co m m e nt a r y
It should be noted that not all of the inter-
actions between the slit diaphragm pro-
teins and the cytoskeleton are regulated
by phosphorylation. For example, nephrin
and NEPH1 associate with CD2AP and
ZO1, respectively, both of which are pro-
teins associating with actin (Figure 1) (for
references, see Johnstone and Holzman2),
and phosphorylation-dependent regula-
tory mechanisms have not been identi-
fied for these interactions thus far. Thus
the nephrin–NEPH1 protein complex
has multiple ways to associate with and
regulate the actin cytoskeleton. Further
studies using cell culture and especially
animal models are necessary to define the
cooperative functions of the slit diaphragm
protein complexes.
Finally, the key task is to define the
molecular mechanisms regulating podo-
cyte actin dynamics during development
of podocyte injury and proteinuria in var-
ious human glomerular diseases. Uchida
et al. analyzed by immunofluorescence
microscopy seven minimal-change neph-
rosis patients for the presence of tyrosine-
phosphorylated nephrin in podocytes
and found that the level of nephrin Y1228
phosphorylation was significantly lower
than in control patients.3 Although the
patient number was quite small, the
result, together with studies on animal
models and cultured cells, supports a
role for the nephrin protein complex not
only as a structural component of the
slit diaphragm structure, but also as an
active signaling scaffold modulating the
structural and functional characteristics
of podocytes. A further challenge will be
to define the molecular mechanisms and
signals from within the podocyte that
regulate ligand engagement and func-
tional behavior of the slit diaphragm
proteins.
ACKNOWLEDGMENTS
The Academy of Finland is acknowledged for
financial support.
REFERENCES
1. Kestilä M, Lenkkeri U, Männikko M et al. Positionally
cloned gene for a novel glomerular protein—
nephrin—is mutated in congenital nephrotic
syndrome. Mol Cell 1998; 1: 575–582.
2. Johnstone DB, Holzman LB. Clinical impact of
research on the podocyte slit diaphragm. Nat Clin
Pract Nephrol 2006; 2: 271–282.
3. Uchida K, Suzuki K, Iwamoto M et al. Decreased
tyrosine phosphorylation of nephrin in rat and
human nephrosis. Kidney Int 2008; 73: 926–932.
Kidney International (2008) 73
4. Jones N, Blasutig IM, Eremina V et al. Nck adaptor 8. Verma R, Kovari I, Soofi A et al. Nephrin ectodomain
proteins link nephrin to the actin cytoskeleton of engagement results in Src kinase activation, nephrin
kidney podocytes. Nature 2006; 440: 818–823. phosphorylation, Nck recruitment, and actin
5. Zhu J, Sun N, Aoudjit L et al. Nephrin mediates actin polymerization. J Clin Invest 2006; 116: 1346–1359.
reorganization via phosphoinositide 3-kinase in 9. Li H, Lemay S, Aoudjit L et al. SRC-family kinase Fyn
podocytes. Kidney Int 2008; 73: 556–566. phosphorylates the cytoplasmic domain of nephrin
6. Li H, Zhu J, Aoudjit L et al. Rat nephrin modulates and modulates its interaction with podocin. J Am
cell morphology via the adaptor protein Nck. Soc Nephrol 2004; 64: 404–413.
Biochem Biophys Res Commun 2006; 349: 310–316. 10. Verma R, Wharram B, Kovari I et al. Fyn binds to
7. Garg P, Verma R, Nihalani D et al. Neph1 cooperates and phosphorylates the kidney slit diaphragm
with nephrin to transduce a signal that induces actin component nephrin. J Biol Chem 2003; 278:
polymerization. Mol Cell Biol 2007; 27: 8698–8712. 20716–20723.
see original article on page 940
The treatment of acute interstitial
nephritis: More data at last
GB Appel1
Acute interstitial nephritis (AIN) is an uncommon form of acute renal
failure that is usually medication related. Although the clinical features
and renal histopathology are well recognized, therapy beyond
discontinuing the offending drug has been a challenge. The use of
corticosteroids, although supported by numerous small retrospective
studies and anecdotal case reports, has been controversial. The study by
González et al., although it has limitations, provides solid support for the
early use of corticosteroids in the treatment of drug-related AIN.
