPCTM and
EUFP7REGPOT20101).
The authors declare that there are no competing interests.
Connect to...
For the latest news from the journal and beyond: @NatRevGastroHep
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrgastro/info/newsfeeds.html
ahead of print http://www.nature.com/nrgastro/journal/vaop/ncurrent
http://www.nature.com/nrgastro/focus/index.html & http://www.nature.com/nrgastro/collection/index.html
Article series
A special series of articles that explores a specific theme and comprises Reviews and Perspectives that are published
consecutively over a period of time http://www.nature.com/nrgastro/series
Posters
Providing visual overviews of hot topics http://www.nature.com/nrgastro/posters
Conferences
You can meet the editors of Nature Reviews Gastroenterology & Hepatology at some of the meetings listed on our conferences
page http://www.nature.com/nrgastro/info/info_conf.html
NEPHROLOGY
PODOCYTE BIOLOGY IN 2015
New insights into the

mechanisms of podocyte health
Jeffrey H. Miner
Podocyte biologists can boast of some important advances in 2015. Some
ofthe key developments include defining the transcriptional targets of the
Wilms tumour protein on a genome-wide scale, the identification of new
mitochondria-centred pathways for maintaining podocyte homeostasis,
andnew insights into the regulation and pathogenic activation of TRPC6.
The breadth of outstanding podocyte
research published during the past year
makes the selection of just a few papers or
areas to highlight difficult. As in recent years,
much exciting progress was made in understanding the genetics of podocytop athies
and how podocyte cytoskeletal dynamics influence foot process architecture and
the glomerular filtration barrier. However,
I chose three other areas of research that
caught my attention for their wide-ranging
implications for podocyte biology and their
potential to open up new avenues for future
investigations.
WT1 is a zinc finger transcription factor that binds DNA and RNA and exists in
many different isoforms as a result of complex post-transcriptional regulation, inclu
ding alternative splicing. WT1 exhibits two
phases of expression in the kidney. In the
developing kidney it is expressed throughout the metanephric mesenchyme, and
in the adult it is expressed at high levels in
podocytes, the sole site of expression in the
mature kidney. Because the absence of WT1
causes renal agenesis, and various mutations
in WT1 cause podocytopathies, including DenysDrash syndrome, Frasier syndrome, and non-syndromic focal segmental
glomerulosclerosis (FSGS), the identification of the transcriptional targets of WT1 in
podocytes has been a subject of keen interest over the past two decades. This enthusiasm has led to numerous studies reporting
potential targets that are either activated or
repressed by WT1, but progress has for the
most part been piecemeal.
Two papers published in the past year1,2

used state-of-the-art chromatin immunoprecipitation from isolated podocytes or
glomeruli and next-generation sequencing
(ChIP-Seq) to identify genes that are directly
bound by WT1. The results from these studies are astounding: many of the genes that are
considered to be critical for podocyte and/or
glomerular function (for example, Nphs1,
Nphs2, Actn4, Cd2ap, Arhgap24, Myo1e,
Ptpro/Glepp1, Plce1, Magi2, Itga3, Lama5,
Lamb2, Col4a3, Col4a4, and Col4a5) were
found to be bound by WT1 in at least one of
these studies, suggesting that WT1 directly
regulates their transcription. Importantly,
some of the bound genes, including
Mafb/Kreisler, Tcf21/Pod1, Lmx1b, and Foxc1,
encode transcription factors that are involved
in the regulation of podocyte geneexpression. Thus, WT1 probably promotes
expression of a large subset of the podocyte
proteome by activating the transcription
of other transcriptional regulators that act
either cooperatively with or independently
of WT1. A third paper that used previously
obtained ChIP-Seq data, bioinformatics, and
gene expression profiling reported overall
consistentresults3.
These papers not only demonstrate the
importance of WT1 in the regulation of
known podocyte genes, but also identify a
large number of WT1bound targets that are
yet to be investigated but could be relevant to
podocytopathies and might, therefore, represent potential therapeutic targets. The fact
that WT1 can act as both an activator and
a repressor of transcription in the same cell1
46 | JANUARY 2016
provides additional complexity to the role

of WT1 in the podocyte that also deserves
furtherinvestigation.
The importance of mitochondrial health
for the podocyte has been demonstrated in
both humans and mice4,5. In addition, the
commonly used adriamycin-induced model
of FSGS that was originally thought to be specific to Balb/c and other adriamycin-sensitive
mouse strains is now understood to involve
mitochondrial DNA damage as the mech
anism of podocyte injury. Naturally susceptible strains bear a mutation in Prkdc,
which encodes a critical enzymatic component of a complex that helps to repair DNA
double-strand breaks6. Why the podocyte is
particularly prone to injury by adriamycin
isunknown, but reports of neo-susceptibility
to adriamycin-induced nephropathy associ
ated with gene Xyz mutation has often been
used as evidence of gene Xyz function in the
podocyte, even when no role in mitochondrial biology is known or suspected. It is
certainly possible that mitochondrial DNA
damage coupled with a non-mitochondrial
genetic hit can synergize to cause podocyte injury, but caution should be used in
interpreting such data.
Two 2015 papers 7,8 explored different
aspects of podocyte mitochondrial biology.
Mallipattu etal.7 studied mice lacking the zinc
finger transcription factor KLF6 specifically
in podocytes. Although these mice showed
a clinically insignificant increase in urinary albumin level and no overt glomerular
Key advances
The expression of numerous podocytegenes
thought to be essential forpodocyte or
glomerular function are transcriptionally
regulated by the Wilms tumour protein1,2
Mitochondrial health is important for
podocyte function and providing
protectionfrom apoptosis: KLF6 activation
of SCO2 expression protects podocytes
from apoptosis7
Some mitochondrial disorders of the
podocyte might be amenable to targeted
treatment approaches8
Regulation of calcium influx is required for
the long-term maintenance of podocyte
homeostasis: elevated TRPC6 activity
increases podocyte susceptibility to injury10
N E P H R O LO G Y
Mitochondrion
Podocyte
KLF6
SCO2
PHB2
TRPC6
Ca2+
NFAT
Calcineurin
Injury
Adriamycin
WT1
Actn4
Arhgap24
Cd2ap
Col4a3/4/5
Foxc1
Igta3
Lama5/b2
Lmx1b
Mafb
Magi2
Myo1e
Nphs1
Nphs2
Plce1
Ptpro
Tcf21
GBM
Endothelial cell
Figure 1 | Studies over the past year have provided key insights into
transcriptional
Nature
Reviews | Nephrology
programming, susceptibility to injury and mitochondrial health in the podocyte.
GBM,glomerular basement membrane.
phenotype, upon challenge with adriamycin

they became highly albuminuric and developed FSGS and tubulointerstitial lesions,
despite the fact that the mice were maintained
on a non-susceptible C57BL/6J background.
Importantly, the researchers were able to
show that mitochondria were dysmorphic
in the Klf6 mutant mice treated with adria
mycin but not in the adriamycin-treated
control mice, suggesting a role for KLF6 in
mitochondrial biology. Indeed, promoter
analysis revealed eight KLF6 binding sites
in Sco2, which encodes a protein required
for the proper assembly and functioning of
cytochromec oxidase (complex IV of the
mitochondrial respiratory chain). Binding of
KLF6 to SCO2 in human podocytes was confirmed experimentally by ChIP. Additional
studies showed that KLF6 activation of
SCO2 expression protected podocytes from
apoptosis and that KLF6 levels were reduced
in human HIV-associated nephropathy and
FSGS. These data support a role for KLF6 in
maintaining mitochondrial health and preventing podocyte apoptosis via regulation
ofSCO2.
Ising etal.8 studied mice in which podocytes lack prohibitin2 (PHB2), an important
mitochondrial membrane protein. In contrast
to mice lacking KLF6 in podocytes, deletion
of Phb2 in podocytes was sufficient to cause
foot process effacement, glomerulosclerosis,
albuminuria, and renal failure at 46weeks
of age. Ultrastructural analysis showed that
podocyte mitochondria had disorganized or
missing cristae, a phenotype that is consistent with the known role of prohibitin2 in the
membrane. Interestingly, podocyte-specific
mutations in the insulin and insulin-like

growth factor 1 receptors (Insr and Igf1r,
respectively) extended the lifespan of the
mice, as did treatment with rapamycin. These
studies suggest that some mitochondrial dis
orders of the podocyte could be amenable to
targeted treatment regimens.
Transient receptor potential cation channel, subfamily C, member 6 (TRPC6) is a
podocyte slit diaphragm-associated, primarily Ca2+ channel that is mutated in some
forms of autosomal dominant FSGS9. The
understanding that a change in cytosolic calcium ion concentration can have a dramatic
and rapid impact on cellular behaviour has
led to much interest within the podocyte
community in understanding how TRPC6
is transcriptionally regulated (apparently
not by WT1 (REFS1,2)), how it becomes activated, how the resulting calcium influx affects
downstream signalling pathways, and how
related molecules, such as TRPC5, interact
with and/or counteract TRPC6 (REF.9).
Based on the known ability of TRPC6 to be
activated by G protein-coupled receptor signalling, Wang etal. tested the hypothesis that
activation of the G protein Gq subunit can
cause downstream activation of TRPC6 and
podocyte injury10. The researchers expressed
a constitutively active Gq mutant in podocytes in a doxycycline-inducible fashion.
Gq expression caused no overt podocyte
phenotype on its own, but the mice demonstrated increased susceptibility to puromycin
aminonucleoside (PAN)-induced albumin
uria, which occasionally progressed to FSGS.
Induction of the mutant Gq also resulted in
increased expression of an NFAT-responsive
reporter transgene, as well as increased

expression of Trpc6 mRNA and protein in a
calcineurin-dependent manner. Importantly,
knockout of Trpc6 reduced the PAN-induced
albuminuria and prevented development of
FSGS in the mutant Gqexpressing mice,
demonstrating the pathogenic effect of elevated TRPC6 level and/or activity. These
results suggest that increased G protein
signalling in podocytes increases Trpc6
expression and susceptibility to injury via
calcineurin and NFAT and, together with
previous studies9, reveals multiple pathways
for potential therapeuticinterventions.
As demonstrated in part by the above
studies, our understanding of the podocyte
has come a long way from the days of the
twentieth century when it was thought of as
a mere bystander cell within the glomerular
capillary wall (Fig. 1). As researchers continue
to unlock the mysteries of the podocyte,
additional breakthroughs that can help to
inform the design of rational therapies will
indubitablyoccur.
Jeffrey H. Miner is at the Department of Internal
Medicine, Division of Nephrology, Washington
University School of Medicine, 4523 Clayton Avenue,
St Louis, Missouri 63110, USA.
Correspondence to minerj@wustl.edu
doi:10.1038/nrneph.2015.204
1. Dong,L. etal. Integration of cistromic and
transcriptomic analyses identifies Nphs2, Mafb, and
Magi2 as Wilms tumor 1 target genes in podocyte
differentiation and maintenance. J.Am. Soc. Nephrol.
26, 21182128 (2015).
2. Kann,M. etal. Genome-wide analysis of Wilms tumor
1controlled gene expression in podocytes reveals key
regulatory mechanisms. J.Am. Soc. Nephrol. 26,
20972104 (2015).
3. Lefebvre,J. etal. Alternatively spliced isoforms of
WT1 control podocyte-specific gene expression.
Kidney Int. 88, 321331 (2015).
4. Ashraf,S. etal. ADCK4 mutations promote
steroid-resistant nephrotic syndrome through CoQ10
biosynthesis disruption. J.Clin. Invest. 123,
51795189 (2013).
5. Casalena,G. etal. Mpv17 in mitochondria protects
podocytes against mitochondrial dysfunction and
apoptosis invivo and invitro. Am. J.Physiol. Renal
Physiol. 306, F1372F1380 (2014).
6. Papeta,N. etal. Prkdc participates in mitochondrial
genome maintenance and prevents adriamycininduced nephropathy in mice. J.Clin. Invest. 120,
40554064 (2010).
7. Mallipattu,S.K. etal. Kruppel-like factor 6 regulates
mitochondrial function in the kidney. J.Clin. Invest.
125, 13471361 (2015).
8. Ising,C. etal. Inhibition of insulin/IGF1 receptor
signaling protects from mitochondria-mediated kidney
failure. EMBO Mol. Med. 7, 275287 (2015).
9. Greka,A. & Mundel,P. Balancing calcium signals
through TRPC5 and TRPC6 in podocytes. J.Am. Soc.
Nephrol. 22, 19691980 (2011).
10. Wang,L. etal. Gq signaling causes glomerular injury
by activating TRPC6. J.Clin. Invest. 125, 19131926
(2015).
Acknowledgements
The author thanks Ali Gharavi, Columbia University Medical

Center, for helpful insights regarding the use of adriamycin
inmice.
J.H.M. declares no competing interests.
JANUARY 2016 | 47

N E P H R O LO G Y
I M M U N E R E G U L AT I O N O F K I D N E Y D I S E A S E I N 2 0 1 5
Updates on immunosuppression in
kidney disease
Hans-Joachim Anders
Numerous studies in 2015 focused on therapeutic immune modulation and
immunosuppression. Trials of budenoside in patients with IgA nephropathy
who are unresponsive to supportive therapy, and of low-dose IL2 to
enforce regulatory Tcell-mediated immunosuppression in autoimmune
disease all produced promising results.
The use of immunosuppressive drugs for
chronic primary IgA nephropathy (IgAN)
other than rapidly progressive, crescentic
IgAN, is a matter of debate among physicians. To date, many trials have been underpowered, single centre, and have lacked
a rigorous runin phase with optimized
conventional therapy, and thus have been
insufficient to guide clinical practice1. The
STOP IgAN trial addressed the use of conventional immunosuppressive drugs in
patients with persistent proteinuria >0.75g
per 24h after 6months of an escalating conventional supportive therapy protocol that
included angiotensin-converting-enzyme
(ACE) inhibitors, blood pressure control
<125/75mmHg, statins, dietary counselling
with regard to sodium and protein intake,
and smoking cessation2. A first, and important result was that 106 of 309 (34%) of the
patients who completed the runin phase
achieved a reduction in proteinuria <0.75g
per 24h. These patients were considered
to have responded to conventional therapy and were at low-risk of chronic kidney
disease (CKD) progression. Those who did
not respond to the 6months of aggressive
supportive care (162 patients) were randomized to either continue with supportive
treatment or to receive additional steroid or
cyclophosphamide treatment (depending on
CKD stage) for 3years. At 3years, a greater
number of patients had reached full clinical
remission in the immunosuppressive therapy group than in the matched supportive
care group, but there was no effect on annual
loss ofestimated glomerular filtration rate
(eGFR)of >15 ml/min/1.73m2 during the
3year duration of the study and a considerable number of patients experienced adverse
events such as diabetes mellitus, weight
gain, and infections. Immunosuppression
markedly reduced the disappearance of
microhaematuria. Altogether, these data

support the implementation of rigorous
supportive therapy over the broad use of
immunosuppressive drugs to treat patients
with chronic IgAN.
optimized supportive
therapy is the first choice for
most patients with IgAN
Considerable evidence has accrued to
implicate intestinal mucosal immunity in
the pathogenesis of IgAN 3, forming the
rationale to test a modified release formulation of budenoside (Nefecon; Pharmalink,
Stockholm, Sweden)-an orally administered glucocorticosteroid that can reach the
mucosa of the lower ileum and ascending
colon and suppress mucosal immunity. The
phaseIIb NEFIGAN trial randomly allocated
patients who exhibited persistent proteinuria
>0.75g per 24h after a 6month runin period
with angiotensin blockade using ACE inhibitors, angiotensin-receptor blockers,or both
to a 9month course of either 8mg or16mg
budenoside or placebo4. Amarked reduction
in proteinuria and maintenance of eGFR

after 3months followup was reported in
those administered 16mg budenoside compared to those on placebo. Budenosidewas
generally well-tolerated, consistent with low
systemic exposure. Suppression of mucosal
immunity by oral glucocorticoid treatment might, therefore, help restrain the
progression of CKD in patients with IgAN.
Overall, these data suggest that current
treatment algorithms for IgAN require
refinement. Optimized supportive therapy
is the first choice for most patients with
IgAN, and there is hope that topical steroids might improve the poor benefitrisk
balance of conventional immunosuppressive drugs in those who do not respond to
supportivetherapy.
Interleukin (IL)2 is an autocrine Tcell
growth factor that drives Tcell differentiation and proliferation via the dimeric
low-affinity IL2 receptor (IL2R). The
IL2R consists of three polypeptide chains
CD25(), CD122(), and CD132 ().
Injection of up to 120 million IU of recombinant human IL2 per day can promote effector Tcell-mediated immunity in patients
with renal cell carcinoma. Conversely, blockade of the IL2R chain with basiliximab or
daclizumab can suppress Tcell-mediated
renal allograft rejection. The IL2R , , and
chains can form a high affinity IL2R trimer that can respond to low concentrations
of IL2 (REF.5). Although naive CD4+ Tcells
express a low affinity dimeric IL2R variant
that consists of only the and chains, activated Tcells, and especially CD25+Foxp3+
T regulatory (T REG) cells, can respond to
lower doses of IL2 by selective expansion
and activation of cells that have acquired the
chain6. Hence, only 1.53.0 million IU per
day of recombinant IL2 is required to induce
T REG cell-associated intrinsic immuno
suppression, as demonstrated in patients
with hepatitis C virus-related vasculitis, graftversus-host disease, and other autoimmune
Key advances
Optimized supportive therapy for 6months can suppress proteinuria to <0.75g per 24h in
~34%patients with IgA nephropathy (IgAN)2
Steroids, cyclophosphamide, or azathioprine show little benefit in those with IgAN who do not
sufficiently respond to supportive therapy, when considering the effects of drug toxicity2
Budenoside elicits minimal drug toxicity but has effects on proteinuria and estimated glomerular
filtration rate beyond renin-angiotensinsystem blockade in IgAN, thus supporting a pathogenic
role of mucosal immunity4
Low-dose injection of IL2 can control progression of autoimmune diseases, probably by activating
regulatory Tcells to suppress innate and adaptiveimmunity6
Debate on the utility of mouse models in human inflammation research is ongoing, but a
reanalysis of comparative transcriptome data suggests strong similarities between these species9
48 | JANUARY 2016
N E P H R O LO G Y
Atherosclerosis
By boosting TREG cells, IL-2
could help control local
inammation to reduce
plaque formation
Mild to moderate
constitutional
symptoms such as
asthenia, myalgia, fever
and arthralgia
Systemic lupus
erythematosus
By blocking TFH cells and
stimulating TFR cell
dierentiation, IL-2 could
reduce autoantibody
formation and immune
complex deposition
Type 1 diabetes
By boosting TREGcells,
IL-2 could suppress
eector T cellmediated killing of
insulin-producing cells
Kidney diseases?
Local reaction
at injection site
Rheumatoid arthritis
By boosting TREG cells,
blocking TH17 cells and
favouring pTREG cells, IL-2
could help to control
inammation-dependent
joint destruction
from other sources. Indeed, mouse models

have a strong history in predicting disease
pathophysiology and drug efficacy in humans
but strong arguments exist on both sides of
the debate as to how mouse models should
be used in inflammation research9.
Plenty of data show that the immune
system can have various modulatory effects
on all forms of kidney disease10. Key studies presented in 2015 aimed to translate our
mole cular knowledge on immune regulation into therapeutic options for patients,
and the utility of mouse models of human
inflammation remains important to help
achieve this aim. Specific highlights from
this past year include the finding that
low-dose IL2 can foster intrinsic T REG
cell-mediated immunosuppression without
impairment of host defences, and that optimized supportive therapy for primary IgAN
is almost as effective as toxic conventional
immunosuppressive drugs.
Hans-Joachim Anders is at the Medizinische
KlinikundPoliklinik IV, Klinikum der Universitt
MnchenInnenstadt, Ziemssenstrasse 1,
80336Mnchen, Germany.
Correspondence to hjanders@med.uni-muenchen.de
Potential adverse eects

Potential benecial eects
Figure 1 | Low-dose IL2 therapy for autoimmune diseases. The capacity of low-dose IL2 to
Nature Reviews | Nephrology
activate and expand TREG cells via the high affinity IL2 receptor will not only suppress nonrenal forms
of autoimmunity but also renal manifestations of autoimmune diseases and intra-renal inflammation.
The forms of renal inflammation that might benefit from low-dose IL2 therapy requires exploration.
pT , peripherally induced regulatory T; T , T follicular helper cell; T , T follicular regulatory; T , T helper.
Image adapted from Nat. Rev. Immunol. 15, 283294 (2015), Macmillan Publishers Limited.
REG
FH
FR
doi:10.1038/nrneph.2015.202
Acknowledgements
This work was supported by the Deutsche Forschungs

gemeinschaft (grant numbers AN372/112, AN372/122,
AN372/143, AN372/151, AN372/161, AN372/171).
and inflammatory diseases5 (FIG.1). Reduced

availability of functional and metabolically
active TREG cells was reported in 61 patients
with systemic lupus erythematosus (SLE)
compared to 52 healthy controls, and a TREG
cell deficit was associated with IL2 deficiency and SLE activity6. This defect in TREG
cells could be reversed upon administration
of low-dose IL2 invitro. These findings were
confirmed invivo, whereby five patients with
refractory SLE were treated with 1.5 million
IU of human IL2 (aldesleukin) per day for
5 consecutive days. Selective proliferation
and reconstitution of CD4+Foxp3+CD127lo
TREG cells was observed in the blood, as well
as a strong induction of CD25 expression in
these cells6. The capacity of 1.53.0 million
IU aldesleukin for 5days to control refractory
SLE has also been reported in a patient with
persistent arthritis, myositis, and rash who
did not respond to azathioprine, mycophen
olate, rituximab, cyclophosphamide, bortezomib, or belimumab in combination with
methotrexate7. The efficacy of low-dose IL2
therapy in lupus nephritis and other forms of
immune-mediated kidney diseases remains
to be explored.
Finally, mouse models have long been

considered essential in modern biomedical
research to unravel the pathologic mech
anisms of immune-mediated diseases and to
validate potential novel anti-inflammatory
and immunosuppressive drugs. This paradigm, however, was challenged following the
publication in 2013 of a comparative transcriptome analysis between mice and humans
exhibiting trauma, skin burns, and endotoxaemia8. Very few correlations between the
mouse and human gene expression profiles
were identified and, therefore, the use of mice
for research in translational immunity was
put into doubt. Surprisingly, reevaluation of
this data set in 2015 using a different analytic
algorithm produced the opposite conclusion
with the finding that the gene expression patterns observed in humans during inflammation can be closely recapitulated in mice9. The
modification made to the analytic protocol
had a great effect on the correlation between
the mouse and human gene expression profiles and shows that biostatistical analyses of
large data sets are not trivial. Analyses of big
data should, therefore, be approached with
caution and the findings validated with data
HJ.A. has received consultancy fees from GlaxoSmithKline,

Roche, and Bayer.
Vecchio,M. etal. Immunosuppressive agents for
treating IgA nephropathy. Cochrane Database Syst.
Rev. 8, CD003965 (2015).
2. Rauen,T. etal. Corticosteroid monotherapy versus
combined immunosuppression in IgA nephropathy.
New Engl. J. Med. http://dx.doi.org/10.1056/
NEJMoa1415463 (2015).
3. Kiryluk,K. & Novak,J. The genetics and
immunobiology of IgA nephropathy. J.Clin. Invest.
124, 23252332 (2014).
4. Fellstrom,B. C. etal. The NEFIGAN Trial: NEFECON,
anovel targeted release formulation of budesonide,
reduces proteinuria and stabilizes eGFR in IgA
nephropathy patients at risk of ESRD. J. Am. Soc.
Nephrol. Abstr. 26, HIOR04 (2015).
5. Klatzmann,D. & Abbas,A.K. The promise of low-dose
interleukin2 therapy for autoimmune and
inflammatory diseases. Nat. Rev. Immunol. 15,
283294 (2015).
6. von Spee-Mayer,C. etal. Low-dose interleukin2
selectively corrects regulatory Tcell defects in
patientswith systemic lupus erythematosus.
Ann.Rheum. Dis. http://dx.doi.org/10.1136/
annrheumdis-2015-207776 (2015).
7. Humrich,J.Y. etal. Rapid induction of clinical
remission by low-dose interleukin2 in a patient with
refractory SLE. Ann. Rheum. Dis. 74, 791792
(2015).
8. Seok,J. etal. Genomic responses in mouse
modelspoorly mimic human inflammatory diseases.
Proc. Natl Acad. Sci. USA 110, 35073512 (2013).
9. Takao,K. & Miyakawa,T. Genomic responses in mouse
models greatly mimic human inflammatory diseases.
10. Kurts,C., Panzer,U., Anders,H.J. & Rees,A.J. The
immune system and kidney disease: basic concepts
and clinical implications. Nat. Rev. Immunol. 13,
738753 (2013).
1.
JANUARY 2016 | 49

N E P H R O LO G Y
STEM CELLS AND RENAL DEVELOPMENT IN 2015
Advances in generating and

maintaining nephron progenitors
Ryuichi Nishinakamura
2015 saw the publication of several important studies in the renal stem
celland developmental biology fields. Key studies provided insights into
theageing of nephron progenitors and optimal conditions to stimulate the
expansion of nephron progenitors, and reported the invitro generation of
kidney organoids.
The kidney develops through mutual inter
actions between two precursor tissuesthe
metanephric mesenchyme and the ureteric
bud. The metanephric mesenchyme contains
nephron progenitors that give rise to the epithelia of glomeruli and renal tubules whereas
the ureteric bud gives rise to the collecting
ducts and ureters. Nephron progenitors, which
typically express the markers Six2 and Cited1,
proliferate and are maintained in an immature
state at the tips of the branching ureteric bud;
some nephron progenitors, however, differentiate into nephron epithelia in response to
signals from the ureteric buds (Fig.1). This balanced process between progenitor propagation
and differentiation continues during development, and is a critical determinant of nephron
number, which might affect kidney function in
adult life. All remaining nephron progenitors
undergo differentiation shortly after birth in
mice and before birth in humans. As nephron
progenitors are established at embryonic
day 10 in mice, they are maintained only for
~10days invivo.
A new paper1 shows that nephron progenitors age during development. Single-cell
sequencing revealed that gene expression
profiles of mouse nephron progenitors differ
ES/iPS cells
depending on their developmental stage, and

that older progenitor populations (postnatal
day 0 in this case) contain a more hetero
geneous population of cells than do younger
populations (embryonic day 12.5). Agedependent increases in mTor activity and
ribosome biogenesis and a reduction in Fgf20
levels were also observed. When the researchers transplanted nephron progenitors into the
progenitor region (niche) located at the periphery of the developing kidney, they found that
the older progenitors tended to exit fromthe
niche and differentiate more quickly than
the younger progenitors. Interestingly, some
ofthe older progenitors remained in the niche
(that is, they behaved more like the younger
progenitors), when transplanted into a niche
comprising young progenitors. This alteration
in behaviour of older progenitors might in
part be explained by an abundant secretion of
Fgf20 by the young progenitors. These results
suggest that intrinsic age-dependent changes
affect cellcell interactions, which may cause
the cessation of nephrogenesis around the time
of birth. In other words, nephron progenitors
invivo are not self-renewing like age-resistant
stem cells, but rather should be considered to
be transient amplifying cells.
Nephron
progenitors
Progenitor
propagation
Collecting duct
Dierentiation
S-shaped
body
Eerent arteriole
Glomerular capillaries
Podocytes
Parietal epithelium
(Bowmans capsule)
Progenitor
ageing
Ureteric
bud
Renal
vesicle
These findings lead to the question

of whether nephron progenitors can be
induced to self-renew artificiallyan issue
that was tackled, and partially resolved, by
Brownetal.2. These researchers found that
inhibition of Bmp-dependent Smad activity
through use of a small molecule inhibitor
could maintain murine nephron progenitors.
Activation of several other pathways, including Wnt and Fgf-regulated pathways, led to
further expansion of the nephron progenitors for >20days invitro, even from neonatal
progenitors that readily differentiate invivo.
The invitro-expanded cells retained expression of progenitor-related markers, such as
Cited1 and Six2, and retained the potential to
differentiate into renal tubule epithelia with
features of proximodistalnephron segments.
These findings show that nephron progenitors
can be expanded inthe undifferentiated state
invitro. Of note, theobservation that expansion of nephron progenitors is stimulated
by Fgf9 is partly consistent with the above-
mentioned findings by Chen etal.,1 as Fgf9
and Fgf20 are known to cooperate to maintain progenitors in an undifferentiated state
invivo3. Although theresults from these two
studies suggest that the environment (niche)
can override the intrinsic clock of nephron
progenitors to extend their lifespan, the similarity of the invitro-expanded cells to endogenous nephron progenitors at particular stages
invivo is unclear. We dont yet know whether
these invitro-expanded cells are halted at an
immature stage or whether they continue to
age. The latter scenario is more likely than the
first, because Brown etal. found that invitro
expansion of the progenitors for >30days was
difficult and that glomeruli were not formed
from cells that had been expanded even for a
short time in culture. Single-cell sequencing
of expanded cells will be useful to address this
point by comparing their heterogeneity and
ageing status with those invivo. Brownetal.2
also reported the short-term expansion of
Loop of
Henle
Figure 1 | Strategies towards the generation, propagation and differentiation of nephron progenitors. Nephron progenitors have been shown
Naturestem
Reviews
Nephrology
to age invivo, and interact with the ureteric bud to form the nephron. Nephron progenitors can now be induced from pluripotent
cells,| and
propagated invitro with at least partial differentiation potential. ES, embryonic stem; iPS, induced pluripotent stem.
50 | JANUARY 2016
N E P H R O LO G Y
Key advances
Nephron progenitors age during
development invivo, but old progenitors
can be rejuvenated when surrounded by
younger progenitors1
Nephron progenitors can be expanded
invitro beyond their physiological lifespan2.
Kidney organoids containing multiple
lineages can be induced from human
induced pluripotent stem cells invitro6
nephron progenitors derived from human

embryonic stem (ES) cells. However, the purity
and competence of the starting material may
not have been sufficient to induce the formation of glomeruli, as discussed in further
detailbelow.
In 2014, two groups reported the induction of human nephron progenitors from
pluripotent stem cells. We identified that the
precursors of nephron progenitors in mice are
maintained in an immature state and transition
to the posterior intermediate mesoderm 1 day
later than precursors of the ureteric bud transition to the anterior intermediate mesoderm4.
Using these findings, we successfully generated three-dimensional glomeruli and renal
tubules invitro from both mouse ES cells and
human induced pluripotent stem (iPS) cells.
Meanwhile, Takasato etal.5 reported a protocol to induce both renal tubules and a ureteric
bud-like lineage from human ES cells, showing
that the kidney structure can be self-organized
invitro to some extent. However, the nephron
progenitors induced using this protocol rarely
formed glomeruli. Of note, Brown etal.2 used
this same protocol to expand their nephron
progenitors, which may in part explain why
their expanded cells did not form glomeruli.
In 2015, Takasato etal.6 used a revised protocol to show that human iPS cells exposed to a
longer period of Wnt developed the posterior
nephron progenitor lineage, whereas shorter
Wnt exposure stimulated development of the
anterior ureteric bud lineage, consistent with
our previous findings4. Using this optimized
protocol, kidney organoids, containing glom
eruli, renal tubules, collecting ducts, vascular
endothelial cells, and interstitial cells, were
generated from human iPS cells invitro6. The
induced proximal tubules had resorbing capabilities and were sensitive to the nephrotoxic
anticancer drug, cisplatin. Thus, these kidney
organoids could be useful for the screening
of drugs for toxicity, modelling genetic kidney diseases, and providing a cell source for
research and future therapy.
Although the above-described studies represent notable advances, the precise competence of each component induced using these
invitro approaches needs to be determined

and improved. For example, the ureteric bud
should branch extensively, support propagation (not self-renewal) and differentiation of
nephron progenitors at each branched tip, and
form a single urinary exit tract at the opposite
end. In addition, the glomeruli should be vascularized so that urine can be produced and
flow through the kidney. For human nephron
progenitor expansion, which is critical for
securing sufficient numbers of cells for future
regenerative medicine approaches, a method
to purify the progenitors needs to be established, and a more complete expansion protocol needs to be developed. Although much
remains to be achieved, the renal stem cell and
developmental biology fields are progressing
rapidly. The combined approaches of working
towards a precise understanding of the mechanisms of kidney development invivo and
development of improved protocols for producing kidney tissues invitro will accelerate
research toward the recreation of a genuine,
functioning kidney in the future.
Ryuichi Nishinakamura is at the Department of