Kidney International (2008) 73, 905–907. doi:10.1038/ki.2008.53
Although the spectrum of acute interstitial The classic clinical triad of rash, fever, and
nephritis (AIN) encompasses many enti- eosinophilia in a patient with acute renal
ties, including sarcoidosis, tubulointerstitial failure, especially if non-oliguric, would
nephritis and uveitis syndrome, lupus, and prompt a search for urinary eosinophils
other autoimmune interstitial nephritides, and a discontinuation of methicillin or
medication-related AIN remains the most any other potential offending medica-
common and clinically relevant form tion.3 Recent studies document that the full
in native kidneys.1–3 Although numer- hypersensitivity triad is not often present,
ous medications have been incriminated, and suspicion of AIN must be present
methicillin and other β-lactam antibiot- with any of these features in a patient with
ics were the prototype offending agents renal failure on suspect medications.1–4
for many years.3 Studies documented an When nonsteroidal anti-inflammatory
epidemiologic relationship between the drugs (NSAIDs) were reported to have
renal lesion and the penicillin, there were unique clinical features, such as onset of
cases with recurrence on rechallenge, and the nephrotic syndrome, in association
remissions of the clinical disease occurred with acute renal failure and AIN, this was
when the offending agent was stopped. rapidly absorbed by clinicians.3 Thus, clini-
cal criteria for medication-induced AIN
1–4Columbia University College of Physicians and have been established for years. Likewise,
Surgeons, New York, New York, USA nephropathologists have become adept at
Correspondence: GB Appel, Department of
diagnosing AIN and even predicting medi-
Medicine, Room 2124 Presbyterian Hospital,
Columbia Medical Center, 622 West 168th St., cations as the etiology of the AIN by noting
New York, New York 10032. not only ‘tubulitis,’ but eosinophilic infil-
E-mail: gba2@columbia.edu trates and at times granulomatous changes
905
co mmentar y
(Figure 1).3,5 The problem for the clinician
has been therapy for this lesion.
In some anecdotal cases and brief series,
addition of high-dose corticosteroids has
led to dramatic improvement in the acute
renal failure and return of renal function to
almost baseline.2,3 However, skeptics can
show equally impressive results in anecdo-
tal cases and small series with just discon-
tinuation of the medication alone.2–4 The
debate is not trivial, as patients are often
severely ill with their underlying infec-
tious and other medical conditions as well
as the superimposed renal failure. Adding
the potential risk of immunosuppression
with corticosteroids is a major clinical deci-
sion far greater than just discontinuing or
replacing the potential offending drug.
If there were solid data from controlled
randomized trials, the answer to two cru-
cial questions might be apparent. The first
question deals with speed and complete-
ness of clinical recovery from the acute
renal failure. Prolonged renal failure surely
increases risks for morbidity and mortal-
ity.3,7 The second question deals with the
pathogenesis of the lesions. At what stage
does the inflammatory infiltrate lead to
interstitial fibrosis and irreversible dam-
age?7,8 Although not always apparent
immediately after the episode of acute
renal failure resolves, this residual dam-
age may relate to chronic kidney disease,
Figure 1 | Low-power micrograph (hematoxylin and eosin) showing inflammatory interstitial
infiltrate in medication-induced acute interstitial nephritis. Tubular architecture is focally
obliterated by infiltrate.
906
hypertension, and other adverse effects
years later. The question, then, is why
there are no large controlled randomized
trials to guide us here.
The answer is multifactorial. First, AIN
is a relatively uncommon form of acute
renal failure. Only a small minority of
patients with acute renal failure come to
renal biopsy, and even here, only a small
proportion of biopsied patients have AIN.
No single center can easily perform the
large randomized trial necessary to give
us the answers to those pressing ques-
tions. Second, many cases of AIN never
come to biopsy. The diagnosis seems firm
on clinical grounds, the offending agent
is stopped, and the patient recovers from
the acute renal failure. How many of these
patients truly would have AIN on biopsy
is never known. Even in clear clinical cases
the offending drug may not be clear, with
multiple potential agents that may cause
AIN discontinued at the same time. A third
problem is the concept of randomization to
different therapeutic strategies. It is easier
for clinicians to randomize patients to two
active currently used treatment regimens
for lupus nephritis, than to delegate their
patients to placebo versus corticosteroids.