KidneyDevelopment, Institute of Molecular
Embryology &Genetics, Kumamoto University, 221
Honjo, Chuoku, Kumamoto 8600811, Japan.
Correspondence to ryuichi@kumamotou.ac.jp
doi:10.1038/nrneph.2015.203
1. Chen,S. etal. Intrinsic age-dependent changes and
cellcell contacts regulate nephron progenitor lifespan.
Dev. Cell 35, 4962 (2015).
2. Brown,A.C., Muthukrishnan,S.D. & Oxburgh,L. A
synthetic niche for nephron progenitor cells. Dev. Cell
34, 229241 (2015).
3. Barak,H., Huh,S., Chen,S. & Jeanpierre,C.
FGF9 and FGF20 maintain the stemness of nephron
progenitors in mice and man. Dev. Cell 22,
11911207 (2012).
4. Taguchi,A. etal. Redefining the invivo origin of
metanephric nephron progenitors enables generation
of complex kidney structures from pluripotent stem
cells. Cell Stem Cell 14, 5367 (2014).
5. Takasato,M. etal. Directing human embryonic stem
cell differentiation towards a renal lineage generates
aself-organizing kidney. Nat. Cell Biol. 16, 118126
(2014).
6. Takasato,M. etal. Kidney organoids from human
iPScells contain multiple lineages and model
humannephrogenesis. Nature 526, 564568
(2015).
RENAL FIBROSIS IN 2015
Understanding the
mechanisms of kidney fibrosis
Dong Zhou and Youhua Liu
The year 2015 has seen great progress in the renal fibrosis field, as key
studies began to build a consensus on the importance of epithelial-to
mesenchymal transition, cell cycle arrest, and defective metabolism in the
pathogenesis of kidney fibrosis. New findings also point to a role of
developmental signalling in renal fibrogenesis.
Renal fibrosis is an important topic that
attracts broad interest in nephrology owing to
its status as a hallmark and common outcome
across all kinds of progressive chronic kidney
disease (CKD). The year 2015 saw much progress in the renal fibrosis field, with major
breakthroughs and new findings markedly
advancing our understanding of the fibrogenic process. These studies have laid strong
foundations for the future development of
novel treatments for fibrotic CKD. For the
first time in more than a decade, scientists
in the field have begun to build a consensus
on several key issues such as the importance
of partial epithelial-tomesenchymal transition (EMT), cell cycle arrest, and defective
cellular metabolism in the development and
progression of kidneyfibrosis.
The process of renal fibrosis is characterized by an excessive deposition of extracellular

matrix in the interstitial compartment, leading to scar formation. An activated form
of interstitial fibroblast the smooth
muscle actin-positive myofibroblast is
widely recognized as the major type of
matrix-producing cell in the fibrotic kidney.
However, tubular epithelial cells, which are
the main constituent of renal parenchyma,
often localize at the epicentre of damage and
are especially vulnerable to damage after kidney injury. In this context, a key question is
how tubular injury drives fibroblast activation
and matrix overproduction. One hypothesis is
that kidney tubular cells undergo EMT after
injury, a phenotypic conversion programme
that is characterized by the loss of epithelial
JANUARY 2016 | 51

N E P H R O LO G Y
markers and gain of mesenchymal features.
Such a notion, however, has been intensely
contested as studies using genetic cell lineage
tracing could not find evidence of a direct
contribution of epithelial cells to the myo
fibroblast population in the fibrotic kidney1,
instigating a controversy over the relative contribution of EMT to fibroblast activation that
has lasted several years.
In 2015, two backtoback studies
addressed this dispute and offered new
insights into the potential role of tubular
EMT in the development and progression
of renal fibrosis2,3. Thesestudies tackled the
issue by generating genetically modified
mice, in which Snail or Twist, two key transcription factors that regulate the EMT programme, were ablated specifically in tubules.
As a result, the EMT programme is specifically inhibited in the renal tubular epithelium
invivo. Both studies demonstrated that inhibition of the EMT programme by conditional
deletion of Snai1 or Twist1 in tubular epithelial cells reduced interstitial fibrosis in numerous CKD models, including unilateral ureteral
obstruction, nephrotoxic serum-induced
nephritis, and folic acid-induced nephro
pathy. Not surprisingly, inhibition of an EMT
programme in the kidney also led to preservation of tubular cell integrity and function,
restoration of tubular repair and regeneration,
and a reduction in myofibroblast accumu
lation, suggesting that the EMT programme
is crucial and required for initiating tubular
dysfunction and driving fibrosis development
after various insults.
The mechanism of EMT involvement in
renal fibrosis revealed by these studies is particularly intriguing. Both studies found that
tubular epithelial cells only undergo a partial
EMT during renal fibrosisthe cells express
markers of both epithelial and mesenchymal
cells and remain associated with their basement membrane. In this respect, these observations are in harmony with earlier genetic cell
linage tracing studies1, and demonstrate that
a complete phenotypic conversion of tubular
epithelial cells to a myofibroblast phenotype is
extremely rare, if occurring at all. Nevertheless,
this partial EMT is sufficient to induce tubular function impairment, triggering cell cycle
arrest and promoting the release of critical
fibrogenic cytokines. Lovisaetal. further
demonstrated that one of the functional consequences of partial EMT is the induction of
arrest in the G2 phase of the cell cycle, which
compromises the potential of tubular epithelial cells to repair and regenerate3. As cell cycle
arrest has been postulated as a mechanistic
pathway that leads to kidney fibrosis, the linkage of EMT to cell cycle arrest is especially
PGC-1
CPT-1
PPAR
Secretory
phenotype
Shh
TGF-
Wnts
+ RAS
Snail
or Twist
Partial EMT
Cell cycle arrest
FAO
Dysregulated
FA metabolism
ATP depletion
Dedierentiation
Lipid deposition
TEC
Basement membrane
Fibroblast
+ Proliferation
+ Activation
+ ECM production
ECM
Figure 1 | Studies have highlighted important roles for varying tubular

such
Natureresponses
Reviews | Nephrology
aspartial epithelial-tomesenchymal transition (EMT), cell cycle arrest and defective
metabolism in driving renal fibrosis. ECM, extracellular matrix; FA, fatty acid; FAO, fatty acid
oxidation; TEC, tubular epithelial cell. Adapted from Nature Publishing Group Ovadya, Y.
&Krizhanovsky, V. Nat. Med. 21, 975977 (2015).
appealing, as it helps to form a consensus

on our understanding of the mechanism of
renalfibrosis.
Damage to the tubular epithelium might
induce renal fibrosis via other mechanisms
as well. In 2015, a landmark study used a
genome-wide transcriptome approach to
demonstrate that defects in fatty acid metabolism in tubular epithelial cells have a crucial
role in the pathogenesis of kidney fibrosis4.
This metabolic reprogramming is characterized by the decreased expression of key
enzymes and regulators of fatty acid oxidation
(FAO) and increased intracellular lipid deposition4. As FAO is the preferred energy source
for kidney proximal tubular epithelial cells, a
reduction in FAO would affect lipid metabolism by disrupting the balance between fatty
acid synthesis, uptake, and consumption, leading to dysregulated intracellular lipid accumulation. Inhibition of FAO in tubular epithelial
cells invitro indeed causes ATP depletion,
cell death, dedifferentiation, and intracellular
lipid deposition4. Conversely, restoring fatty
acid metabolism by genetic or pharmacologic
approaches protects against renal fibrosis, suggesting that stimulation of metabolic pathways
could be a novel strategy for preventing and
treating fibrotic CKD.
Kang etal. also investigated the mech
anisms behind the depressed metabolic pathways in fibrotic kidney disease4. Transforming
growth factor1 (TGF1), the most potent
profibrotic cytokine, inhibits the expression
of carnitine palmitoyltransferase 1 (CPT1),
52 | JANUARY 2016
the rate-limiting enzyme in FAO, and thereby

decreases fatty acid metabolism. Furthermore,
TGF1 also represses mRNA expression of
upstream regulators of CPT1 that encode the
peroxisome proliferator-activated receptor
(PPAR) and PPAR coactivator1(PGC1).
A separate study in 2015 also showed that inhibition of microRNA21 (miR21) enhances
mitochondrial function, reduces production
of reactive oxygen species, preserves tubular integrity, and attenuates renal fibrosis5.
Therefore, miR21, a downstream target of
TGF1 signalling, contributes to kidney
fibrosis by silencing metabolicpathways.
Key advances
Renal tubular epithelial cells undergo a
partial epithelial-tomesenchymal transition
after injury, which impairs tubular repairand
regeneration, induces cell cycle arrest,
anddrives interstitial fibroblast activation2,3
Kidney fibrosis is associated with defective
cellular metabolism and mitochondrial
dysfunction in renal tubular epithelial cells;
approaches to restore fatty acid metabolism
or stimulate metabolic pathways might
represent new therapeutic strategies to
combat fibrotic kidney disease4,5
Activation of key developmental pathways
such as Wnt and hedgehog signalling after
injury has a critical role in triggering
fibroblast proliferation, as well as in
activating the intrarenal reninangiotensin
system810
N E P H R O LO G Y
One common outcome of partial EMT, cell
cycle arrest, and depressed metabolism following kidney injury is the conversion of tubular
cells to a pathologic secretory phenotype. Our
understanding of the secretome by injured
tubular epithelial cells continued to advance in
2015. Emerging evidence suggests that injured
tubular cells produce and secrete the ligands of
key developmental signalling pathways, such
as Wnts and sonic hedgehog (Shh)6. Tubulederived Shh mediates epithelialmesenchymal
communication by selectively targeting interstitial fibroblasts in a paracrine manner, and
induces fibroblast proliferation and myo
fibroblastic activation, leading to kidney fibro
genesis6. An interesting study in 2015 showed
that hedgehog-responding Gli1+ cells possess
features of mesenchymal stem cells invitro,
and proliferate and expand following kidney
injury7. Genetic ablation of these cells substantially ameliorates kidney fibrosis. Consistent
with these findings, pharmacologic inhibition
of Shh/Gli signalling reduces the size of the
myofibroblast population and inhibits fibrosis
after injury6,8.
Wnt ligands are induced in many cell types
of the injured kidney and can target both interstitial fibroblasts and tubular epithelial cells via
autocrine or paracrine mechanisms. Similar to
Shh, Wnts induce fibroblast proliferation and
myofibroblastic activation, leading to matrix
overproduction and the development of fibrosis9. In 2015, we showed that multiple genes of
the reninangiotensin system are direct targets
of Wnt/catenin signalling in tubular epithelial cells10. These studies provide a novel mechanistic link between Wnt upregulation and
activation of the intrarenal reninangiotensin
system, hypertension, and kidney fibrosis.
In summary, important studies published
in 2015 highlight a central role for varying
tubular responses such as partial EMT, cell
cycle arrest, and defective metabolism in driving renal fibrosis (FIG.1). Intriguingly, these
tubular responses eventually converge on the
acquisition of a secretory phenotype in tubular epithelial cells, leading to the release of
pathological mediators that sustain fibroblast
activation and inflammation. In our opinion,
advances in understanding the pathogenesis
of renal fibrosis in 2015 will be memorable,
and should inspire more intensive studies for
many years to come.
Dong Zhou and Youhua Liu are at the Department
ofPathology, University of Pittsburgh School of
Medicine, 200 Lothrop Street, Pittsburgh,
Pennsylvania 15261, USA.
Youhua Liu is also at the State Key Laboratory of
Organ Failure Research, Nanfang Hospital, Southern
Medical University, 1838 North Guangzhou Avenue,
Guangzhou 510515, China.
Correspondence to Y.L.
liuy@upmc.edu
doi:10.1038/nrneph.2015.215
1.
2.
3.
4.
5.
Humphreys,B.D. etal. Fate tracing reveals the pericyte

and not epithelial origin of myofibroblasts in kidney
fibrosis. Am. J.Pathol. 176, 8597 (2010).
Grande,M.T. etal. Snail1-induced partial epithelialtomesenchymal transition drives renal fibrosis in mice
and can be targeted to reverse established disease.
Nat. Med. 21, 989997 (2015).
Lovisa,S. etal. Epithelial-tomesenchymal transition
induces cell cycle arrest and parenchymal damage in
renal fibrosis. Nat. Med. 21, 9981009 (2015).
Kang,H.M. etal. Defective fatty acid oxidation in renal
tubular epithelial cells has a key role in kidney fibrosis
development. Nat. Med. 21, 3746 (2015).
Gomez,I.G. etal. Anti-microRNA21 oligonucleotides
prevent Alport nephropathy progression by stimulating
metabolic pathways. J.Clin. Invest. 125, 141156
(2015).
Zhou,D. etal. Sonic hedgehog is a novel tubule-derived

growth factor for interstitial fibroblasts after kidney
injury. J.Am. Soc. Nephrol. 25, 21872200 (2014).
7.
Kramann,R. etal. Perivascular Gli1+ progenitors
arekey contributors to injury-induced organ fibrosis.
Cell Stem Cell 16, 5166 (2015).
Kramann,R. etal. Pharmacological GLI2 inhibition
8.
prevents myofibroblast cell-cycle progression and reduces
kidney fibrosis. J.Clin. Invest. 125, 29352951 (2015).
Xiao,L. etal. Sustained activation of Wnt/-catenin
9.
signaling drives AKI to CKD progression. J.Am. Soc.
Nephrol. 27, http://dx.doi.org/10.1681/
ASN.2015040449 (2015).
10. Zhou,L. etal. Multiple genes of the reninangiotensin
system are novel targets of Wnt/-catenin signaling.
J.Am. Soc. Nephrol. 26, 107120 (2015).
6.
Acknowledgements
The authors work is supported by NIH grants DK064005,

DK091239 and DK106049, and National Science of
Foundation of China grants 81130011 and 81521003.
H Y P E RT E N S I O N I N 2 0 1 5
Resistant hypertension: impact

and evolving treatment options
Lilach O. Lerman and Stephen C. Textor
Combination therapy with optimal doses of multiple antihypertensive drugs
fails to achieve blood pressure (BP) control in up to 15% of hypertensive
patients. Key studies in 2015 highlighted the risks of uncontrolled
hypertension and evaluated new therapeutic modalities designed to achieve
satisfactory BP control in patients with treatment-resistant hypertension.
Arterial hypertension substantially increases
the risks of adverse cardiovascular and renal
outcomes, including end-stage renal disease
(ESRD), ischaemic cardiac and cerebro
vascular events, accelerated atherosclerosis,
congestive heart failure and all-cause mortality. Although most of these risks can be
reduced by interventions that lower blood
pressure (BP), some hypertensive individuals
are treatment resistant, often owing to poor
adherence or intolerance to antihypertensive
therapies. The finding that combination regi
mens with multiple antihypertensive drugs at
optimal doses fail to achieve goal BP levels in
up to 15% of hypertensive patients1 has led
to a relentless search for novel BPlowering
therapies. Four key studies published in
2015 provided new insights into the risks of
uncontrolled hypertension and the efficacies of novel approaches designed to address
the unmet needs of patients with resistant
hypertension(RH).
The risks of uncontrolled and difficulttocontrol hypertension were reinforced by
the findings of a 5year, retrospective, longitudinal cohort study of 470,386 hypertensive
patients enrolled in the Kaiser Permanente

health-care system2. In this ethnically-diverse
population, 12.8% (60,327) of individuals
were identified as having RH, of whom 4.9%
had controlled RH on four or more anti
hypertensive drugs and 7.9% had uncontrolled RH, based on a goal systolic BP of
140mmHg and/or diastolic BP of 90mmHg.
Individuals were followed until they experienced any outcome or until the end of the
observation period. The researchers found
that patients with RH had increased risks of
ESRD, ischaemic cardiac events, congestive
heart failure, and cerebrovascular accidents
compared with those without RH (multi
variable adjusted HRs of 1.32, 1.24, 1.46,
1.14 and 1.06, respectively). Importantly,
patients with uncontrolled RH had similar
rates of comorbidities at baseline, but a 25%
greater risk of ESRD and 23% greater risk of
cerebrovascular accidents than those with
controlled RH.
Activation of the sympathetic nervous
system has a role in some forms of RH and
target organ injury. In the past decade, renal
denervation (RDN) has emerged as a novel,
JANUARY 2016 | 53

but controversial, approach to decrease BP via

attenuation of sympathetic nerve activity in
the renal artery. Although two large clinical
trials of RDNSymplicity HTN1 (REF.3)
and Symplicity HTN2 (REF. 4)reported
decreases in BP with RDN and established
the overall safety of the procedure, the
prospective, single-blind, randomized, shamcontrolled Symplicity HTN3 trial, published in 2014, did not meet its efficacy end
points5a failure that has been partly attributed to technical confounders. Consequently,
the future of RDN in the USA remains
uncertain, although the procedure is used
extensively in Europe forthetreatment of RH.
A key study published in 2015 provided insight into the efficacy of RDN for
BP reduction in the hypertensive population. The Global Symplicity Registry (GSR)
was established to address the safety and
cost:benefit ratio of RDN in an uncontrolled real-world population 6. The GSR
has now reported 6month followup data
from almost 1,000 patients with hyper
tension who underwent RDN at 134 centres
in five continents. Among these patients,
323 individ uals had severe RH, defined
as a pretreatment ambulatory systolic BP
>135mmHg (mean 17916mmHg) despite
treatment with at least three antihypertensive
drugs. In this severe RH cohort, office systolic BP decreased by 20.320.8mmHg and
24h mean systolic BP by 8.916.9mmHg
at 6months after RDN; 18.6% of patients
achieved an office systolic BP <140mmHg. A
small reduction in the number of antihypertensive medications and low rates of complications after RDN (<1%) were also reported.
These data demonstrate that RDN provides
fairly safe BP reduction in patients with
RH beyond that achieved using intensive
pharmacologicaltherapies.
An alternative approach to treating
drug-resistant hypertension was examined
in a prosp ective trial undertaken by the
PATHWAY Studies Group of the British
Hypertension Society. These researchers investigated the optimal choice of a
fourth-line drug treatment for RH added
to a regimen that included an angiotensin-
converting-enzyme inhibitor or an angio
tensinII-receptor blocker, a calcium-channel
blocker and a thiazide-like diuretic. They
reasoned that if RH is primarily caused by
sodium retention, a diuretic would be the
most effective fourth-line therapy, whereas if
the underlying pathogenic mechanisms are
heterogeneous, drugs with alternative mech
anisms would be equally effective and best
stratied using biomarkers of sodium and/or
volume status, such as plasma renin levels7.
In their 12month double-blind, placebocontrolled, crossover phase IV trial, the

researchers compared spironolactone, which
has diuretic action based on mineralocorticoid receptor antagonism, with the blocker
doxazosin, which reduces peripheral vascular resistance, and the blocker bisoprolol,
which reduces renin release and cardiac output. Medication adherence was established
among the participants before randomization and treatment cycles were initiated at
low doses, which were doubled after 6weeks.
The intention-totreat analysis included
314patients, of whom 230 completed all treatment cycles. The average reduction in home
systolic BP with spironolactone was superior
to placebo by 8.70mmHg, to doxazosin by
4.03mmHg, and to bisoprolol by 4.48mmHg.
Spironolactone controlled BP in 58% of
patients, and reduced BP even among those
who achieved BP control with doxazosin or
bisoprolol. The degree of home systolic BP
reduction with spironolactone showed an
inverse relationship with plasma reninlevels; this drug was superior to bisoprolol and
doxazosin across most of the plasma renin
distribution. These findings indicate that
spironolactone is a well-tolerated and effective fourth-line treatment for patients with
RH, and implicate sodium retention as the
predominant pathophysiological mechanism
underlying RH.
In the fourth key study, Beaussier and
colleagues evaluated the role of medication adherence in BP control and target
organ injury among the participants of a
12week, single-centre, prospective, randomized, open-blinded end point trial 8.
The primary study reported that among
patients whose BP was not adequately controlled by a 4week triple-therapy regimen,
sequential nephron blockade (SNB) using
spironolactone, furosemide and amiloride
was more effective at reducing BP than
sequential reninangiotensin system (RAS)
blockade using ramipril and bisoprolol 9.
In their 2015 analysis, the investig ators
evaluated medication adherence among
Lara Crow/NPG
N E P H R O LO G Y
the study participants using plasma irbes

artan levels, urinary Nacetylseryl-aspartyllysyl-proline:creatinine ratio (a measure of
ramipril effect), last medication intake >24h
before clinic visit and detailed pill counts.
Their results indicated that low medication
adherence was prevalent (observed among
18.3% of patients with RH) and had a differential effect on the efficacy of the drug
regimens. In patients with acceptable medication adherence, but not in those with low
adherence, SNB resulted in greater decreases
in systolic BP (by a mean of 11.5mmHg),
pulse-wave velocity and left-ventricular mass
than did sequential RAS blockade. Hence,
optimizing diuretic therapy using SNB was
the most effective strategy for controlling
BP and decreasing target organ damage in
patients with RH.
Together these key studies have advanced
our understanding of the risks, prevalence
and clinical significance of RH. The findings also better delineate the pathophysiology underlying RH and optimal strategies
to improve BP control. Clearly, adequate
BP control remains elusive and clinically
consequential in a signif icant fraction of
hypertensive patients. A muti-pronged
strategy involving behavioural and lifestyle
modifications, effective broad-range drugs,
and other novel approaches is likely required
to decrease morbidity and mortality in the
RHpopulation.
Key advances
Uncontrolled resistant hypertension (RH) increases the risks of renal, cardiac and cerebrovascular
complications; blood pressure (BP) control reduces these risks2
In a real-world population of patients with RH, renal denervation provided fairly safe BP
reduction beyond that achieved using intensive pharmacological therapies6
The diuretic spironolactone is an effective fourth-line antihypertensive drug for patients with RH,
suggesting that sodium retention might be the predominant pathophysiological mechanism in
these patients7
In patients with RH, poor medication adherence prevents effective BP control and regression of
target organ damage and has differential effects on the efficacy of antihypertensive regimens8
54 | JANUARY 2016
N E P H R O LO G Y
Lilach O. Lerman and Stephen C. Textor are at
theDivision of Nephrology & Hypertension,
MayoClinic, 200 First Street SW, Rochester,
Minnesota55905, USA.
Correspondence to L.O.L.
lerman.lilach@mayo.edu
doi:10.1038/nrneph.2015.199
1.
2.
Oparil, S. & Schmieder, R. E. New approaches in the

treatment of hypertension. Circ. Res. 116,
10741095 (2015).
Sim, J. J. et al. Comparative risk of renal,
cardiovascular, and mortality outcomes in controlled,
uncontrolled resistant, and nonresistant hypertension.
Kidney Int. 88, 622632 (2015).
3.
4.
5.
6.
7.
Krum, H. et al. Catheter-based renal sympathetic

denervation for resistant hypertension: a multicentre
safety and proofofprinciple cohort study. Lancet 373,
12751281 (2009).
Symplicity HTN-2 Investigators. Renal sympathetic
denervation in patients with treatment-resistant
hypertension (The Symplicity HTN2 Trial):
arandomised controlled trial. Lancet 376,
19031909 (2010).
Bhatt, D. L. et al. A controlled trial of renal
denervation for resistant hypertension. N. Engl.
J.Med. 370, 13931401 (2014).
Bohm,M. etal. Registry on the effect of renal artery
denervation in patients with uncontrolled
hypertension. Hypertension 65, 766774 (2015).
Williams, B. etal. Spironolactone versus placebo,
bisoprolol, and doxazosin to determine the optimal
treatment for drug-resistant hypertension
8.
9.
(PATHWAY2): a randomised, double-blind, crossover

trial. Lancet 386, 20592068 (2015).
Beaussier, H. etal. True antihypertensive efficacy of
sequential nephron blockade in patients with resistant
hypertension and confirmed medication adherence.
J.Hypertens. 33, 25262533 (2015).
Bobrie, G. etal. Sequential nephron blockade versus
sequential renin-angiotensin system blockade in
resistant hypertension: a prospective, randomized,
open blinded endpoint study. J. Hypertens. 30,
16561664 (2012).
Acknowledgements
The authors work is partly supported by NIH grants

DK100081, DK73608, HL123160, and DK102325.
JANUARY 2016 | 55

Connect to...
For the latest news from the journal and beyond: @NatRevNeph
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrneph/info/newsfeeds.html
ahead of print http://www.nature.com/nrneph/journal/vaop/ncurrent
http://www.nature.com/nrneph/focus/index.html & http://www.nature.com/nrneph/collections/index.html
Posters
Providing visual overviews of hot topics http://www.nature.com/nrneph/posters
Conferences
You can meet the editors of Nature Reviews Nephrology at some of the meetings listed on our conferences page
http://www.nature.com/nrneph/info/info_conf.html
NEUROLOGY
PA R K I N S O N D I S E A S E I N 2 0 1 5
Evolving basic, pathological

and clinical concepts in PD
Lorraine V. Kalia and Anthony E. Lang
Important advances in Parkinson disease (PD) in 2015 included a revision of
its clinical diagnostic criteria and a proposal for research criteria defining
prodromal PD. Research published in the past year has also continued to
expand our understanding of the roles of Lewy pathology and synuclein
inthe pathobiology of PD.
Parkinson disease (PD) is a common neuro
degenerative disorder that has long been
defined by the clinical motor features of
parkinsonism, and the pathological findings
of prominent dopaminergic neuron loss with
synuclein-containing Lewy bodies and Lewy
neurites. Although these aspects remain at the
core of PD, our conceptualization of the dis
ease is changing, with growing evidence that
nonmotor symptoms are a critical component,
often preceding motor symptoms and involv
ing central and peripheral nondopaminergic
neurons, and that synuclein aggregates
distinct from Lewy pathology are key players
in disease pathogenesis1. Notable research
published in 2015 focused on these evolving
concepts torevise how we define and diagnose
PD, and to improve our understanding of how
synuclein mediates disease development
andprogression.
The International Parkinson and Move
ment Disorder Society (MDS) established a
task force of movement disorders experts to
redefine clinical PD in light of new insights
into the disease. The task force proposed that
early PD should be considered to occur in three
stages: preclinical PD, when neurodegenera
tion has started but no clinical signs or symp
toms are present; prodromal PD, when clinical
signs and/or symptoms are present but are
insufficient to diagnose clinical PD; and clinical
PD, when the diagnostic criteria are fulfilled.
In 2015, this task force published criteria for
prodromal2 and clinical3PD.
The new diagnostic criteria for clinical PD
were drafted by considering the task forces
refined PD definition and previously published
criteria, and were revised on the basis of input

from other neurologists who trialled the cri
teria with patients. The resulting MDSPD
Criteria3 (FIG.1) are intended for use in research,
but can also guide diagnosis of PD in the clinic.
Presence of parkinsonism remains the cen
tral feature, as in earlier criteria, but explicit
instructions for defining these motor features
are alsoincluded.
Novel aspects of the criteria include con

sideration of nonmotor symptoms (hyposmia,
autonomic dysfunction, sleep dysfunction and
psychiatric dysfunction) and ancillary diag
nostic tests (olfactory tests, cardiac metaiodo
benzylguanidine scintigraphy and presynaptic
dopaminergic imaging), as well as incorpora
tion of absolute exclusion criteria that rule out
PD (with >97% sensitivity). Importantly, early
dementia is no longer considered an exclusion
criterion for PD, thereby eliminating the dis
tinction between PD and dementia with Lewy
bodies (DLB). Under the new definition and
criteria, DLB is classified as a PD subtype if all
other criteria for PD are met. A study testing
the validity of the MDSPD Criteria against
gold-standard clinical diagnosis is underway.
Prior to receiving their diagnosis, patients
with PD have an increased incidence of sev
eral nonmotor symptoms and mild motor
symptoms compared with people who do not
develop PD4. During this prodromal phase,
neurodegeneration is less severe and less exten
sive than at the time of diagnosis of clinical
PD. Currently, no neuroprotective or disease-
modifying therapies are available for PD, but it
STEP 1
Is parkinsonism (bradykinesia plus rest tremor
and/or lead-pipe rigidity) present?
STEP 2
Not clinical PD
or
Consider prodromal PD
Yes
Are all absolute exclusion criteria* absent?
STEP 3
No
No
Not clinical PD
Yes
Are red ags and/or supportive criteria present?
0 red ags + 2
supportive criteria
1 red ag + 1
supportive criteria
or
2 red ags + 2
supportive criteria
Clinically established PD
Clinically probable PD
>2 red ags
Not clinical PD
Figure 1 | New clinical diagnostic criteria for Parkinson disease (PD).

The International
Nature
Reviews | Neurology
Parkinson and Movement Disorder Society Clinical Diagnostic Criteria for PD3 provide detailed
descriptions of parkinsonism, absolute exclusion criteria, red flags, and supportive criteria, as well
asexplicit instructions for the application of the criteria. *The nine absolute exclusion criteria
stronglysuggest an alternative cause of parkinonsim. The 10 red flags are features that imply
another causeof parkinsonism, but can be seen in pathologically proven PD. The four supportive
criteriaa dramatic beneficial response to levodopa, development of levodopa-induced
dyskinesia, presenceof rest tremor of a limb, and evidence of olfactory loss or cardiac denervation
on metaiodobenzylguanidine scintigraphyare highly suggestive of a diagnosis of PD.
JANUARY 2016 | 56

N E U R O LO G Y
Key advances
A task force established by the International Parkinson and Movement Disorder Society (MDS)
proposed new clinical diagnostic criteria for Parkinson disease (PD), which incorporate nonmotor
features, ancillary tests and absolute exclusion criteria3
The MDS task force also proposed research criteria for prodromal PD, which are primarily intended
to identify candidates for enrolment into clinical trials of putative neuroprotective therapies2
Animal studies demonstrated distinct functional effects of small aggregates of synthetic
synuclein with differing structural characteristics, supporting the hypothesis that
differentsynuclein strains underlie development of PD versus multiple system atrophy8
Direct detection of synuclein oligomers in post-mortem brain tissue from patients with PD
wasachieved by a novel approach based on the proximity ligation assay, revealing previously
undetected synuclein pathology in PD9
is believed that interventions with the poten

tial to slow or stop neurodegeneration will be
most effective if initiated as early as possible in
theprocess.
To enable identification of people at the
earliest stage of PD for enrolment into clinical
trials of putative neuroprotective therapies, the
MDS task force developed research criteria for
prodromal PD2. These criteria are designed to
identify individuals with >80% probability of
having prodromal PD, on the basis of an algo
rithm incorporating risk and protective factors
(that is, factors known to influence the likeli
hood of developing clinical PD) and prodromal
markers (that is, markers of ongoing neurode
generation). This is a novel approach to defining
prodromal PD, and the diagnostic performance
of the criteria will need to betested.
The new criteria for prodromal PD2 and
clinical PD3 are recognized to be works in pro
gress, and will require continuous revision to
accommodate the discovery of new risk fac
tors and prodromal markers, development of
additional diagnostic tests and biomarkers,
and advances in our understanding of the
pathobiology of PD.
2015 has also seen rapid advances in
synuclein research. A close link between
synuclein and PD was originally established
through the discovery that synuclein gene
(SNCA) mutations cause PD, and that Lewy
bodies are composed primarily of synuclein.
However, the notion that the large insoluble
synuclein aggregates within Lewy bodies are
the causeor even a surrogate markerof
neurodegeneration in PD continues to be
challenged. Neuronal loss has been shown to
precede the development of Lewy pathology
inthe substantia nigra5, and Lewy body den
sity in PD does not correlate with the degree
of loss of nigral neurons6. This lack of corre
lation was reinforced by the discovery that
a subset of people with LRRK2 gene muta
tionswhich are usually associated with
Lewy pathologycan manifest typical clinical
features of PD but have no Lewy bodies in the
substantia nigra at autopsy7. However, Lewy
body density does correlate with neuronal loss

in some extranigral regions (olfactory bulb,
cingulate cortex, temporal cortex and parietal
cortex) in sporadic PD6, and presence of Lewy
bodies correlates with certain nonmotor fea
tures, including cognitive impairment, anxiety
and orthostatic hypotension, in LRRK2related
PD7. Nevertheless, evidence is mounting that
synuclein aggregates other than Lewy bodies
and Lewy neurites are important mediators of
neurodegeneration inPD.
A particular focus of recent research has
been on small synuclein aggregates or
assembliesfor example, oligomers, ribbons
and fibrilswhich have been shown to cause
neuronal toxicity by impairing numerous cellu
lar processes. Laboratory studies have demon
strated that small synuclein aggregates can
be released by neurons, be taken up by neigh
bouring cells, and subsequently induce Lewylike pathology via recruitment of endogenous
host synuclein, thereby spreading pathology
within the brain in a prion-like manner. In ani
mal models, injection of synthetic synuclein
fibrils into the striatum results in loss of sub
stantia nigra neurons, as well as propagation of
Lewy pathology to distal brain regions.
In a letter published in Nature in 2015,
Peelaerts etal.8 demonstrated that in a rodent
model, synthetic synuclein aggregates with
different structural characteristics (strains)
could have distinct pathological effects.
Oligomers had the greatest propensity to
spread, ribbons and fibrils induced Lewy-like
pathology, and ribbons also induced accumu
lation of phosphorylated synuclein in oligo
dendroglia (reminiscent of the pathology of a
related neurodegenerative synucleinopathy,
multiple system atrophy).
Direct immunohistochemical detection of
small synuclein aggregates in post-mortem
brain tissue has been hampered by a lack of
antibodies that can distinguish synuclein
aggregates from other protein aggregates (for
example, amyloid aggregates). In a study
published in Brain in 2015, Roberts etal.9 devel
oped a novel approach, based on the proximity
57 | JANUARY 2016
ligation assay, which utilizes antibodies that are

specific for synuclein, and oligonucleotide-
labelled secondary antibodies that generate
a detectable signal only when at least two
synuclein molecules are in close proximity.
Using this technique on post-mortem PD
brain tissue, the researchers demonstrated the
presence of synuclein oligomers not only in
neurons with Lewy bodies, but also in neurons
with pale bodies (precursors of Lewy bodies)
and in seemingly unaffected neurons.
In a rodent model, Peelaerts etal.8 demon
strated that synuclein aggregates injected
intravenously crossed the bloodbrain barrier
and were distributed to the CNS. Previous
animal studies have also shown spread of
synuclein from the periphery to the brain after
intramuscular or gastric injection. Preliminary
evidence that these findings are relevant to PD
has come from a recently published population-
based cohort study, which showed a decreased
risk of PD in people who had undergone trun
cal vagotomy10. These findings do not directly
demonstrate spread of synuclein, but they
suggest that an intact vagus nerve, which may
act as a conduit from the gut to the brain,
increases the risk of PD. Collectively, this work
informs our understanding of the fundamental
aspects of PD pathogenesis, and is expected to
have important implications for how we define
and diagnose PD in the future.
Lorraine V. Kalia and Anthony E. Lang are at the Krembil
Research Institute and the Edmond J. Safra Program in
Parkinsons Disease, Toronto Western Hospital, 399
Bathurst Street, Toronto, OntarioM5T2S8, Canada.
Correspondence to A.E.L.
lang@uhnresearch.ca
doi:10.1038/nrneurol.2015.249
Kalia,L.V. & Lang,A.E. Parkinsons disease. Lancet
386, 896912 (2015).
2.
Berg,D. etal. MDS research criteria for prodromal
Parkinsons disease. Mov. Disord. 30, 16001611
(2015).
Postuma,R.B. etal. MDS clinical diagnostic criteria for
3.
Parkinsons disease. Mov. Disord. 30, 15911601
(2015).
4. Schrag,A., Horsfall,L., Walters,K., Noyce,A.
&Petersen,I. Prediagnostic presentations of
Parkinsonsdisease in primary care: a casecontrol
study. Lancet Neurol. 14, 5764 (2015).
5.
Dijkstra,A.A. etal. Stage-dependent nigral neuronal
loss in incidental Lewy body and Parkinsons disease.
Mov. Disord. 29, 12441251 (2014).
Iacono,D. etal. Parkinson disease and incidental Lewy
6.
body disease: just a question of time? Neurology 85,
16701679 (2015).
Kalia,L.V. etal. Clinical correlations with Lewy
7.
bodypathology in LRRK2related Parkinson disease.
JAMANeurol. 72, 100105 (2015).
Peelaerts,W. etal. Synuclein strains cause distinct
8.
synucleinopathies after local and systemic
administration. Nature 522, 340344 (2015).
9. Roberts,R.F., Wade-Martins,R. &Alegre-Abarrategui,J.
Direct visualization of alpha-synuclein oligomers reveals
previously undetected pathology in Parkinsons disease
brain. Brain 138, 16421657 (2015).
10. Svensson,E. etal. Vagotomy and subsequent risk of
Parkinsons disease. Ann. Neurol. 78, 522529 (2015).
1.
N E U R O LO G Y
STROKE IN 2015
Acute endovascular recanalization

therapy comes of age
Alejandro A. Rabinstein
The past year has seen practice-changing findings in stroke research.
Strongevidence now supports endovascular thrombectomy as the new
gold standard of care in acute ischaemic stroke, and a pragmatic trial
raisedconcerns over early intensive mobilization after stroke. Moreover,
new insights were gained into the trajectory of stroke-associated
cognitivedecline.
The year 2015 will be remembered for the
publication of five major trials that con
clusively demonstrated that endovascular
thrombectomy improves outcomes in patients
with disabling acute ischaemic strokes caused
by a proximal intracranial vessel occlusion15.
Good research was published concerning
other stroke-related topics, but none was as
influential as these trials on endovascular ther
apy, which have catapulted this intervention to
the status of a new standard of care.
Everything started with the presenta
tion of the results of the Dutch MR CLEAN
trial1, which was a pragmatic, multicentre,
randomized clinical trial that compared
endovascular (intra-arterial) recanalization
therapy plus usual care (including intravenous
thrombolysis when indicated) versus usual
care alone (control group). Inclusion to MR
CLEAN required demonstration of an occlu
sion of a proximal intracranial artery in the
anterior circulation (usually distal intracranial
internal carotid artery or M1/M2 branches of
the middle cerebral artery) by CT angiography.
Initiation of treatment had to be possible
within 6h after stroke onset. Although any
patient with an NIH Stroke Scale (NIHSS)
score of 2 could be enrolled, patients typi
cally had disabling deficits (median NIHSS
score 17, interquartile range 1421).
The MR CLEAN trial1 was completed
according to plan, with 500 patients enrolled.
The results showed that the patients treated
with endovascular intervention had better
functional outcomes at 90days: the adjusted
common odds ratio for the shift of the mod
ified Rankin Scale (mRS) score toward
lower values was 1.67 (95% CI, 1.212.30).
Functional independence (mRS score 02)
was seen in 33% of patients in the endo
vascular arm versus 19% in the control group.
In both groups, the majority of patients

received intravenous thrombolysis (87% in
the endovascular group and 91% in the control
group). Retrievable stents were used in all but
six of the patients treated with endovascular
intervention. Good or excellent reperfusion
was achieved in 58% of the patients. Rates of
symptomatic intracranial haemorrhage were
low and comparable in both groups, and no
other safety concerns related to the use of
endovascular intervention were observed.
The remarkable benefit from endovascu
lar therapy shown by MR CLEAN prompted a
review of various ongoing explanatory trials,
namely, ESCAPE2, EXTENDIA 3, SWIFT
PRIME4 and REVASCAT5. In all these trials,
the interim analyses demonstrated a clear dif

ference in favour of endovascular thrombec
tomy, leading to discontinuation of enrolment.
Although the five trials differed in some
aspects, most notably in the imaging criteria
used for inclusion, the similarities in their
design and main results overshadow these
methodological differences. The five trials
enrolled patients with severe anterior circu
lation strokes and good prestroke function.
They all had the inclusion criteria of absence
of extensive early ischaemic changes on CT
scan, and proof of a proximal intracranial
vessel occlusion by CT angiogram. The trials
treated most patients in both arms with intra
venous thrombolysis, and relied primarily on
a retrievable stent to achieve reperfusion. In
the great majority of patients, the endovascu
lar intervention was started less than 6h after
symptom onset. All trials consistently showed
higher reperfusion rates and better functional
recovery in patients treated with endovascu
lar thrombectomy than in patients receiving
standard care.
In combination, these five trials have
shown that endovascular therapy increases
the absolute proportion of patients regain
ing functional independence at 90days by
1331%, which translates to a number needed
to treat ranging from 37 across the five trials.
Therefore, endovascular reperfusion with a
retrievable stent should now be considered
the standard of carefor patients with severe
strokes, and health-care systems should be
adjusted to enable timely intervention in
thesepatients6.
CT perfusion
imaging
Endovascular
thrombectomy
IV rtPA
CT scan
CT angiography
Preoperative DSA
Postoperative DSA
Figure 1 | Gold-standard treatment of acute ischaemic stroke. In patients

withReviews
severe stroke
and
Nature
| Neurology
documented proximal intracranial artery occlusion, intravenous recombinant tissue plasminogen
activator (IV rtPA) should be promptly followed by endovascular therapy. Use of a perfusion scan
(such as CT perfusion imaging) to assess the core and the penumbra is currently considered
optional. After a CT scan to exclude haemorrhagic stroke, patient should receive IV rtPA as soon as
possible, and undergo CT angiography. Perfusion can be assessed with CT perfusion imaging. The
success of endovascular thrombectomy should be evaluated with preoperative and postoperative
digital subtraction angiography (DSA).
JANUARY 2016 | 58

N E U R O LO G Y
Key advances
MR CLEAN led the way to a series of randomized controlled trials that have conclusively
demonstrated that endovascular thrombectomy improves functional outcomes in patients with
disabling ischaemic stroke1
Endovascular thrombectomy benefits patients with stroke resulting from persistent proximal
occlusion of an intracranial vessel in the anterior circulation, if the intervention can be started
within 6h of stroke onset1
Very early and intensive mobilization is not beneficial and can be detrimental after a stroke,
andshould probably be avoided in patients with severe infarcts or intracerebral haemorrhage; 8
A conservative mobilization strategy during the first 2448h is more advisable than early
intensive mobilization
Cognitive decline can be accelerated after a clinical stroke, not only acutely but also over the
ensuing years10; the mechanisms underlying cognitive deterioration after stroke need to be
better elucidated
New learning and verbal memory tend to be more affected in the acute phase after stroke,
whereas executive dysfunction predominates in the later phase10
Nevertheless, important questions remain

unanswered. The main question concerns
the ideal imaging method for patient selec
tion (FIG.1). Some trials (MR CLEAN, SWIFT
PRIME and REVASCAT) only mandated
CT for identification of the ischaemic core
and CTangiogram for demonstration of the
arterial occlusion amenable to endovascular
therapy. However, ESCAPE included evalu
ation of collaterals on CT angiography, and
EXTENDIA required CT perfusion imag
ing to define the core and estimate the extent
of salvageable tissue. Furthermore, a many
patients enrolled into MR CLEAN and SWIFT
PRIME also underwent CT perfusionimag
ing prior to enrolment, even though this
imaging modality was not mandated by the
protocol of these studies. Moreover, rapid MRI
strategies have become increasingly available,
although their use is less widespread than
that of CT. Future trials are likely to compare
these various imaging techniques to iden
tify the optimal candidates for endovascular
intervention.
Early mobilization after stroke has long
been regarded as a cornerstone of neuro
rehabilitation, and as one of the keys to the
benefits of treatment in specialized stroke
units. Although this intervention is broadly
recommended in practice guidelines7, its value
has remained largely unproven. AVERT was a
pragmatic trial in which 2,104 patients were
randomly assigned to very early mobiliza
tionor usual care8. The very early mobilization
intervention consisted of frequent outofbed
activity initiated within 24h of stroke onset.
The results were surprising to many: favour
able functional outcome at 90days was less
common in the very early mobilization group
than in the usual care group (46% versus 50%;
adjusted OR 0.73; 95% CI 0.590.90), and the
early mobilized patients did not have reduced
rates of immobilization-related complica

tions or an increased chance of walking 50m
unassisted by 3months. The evidence against
very early mobilization was most prominent
in patients with severe ischaemic strokes and
those with intracerebral haemorrhage.
AVERT does not show

that early mobilization is
detrimental, but it should be
pursued gradually
follow-up (mean 6.1years). Stroke was associ

ated with acute decline in global cognition,
new learning and verbal memory. However,
the greatest contribution of this study was the
recognition that stroke was also associated
with accelerated decline in global cognition
after the event. In fact, the occurrence of a
stroke increased the risk of cognitive impair
ment at 6years by 12%. The predominant
impairment over time was in executive func
tion; new learning and verbal memory tended
to recover. These results highlight the fact that
the cognitive consequences of a stroke are
complex, dynamic and sometimes persistent.
Although they enhance our knowledge on vas
cular cognitive impairment, they still tell only
part of the story: this study could not account
for the effects of subclinical strokes, which
are also known to contribute to progressive
cognitivedecline.
The year 2015 brought us plenty of good
research that advanced our understanding of
stroke diagnosis, treatment and prognosis.
However, it will be mostly remembered from
the long-awaited evidence that endovascular
therapy can markedly improve outcomes in
patients with severe acute ischaemic strokes.
Alejandro A. Rabinstein is at the Cerebrovascular
Division of the Department of Neurology, Mayo West
8B, 200 First Street SW, Mayo Clinic, Rochester,
Minnesota 55905, USA.
rabinstein.alejandro@mayo.edu
doi:10.1038/nrneurol.2015.241
An important caveat must be taken into

account when interpreting the results of
AVERT. Mobilization during the first 24h
also occurred in 59% of patients in the usual
care group, although this mobilization was
substantially less intense than in the inter
vention group. Moreover, only 7% of patients
in the usual care group stayed in bed for more
than 48h, which might explain the low rate of
immobility-related complications observed in
this study. Thus, AVERT does not show that
early mobilization is detrimental but, rather,
that it should be pursued gradually and with
moderation, especially in patients with severe
infarcts or intracerebral haemorrhage.
Our understanding of the vascular contri
bution to cognitive decline has been growing
exponentially in recent years9. However, our
insight into the trajectory of cognitive decline
after a stroke has been hampered by the lack of
a longitudinal study with cognitive assessment
before and serially after the stroke. A study by
Levineetal., which analysed the cognitive
data from the prospective cohort enrolled
in REGARDS, attempted to fill this gap10.
Among 23,572 patients included in the ana
lysis, 515(2%) had an incident stroke during
59 | JANUARY 2016
Berkhemer,O.A. etal. A randomized trial of

intraarterial treatment for acute ischemic stroke.
N.Engl. J.Med. 372, 1120 (2015).
2.
Goyal,M. etal. Randomized assessment of
rapidendovascular treatment of ischemic stroke.
N.Engl. J.Med. 372, 10191030 (2015).
3.
Campbell,B.C. etal. Endovascular therapy for
ischemicstroke with perfusion-imaging selection.
N.Engl. J.Med. 372, 10091018 (2015).
4.
Saver,J.L. etal. Stent-retriever thrombectomy
afterintravenous tPA versus tPA alone in stroke.
N.Engl. J.Med. 372, 22852295 (2015).
5.
Jovin,T.G. etal. Thrombectomy within 8hours after
symptom onset in ischemic stroke. N.Engl. J.Med.
372, 22962306 (2015).
6.
Powers,W.J. etal. American Heart Association/
American Stroke Association focused update of the
2013 guidelines for the early management of patients
with acute ischemic stroke regarding endovascular
treatment: a guideline for healthcare professionals
from the American Heart Association/American Stroke
Association. Stroke 46, 30203035 (2015).
7. Bernhardt,J., English,C., Johnson,L. &
Cumming,T.B. Early mobilization after stroke: early
adoption but limited evidence. Stroke 46, 11411146
(2015).
8.
Bernhardt,J. etal. Efficacy and safety of very early
mobilisation within 24h of stroke onset (AVERT):
arandomised controlled trial. Lancet 386, 4655
(2015).
9.
Gorelick,P.B. etal. Vascular contributions to cognitive
impairment and dementia: astatement for healthcare
professionals from the American Heart Association/
American Stroke Association. Stroke 42, 26722713
(2011).
10. Levine,D.A. etal. Trajectory of cognitive decline after
incident stroke. JAMA 314, 4151 (2015).
1.
N E U R O LO G Y
NEURO-ONCOLOGY IN 2015
Progress in glioma diagnosis,

classification and treatment
Patrick Y.Wen and David A.Reardon
Gliomas are the most common form of malignant primary brain tumour.
Inthe past year, substantial progress has been made in the classification and
treatment of lower-grade gliomas (WHO grades II and III), and the FDA has
approved a new therapy for newly diagnosed glioblastomas.
Gliomas account for 80% of all malignant pri
mary brain tumours, with an annual incidence
of approximately 6 per 100,000 in the USA1.
Under the current WHO classification, glio
mas are divided into four histological grades2.
Grade I gliomas, such as pilocytic astrocyto
mas, are very slow-growing tumours that are
potentially curable if completely resected.
Grade II gliomas (astrocytomas, oligodendro
gliomas and oligoastrocytomas) and the more
aggressive grade III gliomas (including ana
plastic astrocytomas, anaplastic oligodendro
gliomas and anaplastic oligoastrocytomas)
have an intermediate clinical course, whereas
grade IV gliomas (glioblastoma) have the most
aggressive clinical course (median survival
between 14.5 and 16.6months3,4).
Clear classification of the different types of
glioma is crucial to ensure that they are diag
nosed accurately and treated with the appro
priate therapy. Such classification is particularly
important for grade II and III gliomas (lower-
grade gliomas). The diagnosis and surgical
removal of these tumours is hindered by their
diffuse nature, and they generally progress to
higher-grade tumours over time2. Despite its
widespread use, the WHO histopathological
classification is limited by substantial inter
observer variability and poor correlation with
clinical outcome. Three studies published in
2015 clarified the classification of lower-grade
gliomas, thereby potentially improving the
prediction of tumour outcome, and guiding
therapeutic strategies to treatgliomas57.
In addition to histopathological character
istics, a number of molecular alterations have
been identified in lower-grade gliomas over
the past two decades. Such genomic modifi
cations correlate with prognosis and, in some
cases, predict response to therapy. Codeletion
of chromosome arms 1p and 19q is associated
with oligodendrogliomas and with sensitivity
to chemotherapy5. Mutations in the isocitrate
dehydrogenase 1 and 2 genes (IDH1 and IDH2)
are present in the majority of lower-grade
gliomas and predict an improved outcome5.

TP53 and ATRX mutations are present in
astrocytomas, and mutations in the promoter
of TERT, a gene that encodes a subunit of the
telomerase enzyme, are seen in a subset of
oligodendrogliomas andglioblastomas6.
Using unsupervised clustering of muta
tions and various whole-genome molecular
analyses (including exome and RNA sequenc
ing, and DNA copy number and DNA methy
lation profiling), the Cancer Genome Atlas
Research Network5 and a group from Japan6
classified lower-grade gliomas into three
non-overlapping and prognostically relevant
subtypes (types IIII) (FIG.1), according to
the presence or absence of IDH mutations
and 1p/19q codeletions. 80% of lower-grade
gliomas had IDH mutations, which were associ

ated with improved prognosis. The 30% of
patients who had lower-grade gliomas with an
IDH mutation and 1p/19q codeletion (typeI)
had the most favourable clinical outcome. Most
of these tumours were oligodendrogliomas
and harboured mutations in CIC, FUBP1,
NOTCH1 and/or the TERT promoter. 50% of
lower-grade gliomas harboured IDH muta
tions but no 1p/19q codeletion (type II),
were usually classified hisologically as astro
cytomas, and had mutations in TP53 (94%)
and/or ATRX(86%). The 20% of lower-grade
gliomas without IDH mutations (typeIII)
had genomic aberrations and a clinical course
that closely resembled thatof glioblastoma. It
is interesting to note that oligoastrocytomas,
which have histopathological features of both
astrocytomas and oligodendrogliomas, were
found among all three subtypes and did not
exhibit a specific molecular signature, thereby
providing support for the eventual elimination
of oligoastrocytoma as a diagnostic entity.
The third study classified lower-grade glio
mas on the basis of IDH and TERT promoter
mutations and codeletion of chromosome arms
1p and 19q7. The authors categorized these
tumours into five molecular subgroups and,
importantly, they confirmed the three main
subgroups found by the other twostudies.
These three studies 57 have important
implications for patients with lower-grade
gliomas. Division of lower-grade gliomas into
Lower-grade gliomas
(WHO grades II and III)
No IDH mutations
20%
IDH1/2 mutations
80%
Histological
classication
Additional
mutated genes
Clinical prognosis
1p/19q codeletion
30%
Intact 1p/19q
50%
Oligodendroglioma
Astrocytoma
TERT, CIC, FUBP1,

NOTCH1
TP53, ATRX
TERT, EGFR,
CDKN2A, MDM4,
PTEN, NF1
Good
Intermediate
Poor (similar
to glioblastoma)
Type I
Type II
Type III
Figure 1 | Molecular alterations and classification of lower-grade gliomas.