There are also few reports of alternate ther-
apies for AIN other than corticosteroids.9
Some clinicians already have preconceived
thoughts on whether corticosteroids are
the optimal therapy for a given patient at
a given time and do not feel it is ethical to
use placebo in one arm of a study. Finally,
how is long-term benefit to be proved?
Only rarely is a group of AIN patients fol-
lowed long enough to check for changes in
residual renal function. Even less common
is the use of repeat biopsy to check for the
development of renal interstitial fibrosis.
Such a biopsy would rarely contribute to a
change in therapy of an individual patient
but might provide tremendously valuable
information for therapeutic intervention
for others in the future.
Given these outstanding questions and
problems with the ideal study design, the
clinician still needs additional data to treat
his or her patient appropriately. The study
by González et al.10 (this issue) provides
some of those data. This is a multicenter
retrospective study of the influence of
corticosteroids in 61 patients with biopsy-
proven AIN. As in many series, antibiotics
and NSAIDs were the two primary offend-
ing classes of drugs, comprising 93% of the
patients. The majority of patients (52/61)
received corticosteroids and experienced
a significantly lower serum creatinine as
well as recovery and ability to discontinue
dialysis. Of great interest is that the cor-
ticosteroid-treated patients who showed
a complete return to baseline serum cre-
atinine, although similar at baseline to
those who did not fully recover, differed
in time of onset of corticosteroid therapy.
An earlier onset of use of corticosteroids
after discontinuing the offending drug (13
versus 34 days) was associated with a bet-
ter recovery of renal function. The etiology
of the medication-related AIN (antibiotic
versus NSAID) did not appear to influence
the outcome.
This study provides valuable informa-
tion but is far from conclusive. It is still
small in numbers and retrospective in
nature. What proportion of all clinically
diagnosed AIN with acute renal failure
it applies to is unknown. Does it apply to
drug classes other than antibiotics and
NSAIDs, such as the proton pump inhibi-
tors? Treatment with corticosteroids was
not uniform as regards dose and duration
of therapy. Repeat biopsies were done in
only a handful of patients. Nevertheless,
a clear message is available to the clini-
cian. At the time of discontinuation of the
Kidney International (2008) 73
 co m m e nt a r y

putative offending medication, the use of
corticosteroids should be considered in
all patients with AIN. This consideration
should take into account not only rapidity
and completeness of return of renal func-
tion to normal but also potential long-term
benefits in avoiding interstitial fibrosis and
eventual chronic kidney disease.
REFERENCES
1. Baker RJ, Pusey CD. The changing profile of
acute tubulointerstitial nephritis. Nephrol Dial
Transplant 2004; 19: 8–11.
2. Rossert J. Drug-induced acute interstitial
nephritis. Kidney Int 2001; 60: 804–817.
3. Bhatt P, Appel GB. Tubulo-interstitial diseases.
In: ACP Medicine. WebMD Inc.: New York, 2006,
2027–2043.
4. Clarkson MR, Giblin L, O’Connell FP et al. Acute
interstitial nephritis: clinical features and
response to corticosteroid therapy. Nephrol Dial
see original article on page 933
Transplant 2004; 19: 2778–2783.
5. Joss N, Morris S, Young B, Geddes C.
Granulomatous interstitial nephritis. Clin J Am Soc
Nephrol 2007; 2: 222–230.
6. Buysen JG, Houthoff HJ, Krediet RT, Arisz L. Acute
interstitial nephritis: a clinical and morphological
study in 27 patients. Nephrol Dial Transplant 1990;
5: 94–99.
7. Schwarz A, Krause PH, Kunzendorf U et al. The
outcome of acute interstitial nephritis: risk factors
for the transition from acute to chronic interstitial
nephritis. Clin Nephrol 2000; 54: 179–190.
8. Bhaumik SK, Kher V, Arora P et al. Evaluation of
clinical and histological prognostic markers in
drug-induced acute interstitial nephritis. Ren Fail
1996; 18: 97–104.