The analysis
of
Nature Reviews
| Neurology
genomic modifications in gliomas has allowed the classification of such tumours into three robust
subtypes (typeIIII, yellow boxes)57. These molecular categories have a better correlation with
clinical prognosis (blue boxes) than do the histopathological WHO grades, and will potentially
facilitate diagnosis and prognostication for lower-grade gliomas. IDH, isocitrate dehydrogenase.
JANUARY 2016 | 60

N E U R O LO G Y
Key advances
Recent genome-wide analyses allowed the
precise classification of lower-grade
gliomas into three subtypes (typeIIII),
according to their genomic alterations57
Combining radiotherapy with procarbazine,
lomustine and vincristine chemotherapy to
treat high-risk grade II gliomas significantly
improved patient survival8,9
In October 2015, the FDA approved the use
of tumour treating fields in conjunction with
standard chemotherapy to treat patients
with glioblastoma
The combination of chemotherapy
andtumour treating fields to treat patients
with glioblastoma improved their survival
while preserving life quality10
three robust subgroups (FIG.1) will allow more-

accurate determination of diagnosis and prog
nosis, and assignment of patients tothe most
appropriate therapies. For example, the20%
of patients with gliomas that have no IDH
mutations would probably benefit from more-
aggressive therapies, as their tumours exhibit
molecular features and behaviour that closely
resemble those of glioblastoma.
In addition to the progress made in the
classification of lower-grade gliomas, 2015 has
seen an important advance in the treatment
of gradeII gliomas8,9. The Radiation Therapy
Oncology Group conducted a study (RTOG
98-02) involving patients with grade II glio
mas who were considered to be at high risk
of recurrence (age 40years or age <40years
with subtotal resection)8. These patients were
randomly assigned to radiotherapy alone or
radiotherapy followed by up to six cycles of
PCV (procarbazine, lomustine and vincristine)
chemotherapy. The patients treated with radio
therapy and PCV had a significantly longer
median survival time (13.3versus 7.8years)
and a longer median progression-free survival
time (10.4 versus 4.0years)8 than those treated
with the radiotherapy alone. The benefits of
the combined treatment were most promi
nent in patients with oligodendrogliomas and
oligoastrocytomas, although some efficacy was
also observed in patients with astrocytomas9.
These results suggest a new approach to the
treatment of high-risk lower-grade gliomas,
with chemotherapy being introduced earlier
in the course of treatment, after initial resec
tion. Given the toxicity of PCV chemotherapy,
some debate remains as to whether the better
tolerated temozolomide is preferable to PCV.
The past year has also seen progress
inthe treatment of glioblastoma. Since 2005,
the standard therapy for glioblastoma has
consisted of radiotherapy with temozolomide
chemotherapy3, and all attempts to develop

more-effective therapies have been unsuccess
ful. In October 2015, the FDA approved the
use of low-intensity, intermediate-frequency
alternating electric fields (tumour treating
fields, or TTF), in conjunction with standard
chemoradiotherapy to treat patients newly
diagnosed with glioblastoma. This electro
magnetic therapy is believed to disrupt cell
division in glioblastoma cells. In a phaseIII trial
in patients newly diagnosed with glioblastoma,
the addition of TTF to standard chemoradi
ation improved progression-free survival to
7.1months, compared with 4.2months for
the chemoradiation-only control group. Most
importantly, the use of TTF increased the
overall survival time to 19.4months compared
with 16.6months for the control group10. TTF
treatment was well tolerated with no significant
added toxicity except for scalp rash related to
the electrodes. Quality of life and gross cognitive
function were also comparable in the two arms
of the trial10. However, some controversy persists
regarding this therapy, and its role in the stand
ard treatment of patients with glioblastoma and
in clinical trials remain to bedetermined.
In a field where progress has been unaccept
ably slow, the past year has seen important
advances in the classification and treatment
of gliomas which, hopefully, will translate into
improved outcomes for patients. Future chal
lenges include establishment of strategies to
overcome the extensive spatial and temporal
heterogeneity of molecular alterations in glio
mas, and development of agents that cross the
bloodbrain barrier effectively.
Patrick Y.Wen and David A.Reardon are at the Centre

for Neuro-Oncology, Dana Farber/Brigham and
Womens Cancer Centre, 450 Brookline Avenue,
Boston, Massachusetts 02215, USA.
Correspondence to P.Y.W.
pwen@partners.org
doi:10.1038/nrneurol.2015.242
Ostrom,Q.T. etal. CBTRUS statistical report: primary
brain and central nervous system tumours diagnosed
in the United States in 20082012. Neuro Oncol.
17(Suppl. 4), iv1iv62 (2015).
2.
Louis,D.N. etal. The 2007 WHO classification
oftumours of the central nervous system.
ActaNeuropathol. 114, 97109 (2007).
Stupp,R. etal. Effects of radiotherapy with
3.
concomitant and adjuvant temozolomide versus
radiotherapy alone on survival in glioblastoma in a
randomised PhaseIII study: 5year analysis of the
EORTC-NCIC trial. LancetOncol. 10, 459466
(2009).
4.
Gilbert,M.R. etal. Dose-dense temozolomide for newly
diagnosed glioblastoma: a randomized PhaseIII clinical
trial. J.Clin. Oncol. 31, 40854091 (2013).
5. CancerGenome AtlasResearch Network.
Comprehensive, integrative genomic analysis of
diffuselower-grade gliomas. N.Engl. J.Med. 372,
24812498 (2015).
Suzuki,H. etal. Mutational landscape and clonal
6.
architecture in grade II and III gliomas. Nat. Genet. 47,
458468 (2015).
7.
Eckel-Passow,J.E. etal. Glioma groups based on
1p/19q, IDH, and TERT promoter mutations in
tumours. N.Engl. J.Med. 372, 24992508 (2015).
8.
Buckner,J. etal. Phase III study of radiation
therapy(RT) with or without procarbazine, CCNU, and
vincristine (PCV) in low-grade glioma: RTOG 9802 with
Alliance, ECOG, and SWOG [abstract]. J.Clin. Oncol.
32 (Suppl. 5), 2000 (2014).
9.
Buckner,J. etal. R9802: Phase III study of radiation
therapy (RT) with or without procarbazine, CCNU, and
vincristine (PCV) in low-grade glioma: results by
histologic type. Neuro Oncol. 16 (Suppl. 5), v11 (2014).
10. Stupp,R. etal. Maintenance Therapy With TumorTreating Fields Plus Temozolomide vs Temozolomide
Alone for Glioblastoma: A Randomized Clinical Trial.
JAMA 314, 2535-2543 (2015).
1.
P.Y.W. and D.A.R. serve on the advisory board of Novocure.
M U LT I P L E S C L E R O S I S I N 2 0 1 5
Managing the complexity of

multiple sclerosis
Olga Ciccarelli and Alan Thompson
The application of imaging biomarkers has provided new insights into the
mechanisms of damage in multiple sclerosis (MS) and the risk of MS
development and progression. The goal of eliminating all disease activity
requires a timely escalation of treatment. This increasing complexity is
compounded by the need to treat comorbidities.
Multiple sclerosis (MS) is an increasingly
complex disease in terms of its pathogen
esis, comorbidities, prognosis and treat
ment, and successful patient management
requires knowledge of this complexity. 2015
saw advances in our understanding of the
61 | JANUARY 2016
mechanisms of the disease, ways in which

to formulate patient prognoses, how best to
escalate treatment, and the role of comorbid
ities. All of these aspects need to be incorpo
rated into an effective patient management
plan (FIG.1).
N E U R O LO G Y
The mechanisms that underlie the patho
genesis of MS are yet to be fully elucidated, but
they are known to include a cascade of events
that induce physical and cognitive deficits. A
reduction in neuronal integrity and function
that affects the grey matter compartment is
thought to be the key pathological process
that leads to cognitive impairment in MS.
Findings of a study published by Freeman
etal.1 in 2015 suggest that synaptic and/or
dendritic damage occurs prior to quantifiable
grey matter volume loss, and might reflect
neuronal and axonal loss that contributes to
clinical deficits. In this study, Freeman and
colleagues1 used [11C]flumazenil ([11C]FMZ)
PET, which quantifies aminobutyric acid
typeA (GABAA) receptor density invivo, to
identify grey matter damage beyond cortical
lesions2. FMZ is an antagonist of the central
benzodiazepine receptor, a component of
the GABAA receptor complex that is present
on axosomatic and axodendritic synapses
throughout the cortical and subcortical grey
matter. The number of [11C]FMZ binding sites
per grey matter region was lower in several
cortical areas (the parietal, cingulate andinsu
lar cortices and the left frontal cortex) and
subcortical regions (the thalamus, hippocam
pus and amygdala) in patients with MS than
in healthy controls. Greater amounts of neu
ronal damage were seen in patients with sec
ondary progressive MS than in patients with
relapsingremitting MS (RRMS), but the most
striking result was that [11C]FMZ binding was
lower in patients with RRMS than in healthy
controls even in the absence of notable grey
matter atrophy. A significant relationship was
found between the level of cortical [11C]FMZ
binding and performance on several cognitive
tests. A goal of future research is to provide
neuroprotective and reparative therapies that
could be applied at such early stages of MS
to stop or at least slow down neurodegener
ation and reduce cognitive impairment in
progressive MS3.
Clinically isolated syndrome (CIS) rep
resents a patients first neurological episode
that is suggestive of MS. Most patients with
CIS develop RRMS within 5years of onset,
and most patients with RRMS develop pro
gressive MS 1015years after onset of MS.
Although challenging, formulating a prog
nosis that accurately predicts the development
of MS and the accumulation of neurological
disability is crucial for designing successful
treatment plans for individual patients. A key
step towards such individualized treatment
of patients with CIS is to stratify them into
groups according to demographic, clinical,
radiological and biological characteristics.
Patients in such groups are likely to have a
Identifying and managing

comorbidities
Clarifying and targeting

pathogenic mechanisms
contributing to clinical
disability and progression
Maximizing
physical and
cognitive
ability in MS
Early treatments and

timely, appropriate
treatment escalation
Identifying therapies
for progressive MS
Figure 1 | Towards successful management of the complexity of multiple sclerosis (MS).

Nature Reviews
Neurology
Anunderstanding of the pathogenic mechanisms, progression, comorbidities
and |optimal
treatment escalation is necessary to formulate a treatment plan for individuals with MS.
different risk of developing MS and longterm disability, so will benefit from different
treatments at different time points.
A study published by Tintore etal.4 in 2015
illustrates the importance of this stratifica
tion step. This study included a single-centre
cohort of 1,015 patients with CIS who were
clinically and radiologically followed up for
a mean of 6.8years4. The results showed that
10 brain lesions visible with MRI at the onset
of CIS was highly predictive (classified in the
study as a high-impact prognostic factor) of
development of MS and disability. The pres
ence of oligoclonal bands in the cerebrospinal
fluid was also predictive, but to a lesser extent
(classified as a medium-impact prognostic
factor). Presentation of CIS with optic neuritis
and the use of a disease-modifying treatment
had a (probably marginal) protective effect
against the development of MS and disability.
Other demographic factors, such as gender
and age at onset, were low-impact prognos
tic factors. The study had some methodologi
cal limitations, such as the number of patients
who dropped out during followup (unavoid
able in this type of longitudinal study) and
the fact that the latest 2010 McDonald diag
nostic criteria were not used. Nevertheless,
the results provide further clarity about the

heterogeneous outcomes of patients with CIS
and help to inform their prognosis.
The ability to predict disease course in
RRMS is particularly important now that we
have a range of new treatments for the disease;
although there is now a consensus that early
treatment is optimum, the nature and timing
of escalation to second-line treatment remains
a challenge. The ultimate goal of such esca
lation is to increase the chance that patients
reach a long-term status of no evidence of
disease activity (NEDA)5.
A central question that surrounds second-
line treatment is whether natalizumab is
more effective than fingolimod once failure
of first-line treatment has been established. A
headtohead comparison between these two
drugs is unlikely to be conducted, but Kalincik
etal.6 in 2015 extracted data from the MSBase
registry to compare the outcomes of treat
ment escalation to natalizumab or fingolimod
in patients with MS who had experienced
disease activity while receiving injectable
disease-modifying treatments. The relapse
rate after switching to natalizumab was 50%
lower than after switching to fingolimod, with
a corresponding increase in the proportion
Key advances
Synaptic and/or dendritic loss might be an early pathological abnormality in multiple sclerosis
(MS) and precede MRI-detectable volume loss1
In clinically isolated syndrome, oligoclonal bands and the lesions detected with brain MRI are
prognostic factors for the development of MS and early disability4
The relapse rate was 50% lower when patients switched from injectable disease-modifying
treatment to natalizumab than when they switched to fingolimod, but no difference was
observed in disability progression6
Although MS patients live longer than before, their life expectancy remains ~7years shorter than
that of a matched healthy population; treatment of comorbidities might improve survival8
The Progressive MS alliance is driving an initiative to identify new treatments to slow or stop
progression of MS3
JANUARY 2016 | 62

N E U R O LO G Y
of relapse-free patients on natalizumab.
Importantly, however, 6month sustained dis
ability progression rates did not differ between
the two treatments. This finding highlights the
need to identify new treatments that can slow
or stop progression of MS, a major initiative
being driven by the Progressive MS Alliance,
who, in 2015, published an appraisal of current
knowledge in this area and suggested future
steps3. One other important consideration is
that drug efficacy is only one factor that is
considered by doctors and patients when dis
cussing treatment escalation; treatment safety
and tolerability, together with risk assessment7,
are additional important elements, particularly
from the patients perspective.
In addition to specific treatment
approaches, a holistic approach to manage
ment, including a focus on well-being, is
paramount, and identifying and managing
comorbidities is an important element of
this approach. The impact of comorbidities
on clinical symptoms and disability progres
sion in MS is becoming clear, and knowledge
of how physical and mental comorbidities
affect MS will improve management of the
complexity of the disease.
In 2015, Marrie etal.8 addressed the ques
tion of whether comorbidities are responsible
for the reduced survival associated with MS.
They used population-based administrative
data to study 5,797 people with MS and 28,807
healthy controls who were matched for sex,
year of birth and geographical region. Median
survival from birth was 75.9years in the MS
population, and 83.4years in the control popu
lation, which corresponded to a twofold unad
justed increase in the hazard of death in the
MS population. Comorbidities (depression,
diabetes and ischaemic heart disease) were
associated with increased mortality in MS,
but did not confer a greater risk of mortality
in the MS population than in the control pop
ulation. Mortality from infectious diseases and
diseases of the respiratory system was higher
in the MS population than in thecontrol
population. These findings extendthe results
ofprevious studies that reported an effect of
comorbidity on the diagnosis of and disability
in MS9, suggesting that treatment and preven
tion of comorbidities improves survival in MS.
Future research will fill important gaps in our
knowledge about the worldwide epidemiology
of comorbidity in MS10.
These recent advances in MS research and
clinical trials will help clinicians to manage the
complexity of MS in clinical practice and will
inform future research in the field. We antici
pate that 2016 will bring major advances in the
treatment of progressive MS, which remains a
substantial unmet need3.
Olga Ciccarelli and Alan Thompson are at Queen

Square Multiple Sclerosis Centre, University College
London Institute of Neurology and National Institute
for Health Research UCL HospitalsUCL Biomedical
Research Centre, Queen Square, BOX9,
LondonWC1N3BG, UK.
Correspondence to A.T.
alan.thompson@ucl.ac.uk
doi:10/1038/nrneurol.2016.2
1.
2.
3.
4.
5.
6.
Freeman,L. etal. The neuronal component of gray

matter damage in multiple sclerosis: a [11C]flumazenil
positron emission tomography study. Ann. Neurol. 78,
554567 (2015).
Louapre,C. etal. Beyond focal cortical lesions in MS:
aninvivo quantitative and spatial imaging study at 7T.
Neurology 85, 17021709 (2015).
Salvetti,M. etal. Progressive MS: from pathophysiology
to drug discovery. Mult. Scler. 21, 13761384 (2015).
Tintore,M. etal. Defining high, medium and low impact
prognostic factors for developing multiple sclerosis.
Brain 138, 18631874 (2015).
De Stefano,N. etal. Long-term assessment of no
evidence of disease activity in relapsingremitting MS.
Neurology 85, 17221723 (2015).
Kalincik,T. etal. Switch to natalizumab
versusfingolimod in active relapsingremitting
multiple sclerosis. Ann. Neurol. 77, 425435

(2015).
Clanet,M.C. etal. Risk evaluation and monitoring
7.
inmultiple sclerosis therapeutics. Mult. Scler. 20,
13061311 (2014).
8.
Marrie,R.A. etal. Effect of comorbidity on mortality
in multiple sclerosis. Neurology 85, 240247 (2015).
9.
Marrie,R.A. etal. Comorbiditydelays diagnosis and
increases disability at diagnosis inMS. Neurology 72,
117124 (2009).
10. Marrie,R.A. etal. A systematic review of the
incidence and prevalence of comorbidity in multiple
sclerosis: overview. Mult. Scler. 21, 263281 (2015).
Acknowledgements
The NMR Unit and its staff are supported by the UK Multiple
Sclerosis Society, National Institute for Health Research
University College London HospitalsUniversity College
London Biomedical Research Centre, the Engineering and
Physical Sciences Research Council, and University College
London internal funding schemes.
A.T. has received honoraria and support for travel for consultancy from Biogen Idec, Eisai, Genzyme, Novartis and
Medday, and for speaking from EXCEMED, Novartis,
Remedica and Teva. He receives an honorarium from Sage
Publications as Editor-in-Chief of Multiple Sclerosis Journal.
O.C. serves as a consultant for Biogen, GE Healthcare and
Novartis, and payments are made to the institution; she
receives an honorarium as Associate Editor of Neurology.
EPILEPSY IN 2015
Classic antiepileptic drugs under

fire, and new options emerge
Christian E. Elger
Diagnosis and management of epilepsy remain challenging, particularly
among women of child-bearing age. In 2015, notable steps were taken in the
right direction, with work that provided insight into the diagnosis of epilepsy,
management of this condition during pregnancy, and new treatment options.
The term epilepsy encompasses a complex
group of disorders that are unified by the
occurrence of epileptic seizures. Variation in
the presentation of seizures is considerable,
however, and many features of seizures are
not recognized as such by patients, particu
larly if these features have not been adequately
explained by their physicians. The question of
how many seizures warrant a diagnosis of epi
lepsy and the initiation of a suitable therapy is
important, as these clinical decisions can have
considerable social consequences, such as the
patients choice of job, whether or not a driv
ers license can be obtained, and difficulties
establishing personal relationships1.
One important study published in 2015
went some way towards addressing ques
tions about the occurrence of seizures prior
to the primary diagnosis. This study included
220patients who were newly diagnosed with
epilepsy, and examined the number of seizures
that these patients had experienced prior to
the index seizure that led to a consultation2.
63 | JANUARY 2016
The delay between prior seizures and the

index seizure, and the underlying causesof
seizures, were also analysed. Almost half
ofthe patients had experienced prior sei
zures within 4weeks (36%), 6months (21%)
or 2years (14%) of the index seizure. One
of themain reasons for a delay in diagnosis
wasthe nature of the previous seizures: con
vulsive episodes were more likely than non
convulsive (nondisruptive) episodes to result
in a consultation. The principle of these find
ings is not new, but the study showed for the
first time that undiagnosed seizures can have
socioeconomic consequences (odds ratios
up to 2.5)2. Furthermore, the work indicates
that studies that focus on first-onset seizures
should differentiate between thosethat truly
represent the first onset and those thatare
simply newly diagnosed.
The causes of epilepsy, like the types of sei
zures, are diverse and partially determine the
choice of antiepileptic drug (AED). For genetic
epilepsies (idiopathic generalized epilepsies)
N E U R O LO G Y
with tonicclonic convulsions, for example,
valproate is often the most effective AED3.
However, other factors influence the choice of
AED, including pregnancya factor that has
caused controversy over the past 12months.
At the end of 2014, the European Medicines
Agency (EMA) published a report advising
that valproate should not be administered to
women who can become pregnant or girls
unless other treatments are ineffective or not
tolerated (see the EMA website). The rec
ommendation was based on studies showing
that valproate can cause fetal malformations
and affect fetal brain development if taken at
a considerable range of doses during preg
nancy4,5. Nevertheless, a large number of
girls and women have idiopathic general
ized epilepsy with tonicclonic seizures and
would, therefore, benefit from treatment
with the cheap and effective valproate. The
fact that valproate is the most effective option
for controlling tonicclonic convulsions in
certain types of epilepsy means that use of
alternative AEDs increases the risk of inju
ries and sudden unexpected death in epilepsy
(SUDEP). The EMA recommendation, there
fore, makes thecounselling of women with
epilepsydifficult.
undiagnosed seizures
can have socioeconomic
consequences
Two important papers published in 2015
addressed the problem of the teratogenic
risks of valproate use during pregnancy. The
first was a prospective observational study
of IQ in children whose mothers had epi
lepsy and received different AEDs during
pregnancy6. 408 children aged 6years were
included. Requirements for educational
interventions were assessed in addition to IQ
so as to provide a real-world measure of the
effects of IQ differences. Children of moth
ers who received >800mg of valproate daily
had significantly lower IQs than did children
of mothers who received carbamazepine or
lamotrigine. In children who were exposed
to <800mg of valproate daily during preg
nancy, IQ was not affected, although four
of 17 children in this group had impaired
verbal abilities that necessitated educational
support. In contrast to an earlier study by
the NEAD (Neurodevelopmental Effects
of Antiepileptic Drugs) group7, this study
revealed no influence of periconceptional
folate on the cognitive development of the
offspring, so the relevance of this intervention
is still undetermined. The authors concluded
Key advances
In almost 50% of patients with epilepsy, theindex seizure that leads to the first consultation is not
the first seizure experienced2
Valproate, a widely used drug for tonicclonic seizures in idiopathic generalized epilepsy, affects
the IQ of children only when used by their mother at a dose of >800mg daily during pregnancy6
The risk of major congenital malformations is relatively low if the daily dose of valproate during
pregnancy does not exceed 700mg8
Perampanel is a powerful new drug for thetreatment of tonicclonic seizures in idiopathic
generalized epilepsy9
The zinc-dependent metal-regulatory transcription factor1 has been identified asa target for
preventing epileptogenesis inexperimental epilepsy10
that for women who might become pregnant,

low-dose valproate (500600mg) is a better
option than switching to other AEDs. This
finding agrees with a recommendation by
the Committee on European Affairs of the
International League Against Epilepsy.
The second paper analysed data from
the European and International Registry on
Antiepileptic Drugs in Pregnancy to assess
the risk of major congenital malformations
(MCMs) in children if their mother used val
proate during pregnancy8. Data from 1,224
women with epilepsy who received valproate
during pregnancy clearly showed thatthe
incidence of MCMs is dose-dependent:
therate was just 5.9% when the daily dose of
valproate was <700mg.
These two studies provide relevant advice
about the use of valproate if other drugs are
not suitable for women who might become
pregnant. In particular, both suggest that
low doses of valproate are relatively safe and
can be used with low risk in women during
pregnancy. In light of these new findings,
a complete ban of valproate for women
with idiop athic generalized epilepsy and
tonicclonic convulsions could be seen as
throwing the baby out with the bath water.
The past year has also seen the arrival of a
long-awaited new treatment option for idio
pathic generalized epilepsy. This form of epi
lepsy needs to be treated with specific AEDs,
such as valproate, lamotrigine, topiramate,
phenobarbital and primidone 3. However,
the availability of a number of these drugs is
limited, and adverse effects are common. The
most recently developed AEDs were intended
primarily for the treatment of focal epilep
sies, but one of these drugsperampanel,
a noncompetitive -amino-3-hydroxy-5-
methyl-4-isoxazolepropionic acid (AMPA)
receptor antagonistwas tested in patients
with drug-resistant idiopathic generalized
epilepsy with tonicclonic convulsions 9.
Perampanel was clearly effective for the
difficult-totreat tonicclonic convulsions:
theproportion of patients who achieved a
>50% reduction in seizure frequency was

64.2% among those who received param
panel and 39.5% among those who received
the placebo. It should be emphasized that the
patients had pharmacoresistant, primarily
generalized, tonicclonic seizures within the
syndrome of idiopathic generalized epilepsy.
Perampanel does not solve the problem of
treating pregnant women with valproate, as
its safety profile in this context is unknown,
and several years of work will be needed to
collect the relevant data. Nevertheless, peram
panel seems to hold the promise of eventually
replacing valproate with a safer drug that has
fewer adverse effects (such as severe weight
gain, polycystic ovarian syndrome, hair loss
and tremors).
One major aim in epileptology is to be
able to control epileptogenesis, the process
by which, following an initial hit, the brain
develops epilepsy that is difficult to treat.
Work published in 2015 has identified the first
clear pharmacological target for an interven
tion in epileptogenesis. Previous studies in
animal models of temporal lobe epilepsy had
revealed that transient upregulation of the
CaV3.2 calcium channel has a role in epilepto
genesis and generates thefeatures of Ammons
horn sclerosis after the initial hit (in this case,
induced status epilepticus). A recent study
in mice revealed that an increase in intra
cellular zinc levels is the basis for this process,
as it increases activity of metal-regulatory
transcription factor 1 (MTF1), which in
turn alters the expression of CaV3.2(REF.10).
Experimental models of epilepsy might
not fully recapitulate the human condition,
although these data were confirmed in a
specimen of a human hippocampus that
had been surgically removed to control sei
zures in pharmacoresistant temporal lobe
epilepsy. Even if the mechanism does trans
late to humans, however, the initial hit that
initiates epileptogenesis could happen at any
time and even go unnoticed, so intervention
during this time-sensitive process is likely to
remain difficult.
JANUARY 2016 | 64

N E U R O LO G Y
In combination, the most prominent
epilepsy studies published in 2015 high
lightedthe socioeconomic consequences of
epilepsy, the importance of diagnosing epi
lepsy at its onset, and managing the condi
tion in this context. In particular, studies that
examined epilepsy management in women
who are of child-bearing age or who wish
to become pregnant have provided reassur
ance that valproate is effective with limited
risks when administered at low doses, despite
use of the drug becoming increasingly con
troversial. Parampanel could offer new pos
sibilities for seizure management, but time
will be needed to gather data about its use in
pregnancy. The ideal solution for managing
epilepsy and minimizing its socioeconomic
consequences would be to prevent epilepto
genesis; experimental data published in this
field of research in 2015 are promising, but
translational aspects remain problematic and
pose a challenge for the future.
Christian E. Elger is in the Department of Epileptology,
University of Bonn, Sigmund-Freud-Strasse 25,
53127Bonn, Germany.
christian.elger@ukb.uni-bonn.de
doi:10.1038/nrneurol.2016.1
1.
de Boer,H.M., Mula,M. & Sander,J.W. The global

burden and stigma of epilepsy. Epilepsy Behav. 12,
540546 (2008).
2. Firkin,A.L. etal. Mind the gap: multiple events and

lengthy delays before presentation with a first
seizure. Epilepsia 56, 15341541 (2015).
3. Marson,A.G. etal. The SANAD study of
effectivenessof valproate, lamotrigine, or topiramate
for generalised and unclassifiable epilepsy: an
unblinded randomised controlled trial. Lancet 369,
10161026 (2007).
4. Campbell,E. etal. Malformation risks of antiepileptic
drug monotherapies in pregnancy: updated results
from the UK and Ireland Epilepsy and Pregnancy
Registers. J.Neurol. Neurosurg. Psychiatry 85,
10291034 (2014).
5. Bromley,R. etal. Treatment for epilepsy in
pregnancy:neurodevelopmental outcomes in the
child. CochraneDatabase Syst. Rev. 10, CD010236
(2014).
6. Baker,G. A. etal. IQ at 6years after inutero
exposure to antiepileptic drugs: a controlled cohort
study. Neurology 84, 382390 (2015).
7. Meador,K.J. etal. Fetal antiepileptic drug exposure
and cognitive outcomes at age 6years (NEAD study):
a prospective observational study. Lancet Neurol. 12,
244252 (2013).
8. Tomson,T. etal. Dose-dependent teratogenicity of
valproate in mono- and polytherapy: an observational
study. Neurology 85, 866872 (2015).
9. French,J.A. etal. Perampanel for tonicclonic
seizures in idiopathic generalized epilepsy:
arandomized trial. Neurology 85, 950957 (2015).
10. van Loo,K. M. etal. Zinc regulates a key
transcriptional pathway for epileptogenesis via metalregulatory transcription factor1. Nat. Commun. 6,
8688 (2015).
FURTHER INFORMATION
European Medicines Agency, Valproate and related
substances: http://www.ema.europa.eu/ema/index.
jsp?curl=pages/medicines/human/referrals/Valproate_and_
related_substances/human_referral_prac_000032.
jsp&mid=WC0b01ac05805c516f
ALL LINKS ARE ACTIVE IN THE ONLINE PDF
N E U R O D E G E N E R AT I V E D I S E A S E I N 2 0 1 5
Targeting tauopathies for

therapeutic translation
Julio C. Rojas and Adam L. Boxer
Tau protein abnormalities are key pathogenic features of Alzheimer disease
and other neurodegenerative diseases. In 2015, new studies of the less
common tauopathies, including progressive supranuclear palsy, chronic
traumatic encephalopathy and frontotemporal lobar degeneration, have
identified invivo biomarkers and mechanisms that initiate tau pathology.
The microtubule-associated protein tau
(encoded by the MAPT gene) regulates
microtubule structure and function in neu
rons1. In2015, research rapidly advanced
the prospect of tau-based therapeutics for
neurodegenerative disorders (FIG.1).
Tau-dependent cytotoxic mechanisms are
pervasive in neurodegenerative disorders,
occurring not only in Alzheimer disease
(AD), but also in conditions such as chronic
traumatic encephalopathy (CTE) and parkin

sonian disorders1. These mechanisms involve
post-translational modifications of tau,
including hyperphosphorylation, cleavage
and aggregation.
Tauopathies are diseases in which insol
uble deposits of tau protein aggregate in
neurons and glia, as shown by autopsy stud
ies of human patients. When animal or cell
culture models are inoculated with brain
65 | JANUARY 2016
tissue from patients with tau pathology, tau