9. Preddie DC, Markowitz GS, Radhakrishnan J
et al. Mycophenolate mofetil for the treatment of
interstitial nephritis. Clin J Am Soc Nephrol 2006;
1: 718–722.
10. González E, Gutiérrez E, Galeano C et al. Early
steroid treatment improves the recovery of renal
function in patients with drug-induced acute
interstitial nephritis. Kidney Int 2008; 73: 940–946.
System show that the proportion of sud-
den deaths in hemodialysis patients was
significantly higher on Mondays and
Tuesdays compared with other days.5
No other risk factors were identified.
In a separate study, the same authors
found a threefold increased risk of sud-
den cardiac arrest in the 12 hours before
hemodialysis at the end of the weekend
interval (Figure 1).6 The exact reasons
for this phenomenon are not clear, but
accumulation of fluid and electrolytes
may potentially play a role.
Cardiac arrests in the dialysis unit are
relatively rare events but carry a poor
prognosis. Karnik et al. showed a fre-
quency of 7 cardiac arrests per 100,000
hemodialysis sessions. 7 Affected
patients were older, more likely to have
diabetes, and more likely to dialyze via
a catheter than the general hemodialysis
population. Eighty-one percent of car-
diac arrests occurred while the patient

Cardiac arrests in hemodialysis was on dialysis. The vast majority of

patients had no overt symptoms prior

patients: An ongoing challenge to the event. An abnormal rhythm was

documented in only 17% of patients,
with ventricular fibrillation/tachycar-
M Ostermann1 dia being the predominant one (62%).
Again, cardiac arrests were more fre-
Hemodialysis patients have significant cardiac-related mortality. quent on Mondays than on other days. It
Sudden cardiac arrests in the dialysis unit are infrequent events was also noted that 37% of patients had
but carry a poor prognosis. The predominant rhythm is ventricular been hospitalized within the previous 30
tachycardia/fibrillation. Although the exact etiologies are not clear, days, but there were no data on recent
several studies have confirmed an increased incidence on the first day laboratory results, changes in medica-
tion, or reason for hospital admission.
after the weekend interval. Use of cardioprotective drugs and possibly Outcome was poor: 60% of patients died
an implantable cardioverter defibrillator may improve the prognosis of within 48 hours of the arrest.
survivors after a cardiac arrest. More research is needed in this field. Pun et al. showed that traditional
Kidney International (2008) 73, 907–908. doi:10.1038/ki.2008.40 risk factors, including cardiovascular
comorbidities, diabetes, hemoglobin,
and dialysis adequacy, did not predict
Dialysis patients have an incidence sudden death accounted for 25% of the prognosis in hemodialysis patients after
of cardiac-related death 10–20 times observed total mortality. Analysis of the a cardiac arrest.8 Only use of β-blockers,
higher than that of the general popu- United States Renal Data System data- calcium-channel blockers, angiotensin-
lation. 1 They are particularly vulner- base showed similar results: between converting enzyme inhibitors, and angio-
able to a sudden cardiac arrest. In the 2001 and 2003, 32% of all deaths tensin receptor blockers at the time of
Hemodialysis (HEMO) Study2 and Die among hemodialysis patients were due the event was significantly associated
Deutsche Diabetes Dialyse Studie (4D),3 to sudden in- or out-of-hospital car- with a better outcome. According to
diac arrests.4 Interestingly, this risk was data from Herzog et al., dialysis patients
1Guy’s Hospital, Department of Nephrology, high in the first 6 months after starting who survive a cardiac arrest may also
London, United Kingdom dialysis, dropped to its lowest point by benefit from an implantable cardioverter
Correspondence: M Ostermann, Guy’s and 6 months, and then progressively rose defibrillator (ICD).9 In this retrospec-
St Thomas’ Hospital, Department of Nephrology,
St Thomas Street, London SE1 9RT, again with each year on dialysis. Data tive cohort study, dialysis patients with
United Kingdom. from the System Case Mix Adequacy an ICD after a cardiac arrest had a 42%
E-mail: Marlies.Ostermann@gstt.nhs.uk Study of the United States Renal Data reduction in mortality compared with

Kidney International (2008) 73 907