can self-template and spread transcellularly
in a prion-like fashion2. This finding suggests
that the tau protein is an excellent target for
therapeutic molecules that reduce its levels,
alter its post-translational modification, or
block itsspread.
The mechanisms that initiate tau pathol
ogy are a major unresolved question. AD is
the most common tauopathy, but translational
studies designed to elucidate mechanisms of
tau-dependent neurodegeneration are com
plicated by the presence of copathologies
that influence the clinical phenotype. In 2015,
Nagetal. highlighted this challenge by show
ing that the TAR DNA binding protein43
(TDP43) and hippocampal sclerosis make
major contributions to the typical amnestic
AD phenotype3. Autopsy specimens from
636patients3 showed that hippocampal sclero
sis and TDP43 were important contributors
to cognitive impairment. A previous study
from the same authors had demonstrated
that comorbid synucleinand vascular pathol
ogies are common in AD, andalso contribute
to the clinical phenotype4. Together, these
comorbid pathologies could influence clini
cal assessments, making correlations with tau
levels problematic. Initial translational efforts
might, therefore, be more successful in dis
orders with more homogeneous underlying
pathology, such as the primary tauopathies,
progressive supranuclear palsy (PSP) or
autosomal dominant frontotemporal lobar
degeneration (FTLD) associated with MAPT
mutations; in all of these disorders, abnormal
tau is the mainculprit.
Zhao et al. 5 demonstrated the advan
tages of focusing on PSP for understanding
molecular mechanisms of tau-mediated
neurodegeneration. Previous studies have
identified strong genetic risk factors for PSP,
particularly polymorphisms in MAPT itself6,
but the molecular mechanisms underlying
these risk factors were poorly understood.
Zhao and colleagues confirmed that a genetic
variant that increases the risk of PSP is associ
ated with upregulated appoptosin expression
and subsequent high levels of cleaved tau5. In
cell culture models, cleaved tau had a high
tendency to dissociate from microtubules
and aggregate5. Appoptosin also promoted
tau cleavage via caspase3 activation,5 and
its overexpression promoted synaptic abnor
malities and prion-like transcellular spread
of cleaved tau5. When mice received injec
tions of an appoptosin-expressing virus into
the globus pallidus, they developed brady
kinesia and tau pathology similar to that
seen in patients with PSP5. Overexpression of
appoptosin was also observed in the brains
N E U R O LO G Y
ofpatients with AD or FTLD5, providing sup
port for links between the molecular mech
anisms of tau-dependent neurodegeneration
in PSP, AD and FTLD.
Another molecule that regulates tau
structure and function is the peptidyl-prolyl
cistrans isomerase NIMA-interacting 1
(Pin1). PIN1 mutations have previously
been implicated as a genetic risk factor for
AD. Pin1 exerts a neuroprotective effect by
modulating the chirality of tau phosphory
lated at residue 231 (p231 tau), maintaining
it in a trans conformation and preventing the
formation of toxic cis p231 tau (cis ptau), a
process that Kondo etal. termed cistauosis
(REF.7). Kondo and coworkers discovered that
cis ptau was prominent in autopsy specimens
from humans with CTE7, in line with the pre
vious reports that have described fulminant
tau pathology in CTE. The researchers were
also able to induce cistauosis in mouse models
of single or repetitive traumatic brain injury
(TBI), and in cultured neurons by inducing
neuronal stress7. In these models, cistauosis
led to disruption of microtubule assembly,
impaired axonal transport, and spread of cis
ptau to contiguous neurons via a prion-like
mechanism with resulting induction of apop
tosis. Remarkably, this pathogenic cascade
was prevented by monoclonal antibodies to cis
ptau, which decreased cellular neurotoxicity,
histopathological changes and behavioural
deficits in rodent models of TBI7. Similar to
appoptosin-induced tau cleavage, cistauosis
occurred before insoluble tau deposition, and
within days of the induction of of TBI7. The
findings suggest that monoclonal antibodies
that neutralize toxic forms of tau, such as cis
ptau, might be effective therapies.
An intriguing study by Olivera etal.8 sug
gests that tau could spread into the periphery
in individuals with TBI. In military personnel
Genetic
risk factor
rs1768208
T-allele
Protein
Conformational Amplication
change
and spread
Phenotype Biomarker
Appoptosin
PSP
No biomarker
available
AD
Tau PET
Caspase-3
Tau
MAPT H1/H1
haplotype
?
Tau fragments
Tau
?
FTLD-tau Volumetric
MRI
Cis-p-tau
Pin1
PIN1
Trans-p-tau
Nuclear
membrane
Microtubule
CTE
No biomarker
available
TBI
Plasma tau
Neuronal
membrane
Normal tau
Toxic form of tau
Toxic cis-phospha-tau
Phosphate
Figure 1 | Advances in tau research in 2015. The rs1768208 polymorphism,

linkedReviews
to progressive
Nature
| Neurology
supranuclear palsy (PSP), elevates the level of appoptosin and promotes caspase-3mediated tau
cleavage. In PSP and Alzheimer disease (AD), the cleaved tau is thought to spread trans-synaptically,
resulting in the disease5. In AD, tau-PET signal is elevated10. One of the functions of tau, encoded by the
MAPT gene, is to stabilize microtubules. The MAPT H1/H1 haplotype is linked to PSP6, and MRI can
reveal brain atrophy in asymptomatic MAPT mutation carriers9. PIN1 polymorphisms promote toxic
cisptau formation. In animal models, traumatic brain injury (TBI) induces cisptau formation that
leads to neuronal dysfunction. In human patients with TBI, tau pathology is reflected by elevated blood
levels of tau 8. CTE, chronic traumatic encephalopathy; FTLD, frontotemporal lobar degeneration.
who reported having three or more TBIs,

plasma tau concentrations were higher than
in individuals with a single documented TBI,
who in turn had higher plasma tau concentra
tions than controls without TBI. Plasma tau
concentrations correlated with the severity of
post-concussive symptoms. These findings
Key advances
Cistauosis is a conformational change in phosphorylated tau, observed in chronic traumatic
encephalopathy and Alzheimer disease, that contributes to microtubule dysfunction, prion-like
spread of abnormal tau and apoptosis7
In cell culture and animal models, the neuropathological effects of cistauosis can be blocked
byatherapeutic monoclonal antibody against the cis-phospho-tau epitope7
Overexpression of the mitochondrial carrier protein appoptosin, associated with a strong
genetic risk factor for progressive supranuclear palsy (PSP), promotes caspase-mediated tau
cleavage, motor dysfunction and tau neuropathology5
In asymptomatic carriers of mutations linked with frontotemporal dementia, cognitive and
structural MRI changes are detected years before expected disease onset, suggesting the
possibility of clinical prevention trials in MAPT mutation carriers9
Elevated tau levels are measurable in the peripheral blood of individuals with a history
ofrepetitive traumatic brain injury, and correlate with the severity of clinical symptoms8
Pure tauopathies, such as PSP, constitute ideal human patient models for translational studies
oftau, including clinical trials
complement the cistauosis model and sug

gest the possibility of monitoring tau-related
neurodegeneration with a blood test.
If tau cleavage and cistauosis are initiating
factors for tau-mediated neurodegeneration,
intervention with anti-tau antibodies or other
therapies early in the disease course (ideally
before the onset of symptoms) is paramount.
The Genetic FTD Initiative (GENFI) is a
multicentre natural history study of carriers
of autosomal dominant FTLD mutations,
including asymptomatic individuals with
MAPT mutations recruited in the UK, EU
and Canada. In the first report of GENFI
results9, neuropsychological abnormalities
could be detected up to 5years and atrophy
at least 10years prior to the estimated time
of symptom onset. A pattern of sequential
atrophy was demonstrated, with insular and
temporal cortices affected first, followed by
frontal and subcortical areas and, around the
time of symptom onset, parietal and cingu
late cortices. These results suggest that clinical
trials of interventions to prevent the onset of
tau-related neurodegeneration in asympto
matic MAPT carriers are feasible. In North
JANUARY 2016 | 66

N E U R O LO G Y
America, the ARTFL and LEFFTDS projects,
which are similar to GENFI, will further
enable such studies.
Advances in understanding the patho
physiology of tau-dependent neurodegen
eration have sharpened the rationale for
new therapies aimed at the tau protein itself.
Insights from PSP and CTE have suggested
mechanisms that initiate tau pathology, new
targets for tau directed therapeutics and
potential biomarkers to assess therapeutic
effects in humans.
Julio C. Rojas and Adam L. Boxer are at the Clinical
Trials Program, Memory and Aging Center, Department
of Neurology, University of California, San Francisco,
675 Nelson Rising Lane, Suite 190, MC 1207,
SanFrancisco, California 94143, USA.
Correspondence to A.L.B.
adam.boxer@memory.ucsf.edu
doi:10/1038/nrneurol.2016.5
1.
2.
3.
4.
5.
6.
7.
8.
Wang,Y. & Mandelkow,E. Tau in physiology and

pathology. Nat. Rev. Neurosci. 17, 2235 (2015).
Goedert,M.Alzheimers and Parkinsons diseases:
theprion concept in relation to assembled A, tau,
and -synuclein. Science 349, 1255555 (2015).
Nag,S. etal. Hippocampal sclerosis and TDP43
pathology in aging and Alzheimer disease.
Ann.Neurol. 77, 942952 (2015).
Schneider,J.A., Arvanitakis,Z., Leurgans,S.E.
&Bennett,D.A. The neuropathology of probable
Alzheimer disease and mild cognitive impairment.
Ann. Neurol. 66, 200208 (2009).
Zhao,Y. etal. Appoptosin-mediated caspase cleavage
of tau contributes to progressive supranuclear palsy
pathogenesis. Neuron 87, 963975 (2015).
Hoglinger,G.U. etal. Identification of common
variants influencing risk of the tauopathy progressive
supranuclear palsy. Nat. Genet. 43, 699705
(2011).
Kondo,A. etal. Antibody against early driver of
neurodegeneration cis Ptau blocks brain injury and
tauopathy. Nature 523, 431436 (2015).
Olivera,A. etal. Peripheral total tau in military
personnel who sustain traumatic brain injuries during
67 | JANUARY 2016
deployment. JAMA Neurol. 72, 11091116

(2015).
9.
Rohrer,J.D. etal. Presymptomatic cognitive
andneuroanatomical changes in genetic
frontotemporal dementia in the Genetic
Frontotemporal dementia Initiative (GENFI) study:
across-sectional analysis. Lancet Neurol. 14,
253262 (2015).
10. Johnson,K.A. etal. Tau positron emission
tomographic imaging in aging and early Alzheimer
disease. Ann. Neurol. http://dx.doi.org/10.1002/
ana.24546 (2015).
Acknowledgements
J . C . R . a n d A . L . B. a re s u p p o r t e d by N I H ( g ra n t s
U 5 4 N S 0 9 2 0 8 9 a n d R 01 AG 0 3 8 7 91 ) a n d t h e Ta u
Consortium.
A.L.B. has received research support from Avid, Biogen,

Bristol Myers Squibb, C2N Diagnostics, Cortice Biosciences,
Eli Lilly, Forum Pharmaceuticals, Genentech and TauRx. He
has served as a consultant for Asceneuron, Ipierian, Ionis (formerly Isis) Pharmaceuticals, Janssen and Merck. He serves on
a Data and Safety Monitoring Board for Neurogenetics
Pharmaceuticals. He has stock and/or options in Alector and
Delos. J.C.R. declares no competing interests.
Connect to...
For the latest news from the journal and beyond: @NatRevNeurol
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrneurol/info/newsfeeds.html
ahead of print http://www.nature.com/nrneurol/journal/vaop/ncurrent
http://www.nature.com/nrneurol/focus/index.html & http://www.nature.com/nrneurol/collection/index.html
Article series
consecutively over a period of time http://www.nature.com/nrneurol/series
Conferences
You can meet the editors of Nature Reviews Neurology at some of the meetings listed on our conferences page
http://www.nature.com/nrneurol/info/info_conf.html
RHEUMATOLOGY
G L U C O C O RT I C O I D S I N 2 0 1 5
New answers to old problems

Sarah A.Jones and Eric F.Morand
Since glucocorticoids were first used to treat patients with rheumatoid
arthritis in 1949, they have become the most common therapy for
inflammatory disorders; however, their use is associated with major
metabolic adverse events. Here, we review three 2015 reports with
clinicaland fundamental implications for the use of glucocorticoid therapy
inrheumatology.
Fouryears after the end of World War II, Dr
Philip Hench of the Mayo Clinic (Rochester,
Minnesota, USA) proposed to use glucocorticoids to treat rheumatoid arthritis (RA)1.
As glucocorticoids became widely used as
therapy for patients with inflammatory diseases, it became clear that major, predictable,
dose-dependent metabolic adverse events were
associated with their use. The negative consequences of glucocorticoid treatment provided
an impetus for research into glucocorticoids
biology that is still strong today.
Long-term use of glucocorticoids, for
example in systemic lupus erythematosus
(SLE), is known to be particularly harmful. In 2014, Ruiz-Arruza etal.2 reported a
positive and dose-dependent association of
glucocorticoid use with organ damage, measured by using the validated Systemic Lupus
Collaborating Clinics (SLICC)-ACR damage
index in a major study of damage outcomes
in 230 patients with SLE. The association was
maintained even after controlling for disease
activity and patient-related variables. These
detrimental consequences of glucocorticoid
use, consistent with earlier studies3, were
not limited to traditional glucocorticoid
adverseeffects.
Although physicians have long known
about the morbidity burden associated with
glucocorticoid treatment, its potential healthcare costs have not been quantified until
recently: in 2015, Chen etal.4 reported the
largest study performed to date of the determinants of healthcare costs in SLE. Using a US
insurance claims database that included clinical and prescription data from >30 million
patients, >50,000 patients with SLE were monitored for 1year and categorized according to
glucocorticoid use. Importantly, this analysis
included patients from all care settings covered
by the database and not just tertiary centres.
The study showed that glucocorticoid use was
pervasive, with more than half of all patients
exposed to glucocorticoids, and the majorityexposed to high doses. Glucocorticoid use
was associated with severe SLE phenotypes,
including renal disease, and was associated
with greater need for healthcare resources,
including hospitalization and coprescription,
which ultimately result in high costs a relationship that was dependent on glucocorticoid
dose. Although the patients included in this
study were diagnosed with SLE via diagnostic
codes only, and despite the fact that this study
could not discriminate between costly medical
events associated with disease activity versus
those directly attributable to glucocorticoid
use perse4, the relationship between glucocorticoid use and healthcare costs remained
after adjustment for known patient-related
variables. Total costs over 1year for patients
treated with high-dose glucocorticoids were,
at USD45,000 per year, threefold those of
patients not treated with glucocorticoids. An
urgent need to better understand the link
between glucocorticoid use in the context
of SLE and healthcare costs has thus been
identified (FIG.1).
Given the known problems of gluco
corticoid use, the search for metabolically
inert glucocorticoid mimetics with a reduced
health and economic burden has been
described as the holy grail of inflammation
68 | JANUARY 2016
pharmacology 1, but this quest requires a

better understanding of glucocorticoid
mechanisms of action. This area of research
is complicated by the fact that ~1% of the
genome is regulated by glucocorticoids, but
important progress has been made in 2015
with several reports showing that glucocorticoid-induced leucine zipper protein (GILZ,
also known as TSC22 domain family protein3) is protective in multiple autoimmune
disorders. GILZ seems to be one of the main
mediators of the immune effects of glucocorticoids, but functions independently of the
glucocorticoid receptor and does not seem to
replicate its adverse metabolic effects5. GILZ
is protective against inflammation in the gut,
central nervous system and airways, and is an
effective anti-inflammatory agent in experi
mental models of RA5. In 2015, psoriasis
and SLE were added to the list of diseases in
which GILZ has a protective role this protein inhibits IL17 production and T helper
(TH)17 cell differentiation6 and is an important endogenous mechanism in the prevention of spontaneous Bcell activation and
CsA
Glucocorticoids
GILZ
Financial
burden
Healthcare
costs
Morbidity
TH17 B cell Other
Inammatory
disease remission
Figure 1 | Beneficial
anti-inflammatory
effects
Nature
Reviews | Rheumatology
versus harmful metabolic consequences of
glucocorticoids. New data has made clear that
considerable economic cost ensues from the
use of glucocorticoids in diseases such as
systemic lupus erythematosus (SLE), even
though the cost-per-pill of this treatment is
low. The endogenous protein glucocorticoidinduced leucine zipper (GILZ) exerts important
inhibitory effects on Bcell activation and T
helper (TH)17 pathways, suggesting that
GILZisa potential glucocorticoid mimetic.
Ciclosporin (CsA) has been shown to selectively
regulate TH17 pathways even in situations
where glucocorticoids do not, suggesting
theexistence of immune-pathway selective
mechanisms of glucocorticoid resistance
thatmight be targeted directly.
R H E U M AT O L O G Y
Key advances
The healthcare costs associated with
resorting to glucocorticoid use to control
autoimmunity in patients with systemic
lupus erythematosus (SLE) has been defined
as USD45,000 per year three times the
cost for patients with SLE not treated
withglucocorticoids4
In the quest for safe and effective
glucocorticoid alternatives,
glucocorticoid-induced leucin zipper
(GILZ)stands out as a strong contender,
particularly as GILZ has now been shown
tocontrol Bcell survival and activation
inthe context ofSLE7,8
A pathway responsible for glucocorticoid
insensitivity has been identified in T helper
type17 (TH17) cells which could be targeted
in novel therapeutic approaches10
autoantibody production7,8. In their study of

GILZ-mediated regulation of Bcells, Bruscoli
etal. reported a cell-intrinsic Bcell apoptosis defect in the absence of GILZ that led to
accumulation of Bcells in lymphoid tissues
in mice, probably owing to decreased suppression of nuclear factorB (NFB) pathways
and Bcl2 expression7. We independently
observed that GILZ-deficient Bcells showed
dysregulated expansion and participation in
germinal centre reactions after immunization,
proliferating even in conditions of insufficient stimulation that would normally induce
apoptosis8. Importantly, these effects can be
observed in both mice and humans, and treatment with a cell-permeable GILZ fusion protein restored regulation of proliferation and
effector gene expression in both the TH17-cell
and B-cell examples6,8. GILZ-deficient mice
developed spontaneous antibody-mediated
autoimmunity. Furthermore, the fact that
patients with SLE with lower disease activity
had stronger induction of GILZ expression
in myeloid and lymphoid cells in response
to glucocorticoid treatment is particularly
interesting and further supports a protective
role for this protein. If the status of GILZ as
an anti-inflammatory glucocorticoid mimetic
with no metabolic effects is upheld, therapeutic use of this molecule could be pursued as
an alternative to glucocorticoid treatment for
autoimmune disease (FIG.1).
Up to one-third of all patients with
inflammatory disorders require intolerably high doses of glucocorticoids to achieve
disease control, a state referred to as steroid
resistance. TH17 cells, which produce pathogenic IL17 in human autoimmune diseases9,
seem to be at least partially responsible for
this effect. The discovery of IL17 led to a
burst of research focused on the cytokine

itself, the cells that produce it, its modes of
action and its contribution to autoimmune
disease9. New evidence suggests that ciclosporin, a calcineurin inhibitor that has been
used historically to treat glucocorticoidrefractory inflammation, leads to remission
by acting specifically on glucocorticoidresistant TH17 cells10. Schewitz-Bowers etal.
showed that while the glucocorticoid dexamethasone led to only small changes in the
proliferation or gene expression profiles of
TH17 cells (an effect apparently independent
of the glucocorticoid receptor), ciclosporin
had a strong suppressive effect on TH17 cells
specifically, and this effect did not occur in
TH0 cells10. In a model of experimental autoimmune uveitis driven by T H1 and T H17
cells, both dexamethasone and ciclosporin
suppressed disease, but dexamethasone did
so by ablating TH1 cells (not affecting TH17
cells) whereas ciclosporin selectively targeted
TH17 cells and their gene expression, an effect
recapitulated in human TH17 cells polarized
in vitro. Whereas calcineurin inhibition
might be inappropriate for routine use in the
clinic, the study by Schewitz-Bowers etal.
demonstrated that distinct, therapeutically
susceptible biochemical pathways operate
in glucocorticoid-resistant TH17 cells in an
autoimmune setting10. Probing such pathways
not only may reveal targets for new therapies
for the many IL17mediated autoimmune
diseases, but also might improve sensitivity
to glucocorticoids while the search for their
ultimate replacement continues.
Sarah A.Jones and Eric F.Morand are at Monash

University Centre for Inflammatory Diseases, School of
Clinical Sciences at Monash Health, 246 Clayton Road,
Clayton, Melbourne, Victoria 3168, Australia.
Correspondence to E.F.M.
eric.morand@monash.edu
doi:10.1038/nrrheum.2015.176
Hillier,S.G. Diamonds are forever: the cortisone
legacy. J.Endocrinol. 195, 16 (2007).
2. Ruiz-Arruza,I. etal. Glucocorticoids and irreversible
damage in patients with systemic lupus erythematosus.
Rheumatology (Oxford) 53, 14701476 (2014).
3. Thamer,M., Hernan,M.A., Zhang,Y., Cotter,D. &
Petri,M. Prednisone, lupus activity, and permanent
organ damage. J.Rheumatol. 36, 560564 (2009).
4. Chen,S.Y. etal. Glucocorticoid use in patients with
systemic lupus erythematosus: association between
dose and health care utilization and costs.
ArthritisCare Res. (Hoboken) 67, 10861094 (2015).
5. Beaulieu,E. & Morand,E.F. Role of GILZ in immune
regulation, glucocorticoid actions and rheumatoid
arthritis. Nat. Rev. Rheumatol. 7, 340348 (2011).
6. Jones,S.A. etal. GILZ regulates TH17 responses
andrestrains IL17mediated skin inflammation.
J.Autoimmun. 61, 7380 (2015).
7. Bruscoli,S. etal. Lack of glucocorticoid-induced
leucine zipper (GILZ) deregulates Bcell survival and
results in Bcell lymphocytosis in mice. Blood 126,
17901801 (2015).
8. Jones,S.A. etal. Glucocorticoid-induced leucine
zipper (GILZ) inhibits Bcell activation in systemic lupus
erythematosus. Ann. Rheum. Dis. http://dx.doi.org/
10.1136/annrheumdis-2015-207744 (2015).
9. Jones,S.A., Sutton,C.E., Cua,D. & Mills,K.H.
Therapeutic potential of targeting IL17.
Nat.Immunol. 13, 10221025 (2012).
10. Schewitz-Bowers,L.P. etal. Glucocorticoid-resistant
TH17 cells are selectively attenuated by cyclosporine A.
1.
Acknowledgements
The authors wish to thank the Lupus Research Institute, the

National Health and Medical Research Council of Australia
and the Ulysses Club Arthritis Research Fund for funding
theirwork.
S Y S T E M I C L U P U S E RY T H E M AT O S U S I N 2 0 1 5
Cellular and metabolic

requirements of effector T cells
George C.Tsokos
Key advances in lupus research in 2015 highlight the contribution of Tcells
tothe pathogenesis of the disease. The findings not only shed light on
theregulation and activity of these cells, but also suggest several novel
therapeutic targets.
Immune-cell metabolism, the roles of follicular helper T (TFH) cells and the contribution
of Tcells to tissue damage are topics that have
received considerable interest within the field
of lupus research. A number of stellar findings reported in 2015 draw attention to the
ability of myeloid antigen-presenting cells
(APCs) to stimulate the differentiation of

TFH cells and sustain effector Tcells at sites
of inflammation. These findings also demonstrate that activated Tcells in lupus use both
aerobic glycolysis and oxidative phosphorylation to meet their energy requirements
(FIG.1).
JANUARY 2016 | 69

Key advances
Myeloid cells contribute to aberrant
follicular helper T (TFH) cell activity in SLE
viathe OX40LOX40 axis2
Expression of ICOSL in myeloid cells
promotes local inflammation by sustaining
effector Tcells in inflamed tissues4
Aberrant mitochondrial oxidative
phosphorylation in lupus Tcells can be
pharmacologically inhibited to suppress
autoimmunity and nephritis10
TFH cells, a subset of CD4+ T cells, have

been credited with having a major role in the
expression of autoimmunity because they provide expert help to Bcells to form germinal
centres, class-switch and produce high-affinity
antibodies. TFH cells produce IL21 and express
inducible Tcell costimulator (ICOS); the
successful blockade of either of these inflammatory mediators limits autoimmunity and
lupus-like disease. Furthermore, TFH cells seem
to have an additional role in human systemic
lupus erythematosus (SLE): these cells are
expanded in the peripheral blood of patients
with SLE (more so when the disease is active)
and are present in the kidneys of patients
with lupus nephritis. Therefore, TFH cells are
plausible treatmenttargets1.
Jacquemin etal.2 presented evidence that the
OX40 ligand (OX40L)OX40 axis is important
in the generation of TFH cells in both young
patients and adults with SLE. OX40L (also
known as TNF ligand superfamily member
4) was found to be expressed on the surface of
myeloid (CD11c+) APCs in inflamed kidneys
from patients with lupus nephritis and in the
peripheral blood of patients with active SLE.
By contrast, peripheral blood Bcells expressed
minimal, if any, OX40L on their surface. Soluble
OX40L, in the presence of canonical Tcell
stimulation invitro, promoted the expression of TFH cell signature molecules including
CXC-chemokine receptor 5 (CXCR5), Bcell
lymphoma 6 protein (BCL6) and IL21.
Interestingly, the generated TFH cells could support autologous Bcells to produce immunoglobulin. Engagement of the T-cell receptor in
SLE Tcells is known to generate a strong signal1,
and this signal seems to be important in the
generation of TFH cells. Jacquenin etal.2 showed
that engagement of OX40 on TFH cells enhances
this signal further. Lastly, the study showed that
ribonucleoprotein (RNP)anti-RNP immune
complexes, which are abundant in patients with
SLE, stimulate myeloid APCs to express OX40L
through Toll-like receptor7activation.
So, the OX40LOX40 axis seems to be
part of a linear pathogenic course: immune
complexes induce myeloid APCs to express
OX40L that promotes the differentiation of

TFH cells, which lead to antibody production
and homing of Tcells to inflamed tissues.
Given the strong clinical correlations obtained
in cross-sectional patient samples, prospective
studies to determine whether OX40L+ APCs
precede disease activity, or whether they
become part of a grossly dysregulated immune
system, are obviously required. In mice, the
activity of TFH cells is regulated by follicular
regulatory T (TFR) cells3. Although such cells
are thought to operate in humans as well, it is
not known whether the activity of TFH cells is
simply boosted bythe OX40LOX40 axis or
is unchecked by TFR cells; probably both hold
true. The work by Jacquenin etal.2 provides
a rationale for targeting the OX40LOX40
axis with biologic agents in thetreatment
of SLE. Such an approach should be taken
judiciously, firstly by ensuring that other
OX40Lexpressing cells would not be affected
(pertinently, blockade of the CD40LCD40
axis causes activated CD4OLexpressing platelets to aggregate) and secondly by identifying
patients in whom the OX40LOX40 axis is
dominant in the expression of the disease.
The role of myeloid cells in the expression
of lupus pathology was further elaborated in
2015 by Teichman etal.4, who demonstrated
that ICOS ligand (ICOSL) expressed on the
surface of myeloid cells, but not Bcells, promotes tissue inflammation, by promoting
the survival of effector Tcells in inflamed
tissues. For this study, the researchers generated lupus-prone MRLlpr mice with specific
OX40
Immune
complexes
B cell
Autoantibody
production
OX40L
Altered T cell metabolism:

Aerobic glycolysis
Oxidative phosphorylation
Myeloid APC
Other
stimuli
TFH
ablation of ICOSL expression on myeloid

(CD11c+) cells or on Bcells. Inflammation of
the kidneys and of the lungs was reduced only
in MRLlpr mice lacking ICOSL expression in
myeloid cells. Interestingly, skin inflammation was not affected, suggesting that different mechanisms might be involved in the
expression of pathology in various organs
even within individual subjects. A similar
concept was proposed when TNF receptor1
was found to be required for the development
of skin inflammation in the MRLlpr mouse,
but was protective against kidney disease5.
Lack of ICOSL expression on myeloid cells
in MRLlpr mice led to decreased numbers
of CD8+ terminal-effector cells and CD8+
effector-memory cells without affecting
overall numbers of Tcells or other subsets,
including CD3+CD4CD8 cells.
An unexpected finding in this study,
other than the fact that MRLlpr mice lacking
ICOSL in Bcells developed full-blown disease, was that autoantibody titres in the sera
of MRL/lpr mice lacking ICOSL in myeloid
cells remained unchanged. This observation
underscores the separation of autoimmunity
and organ damage a concept demonstrated
previously in NZM2328 lupus-prone mice, in
which replacement of a locus associated with
chronic glomerulonephritis (Cgnz1) with a
region from a normal (wild-type) mouse prevented kidney failure despite the presence of
abundant autoimmunity 6. Tcells are present
in the nephritic kidney; they produce inflammatory cytokines such as IL17 (REFS1,7), and
ICOSL
ICOS
TEFF
TMEM
Tissue
inammation
Inamed tissue:
Kidney
Skin
Figure 1 | Cellular and metabolic requirements for the generation of effector Tcells in patients
with SLE. Immune complexes and other stimuli lead to increased expression
of OX40
ligand
Nature Reviews
| Rheumatology
(OX40L) on myeloid antigen-presenting cells (APCs), which engage OX40 on CD4+ cells to
propagate the generation of follicular helper T (TFH) cells.Myeloid APCs express inducible Tcell
costimulator (ICOS) ligand (ICOSL), which facilitates the maintenance of effector and memory
Tcells at sites of inflammation.These activated cells use both aerobic glycolysis and oxidative
phosphorylation to meet their energy requirements. SLE, systemic lupus erythematosus;
TEFF,effector T cell; TMEM, memory T cell.
70 | JANUARY 2016
clinical research has shown that the magnitude
of the cell infiltrate dictates future kidney function. Teichmann etal.4 show that the mainten
ance of T effector cells in the inflamed kidney
depends on the presence of ICOL-expressing
myeloid cells. The ICOSLICOS axis leads to
the activation of AKT, which, in turn, stimulates the rapamycin-sensitive mechanistic
target of rapamycin (mTOR) complex 1, a
pathway of considerable interest as a target for
the treatment ofSLE8.
Tcells, like all active cells, need energy to
execute their tasks. Stimulation of naive Tcells
introduces a metabolic remodelling process
(aerobic glycolysis), which endows the cells with
the energy required to accomplish their effector
functions. Although glycolysis yields considerably less energy (stored in ATP) than oxidative phosphorylation, the former is preferred
because it generates metabolic intermediates
(precursors for nucleotides and amino acids)
necessary for cell growth and proliferation9.
Yin etal.10 present data demonstrating that
both aerobic glycolysis and mitochondrial oxidative phosphorylation are elevated in Tcells
from lupus-prone mice and from patients with
SLE, as compared with nonautoimmune mice
and healthy controls, respectively. The activation of SLE Tcells (and many of them are
already activated) might have been expected
to promote only aerobic glycolysis, as happens in normal cells. The observed increase in
mitochondrial metabolism is consistent with
previously reported mitochondrial abnormalities in patients with SLE8, which contribute
to aberrant Tcell function. The reason for
this partial diversion to oxidative phosphoryl
ation in SLE Tcells is unclear. In preclinical
studies, Yin etal.10 found that treatment with
2deoxyDglucose (an aerobic glycolysis
inhibitor) and metformin (a mitochondrial
metabolism inhibitor) resulted in suppression
of autoimmunity and nephritis in two different strains of lupus-prone mice (B6.Sle1.Sle2.
Sle3 and NZB/W). Although neither of these
drugs would be used as monotherapy in the
treatment of active SLE, they would probably
serve as valuable adjuvants to minimize the use
of immunosuppressive agents. Metabolomics is
a new field within the study of autoimmunity,
and it is certain that this report will be followed
by manyothers.
From the clinical point of view, the three
reports discussed here present a number of
novel targets for the treatment of SLE: the
OX40LOX40 and ICOSLICOS axes, myeloid
CD11c+ APCs, and two metabolic pathways
active in lupus Tcells (FIG.1). As mentioned in
this commentary, however, the contribution of
these pathways to the expression of disease in
individual patients couldvary.
George C.Tsokos is at the Division of Rheumatology,

Beth Israel Deaconess Medical Centre, Harvard
Medical School, Harvard University, 330 Brookline
Avenue, CLS937 Boston, Massachusetts 02215, USA.
gtsokos@bidmc.harvard.edu
doi:10.1038/nrrheum.2015.178
1.
2.
3.
4.
Moulton,V.R. & Tsokos,G.C. Tcell signaling

abnormalities contribute to aberrant immune cell
function and autoimmunity. J.Clin. Invest. 125,
22202227 (2015).
Jacquemin,C. etal. OX40 ligand contributes to
human lupus pathogenesis by promoting T
follicularhelper response. Immunity 42, 11591170
(2015).
Sage,P.T., Paterson,A.M., Lovitch,S.B. &
Sharpe,A.H. The coinhibitory receptor CTLA4
controls Bcell responses by modulating T follicular
helper, T follicular regulatory, and T regulatory cells.
Immunity 41, 10261039 (2014).
Teichmann,L.L. etal. Local triggering of the ICOS
coreceptor by CD11c+ myeloid cells drives organ
inflammation in lupus. Immunity 42, 552565

(2015).
5. Deng,G.M. & Tsokos,G.C. Pathogenesis and
targeted treatment of skin injury in SLE. Nat. Rev.
Rheumatol. 11, 663669 (2015).
6.
Ge,Y. etal. Cgnz1 allele confers kidney resistance to
damage preventing progression of immune complexmediated acute lupus glomerulonephritis. J.Exp. Med.
210, 23872401 (2013).
7.
Crispin,J.C. etal. Expanded double negative Tcells
inpatients with systemic lupus erythematosus produce
IL17 and infiltrate the kidneys. J.Immunol. 181,
87618766 (2008).
8. Perl,A. Activation of the mechanistic target of
rapamycin in rheumatic diseases. Nat. Rev. Rheumatol.
(in press).
9. Buck,M.D., OSullivan,D. & Pearce,E.L. Tcell
metabolism drives immunity. J.Exp. Med. 212,
13451360 (2015).
10. Yin,Y. etal. Normalization of CD4+ Tcell
metabolismreverses lupus. Sci. Transl. Med. 7,
274ra18 (2015).
P S O R I AT I C A RT H R I T I S I N 2 0 1 5
Advancement continues
in imaging, tight control
and newdrugs
Ignazio Olivieri and Salvatore DAngelo
In 2015, a EULAR task force released evidence-based recommendations
onthe use of imaging in the clinical management of spondyloarthritis,
including psoriatic arthritis. These recommendations, together with
articles dealing with tight control strategies and use of the IL17A inhibitor
secukinumab, have consolidated progress in the management of
psoriaticarthritis.
The progressive improvement in the management of psoriatic arthritis (PsA)1 that began in
earnest at the onset of the new millennium has
continued apace during 2015 (FIG.1). The main
areas of progress include improvements in
treatment strategies, the evolution of outcome
assessment due to the use of validated instruments for several manifestations of psoriatic
disease and composite disease activity indices, and advances in pharmacologic therapy
forPsA1.
The clinical spectrum of PsA is broad, and
the course of the disease is variable. In addition to the skin and nails, PsA affects the joints,
entheses, synovial sheaths of tendons, and axial
skeleton. Currently, PsA is classified as a form
of spondyloarthritis (SpA) alongside ankylosing spondylitis, arthritis related to inflammatory bowel disease, reactive arthritis and
undifferentiated SpA. Patients with SpA can

also be grouped according to whether they
present with predominant axial involvement,
characterized by spondylitis and sacroiliitis, or
prevalent peripheral features such as peripheral
arthritis, enthesitis and dactylitis.
Imaging has a prominent role in the early
diagnosis of PsA and in the management of
patients with the disease. Different imaging
techniques can be used, including conventional radiography, ultrasonography and MRI,
each of which has advantages and limitations.
2015 saw the publication of evidence-based
recommendations on the use of musculoskeletal imaging in the clinical management
of both axial and peripheral SpA, formulated
by a EULAR task force and intended for use
by rheumatologists, radiologists and general
practitioners2.
JANUARY 2016 | 71

Key advances
The EULAR recommendations for the use of
imaging in the diagnosis and treatment
ofspondyloarthritis including psoriatic
arthritis (PsA) are practical and valuable in
daily clinical practice2
Tight control of disease activity using a
treattotarget approach substantially
improves joint outcomes for patients with
early PsA4
Secukinumab was superior to placebo
forpatients with PsA in two studies, and
canbe seen as an additional option to
anti-TNFtreatment9,10
When peripheral involvement is suspected

in a patient with psoriasis or a family history
of psoriatic disease, the EULAR task force recommends that either ultrasonography or MRI
can be used to detect peripheral enthesitis,
peripheral arthritis, dactylitis, tenosynovitis
and bursitis. Ultrasonography, which can be
performed with ease by trained rheumatologists, is more sensitive than clinical examination in detecting involvement of peripheral
structures. MRI, which has received less attention than ultrasonography because of limited
availability and relatively high costs, can also
reveal bone marrow oedema associated with
enthesitis, synovitis and tenosynovitis.
In diagnosing axial involvement in patient
with PsA, conventional radiography is recommended as the initial imaging method for
the detection of sacroiliitis. MRI of the sacroiliac joints is an alternative first-line imaging
method in particular circumstances, such as
young patients or those with a short duration of symptoms, and is also recommended
if axial involvement is suspected but cannot
be confirmed on the basis of clinical findings or pelvic radiographs. Other imaging
modalities such as CT and scintigraphy are
not generally recommended for the diagnosis of axial involvement in patients with SpA
or PsA. The EULAR recommendations also
deal with theuse of imaging in monitoring
inflammation and damage, predicting outcome response to treatment, and detecting
spinal fractures and osteoporosis. Although
it is now well-known that imaging is central
to the management of patients with PsA, its
overuse can be associated with both harms
and high costs. The dissemination of the
EULAR recommendations could help practicing rheumatologists in choosing the right
test, thus avoiding unnecessary imaging (for
example, scintigraphy for the diagnosis of
sacroiliitis), with the aim of optimizing the
use of resources by reducing both direct and
indirect costs.
In the overall approach to managing PsA,

referral, diagnosis, and initiation of therapy now occur earlier in the disease course;
moreover, management strategies incorporate tight-control principles, and international treatment recommendations have
been disseminated1. As far as treatto-target 3
and tight-control strategies are concerned,
the main event in 2015 was publication
ofthe TICOPA (Tight Control of Psoriatic
Arthritis) study 4. The aim of this openlabel, multicentre, randomized controlled
trial was to evaluate whether tight control
with a target of minimal disease activity
(MDA)5 could improve outcomes, in comparison with standard care, in patients with
early PsA (symptom duration <24months)
who had not been previously treated with any
DMARDs. The investigators chose to enrol
newly diagnosed patients because late consultation (>6months from symptom onset) leads
to substantially worse peripheral joint erosions and long-term physical function compared with early referral to and treatment by a
rheumatologist 6. Patients in the tight-control
group (n=101) received objective assessment
of disease activity every 4weeks and protocolled intensive treatment, with escalation of
therapy if MDA was not reached, whereas
patients in the standard-care group (n=105)
were usually evaluated every 3 months
with no formal outcome measure assessment and no set protocol for treatment.
The primary endpoint was the proportion
of patients achieving an ACR20 response

(indicating 20% improvement according to ACR criteria) at 48 weeks postrandomization; key secondary endpoints
included more rigorous outcome measures
such as ACR50, ACR70 and PASI75 (indicating 75% reduction in Psoriasis Area Severity
Index score). The results at 48weeks postrandomization showed significant improvements in articular, skin and patient-reported
outcomes in thetight-control group in comparison withthe standard-care group, suggesting that the new approach using MDA as
treatment target should be implemented in
clinicalpractice.
Pharmacologic therapy for PsA has
evolved considerably in the past 15years1.
TNF inhibitors have been the standard of
care owing to their efficacy for both skin and
musculoskeletal manifestations of the disease;
however, many patients do not respond to or
tolerate these drugs. In the past 2years,the
IL12 and IL23 inhibitor ustekinumab,
thephosphodiesterase4 inhibitor apremilast,
and the IL17A inhibitor secukinumab have
also proven to be beneficial for patients with
psoriatic disease710. IL17A is now thought
to have a key role in the pathogenesis of
PsA. Two multicentre, randomized, placebocontrolled phaseIII studies on the efficacy
of secukinumab in PsA, FUTURE 1 and
FUTURE 2 (REFS 9,10) ,were published in
2015. In FUTURE 1, patients received either
placebo or intravenous loading doses of
Diagnosis
EULAR imaging
recommendations
Treatment
strategies
Tight
control
Advances
in psoriatic
arthritis
in 2015
versus Standard
care
TICOPA study
New drugs
TH17
IL-17A
Secukinumab
Figure 1 | Recent advances in management of psoriatic arthritis. Key publications in 2015 highlight progress in imaging2, treatment strategies4 and drug development9,10 for psoriatic arthritis.
Nature Reviews | Rheumatology
72 | JANUARY 2016
secukinumab (10mg/kg) at weeks 0, 2 and 4
followed by monthly subcutaneous doses of
75mg or 150mg 9. By contrast, FUTURE 2
compared placebo with subcutaneous secukinumab doses of 75mg, 150mg or 300mg,
administered weekly for 4 weeks during
loading then monthly 10. A combined total
of 1,003 patients were enrolled in the two
studies. Both studies met their primary endpoint (the proportion of patients achieving
an ACR20 response) at week 24. In FUTURE
1, secondary endpoints also included ACR50
and ACR70 responses, PASI75 and PASI90
responses, the change from baseline in
DAS28CRP (28joint disease activity score
calculated with Creactive protein level),
the physical component summary score
of the Medical Outcomes Study 36Item
Short-Form Health Survey (SF36), and
Health Assessment Questionnaire Disability
Index (HAQDI) scores. The percentage
of patients with resolution of dactylitis and
enthesitis was higher in the secukinumab
groups than the placebo groups, and progression of joint structural damage was lessened
with secukinumab treatment. Secukinumab
efficacy was sustained through 52weeks and
was seen both in patients who had received
previous anti-TNF treatment and in those had
not, although to a lesser extent in the former
group. Secukinumab treatment was well tolerated, and the safety profile was consistent
with that observed in previous clinical trials in PsA and psoriasis, including a higher
rate of infections (including candidiasis) and
cardiov ascular events with secukinumab
groups than with placebo. In summary, the
FUTURE studies suggest that secukinumab
can be an option for the treatment of patients
with PsA in whom anti-TNF agents are
inappropriate, ineffective or not tolerated.
In conclusion, developments in 2015 in
diagnosis, treatment strategies and new drugs
promise to continue to improve the management of PsA. Prompt dissemination of the
imaging recommendations and tight-control
strategies, together with the forthcoming
commercialization of new drugs, will considerably improve PsA management within the
next fewyears.
Ignazio Olivieri and Salvatore DAngelo are at the
Rheumatology Department of Lucania, San Carlo
Hospital of Potenza and Madonna delle Grazie Hospital
of Matera, Contrada Macchia Romana, Via Potito
Petrone Snc, Potenza 85100, Italy.
Correspondence to I.O.
ignazioolivieri@tiscali.it
doi:10.1038/nrrheum.2015.184
1.
Olivieri,I., DAngelo,S., Palazzi,C. & Padula,A.

Advances in the management of psoriatic arthritis.
Nat. Rev. Rheumatol. 10, 531542 (2014).
2.
3.
4.
5.
6.
7.
Mandl,P. etal. EULAR recommendations for the use

of imaging in the diagnosis and management of
spondyloarthritis in clinical practice. Ann. Rheum. Dis.
74, 13271339 (2015).
Smolen,J.S. etal. Treating spondyloarthritis,
including ankylosing spondylitis and psoriatic arthritis,
to target: recommendations of an international task
force. Ann. Rheum. Dis. 73, 616 (2014).
Coates,L.C. etal. Effect of tight control of inflammation
in early psoriatic arthritis (TICOPA): a UK multicentre,
open-label, randomised controlled trial. Lancet http://
dx.doi.org/10.1016/S0140-6736(15)00347-5 (2015).
Coates,L.C. & Helliwell,P.S. Validation of minimal
disease activity (MDA) criteria for psoriatic arthritis
using interventional trial data. Arthritis Care Res.
(Hoboken) 62, 965969 (2010).
Haroon,M., Gallagher,P. & FitzGerald,O.
Diagnosticdelay of more than 6months contributes
topoor radiographic and functional outcome in
psoriatic arthritis. Ann. Rheum. Dis. 74, 10451050
(2015).
Ritchlin,C. etal. Efficacy and safety of the
antiIL12/23 p40 monoclonal antibody, ustekinumab,
in patients with active psoriatic arthritis despite
conventional non-biological and biological antitumournecrosis factor therapy: 6month and 1year
results of the phase 3, multicentre, double-blind,

placebo-controlled, randomised PSUMMIT 2 trial.
Ann.Rheum. Dis. 73, 990999 (2014).
8.
Kavanaugh,A. etal. Treatment of psoriatic arthritis in
a phase 3 randomized, placebo-controlled trial with
apremilast, an oral phosphodiesterase 4 inhibitor.
Ann. Rheum. Dis. 73, 10201026 (2014).
Mease,P.J. etal. Secukinumab inhibition of
9.
interleukin17A in patients with psoriatic arthritis.
N.Engl. J.Med. 373, 13291339 (2015).
10. McInnes,I.B. etal. Secukinumab, a human antiinterleukin17A monoclonal antibody, in patients with
psoriatic arthritis (FUTURE 2): a randomized, doubleblind, placebo-controlled, phase 3 trial. Lancet 386,
11371146 (2015).
Acknowledgements
The author wishes to thank Enrico Scarano (Radiology

Department, San Carlo Hospital, Potenza), for assistance with
the images for FIG.1.
The authors have received honoraria from Abbvie, BMS,

Celgene, MSD, Novartis, Pfizer, Roche, and UCB to attend
scientific meetings. The authors have received no payment in
the preparation of this manuscript.
I N F L A M M AT I O N I N R H E U M AT O L O G Y I N 2 0 1 5
New tools to tackle

inflammatory arthritis
Charles A.Dinarello and Leo A.B.Joosten
Novel therapies to reduce the progressive and destructive nature of
inflammatory joint disease are always welcome, provided that safety is not
compromised. Among the many 2015 studies in the field, we highlight a
targetable mechanism of uric-acid-induced inflammation in gouty arthritis
and the expected efficacy but unforeseen safety concerns of
antiIL17antibodies.
In addition to being a painful disease, gouty
arthritis is associated with high costs as patients
are hospitalized to reduce the pain. Gout, in
particular gout flare, is primarily mediated
by IL1, and antibodies that block TNF have
been consistently ineffective. Although treatment with canakinumab, which neutralizes
IL1, has been approved in Europe to treat
patients with recurrent flares of gout, the
FDA has not approved canakinumab for this
indication despite its efficacy. This cautious
stance is due to safety concerns related to an
increased incidence of infections in patients
treated with canakinumab, which are of special concern in the aged population of patients
with gout. Furthermore, many patients with
gout also have type2 diabetes mellitus, further increasing health-care costs. Therefore,
an effective and safe treatment for acute gout
flares would be a welcome step to reduce the
overall burden of the disease, particularly as its
incidenceincreases.
In this context, several groups have

reported promising findings in 2015. The
endogenous serine protease inhibitor 1
antitrypsin (AAT) is found in human blood
at concentrations of ~12mg/ml, making it
one of the most abundant proteins in the circulation. AAT purified from human plasma
has been used for >20years to treat patients
with AAT deficiency, a disorder that leads
to lung and pancreatic disease. AsAAT is a
naturally occurring protein, treating AATdeficient patients with human plasmaderived AAT has a remarkable record of
safety in terms of infections and carcinogenesis. AAT has broad anti-inflammatory
properties that can be dependent or independent of its antiprotease activity; protease-independent functions of AAT include
inhibition of caspase1 and suppression of
TNF and IL1 production. In addition,
AAT induces transcription and secretion
of IL1 receptor antagonist (IL1Ra) and
JANUARY 2016 | 73

angiopoietin-related protein 41. In a mouse
model of gouty arthritis 2 , human AAT
reduced joint inflammation and synovial
levels of IL13. As plasma-derived AAT has
been successfully used in AAT-sufficient
humans to treat acute STsegment elevation
myocardial infarction4 and in children with
type1 diabetes, AAT could be used today to
treat acute flares of gout. A drawback of this
approach is that AAT needs to be administered intravenously. Therefore, a novel form
of recombinant AAT was generated by fusing
AAT to the Fc domain of IgG13. The AAT
Fc molecule was 4050 times more effective
than plasma-derived AAT in reducing uratecrystal-induced joint inflammation in this
mouse model 3. Given its higher potency,
AATFc can be administered subcutaneously
an important advantage over plasmaderived AAT when treating patients. Addi
tionally, the risk of viral transmission associated with treating humans with blood-derived
products from other humans is not a concern with recombinant AATFc. The cost
of recombinant AATFc is also considerably lower than that of plasma-derived AAT,
and production is not limited to the pool of
human plasma. For all these reasons, AATFc
seems to be a promising new agent for the
treatment of inflammatory diseases.
Most studies on gout have focused on
inflammation, owing to the involvement
of monosodium urate (MSU) crystals in
the pathogenesis of the disease. However,
whether uric acid itself can have direct
proinflammatory effects remains mostly
unknown. In addition to gout, several other
diseases are associated with elevated serum
uric acid levels, including type2 diabetes,
atherosclerosis, metabolic syndrome and
chronic kidney disease. In a study published
in 2015, Crisan and colleagues revealed
that uric acid primes human blood monocytes to increase production of IL1 5 .
Monocytes from patients with hyperuricaemia and gout release more IL1 when
exposed to MSU crystals or Toll-like receptor 2 (TLR2) agonists, and monocytes from
healthy individuals without elevated serum
uric acid levels also released more IL1
when primed exvivo with clinically relevant
uric acid concentrations. The unexpected
mechanism explaining these observations
involves uric-acid-mediated reduction of
transcription and production of IL1Ra;
as IL1Ra is highly effective at reducing
gout-related inflammation, uric-acidmediated suppression of endogenous IL1Ra
production shifts the balance towards
IL1induced inflammation. In some ways,
this imbalance mediated by uric acid mimics
what takes place in patients with inherited

deficiency of IL1Ra (DIRA) syndrome:
in DIRA, the lack of effective endogenous
IL1Ra results in systemic inflammation,
with a prominent role for IL176. Uric acid
seems to suppress transcription of IL1Ra
by epigenetic reprogramming at the level of
histone methylation5. These studies suggest
a new concept explaining not only how uric
acid increases active IL1Ra levels, but also
how increased uric acid levels might contribute to inflammation in several other chronic
inflammatory diseases in which IL1 plays a
pathogenetic part.
The role of IL17 in inflammation has
been studied in animal models for several years, but attention is now focused on
using IL17 inhibitors to treat human diseases. A human neutralizing antibody to
IL17A (secukinumab) was tested in two
randomized, placebo-controlled trials of
606patients7 and 400patients with psoriatic
arthritis 8; although TNF-blocking agents
have been approved for psoriatic arthritis,
comparative studies showed that targeting
IL17 seems more effective than blocking
TNF with etanercept. Secukinumab doses of
75, 150 or 300mg were administered intra
muscularly every 4weeks, and the primary
end point of the studies was an ACR20
response at 24weeks. Approximately 50% of
patients receiving secukinumab achieved an
ACR20 response at 24weeks (the frequency
was similar in the three treatment groups),
compared to 1520% of patients in the placebo arm. After 52weeks, an ACR20 response
was observed in 60% of patients treated
with the highest doses of secukinumab.
APsoriatic Area and Severity Index (PASI)
score of 90 was achieved in 45% (150mg)
and 49% (75mg) of antibody-treated
patients, compared to 4% of patients in the
placebo arm8.
In many ways, these data were expected
and encouraging. However, an unexpected
increase was observed in the incidence
of strokes and myocardial infarctions in
patients receiving the antibody that was not
observed in patients treated with placebo.
Increased Candida albicans infections were
observed (and expected) in patients treated
with secukinumab, as IL17 is required
for neutrophil-mediated defence against
Candida, adding further concerns regarding the safety record of IL17 neutralization. AntiIL17 therapy is ineffective in
rheumatoid arthritis (RA), and worsened
symptoms in patients with Crohn disease9.
The reasons for the failure of secukinumab
in Crohn disease are still unknown, but
could relate to reduced capacity to control
74 | JANUARY 2016
Key advances
Naturally occurring 1antitrypsin (AAT)
and a synthetic AATIgG fusion protein are
novel therapies for acute flares of gout3
Clinically relevant elevations in levels of
uricacid suppress production of IL1
receptor antagonist, independently of
monosodium urate crystals a unique
pathogenetic mechanism for uric acid in
inflammatorydiseases5
IL17 neutralization is effective in
psoriaticarthritis but deprives patients
ofneutrophil-mediated inflammation,
which is required for defence against
Candidaalbicans7,8
Candida in the gut microbiome. Candida

has a pathogenetic role in Crohn disease,
and neutralization of IL17 could increase
Candida-induced inflammation and possibly
result in invasivecandidiasis10.
Increased Candida infections have also
been reported in patients receiving brodalumab, a human antibody that blocks the
IL17 receptor (and, therefore, targets all
isoforms of IL17). Brodalumab is highly
effective in treating plaque psoriasis, but
reports of increased suicidal ideation in
patients receiving this antibody were unexpected. Depression and suicides do occur
in patients with psoriasis and are thought
to be associated with low self-image as a
result of the disfiguring nature of the skin
lesions. Initially, the company that developed secukinumab (Amgen) attributed these
observations to the disease rather than the
antibody treatment. But if that was the case,
one would expect patients in the placebo
arm to be more likely to experience suicidal
ideation than patients in the treatment arm,
given that the treatment was so successful.
Intriguingly, patients with psoriatic arthritis
treated with a neutralizing monoclonal antibody specific to the IL17 cytokine do not
report increased suicidal thoughts, so why
would neutralization of the IL17 receptor
increase psychological distress and induce
cognitive problems? Cytokine-receptor
blockade might increase the levels of the
cytokine, which might enable binding to its
receptor or to a related receptor. Antibodies
to the IL17 receptor are unlikely to enter
the brain, but as the concentration of IL17
increases in the periphery as a consequence
of receptor blockade, IL17 might progressively enter the brain and elicit an inflammatory response in the central nervous system
via the IL17 receptor; although speculative,
this mechanism would probably affect the
whole class of biologic inhibitors of IL17.
Charles A.Dinarello is at the Department of Medicine,
University of Colorado Denver, Aurora, Colorado
80045, USA; and at the Department of Medicine,
Radboud University Medical Centre, HB 6500
Nijmegen, Netherlands.
Leo A.B.Joosten is at the Department of Medicine,
Radboud University Medical Centre, HB 6500
Nijmegen, Netherlands.
3.
4.
Correspondence to C.A.D.
cdinare333@aol.com
doi:10.1038/nrrheum.2015.180
1.
2.
Frenzel,E. etal. Acute-phase protein 1antitrypsin

a novel regulator of angiopoietin-like protein 4
transcription and secretion. J.Immunol. 192,
53545362 (2014).
Joosten,L.A. etal. Engagement of fatty acids with
Toll-like receptor 2 drives interleukin1 production
5.
6.
7.
viathe ASC/caspase 1 pathway in monosodium urate

monohydrate crystal-induced gouty arthritis.
ArthritisRheum. 62, 32373248 (2010).
Joosten,L.A. etal. -1anti-trypsinFc fusion protein
ameliorates gouty arthritis by reducing release and
extracellular processing of IL1 and by the induction
of endogenous IL1Ra. Ann. Rheum. Dis. http://
dx.doi.org/10.1136/annrheumdis-2014-206966.
Abbate,A. etal. Effects of Prolastin C (plasmaderived1 antitrypsin) on the acute inflammatory
response in patients with STsegment elevation
myocardial infarction (from the VCU- 1RT pilot
study). Am.J.Cardiol. 115, 812 (2015).
Crisan,T.O. etal. Soluble uric acid primes TLRinduced proinflammatory cytokine production
byhuman primary cells via inhibition of IL1Ra.
Ann.Rheum. Dis. http://dx.doi.org/10.1136/
annrheumdis-2014-206564 (2015).
Aksentijevich,I. etal. An autoinflammatory disease
with deficiency of the interleukin-1receptor antagonist.
N.Engl. J.Med. 360, 24262437 (2009).
Mease,P.J. etal. Secukinumab inhibition
ofinterleukin17A in patients with psoriatic
arthritis.N.Engl. J.Med. 373, 13291339

(2015).
8.
McInnes,I.B. etal. Secukinumab, a human antiinterleukin17A monoclonal antibody, in patients with
psoriatic arthritis (FUTURE 2): a randomised, doubleblind, placebo-controlled, Phase 3 trial. Lancet 386,
11371146 (2015).
9.
Hueber,W. etal. Secukinumab, a human antiIL17A
monoclonal antibody, for moderate to severe Crohns
disease: unexpected results of a randomised, doubleblind placebo-controlled trial. Gut 61, 16931700
(2012).
10. Colombel,J.F., Sendid,B., Jouault,T. & Poulain,D.
Secukinumab failure in Crohns disease: the yeast
connection? Gut 62, 800801 (2013).
Acknowledgements
The authors acknowledge support from the NIH (Grant AI

15614 to C.A.D.) and the Interleukin Foundation and the
Dutch Arthritis Association (Grant NR122303 to L.A.B.J.).
JANUARY 2016 | 75

Connect to...
For the latest news from the journal and beyond: @NatRevRheumatol
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrrheum/info/newsfeeds.html
ahead of print http://www.nature.com/nrrheum/journal/vaop/ncurrent
http://www.nature.com/nrrheum/focus/index.html & http://www.nature.com/nrrheum/collections/index.html
Article series
consecutively over a period of time http://www.nature.com/nrrheum/series
Providing visual and audible overviews of hot topics http://www.nature.com/nrrheum/posters & http://www.nature.com/
nrrheum/multimedia
Conferences
You can meet the editors of Nature Reviews Rheumatology at some of the meetings listed on our conferences page
http://www.nature.com/nrrheum/info/info_conf.html
UROLOGY
S E X U A L DY S F U N C T I O N I N 2 0 1 5
Recovering sex drive in women

progress and opportunities
Rossella E.Nappi and Francesca Albani
In 2015, the approval of flibanserin opened a debate about diagnosis and
treatment of female sexual dysfunction. Designing clinical trials with suitable
end points is difficult, but some studies indicate correlations between
hormone levels and low desire. New research demonstrates opportunities
fora better understanding of this multifaceted condition.
The multifaceted nature of sexual desire has
been a hot topic in 2015, mainly because of
the FDA approval of flibanserin, marketed
as Addyi (Sprout Pharmaceutical, USA;
100mg taken orally once daily at bedtime), to
treat acquired generalized hypoactive sexual
desire disorder (HSDD) in premenopausal
women (FDA News Release). Flibanserin
is a multifunctional serotonin 1A receptor
agonist and a serotonin 2A receptor antago
nist, hypothetically targeting brain circuits
that mediate motivation, interest in and
desire for sex 1. Initially developed as an
antidepressant, flibanserin soon became the
candidate molecule to change the history of
sexual medicine in women by mere seren
dipity, similar to sildenafil in the treatment
of erectile dysfunction (ED). Three large,
Fuse/Thinkstock
randomized, placebo-controlled phaseIII

trials, including 3,548 women with HSDD of
whom 2,310 received flibanserin and 1,238
received placebo, indicated that the drug sig
nificantly (both statistically and clinically)
improved the number of satisfactory sexual
events, level of sexual desire and distress in
comparison with placebo after 24weeks and
up to 52weeks oftreatment 2.
The approval of an effective drug for
HSDD is undoubtedly marking a new era
for health care providers (HCPs), who are
finally able to treat a distressful condition
that affects at least 10% of women. However,
a great debate is taking place in the pub
lic arena. Indeed, flibanserin cannot be the
easy fix for a complex condition that has
biomedical, intrapersonal and interpersonal
determinants. The availability of a psycho
active agent to manage HSDD pharmaco
logically might expose HCPs and patients to
the risk of overlooking psychosocial aspects
that should better be addressed by other,
nonpharmacological strategies.
Some studies have shown that psychologi
cal treatments, including cognitive behaviour
therapy and mindfulness meditation training,
are also effective in alleviating low desire, but
adequate clinical trials are currently lacking 3.
Moreover, in the new, fifth edition of the
Diagnostic and Statistical Manual ofMental
Disorders (DSM5), HSDD was combined
with arousal disorder into the single entity
female sexual interest/arousal disorder
(FSIAD), which demonstrates the overlap of
different components of the sexual response
in women and highlights the difficulties in

correlating subjective and objective measures
of desire and arousal4. By contrast, for male
sexual dysfunction, the two DSM5 diagno
ses remain separate, owing to evidence that
men with HSDD show low levels of subjective
arousal, whereas men with ED have low levels
of objective arousal, measurable, for example,
by a change in genital temperature5.
To better understand low desire and poor
arousal in women, it is mandatory to take
into account measures of clinical signifi
cance that encompass both biomedical and
psychosocial domains and to design compre
hensive and effective strategies. Nevertheless,
at the moment, proposing a medication spe
cifically tested for HSDD is appropriate to
relieve distress in women consulting for low
desire, instead of using off-label remedies
that have no rigorous, science-based evi
dence. Data have shown that flibanserin is
effective and well tolerated, when taken as
directed. Even though changes in trial end
points with flibanserin treatment were small
in comparison with placebo (increase in sat
isfactory sexual events by 0.51 per month
and increase of about 0.3 on a retrospec
tive desire scale)2, their clinical relevance in
everyday life is significant and HCPs should
consider p
rescribing the drug based on
individualassessment.
The approval of an effective

drug for HSDD is undoubtedly
marking a new era
A new psychoactive agent with potential for
benefit in this setting is bremelanotide, a syn
thetic analogue of melanocyte-stimulating
hormone, acting as an agonist at melanocortin
receptors. A phaseIIb dose-ranging study in
premenopausal women with HSDD indicated
a significant improvement in clinical out
comes for ondemand subcutaneously self-
injected bremelanotide at a dose of 1.75mg
in comparison with placebo after 12weeks6.
Following these results, a phaseIII trial of the
drug is currently recruiting. The availability
of a psychoactive agent that targets brain cir
cuitries different to flibanserin should help
clarify some of the complex biological system
underlying sexual desire.
JANUARY 2016 | 76

U R O LO G Y
The results confirm the
importance of assessing
personal distress before
labelling a woman as sexually
dysfunctional
Research into the relationship between sex
hormones and sexual function and behaviour
in women currently receives much attention
potentially also driven by the discovery
and approval of flibanserin. The association
between sexual desire and androgens is bio
logically plausible, as levels of both decline
with age and testosterone treatment of post
menopausal women with sexual dysfunction
due to HSDD is effective and safe in the short
term. However, any attempt to directly link
low sexual desire with circulating low andro
gens (total or free testosterone) or androgen
precursors (androstenedione, dehydroepi
androsterone or its sulfate ester) has failed
to identify a minimum concentration that
could be used to diagnose women with sexual
dysfunction related to androgen deficiency.
In addition, the Endocrine Society Clinical
Practice Guideline confirmed the recommen
dations against making a clinical diagnosis of
androgen deficiency in women, unless using
accurate methods for testosterone quantifica
tion, such as mass spectrometry 7. Data corre
lating measurements of androgen levels and
specific signs and symptoms in women are
inconsistent, and monitoring plasma levels
during testosterone treatment with available
assays seems to beunreliable7.
By contrast, the situation in men is quite
different. A study in 2015 established a cut-off
level of <275ng/dl total testosterone measured
on two occasions as associated with decreased
sexual desire and activity 8. This study con
firms the significant positive correlation of
low endogenous total and free testosterone
concentrations with decreased levels of sexual

desire and activity, as well as ED, in sympto
matic men 65years of age8. Importantly, in
a Danish cross-sectional study in 560 healthy
women aged 1965years, sexual desire overall
correlated with free testosterone and andro
stenedione levels9. The researchers stratified
the women into three age groups and by use
of exogenous hormones and found that, in
women aged 2544years who did not use
hormonal contraception, sexual desire corre
lated with levels of total and free testosterone,
androstenedione and dehydroepiandroster
one, whereas, in women aged 4565years, only
androstenedione levels correlated with sexual
desire. In addition, data from the large and eth
nically diverse longitudinal Study of Womens
Health Across the Nation (SWAN) demon
strated a significant association between
total testosterone levels and sexual desire and
arousal, as well as masturbation, which is the
sexual function domain least dependent on
partner status10.
Interestingly, despite having decreased
sexual desire, most women in the SWAN
cohort reported that they were moderately
or extremely sexually satisfied. The results
confirm the importance of assessing personal
distress before labelling a woman as sexually
dysfunctional10. Reinforcing the importance
of taking into account the biopsychosocial
profile of the individual woman, results from
the Danish study indicated a stronger effect
of androgen levels on sexual desire in women
aged 2544years who were more likely to be
in a stable sexual relationship9. By contrast, in
younger and older women, androgens were
less important in determining sexual desire,
owing to a potentially stronger influence of
other factors, such as partner status, body
image, self-esteem, general health or sexual
pain disorders.
In conclusion, the approval of flibanserin
for the treatment of women with low sexual
Key advances
Flibanserin is the first medication approved to treat hypoactive sexual desire disorder (HSDD) in
premenopausal women, after three large phaseIII trials demonstrated significant improvements
in the number of satisfactory sexual events, level of sexual desire and distress versus placebo2
HSDD is a complex condition and prescribing flibanserin should be based on individual
assessment, particularly as psychological treatments might also be effective in alleviating low
desire; however, adequate trials are lacking3
Bremelanotide is another psychoactive agent with potential for benefit in women with HSDD that
indicated improved clinical outcomes in a phaseIIb trial and is set to be tested in a phaseIII trial6
In men aged 65years, cut-off points of testosterone levels were established correlating with
decreased sexual desire, activity and erectile function8, indicating a more direct relationship
between androgen levels and sexual dysfunction compared with women
Using novel methods, two studies in women found some correlation between androgen levels
and sexual desire and arousal, but also confirmed the importance of personal factors when
assessing sexual dysfunction in women9,10
77 | JANUARY 2016
desire in 2015 leads the way to further devel

opments. Forthcoming research into the mul
tiple facets of sexual desire and, if possible,
defining biological determinants will help us
improve clinical research and, ultimately, man
age female sexual dysfunction more effectively
than currently possible.
Rossella E.Nappi is at the Research Center for
Reproductive Medicine, Gynecological Endocrinology
and Menopause, IRCCS S.Matteo Foundation,
Department of Clinical, Surgical, Diagnostic and
Pediatric Sciences, University of Pavia,
Piazzale Golgi2, 27100 Pavia, Italy.
Francesca Albani is at the Unit of Internal Medicine
and Endocrinology, IRCCS Foundation S.Maugeri,
ViaFerrata 8, 27100 Pavia, Italy.
Correspondence to R.E.N.
renappi@tin.it
doi:10.1035/nrurol.2015.314
Stahl,S.M. Mechanism of action of flibanserin,
amultifunctional serotonin agonist and antagonist
(MSAA), in hypoactive sexual desire disorder.
CNSSpectr. 20, 16 (2015).
2. Thorp,J.Jr, Palacios,S., Symons,J., Simon,J.
&Barbour,K. Improving prospects for treating
hypoactive sexual desire disorder (HSDD):
development status of flibanserin. BJOG 121,
13281331 (2014).
3. Pyke,R.E. & Clayton,A.H. Psychological treatment
trials for hypoactive sexual desire disorder: a sexual
medicine critique and perspective. J.Sex. Med. 12,
24512458 (2015).
4. Sungur,M.Z. & Gndz,A.A comparison of
DSMIVTR and DSM5 definitions for sexual
dysfunctions: critiques and challenges. J.Sex. Med.
11, 364373 (2014).
5. Sarin,S., Amsel,R. & Binik,Y.M. How hot is he?
Apsychophysiological and psychosocial examination
of the arousal patterns of sexually functional and
dysfunctional men. J.Sex. Med. 11, 17251740
(2014).
6. Portman,D.J., Edelson,J., Jordan,R., Clayton,A.
&Krychman,M.L. Bremelanotide for hypoactive
sexual desire disorder: analyses from a phase2B
dose-ranging study. Obstet. Gynecol.123, 31S
(2014).
7. Wierman,M.E. etal. Androgen therapy in women:
areappraisal: an Endocrine Society clinical practice
guideline. J.Clin. Endocrinol. Metab. 99, 34893510
(2014).
8. Cunningham,G.R. etal. Association of sex hormones
with sexual function, vitality, and physical function of
symptomatic older men with low testosterone levels at
baseline in the testosterone trials. J.Clin. Endocrinol.
Metab. 100, 11461155 (2015).
9. Whlin-Jacobsen,S. etal. Is there a correlation
between androgens and sexual desire in women?
J.Sex. Med. 12, 358373 (2015).
10. Randolph,J.F.Jr, Zheng,H., Avis,N.E.,
Greendale,G.A. & Harlow,S.D. Masturbation
frequency and sexual function domains are associated
with serum reproductive hormone levels across the
menopausal transition. J.Clin. Endocrinol. Metab.
100, 258266 (2015).
1.
R.E.N. had financial relationships (lecturer, member of advi

sory boards and/or consultant) with Bayer HealthCare,
Boehringer Ingelheim, Ely Lilly, Endoceutics, Gedeon Richter,
HRA Pharma, Merck Sharpe & Dohme, Novo Nordisk, Pfizer,
Procter & Gamble, Shionogi Limited and TEVA Womens
Health. F.A. declares no conflicts of interest.
FURTHER INFORMATION
FDA News Release Addyi approval: http://www.fda.gov/
NewsEvents/Newsroom/PressAnnouncements/ucm458734.htm
U R O LO G Y
understood pathophysiology central and
peripheral disturbances in the processing of
pain and v iscerosensory signals are thought
to beinvolved.
The Multidisciplinary Approach to
the Study of Chronic Pelvic Pain (MAPP)
Research Network was created to study the
pathophysiology and treatment options
for patients with UCPPS (MAPP Research
Network). As part of the Trans-MAPP neuro
imaging study to characterize differences
inbrain activity and pathway connectivity in
patients with UCPPS structural and resting
functional MRI was used at five MAPP net
work sites across the USA. The most recent
report arising from this study is particu
larly interesting. Woodworth etal.4 focused
on microstructural changes in the brainof
patients with UCPPS. Control groups ofthis
study included healthy individuals, but also
patients with irritable bowel syndrome
(acommon gastrointestinal pain disorder)
to strengthen the studys comparisons and
results. The investigators used detrusor
tension imaging and track density imaging
analysis techniques, concluding that patients
with UCPPS have extensive microstructural
changes within white matter areas of the brain
that are responsible for the processing and
integration of sensory information from pel
vic areas. These alterations seemed specific to
patients with UCPPS and were not observed
in patients with irritable bowel syndrome. The
direct clinical implications of these findings
are not yet entirely clear; however, brain plas
ticity could be related to symptom severity
and pain duration, and might be a target for
specific intervention in the future.
Through the MAPP Research Network,
four landmark articles that use extensive
anatomical and functional neuroimaging
to establish a phenotype pattern of UCPPS
B L A D D E R DY S F U N C T I O N I N 2 0 1 5
Novel findings continue to challenge

researchers and clinicians
Rose Khavari and Tim Boone
Despite urologists best efforts to relieve blad

der outlet obstruction, bothersome symptoms
of urinary frequency, urgency and nocturia
often persist. The reasons for this inadequacy
in treatment are unclear, but an emerging
hypothesis might provide an explanation:
maybe pharmacological interventions are
currently given too late, at a point when the
bladder has already undergone substantial
remodelling and drugs are ineffective.
Current management strategies used in
the treatment of hypertension might provide
a model to improve the treatment paradigm
in bladder dysfunction 1. Both heart and
bladder undergo repeated cycles of filling
and contraction and respond to pathological
pressure overload by developing hypertrophy
and fibrosis, eventually resulting in dys
function. In the bladder, as outlet resistance
increases, distinct phases of decompensation
occur, starting with inflammatory responses
and smooth muscle hypertrophy, resulting
in fibrosis2. For hypertension, physicians are
trained to identify symptoms early and begin
therapy promptly in the hope of delaying
life-threatening consequences of long-term
high blood pressure. In patients with bladder
dysfunction, better identifying prognostic
factors related to bladder problems and inter
vening before reaching a point of no return
might similarly be critical to improving
patient outcomes.
MicroRNAs (miRNAs) might be a
resource for the discovery of such new
prognostic factors and could also be tar
gets for therapeutic intervention. A review
by Koek etal.1 published in 2015 summa
rizes common signalling pathways of the
heart and the bladder, with a focus on the
role of miRNAs. These small, noncoding,
single-stranded RNAs regulate gene expres
sion by a post-transcriptional mechanism
and have critical functions in cell prolifera

tion, differentiation and death. In the blad
der and the heart, miRNAs are involved in
various signalling pathways activated in
pathological smooth-muscle hypertrophy,
hypoxia, the reninangiotensin system, pro
tein remodelling and fibrosis. For example,
miR199a5p, which seems to be one of the
key players upregulated in the bladder and
heart during pathological pressure overload,
contributes to increased urothelial perme
ability, smooth-muscle differentiation and
hypert rophy. Another interesting miRNA
is miR29, which is downregulated in the
heart and bladder during pathological pres
sure overload as a fibrotic response, reducing
collagen expression and bladder wall stiffness.
Restoring or inhibiting expression of spe
cific miRNAs could become a therapeutic
strategy. Polymeric nanoparticles that deliver
miRNAs effectively have already been explored
in reducing fibrosis in cystic fibrosis and renal
fibrosis. In addition, miRNAs circulate freely
in the urine and can be collected noninvasively
and in large quantities, possibly enabling their
use as diagnostic or prognostic factors of blad
der dysfunction. Intriguingly, the bladder is
easily accessible for therapeutic drug instilla
tion and could be the ideal site for exploration
of miRNA-based treatmentoptions.
Interestingly, the first miRNA profiling in
the context of lower urinary tract dysfunc
tion, which was already published in 2010,
was performed to study bladder painsyn
drome 3. Interstitial cystitis/bladder pain
syndrome (IC/BPS) is a chronic condition
characterized by pelvic pain associated with
urinary urgency and frequency. Urological
chronic pelvic pain syndrome (UCPPS)
comprises a combination of pelvic disorders,
including IC/BPS and chronic prostatitis/
chronic pelvic pain syndrome, with poorly
Arseniy45/iStock/Thinkstock
2015 has seen advances in various bladder conditions. MicroRNAs might

become therapeutic targets, the MAPP network characterized central
neurological changes in chronic pelvic pain syndrome, and urologists were
reminded to consider cognitive effects of long-term anticholinergics use.
Astudy of obstetric vesicovaginal fistula repair shows how evidence-based
research might improve public health in poorly resourced countries.
JANUARY 2016 | 78

U R O LO G Y
have been published in the past year 47.
Comprehensive evaluation of these results
will enable clinicians to clearly recognize
the implications of these costly, complex and
technically challenging studies, and help to
develop a clear understanding of how to use
these data in patient phenotyping, diagnosis
and management.
Unfortunately, current treatments offer
limited effectiveness with considera ble
adverse effects, even when bladder dysfunc
tion is appropriately evaluated and charac
terized. Antimuscarinic agents are the most
commonly used oral therapeutics for bladder
dysfunction associated with urgency, fre
quency and urge urinary incontinence, but
these drugs have limited effectiveness and
poor patient compliance. Antimuscarinics
(a subcategory of anticholinergics) block the
effect of acetylcholine at muscarinic recep
tors. Drugs with anticholinergic effects are
found in a variety of medications, includ
ing antihistamines, tricyclic antidepressants
andantipsychotics.
Findings published in 2015 indicate that
cumulative exposure to anticholinergics
might lead to cognitive impairment. Gray
et al. 8 reported on the cumulative use of
anticholinergic medication and its association
with the risk of dementia using a population-
based prospective cohort of individuals
65 years of age. The investigators used
pharmacy dispensing data to define cumula
tive anticholinergic exposure as total stand
ardized daily doses (TSDD) prescribed over
10years. Data on 3,434 patients were ana
lysed and showed that 797 participants (23%)
developed dementia (specifically, 637 partici
pants developed Alzheimer disease) during
a mean followup period of 7.3years. The
adjusted hazard ratio for developing demen
tia was 0.92 for 190TSDD and increased
with increasing cumulative anticholinergic
use to a hazard ratio of 1.54 for >1,095TSDD.
The authors concluded that a sustained and
increased risk of cognitive deficits with high

cumulative use of anticholinergics exists, and
suggest that an individual taking, for exam
ple, 5mg oxybutynin daily is at greater risk of
developingdementia.
This provocative comment should raise
concerns among all clinicians managing
patients with overactive bladder. We com
monly prescribe anticholinergics over long
periods of time (for example, in the treatment
of neurogenic bladder) to manage persistent
bladder problems and we need to be aware of
the unintended adverse effects of these medi
cations. Whether these negative effects are
reversible or differ between different classes of
antimuscarinics remains unclear and should
be investigated in large randomized trials.
The final article of 2015 that we would
like to highlight is the study by Barone etal.9,
which has a direct clinical implication. In their
prospective, open-label, controlled trial, the
investigators randomized 524 women with
simple obstetric vesicovaginal fistulas to 7days
versus 14days of catheterization following fis
tula repair. Patients from eight hospitals across
Africa treated by 17 surgeons were included.
Results of 7day bladder catheterization were
not inferior to results of 14day drainage,
with no risk of significantly increasing fistula
breakdown, urinary retention or incontinence
within 3months following the repair. No
standardization of surgical technique, anaes
thesia, resources used or clinical care prior
to randomization was performed, suggesting
that the study outcomes can be generalized
to all simpleobstetric fistula repairs globally.
Many consider obstetric vesicovaginal fistulas
a historical concern restricted to developing
countries, but, in Africa alone, 30,000130,000
new fistulas are estimated to occur following a
prolonged obstructed labour10. These women
labour for days, experience ischaemic injuries
to their pelvis, receive poorly skilled care dur
ing their delivery and are devastated by still
birth and various forms of fistulas; eventually,
Key advances
Similarities in pressure-overload-induced remodelling of the heart and bladder include changes
inmicroRNA expression that influence tissue permeability, stiffness and smooth muscle
differentiation, indicating novel therapeutic targets1
Results from a MAPP network study show that patients with urological chronic pelvic pain
syndrome have extensive microstructural changes within brain white matter regions responsible
for sensory information from pelvic areas4
Findings that high cumulative use of anticholinergic drugs results in an increased risk of cognitive
deficits and dementia should raise concerns among urologists prescribing these medications
long-term for patients with overactive bladder8
A study of obstetric vesicovaginal fistula repair in Africa shows that 7day catheterization is not
inferior to 14day drainage, illustrating how evidence-based research might have a tremendous
effect on public health in areas of poorest resources9
79 | JANUARY 2016
they might be cast out by their families and

communities. Any evidence-based inter
vention to improve fistula repair outcome,
decrease hospital stay, cost of care and patient
discomfort will have a tremendous effect on
public health in the areas with highest rates
of obstetric fistulas and poorest resources.
The evidence provided by this important trial
exemplifies a global approach to better care
and is welcomed with enthusiasm by surgeons
who treatfistulas.
In 2015, areas of new findings in vari
ous bladder-related conditions ranged from
translation of pathogenetic signalling through
pharmacological treatment to surgical
repair. The discoveries continue to challenge
researchers and clinicians alike to expand
our understanding of the diverse set of dis
orders summarized as bladder dysfunction to
provide best patientcare.
Rose Khavari and Tim Boone are at the Houston
Methodist Hospital, Department of Urology, 6560
Fannin, Suite 2100, Houston, Texas 77030, USA.
Correspondence to T.B.
tboone3@houstonmethodist.org
doi:10.1038.2015.303
Koeck,I., Burkhard,F.C. & Monastyrskaya,K.
Activation of common signaling pathways during
remodeling of the heart and the bladder.
Biochem.Pharmacol. http://dx.doi.org/10.1016/
j.bcp.2015.09.012 (2015).
2.
Metcalfe,P.D. etal. Bladder outlet obstruction:
progression from inflammation to fibrosis. BJU Int. 106,
16861694 (2010).
3. Sanchez Freire,V. etal. MicroRNAs may mediate the
down-regulation of neurokinin1 receptor in chronic
bladder pain syndrome. Am. J.Pathol. 176, 288303
(2010).
4.
Woodworth,D. etal. Unique microstructural changes in
the brain associated with urological chronic pelvic pain
syndrome (UCPPS) revealed by diffusion tensor MRI,
super-resolution track density imaging, and statistical
parameter mapping: a MAPP network neuroimaging
study. PLoS ONE 10, e0140250 (2015).
5.
Farmer,M.A. etal. Brain white matter abnormalities
infemale interstitial cystitis/bladder pain syndrome:
aMAPP network neuroimaging study. J.Urol. 194,
118126 (2015).
6.
Kairys,A.E. etal. Increased brain gray matter in
theprimary somatosensory cortex is associated with
increased pain and mood disturbance in patients
withinterstitial cystitis/painful bladder syndrome.
J.Urol. 193, 131137 (2015).
7.
Kilpatrick,L.A. etal. Alterations in resting state
oscillations and connectivity in sensory and motor
networks in women with interstitial cystitis/painful
bladder syndrome. J.Urol. 192, 947955 (2014).
Gray,S.L. etal. Cumulative use of strong anticholinergics
8.
and incident dementia: a prospective cohort study.
JAMA Intern. Med. 175, 401407 (2015).
9.
Barone,M.A. etal. Breakdown of simple female genital
fistula repair after 7day versus 14day postoperative
bladder catheterisation: a randomised, controlled, openlabel, non-inferiority trial. Lancet 386, 5662 (2015).
10. Wall,L.L. Obstetric vesicovaginal fistula as an
international public-health problem. Lancet 368,
12011209 (2006).
1.
Competing interest statement
FURTHER INFORMATION
MAPP Network: http://www.mappnetwork.org/
U R O LO G Y
of an 8mm asymptomatic lower-pole stone.
Patients preferred shockwave lithotripsy under
conscious sedation (45%) over ureteroscopy
(32%) or active surveillance (23%). This study
shows that patients place differing values on
procedural risk versus success. Accordingly, to
facilitate a shared medical decision, assessing
the degree to which risk or success is of higher
Sapan N. Ambani and Khurshid R. Ghani
priority for a patient is important.
In 2015, population-based studies in patients with urinary stones informed us
Although meta-analyses are a summary of
best available evidence, they are only as good
of the changing trends in contemporary stone management and identified
as the quality of the studies assessed. In 2015,
areas for improvement. Although meta-analyses aim to provide the highest
the findings of the UK National Health Service
level of evidence, a randomized controlled trial of medical expulsive therapy double-blind randomized SUSPEND trial
challenged current paradigms and was the defining publication of the year.
were published4, which challenge the find
ings of previous studies and meta-analyses on
theefficacy of medical expulsive therapy in the
Several key contributions to the scientific uro common procedure for the treatment of management of patients with ureteric colic. In
logical literature in 2015 will result in improve stone disease7. The increase in use of ureter the SUSPEND trial, 1,167 patients with ure
ments in the evaluation and management of oscopy was probably caused by the increasing teral stones were randomized to receive either
urinary tract calculi. Population-based studies availability of holmium lasers, in conjunc tamsulosin, nifedipine or placebo. No benefit
demonstrated a change of paradigms in surgi tion with advances in endoscopic equipment of medical expulsive therapy was demon
cal management1, which was also supported and ancillary instrumentation, which have strated, although there was a trend towards sig
by findings of a meta-analysis that assessed helped to improve the efficacy and cost- nificance versus placebo for stones >5mm and
the clinical effectiveness of surgical ther effectiveness of ureteroscopy in comparison stones in the lower ureter in patients receiving
apy for lower-pole stones2. However, more with s hockwavelithotripsy8.
tamsulosin. Per study protocol, stone passage
research is needed to be able to assess the
Although population-based studies enable did not need to be radiographically confirmed;
importance and effect of patient preferences us to appreciate changes in practice patterns, instead, the primary end point was the need
when determining treatment3. Importantly, definitive evidence is often sought to support for surgical intervention within 4weeks4.
one randomized controlled trial reported changing trends. Donaldson and colleagues2
The pragmatic nature of the endpoint in
potentially practice-changing results, making did just that by undertaking a meta-analysis the SUSPEND trial has been both applauded
the trial one of this years most debated studies to better understand the clinical effective and criticized. The absence of rigorous
in urology4. In addition, new guidelines from ness of different treatment modalities in the
the European Association of Urology were management of lower-pole stones. Limited
Key advances
published, which aim to reduce the complexity to seven randomized controlled trials total
of metabolic evaluation in patients with uri ling 691 patients, the authors were able to
Contemporary trends in the surgical
nary stones5. Indeed, medical management conclude what many in the field have long
management of upper tract calculi reflect
achange of paradigms with ureteroscopy
is one aspect of stone care sorely in need of surmised: for lower-pole stones >10mm, per
overtaking shockwave lithotripsy as the
qualityimprovement6.
cutaneous nephrolithotomy and ureteroscopy
main treatment approach1
Continuous technological innovations in have superior stone-free rates compared
Comparison
of clinical effectiveness of
the surgical management of urinary calculi with shockwave lithotripsy. Regarding stone
different
treatment
modalities demonstrates
make it particularly susceptible to changing clearance, percutaneous nephrolithotomy
that percutaneous nephrolithotomy and
practice patterns. Oberlin etal.1 reviewed outperformed both modalities, but ureter
ureteroscopy have superior stone-free rates
case logs submitted to the American Board oscopy was more effective than shockwave
compared with shockwave lithotripsy in
2
of Urology from 2003 to 2012 and found that lithotripsy for stones measuring 1020mm .
lower-pole stones >10mm2
ureteroscopy has now overtaken shockwave In their systematic review, the authors used
Patient appraisal of procedural risk and
lithotripsy as the main modality for managing Cochrane Collaboration methodology and
success differs and assessing the degree
upper urinary tract calculi in the USA. For quality-ofe vidence assessment according
towhich either is of higher priority for a
the young generation of practicing urologists to GRADE, and the study provides the first
patient is important to facilitate shared
whose training included a substantial amount level1a evidence for the surgical management
medical decision making3
of endoscopic surgery these findings might of lower-pole stones. Importantly, however,
The double-blind randomized SUSPEND
be intuitive, but the trend of increased ureter further study conclusions are limited, owing to
trial established that neither tamsulosin nor
oscopy use was also noted among senior urol heterogeneity among trials and other factors;
nifedipine are effective as medical expulsive
therapies in patients with ureteric colic4,
ogists applying for recertification. By contrast, for example, morbidity, retreatment, costs and
challenging current prescription patterns
the usage rates of percutaneous nephrolitho quality of life were not evaluated. Some of these
tomy have remained stable, accounting for outcomes certainly contribute to decision
Of >200,000 patients with 1 anomaly
atinitial urinary evaluation only 16%
45% of all surgeries during this time period1. making and patient preferences surrounding
underwent repeat testing within 6months;
These results support observations made in treatment options. In this regard, Omar etal.3
those seen by a urologist had 24% lower
a population-based study from the Canadian performed an interesting study in which they
odds of repeat testing than those seen in
state of Ontario published in 2014 in which surveyed 100 patients in a stone clinic and
primary care6
ureteroscopy was also noted to be the most enquired about the hypothetical management
S TO N E S I N 2 0 1 5
Changes in stone management

suspending belief for evidence
JANUARY 2016 | 80

U R O LO G Y
drest/iStock/Thinkstock
radiological followup assessment of stone

clearance might concern some, but the lack
of serious clinical consequences for patients,
even after 12weeks, has provided evidence
that challenges the clinical effectiveness of
medical expulsive therapy. The SUSPEND
trial seems to have had sufficient power to
detect a clinical benefit of pharmacological
therapy for lower ureteric stones; however, two
separate randomized trials published around
the same time support the concept that
tamsulosin might offer a modest benefit in
patients with lower ureteric stones measuring
510mm9,10. Although Pickard and colleagues4
conclude that further studies assessing the
efficacy of medical expulsive therapy would
be futile, time will tell whether their trial will
fundamentally change clinical practice in
patients with ureteric colic. In the meantime,
this widely debated study will continue to
challenge physicians to question the voracity
with which these drugs areprescribed.
Finally, the European Association of
Urology (EAU) published updated guide
lines for the metabolic evaluation and recur
rence prevention in patients with kidney
stones5. These guidelines provide a com
prehensive review of the literature and are
much more proscriptive compared with the
American Urological Association (AUA) or
the American College of Physicians guidelines
on the medical management of kidney stones
that were released in 2014. The recent publica
tion of several guidelines is a result of the need
for better patient care in an area that has long
suffered from suboptimal management. This
notion was confirmed by Dauw etal.6 who
conducted a claims-based analysis of >200,000
patients who had at least one abnormal result
at their initial urinary metabolic evaluation.
The authors found that only 16% of patients
had a subsequent repeat evaluation within a
6month period. What is more worrisome
is that patients who saw a urologist had 24%
lower odds of repeat testing than those who
saw a primary carephysician6.
The EAU guidelines are easy to follow and,

if used in conjunction with the AUA guide
lines, should help urologists improve the
medical management of patients with kidney
stones. In contrast to the AUA guidelines, they
offer algorithms based on the stone composi
tion. However, in select situations, particularly
pertaining to uric acid stones, some discord
ance between the guidelines exists. Whereas
the AUA does not recommend routine use
of allopurinol as first-line therapy in patients
with uric acid stones (instead preferring
alkalization of acidic urine with citrate), the
EAU guidelines recommend allopurinol in all
patients with uric acid stones in the setting of
hyperuricosuria >4.0mmol/day5.
In summary, several studies published in
2015 offer important evidence that guides both
the surgical and medical management of uri
nary tract calculi. Future studies will need to
better define the optimal surgical management
of upper urinary tract stones and include addi
tional outcomes, such as patient-reported out
comes, that might influence surgical decision
making. In addition, considering the publica
tion of several conflicting randomized trials
regarding the role of medical expulsive therapy
for ureteric colic, an updated meta-analysis
might provide further insight into the effec
tiveness of these medications. Finally, future
studies evaluating the quality of medical man
agement should assess whether the updated
EAU and AUA guidelines have had an effect
on the most important end point: urinary
stone recurrence.
Sapan N. Ambani is at the Department of Urology,
University of Michigan, 3875Taubman Center, 1500 E.
Medical Center Drive, AnnArbor, Michigan 48109, USA.
Khurshid R. Ghani is at the Division of Endourology and

Stone Disease, Department of Urology, University of
Michigan, NCRC Building 16, 1st Floor, Room 114W,
2800 Plymouth Road, AnnArbor,
Michigan 481092900, USA.
Correspondence to K.R.G.
kghani@med.umich.edu
doi:10.1038/nrurol.2015.290
Oberlin,D.T., Flum,A.S., Bachrach,L.,
Matulewicz,R.S. & Flury,S.C. Contemporary surgical
trends in the management of upper tract calculi.
J.Urol. 193, 880884 (2015).
2. Donaldson,J. F. etal. Systematic review and metaanalysis of the clinical effectiveness of shock wave
lithotripsy, retrograde intrarenal surgery, and
percutaneous nephrolithotomy for lower-pole renal
stones. Eur. Urol. 67, 612616 (2015).
3. Omar,M., Tarplin,S., Brown,R., Sivalingam,S.
&Monga,M. Shared decision making: why do
patients choose ureteroscopy? Urolithiasis
http://dx.doi.org/10.1007/s00240-015-0806-0
(2015).
4.
Pickard,R. etal. Medical expulsive therapy in adults
with ureteric colic: a multicentre, randomised,
placebo-controlled trial. Lancet 386, 341349
(2015).
5.
Skolarikos,A. etal. Metabolic evaluation and
recurrence prevention for urinary stone patients:
EAUguidelines. Eur. Urol. 67, 750763 (2015).
6. Dauw,C. A. etal. Provider variation in the quality of
metabolic stone management. J.Urol. 193, 885890
(2015).
7.
Ordon,M. etal. The surgical management of kidney
stone disease: a population based time series analysis.
J.Urol. 192, 14501456 (2014).
8. Scales,C. D. Jr etal. Comparative effectiveness of
shock wave lithotripsy and ureteroscopy for treating
patients with kidney stones. JAMA Surg. 149,
648653 (2014).
9. Furyk,J. S. etal. Distal ureteric stones and
tamsulosin: a double-blind, placebo-controlled,
randomized, multicenter trial. Ann. Emerg. Med.
http://dx.doi.org/10.1016/j.annemergmed.2015.
06.001 (2015).
10. Sur,R. L. etal. Silodosin to facilitate passage of
ureteral stones: a multi-institutional, randomized,
double-blinded, placebo-controlled trial. Eur. Urol. 67,
959964 (2015).
1.
K.R.G. is a consultant for Boston Scientific and Lumenis.

S.N.A. declares no competing interests.
INFECTION IN 2015
HIV protection with PrEP

implications for controlling other STIs
Douglas S. Krakower and Kenneth H. Mayer
HIV and bacterial sexually transmitted infection incidence remains
persistently high for men who have sex with men. Results of new studies
show that clinicians might be able to address this challenge by providing
oralHIV pre-exposure prophylaxis and by intensifying efforts to diagnose
and treat other sexually transmitted infections.
The rising incidence of sexually transmit
ted infections (STIs) and persistently high
rates of HIV acquisition among men who
have sex with men (MSM) remain impor
tant public health challenges. Several of the
81 | JANUARY 2016
most noteworthy studies published in 2015

address these challenges. An analysis of data
on syphilis and HIV diagnoses among men
found that 1 in 20 HIV-uninfected MSM
diagnosed with syphilis became infected with
U R O LO G Y
One innovative approach to decreasing
the spread of bacterial STIs is for clinicians to
give patients who are diagnosed with an STI
medication to treat sexual partners. One study
tested whether a public health intervention
that promotes partner services and patient-
delivered partner therapy (PDPT) could
decrease rates of chlamydia and gonorrhoea
infections in Washington, USA7. The primary
outcomes for this study were community-level
rates of chlamydia in a sentinel population of
women aged 1425years and gonorrhoea
infections in all women in Washington.
During the 22month duration of the study,
the use of PDPT increased while rates of
chlamydia and gonorrhoea both decreased
by around 10% when adjusted for temporal
trends; however, this decrease was not sta
tistically significant7. Currently, provision of
PDPT can be limited by legal and logistical
barriers, as PDPT is legally permissible in only
37 states in the USA, and its utility in popula
tions of individuals with multiple partners (for
example, some MSM and sex workers) is not
established. However, the promising results
from the PDPT study in Washington suggest
that PDPT should be legalized where it is cur
rently prohibited and that clinicians should be
encouraged to p
rovide PDPT.
Although PDPT represents a potential step
forward in the control of bacterial STIs, con
cerns about increasing antimicrobial resist
ance among bacterial isolates, particularly
gonorrhoea, suggest that substantial chal
lenges in managing these infections remain.
The prevalence of gonococcal isolates with
decreased susceptibility to multiple anti
microbial agents has been increasing, which
restricts the available treatment options
for gonorrhoea. One report documented a
novel strain of Neisseria gonorrhoeae that
has decreased susceptibility to ceftriaxone,
the first-line treatment for gonorrhoea8. This
Key advances
HIV during the subsequent year1. This study

highlights that syphilis and HIV are linked
syndemics for MSM that require innova
tive public health strategies for dual control.
Several clinical trials have demonstrated that
oral antiretroviral pre-exposure prophylaxis
(PrEP) can decrease HIV transmission in
atrisk individuals24. In 2015, new studies
provided evidence that PrEP can reduce HIV
incidence in MSM when provided in clini
cal settings5,6. These real-world PrEP stud
ies found that MSM who used PrEP derived
substantial protection against HIV infection
despite high rates of incident bacterial STIs,
suggesting that PrEP can uncouple the link
between STIs and HIV transmission. These
recent studies raise concerns that PrEP and
other advances in HIV treatment might be
associated with high rates of bacterial STIs,
posing challenges for frontline clinicians.
In 2015, the link between syphilis and HIV
has been quantified: in an analysis ofHIVinci
dence among 2,805 men diagnosed with pri
mary or secondary syphilis in New York City,
USA, between 2000 and 2010, 15% of the
men subsequently acquired HIV, represent
ing an annual incidence of 3.6% for all men
and an incidence of 5.6% among MSM1. HIV
incidence rates were even higher (nearly 8%)
for men diagnosed with another bacterial STI
(chlamydia, gonorrhoea or lymphogranuloma
venereum) after a diagnosis of syphilis. This
study suggests that clinicians who diagnose or
treat MSM with syphilis and other bacterial
STIs should strongly consider offering PrEP
to these patients.
Bacterial STIs can potentiate HIV trans
mission; therefore, an important question is
whether PrEP is effective in MSM when rates
of bacterial STIs are high. Two studies from
2015 provide evidence that PrEP can protect
MSM against HIV acquisition despite high
rates of syphilis and other bacterial STIs. The
PROUD study enrolled 544MSM who pre
sented to sexual health clinics in the UK and
randomized them to initiate daily oral PrEP

immediately or to wait 1year5. Within 1year,
new HIV infections were less frequent among
participants who were offered PrEP than
among those who deferred initiation (3 versus
20 new HIV infections, translating into 86%
efficacy). Owing to these results, the study was
stopped early and all participants were offered
PrEP. None of the three infections among
men who were provided with PrEP immedi
ately were thought to represent breakthrough
infections during PrEP use. One of these men
was thought to have acquired HIV infection
before starting PrEP (that is, had undiagnosed
HIV infection at the start of the study) and
the other two men had self-discontinued PrEP
before their HIV diagnoses. Around 50% of
participants in the PROUD study were diag
nosed with bacterial STIs during the study
duration, including 10% with syphilis and
>30% with rectal STIs5. Another study in over
650MSM in California, USA, who were pre
scribed PrEP as part of their primary care at a
health maintenance organization found that
30% of MSM were diagnosed with at least one
STI after 6months of PrEP use, but none were
diagnosed withHIV6.
These two studies illustrate that PrEP can
be effective when implemented in sexual
health clinics and primary care settings, and
they provide evidence in support of scaling up
PrEP provision by clinicians. A potential chal
lenge to implementing PrEP in care settings is
that many clinicians do not routinely perform
comprehensive HIV risk assessments, so they
might not identify patients who could benefit
from using PrEP. Thus, programmes designed
to train clinicians to provide PrEP should
include guidance on how to conduct routine,
nonjudgmental sexual health assessments that
include discussions of specific sexual behavi
ours (for example, condomless anal sex). Such
training should help clinicians to identify
individuals who might benefit from receiving
PrEP and/or screening for bacterialSTIs.
TonyBaggett/iStock/Thinkstock
HIV and bacterial sexually transmitted infection (STI) incidence is high among men who have sex
with men (MSM), and syphilis and other bacterial STIs increase the risk of HIV acquisition1
Oral HIV pre-exposure prophylaxis (PrEP), the use of antiretroviral drugs by individuals at risk
ofacquiring HIV, can decrease HIV transmission in MSM in clinical settings, even when rates of
bacterial STIs are high5
An innovative approach to decreasing rates of bacterial STIs is patient-delivered partner therapy,
the practice of treating partners of individuals with STIs without direct medical evaluation7
Increasing drug resistance has limited treatment options for Neisseria gonorrhoeae;
solithromycin, a macrolide antibiotic, shows promise as a novel oral treatment for gonorrhoea9
Advances in HIV prevention rely on training clinicians in providing PrEP to MSM and other
individuals at high risk of acquiring HIV, as well as intensive screening, treatment and counselling
regarding bacterial STIs
JANUARY 2016 | 82

U R O LO G Y
strain was isolated from a European woman
traveling in Australia, and it represents the
third strain of N.gonorrhoeae with signifi
cantly decreased susceptibility to ceftriaxone
globally. Other strains have previously been
isolated in Japan and Europe. These multi
resistant gonococcal isolates indicate a need
for additional treatment options for gonor
rhoea. To address this need, a phaseII trial
tested the efficacy and safety of solithromycin,
a novel macrolide antibiotic, for the treat
ment of uncomplicated gonorrhoea9. Single
dosesof solithromycin (either 1000mg or
1200mg) were administered to 46 patients
with cultures positive for gonorrhoea. All
patients were cured, including patients with
pharyngeal or rectal infections. The advan
tages of solithromycin include one-time oral
dosing, invitro activity against Chlamydia
trachomatis and Mycoplasma genitalium (both
of which are common coinfections of gonor
rhoea) and against gonococcal strains resist
ant to azithromycin. In addition, the structure
of solithromycin differs from lactams, so
that it can be used by patients who are aller
gic to cephalosporins. However, frequent
gastrointestinal intolerance was observed9.
A phase III trial comparing the efficacy
of solithromycin with standard firstline
treatment is currentlyunderway.
Finally, a notable publication from 2015
are the updated Sexually Transmitted Diseases
Treatment Guidelines from the Centers of
Disease Control and Prevention, which rep
resent a comprehensive and user-friendly
resource for all clinicians who diagnose or treat
STIs and are freely available10.
In conclusion, 2015 represents a year in
which studies have provided strong evidence
that provision of PrEP to MSM in clinical
settings is effective. Successful translation
of these findings will require clinicians to
become more comfortable and proficient in
discussing sex with their patients. However,
the provision of PrEP must be accompanied
by intensive screening, treatment and counsel
ling regarding bacterial STIs, so that advances
in HIV prevention are not overshadowed by
challenges in the control and management of
bacterial STIs.
Douglas S. Krakower is at the Division of Infectious
Diseases, Beth Israel Deaconess Medical Center,
110Francis Street, WLMOB, Suite GB, Boston,
Massachusetts 02215, USA.
Kenneth H. Mayer is at the FenwayInstitute,
1340Boylston Street, 8th Floor, Boston,
Massachusetts02215, USA.
Correspondence to K.H.M.
khmayer@gmail.com
doi:10.1038/nrurol.2015.282
Pathela,P., Braunstein,S.L., Blank,S., Shepard,C.

&Schillinger,J.A. The high risk of an HIV diagnosis
following a diagnosis of syphilis: a population-level
analysis of New York City men. Clin. Infect. Dis. 61,
281287 (2015).
2. Baeten,J.M. etal. Antiretroviral prophylaxis for
HIVprevention in heterosexual men and women.
N.Engl.J.Med. 367, 399410 (2012).
3. Grant,R.M. etal. Preexposure chemoprophylaxis
forHIV prevention in men who have sex with men.
N.Engl. J.Med. 363, 25872599 (2010).
4. Thigpen,M.C. etal. Antiretroviral preexposure
prophylaxis for heterosexual HIV transmission in
Botswana. N.Engl. J.Med. 367, 423434 (2012).
5. McCormack,S. etal. Pre-exposure prophylaxis to
prevent the acquisition of HIV1 infection (PROUD):
effectiveness results from the pilot phase of a
pragmatic open-label randomised trial. Lancet http://
dx.doi.org/10.1016/S0140-6736(15)00056-2 (2015).
6. Volk,J.E. etal. No new HIV infections with increasing
use of HIV preexposure prophylaxis in a clinical practice
setting. Clin. Infect. Dis. 61, 16011603 (2015).
7. Golden,M.R. etal. Uptake and population-level
impact of expedited partner therapy (EPT) on
Chlamydia trachomatis and Neisseria gonorrhoeae:
1.
the Washington State community-level randomized

trial of EPT. PLoS Med. 12, e1001777 (2015).
8. Lahra,M.M., Ryder,N. & Whiley,D.M. A new
multidrug-resistant strain of Neisseria gonorrhoeae in
Australia. N.Engl. J.Med. 371, 18501851 (2014).
9. Hook,E.W. 3rd etal. A Phase2 trial of oral
solithromycin 1200mg or 1000mg as single-dose oral
therapy for uncomplicated gonorrhea. Clin. Infect. Dis.
61, 10431048 (2015).
10. Centers for DiseaseControl and Prevention. 2015
Sexually Transmitted Diseases Treatment Guidelines
[online], http://www.cdc.gov/std/tg2015/default.htm
(2015).
Acknowledgements
D.S.K. received funding from the National Institute of Mental

Health (K23 MH098795). The authors acknowledge
DrKathy Hsu from the Ratelle STD/HIV Prevention Training
Center, Massachusetts Department of Public Health in
Boston, Massachusetts, USA, for her invaluable suggestions
regarding this manuscript.
D.S.K. and K.H.M. have conducted research with project

support from Gilead Sciences.
BLADDER CANCER IN 2015
Improving indication, technique

andoutcome of radical cystectomy
J. Alfred Witjes
Radical cystectomy is the standard treatment for patients with bladder
cancer, but the prognosis of patients undergoing this procedure has not
changed for decades. Small steps towards improvement include better
selection of high-risk T1 patients, the use of perioperative chemotherapy
and, maybe, robotic cystectomy.
Looking back on the new developments in
the treatment of bladder cancer, if one would
be pessimistic, the conclusion could be that
we have not seen much improvement in dec
ades: no new drugs in intravesical therapy or
systemic therapy, the operations are still the
same albeit maybe with the robot as a new
tool and, most importantly, survival has
not improved. Thus, several critical unmet
needs remain when treating patientswith
bladder cancer. For example, in those
with non-muscle-invasive bladder cancer
(NMIBC) at high risk of progression, we need
a tool to select those patients who should
receive immediate radical surgery instead of
endoscopic tumour resection and intravesical
therapy. This requirement has become even
more urgent in view of the current shortage
of BCG, which is the drug of choice in this
high-risk category of patients. Furthermore,
it has been shown that high-risk patients who
do progress to muscle-invasive bladder can
cer (MIBC) have very high cancer-specific
mortality (CSM). Van den Bosch etal.1 per
formed a systematic review of over 3,000
83 | JANUARY 2016
patients with T1 bladder cancer and found a

progression rate to MIBC of 21% and a sub
sequent CSM of 65%, which is twice as high
as in a stage-matched group of patients with
denovo MIBC (that is, without a history of
NMIBC andprogression).
How can these patients be identified?
Many markers studied and tested in the
past were not good enough to use in clini
cal practice (because they had insufficient
negativeand/or positive predictive power
to exclude and/orpredict invasive cancer,
respectively) or have never been validated in
large studies including different cohorts of
patients. Next-generation sequencing might
give the answer in the near future but, to date,
series are small and results require validation.
A large meta-analysis addressed exactly
the above question of how the selection cri
teria for early cystectomy in patients with
high-grade T1 bladder cancer could be
improved2. Including pooled data from 73
studies and over 15,000 patients, the authors
looked for prognostic factors of disease pro
gression and CSM. They also found a 5year
progression rate of 21% and CSM of 13%,

suggesting that, indeed, 62% of patients who
had progressed died of bladder cancer. An
analysis for prognostic factors was done in
almost 9,000 patients, revealing that depth of
invasion of a T1 tumour into the lamina pro
pria was the most important prognostic fac
tor of both progression and CSM (P0.001)
among other well-known factors, such as
lymphovascular invasion, the presence of car
cinoma insitu, nonuse of BCG, larger tumour
size and older age2. Although this study has
the standard limitations of a meta-analysis
(no individual patient data, heterogeneity
in definitions and reporting among studies),
the results can certainly be used in clinical
practice: we have to use the information that
should be available after thorough pathologi
cal assessment of resected tissue to help us
save lives by deciding in which patient radi
cal surgery instead of intravesical instillation
therapy should be considered.
If radical surgery is chosen, robot-assisted
radical cystectomy (RARC) is an option
next to the gold standard treatment: open
radical cystectomy (ORC). Two key papers
published in 2015 have compared these two
approaches 3,4. Novara etal. 3 performed a
systematic review ofperioperative outcomes
and complicationsof both approaches. They
concluded that operating time for RARC
was 12h longer than for ORC, but that
blood loss was lower, length of stay (LOS)
was 11.5days shorter and low-grade com
plications were fewer for RARC in compari
son with ORC. However, the quality of the
evidence was low. Of the 105 included papers,
three were small randomized controlled trials
Key advances
Depth of T1 tumour invasion into the lamina
propria has emerged as the most important
prognostic factor of disease progression
and cancer-specific survival2, informing
primary treatment decision making
Robot-assisted radical cystectomy (RARC)
has longer operating times than open radical
cystectomy (ORC), but blood loss, length of
stay and low-grade complications are
reduced, according to a systematic review3
A randomized trial comparing RARC and
ORC finds similar results for 90day
complications, length of stay, pathological
outcome and quality of life, but lower blood
loss and higher cost for RARC4
Data from a large randomized trial in patients
with high-risk MIBC show a potential survival
benefit with immediate adjuvant
chemotherapy after radical cystectomy,
supporting results from previous studies6
(RCTs) and for the remaining 102 reports the

level of evidence was 4, as they were predom
inantly small, retrospective and old, possibly
still reflecting the learning curve of the chal
lenging RARC operation. Also, most included
series still used an extracorporeal reconstruc
tion, which leaves room for improvement.
Finally, a correction for baseline characteris
tics was not performed, leaving room for bias
in, for example, patient selection.
The highest level of evidence is pro
vided by the largest RCT in the field to date.
Bochner etal.4 randomized 60 patients to
RARC and 58 to ORC. The authors have tobe
complemented for accomplishing a RCT. As
their primary end point an advantage of
RARC for 90day grade25 complications
was not met (62% versus 66% for RARC
and ORC, respectively), the trial was closed
prematurely after a planned interim analysis.
In this study, blood loss was lower and oper
ating time longer for RARC, but LOS, pathol
ogy outcome and quality of life were similar
for both approaches. The trial also looked
at costs, which were significantly higher for
RARC. These conclusions about blood loss,
operating time, LOS, costs and outcomes
were confirmed by another large consensus
paper5. Alimitation to all these studies is
their low level of evidence. However, look
ing at this from a logical and clinical point
of view, I would assume that, indeed, the
conclusions above hold true. I would expect
that with growing experience differences
between the two approaches will become
small, except for the costs aspect. Having
said this, I support the consensus panel con
clusion: the key variable driving surgical
and postsurgical outcomes is the skill and
experience of the surgeon or surgical team,
regardless of whether or not robot assistance
is used.(REF.5).
My final key paper of 2015 is addressing
the role of adjuvant chemotherapy after radi
cal cystectomy6. Perioperative chemotherapy
results in a survival benefit in the high-risk
subset of patients with MIBC. Guidelines
advise to use neoadjuvant chemotherapy,
which will result in a survival advantage of
around 6%, although results outside trials
indicate lower values7. However, patients are
often unfit for neoadjuvant chemotherapy
or the urologist choses to do surgery first.
Is adjuvant chemotherapy an alternative in
these patients? Guidelines do not support
such a recommendation. A key paper by
Sternberg etal.6 reports a large RCT com
paring immediate versus deferred cisplatin-
based chemotherapy in pT3pT4 or N+M0
patients after radical cystectomy and node
dissection. After a median followup period
lkunl/iStock/Thinkstock
U R O LO G Y
of 7years in 284patients, overall survival

in the immediate chemotherapy group was
53% versus 43% in the deferred group (HR
0.78; P=0.13). Although this paper did not
show a significant advantage of immedi
ate adjuvant chemotherapy and the study
closed prematurely owing to slow accrual
(660 patients were planned), the hazard
ratio of this study is interesting. Svatek etal.8
published a collaborative series of 3,947 cys
tectomies in which 932 patients received
adjuvant chemotherapy. They found a signifi
cant survival advantage with a hazard ratio
of 0.75 in high-risk (T3 or N+) patients.
Similarly, Leow etal.9 reported an updated
meta-analysis (945 patients, nine RCTs) on
adjuvant chemotherapy and found a signifi
cant survival advantage with a hazard ratio of
0.77, with most benefit in N+ patients. Thus,
although the level of evidence of the individ
ual studies is not high, they all found a similar
survival benefit with adjuvant chemotherapy
after radical cystectomy. Overall, periopera
tive chemotherapy should be discussed with
patients undergoing radical cystectomy and
adjuvant chemotherapy seems an alternative.
Although the use of perioperative chemo
therapy is still limited, it has increased to
almost 40% of patients, especially owing to
the increase of neoadjuvant chemotherapy,
even though access to care still influences
usage rates of chemotherapy10.
In conclusion, these new findings in
stratification, surgical and medical manage
ment of bladder cancer represent small but
important steps on the path to improving
outcome in a population of patients in whom
progress has been stagnant for years.
JANUARY 2016 | 84

U R O LO G Y
J. Alfred Witjes is at the Department of Urology,
Radboud University Nijmegen Medical Centre,
POBox9101, 6500 HB Nijmegen, Netherlands.
fred.witjes@radboudumc.nl
doi:10.1038/nrurol.2015.272
Published online 24 Nov 2015
1.
2.
van den Bosch,S. & Witjes,J.A. Long-term cancerspecific survival in patients with high-risk, nonmuscleinvasive bladder cancer and tumour progression:
asystematic review. Eur. Urol. 60, 493500 (2011).
Martin-Doyle,W., Leow,J.J., Orsola,A., Chang,S.L.
& Bellmunt,J. Improving selection criteria for early
cystectomy in high-grade T1 bladder cancer:
ametaanalysis of 15,215 patients. J.Clin. Oncol. 33,
643650 (2015).
3. Novara,G. etal. Systematic review and cumulative

analysis of perioperative outcomes and complications
after robot-assisted radical cystectomy. Eur. Urol. 67,
376401 (2015).
4. Bochner,B.H. etal. Comparing open radical
cystectomy and robot-assisted laparoscopic radical
cystectomy: a randomized clinical trial. Eur. Urol. 67,
10421050 (2015).
5. Wilson,T.G. etal. Best practices in robot-assisted
radical cystectomy and urinary reconstruction:
recommendations of the Pasadena Consensus Panel.
Eur. Urol. 67, 363375 (2015).
6. Sternberg,C.N. etal. Immediate versus deferred
chemotherapy after radical cystectomy in patients
withpT3pT4 or N+M0 urothelial carcinoma of the
bladder (EORTC 30994): an intergroup, open-label,
randomised phase3 trial. Lancet Oncol. 16, 7686
(2015).
85 | JANUARY 2016
7. Zargar,H. etal. Multicenter assessment of

neoadjuvant chemotherapy for muscle-invasive
bladder cancer. Eur. Urol. 67, 241249 (2015).
8. Svatek,R.S. etal. The effectiveness of off-protocol
adjuvant chemotherapy for patients with urothelial
carcinoma of the urinary bladder. Clin. Cancer Res.
16, 44614467 (2010).
9. Leow,J.J. etal. Adjuvant chemotherapy for invasive
bladder cancer: a 2013 updated systematic review
and meta-analysis of randomized trials. Eur. Urol. 66,
4254 (2014).
10. Reardon,Z.D. etal. Trends in the use of perioperative
chemotherapy for localized and locally advanced
muscle-invasive bladder cancer: a sign of changing
tides. Eur. Urol. 67, 165170 (2015).
Connect to...
For the latest news from the journal and beyond: @NatRevUrol
Web feeds
Receive the latest content as an RSS newsfeed http://www.nature.com/nrurol/info/newsfeeds.html
ahead of print http://www.nature.com/nrurol/journal/vaop/ncurrent
http://www.nature.com/nrurol/focus/index.html & http://www.nature.com/nrurol/collection/index.html
Conferences
You can meet the editors of Nature Reviews Urology at some of the meetings listed on our conferences page
http://www.nature.com/nrurol/info/info_conf.html
Primers in your inboxyour link

to the most recent topics covered
in Nature Reviews Disease Primers.
From mechanisms to management

Nature Reviews Disease Primers aims to cover all
diseases across all medical specialties. Modular in nature,
Primer articles describe all aspects of a condition:
epidemiology; disease mechanisms; diagnosis, screening
and prevention; management; and quality of life.
With such a comprehensive article collection for you to
read, our alerts are the perfect solution to stay updated
on the latest Primers published in the journal.
Visit go.nature.com/m72Xm5 to register for the alert
www.nature.com/nrdp

Key Advances in Medicine 2015

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Key Advances in Medicine 2015

Uploaded by

Copyright:

Available Formats

KEY ADVANCES

2016 Macmillan Publishers Limited. All rights reserved

Nature Publishing Group | Medical Portfolio

From cardiology to urology

CONTENT AVAILABLE ONLINE

Key Advances in Medicine

Designed by Laura Marshall

We hope you enjoy reading Nature Reviews Key Advances in Medicine.

oflong-term outcomes after PCI

CARDIAC RESUSCITATION IN 2015 Improving outcomes

after OHCA targeting the layperson

GENETICS OF CVD IN 2015 Using genomic approaches

to identify CVD-causing variants

DYSLIPIDAEMIA IN 2015 Advances in treatment

HEART FAILURE IN 2015 Better results from prevention

OBESITY IN 2015 Advances in managing obesity

than from additional treatment

METASTATIC PROSTATE CANCER IN 2015 The new and

LUNG CANCER IN 2015 Bypassing checkpoints,

the old that is newagain

MELANOMA IN 2015 Immune-checkpoint blockade

durable cancer control

THYROID CANCER IN 2015 Molecular landscape of

thyroid cancer continues to be deciphered

ENDOCRINE DISRUPTORS IN 2015 Epigenetic

HEPATIC GLUCOSE METABOLISM IN 2015 Nutrient and

ISLET-CELL BIOLOGY IN 2015 Understanding secretion,

ageing and death in cells

PCOS IN 2015 New insights into the genetics of

regenerating the digestive system

LIVER FIBROSIS IN 2015 Crucial steps towards

GUT MICROBIOTA IN 2015 Prevotella in the gut:

HCV IN 2015 Advances in hepatitisC research

GASTROINTESTINAL IMAGING IN 2015 Emerging trends

PANCREATIC CANCER IN 2015 Precision medicine in

2016 Macmillan Publishers Limited. All rights reserved

Key Advances in Medicine

PODOCYTE BIOLOGY IN 2015 New insights into

the mechanisms ofpodocyte health

IMMUNE REGULATION OF KIDNEY DISEASE IN 2015

STEM CELLS AND RENAL DEVELOPMENT IN 2015

Updates on immunosuppression in kidney disease

Advances in generating and maintaining nephron

RENAL FIBROSIS IN 2015 Understanding the mechanisms

PARKINSON DISEASE IN 2015 Evolving basic,

STROKE IN 2015 Acute endovascular recanalization

pathological and clinical concepts in PD

NEURO-ONCOLOGY IN 2015 Progress in glioma

metabolic requirements of effector T cells

PSORIATIC ARTHRITIS IN 2015 Advancement continues

to tackle inflammatory arthritis

in women progress and opportunities

BLADDER DYSFUNCTION IN 2015 Novel findings

suspending belief forevidence

MULTIPLE SCLEROSIS IN 2015 Managing the complexity

INFECTION IN 2015 HIV protection with PrEP

EPILEPSY IN 2015 Classic antiepileptic drugs under fire,

BLADDER CANCER IN 2015 Improving indication,

NEURODEGENERATIVE DISEASE IN 2015 Targeting

and new optionsemerge

implications for controlling other STIs

tauopathies for therapeutic translation

Dont miss out on next years

United Kingdom. No part of this publication may be reproduced, stored in a retrieval