Correspondence
Rather than having the reader try to
interpret fragments of data, it would
be better if the investigators at last
reported their findings according to
initial hsCRP levels. That would help
the reader (and health authorities)
make their minds up about the appro-
priateness of using hsCRP as a screening
tool in the general population.
I declare that I have no conflicts of interest.
Nicolas Danchin
nicolas.danchin@egp.aphp.fr
Hôpital Européen Georges Pompidou, APHP,
Université Paris Descartes, 20 rue Leblanc,
75015 Paris, France
1 Ridker PM, Danielson E, Fonseca FAH, et al.
Reduction in C-reactive protein and LDL
cholesterol and cardiovascular event rates after
initiation of rosuvastatin: a prospective study
of the JUPITER trial. Lancet 2009; 373: 1175–82.
Paul Ridker and colleagues1 conclude
that reductions of LDL cholesterol and
high-sensitivity C-reactive protein
(hsCRP) are indicators of successful
treatment with rosuvastatin. How-
ever, the JUPITER trial did not target
specific concentrations of hsCRP or
LDL cholesterol by means of a strategy
of titration of rosuvastatin. Rather, it
assessed a single dose versus placebo.
No evidence was presented to suggest
that adjusting statin doses to reach
a particular concentration of hsCRP
(or LDL cholesterol) would improve
outcomes.
We were also concerned with the
finding (figure 2)1 that an hsCRP
concentration of less than 1 mg/L with
an LDL cholesterol concentration of at
least 1·8 mmol/L did not significantly
reduce the risk of the primary outcome
(although there seemed to be a
favourable trend), yet the hazard for
hsCRP concentrations of less than
2 mg/L (irrespective of LDL cholesterol
concentration) revealed a significant
improvement in favour of rosuvastatin.
This inconsistency raises concerns
about the usefulness of hsCRP in the
assessment of treatment in the lower-
risk population targeted by JUPITER.
Perhaps if the trial had not been
prematurely halted this trend could
have been more completely assessed
www.thelancet.com Vol 374 July 4, 2009
with longer follow-up and greater
numbers of events.2 Additionally,
early cessation of a clinical trial for
benefit could lead to overestimation
of treatment effects.2 Although the
recommendation to stop the trial
was made according to predefined
criteria by an independent data
and safety monitoring board,3 the
decision to start a low-risk patient on
statins on the basis of hsCRP remains
controversial.4
We declare that we have no conflicts of interest.
*Robert Rosenstein, David Parra
robert.rosenstein@va.gov
Veterans Affairs Medical Center West Palm Beach,
West Palm Beach, FL 33410, USA
1 Ridker PM, Danielson E, Fonseca FAH, et al.
Reduction in C-reactive protein and LDL
cholesterol and cardiovascular event rates
after initiation of rosuvastatin: a prospective
study of the JUPITER trial. Lancet 2009;
373: 1175–82.
2 Bassler D, Montori VM, Briel M, Glasziou P,
Guyatt G. Early stopping of randomized clinical
trials for overt efficacy is problematic.
J Clin Epidemiol 2008; 61: 241–46.
3 Ridker PM, Danielson E, Fonseca FAH, et al.
Rosuvastatin to prevent vascular events in
men and women with elevated C-reactive
protein. N Engl J Med 2008; 359: 2195–207.
4 Myers GL, Christenson RH, Cushman M, et al.
National Academy of Clinical Biochemistry
laboratory medicine practice guidelines:
emerging biomarkers for primary prevention
of cardiovascular disease. Clin Chem 2009;
55: 378–84.
The clinical implications of the JUPITER
study1 are substantial. Therefore
an understanding of the methods
matters even more than usual.
First, the participants were divided
into four mutually exclusive groups
on the basis of an arbitrary cutpoint
for C-reactive protein (CRP) and for
LDL cholesterol. This approach can
produce larger differences than the
more conventional one of continuous
analysis.2 Were there significant
independent continuous relations
between LDL cholesterol and risk
versus CRP and risk?
Second, the treatment group with
high LDL cholesterol and high CRP
received no benefit from rosuvastatin,
whereas in the group with low LDL
cholesterol and low CRP, risk was
reduced by 65%. Outcome after
therapy is the product of the initial as
well as the final risk. Accordingly, what
were the initial and final levels of LDL
cholesterol, apolipoprotein B, and CRP
in the four groups?
Finally, what was the prevalence
in the placebo group of those who
had a CRP of less than 2 mg/L at any
time during the trial? The clearer the
evidence of an independent effect of
CRP, the more rapid will be the clinical
application of these results.
I declare that I have no conflicts of interest.
Allan D Sniderman
allansniderman@hotmail.com
McGill University Health Centre, Royal Victoria
Hospital, Montreal, Quebec H3A 1A1, Canada
1 Ridker PM, Danielson E, Fonseca FA, et al.
Reduction in C-reactive protein and LDL
cholesterol and cardiovascular event rates
after initiation of rosuvastatin: a prospective
study of the JUPITER trial. Lancet 2009;
373: 1175–82.
2 Furukawa TS, Strauss S, Bucher HC, Guyatt G.
Dichotomizing continuous test scores,
sensitivity and specificity, and LR+ and LR-. In:
Guyatt G, Rennie D, Mead MO, eds. Users’
Guides to the Medical Literature: Essentials of
Evidence-Based Clinical Practice, 2nd edn.
Cook: McGraw-Hill, 2008: 212–14.
Although I laud Paul Ridker and
colleagues1 for demonstrating the
vascular anti-inflammatory benefits
of rosuvastatin, we should consider
the less expensive anti-inflammatory
option: aspirin.
Aspirin was used by only 16·6%
of participants in the placebo and
rosuvastatin groups, despite their
elevated cardiovascular risk and thus
indication for aspirin2 (42·1% in the
placebo group and 38·3% in the
rosuvastatin group had metabolic
syndrome). Why aspirin use was not
enforced is unclear, since its use is
indicated (in the absence of specific
contraindications) by age, diabetes,
and hypertension individually, let
alone when in combination. Ironically,
a previous article by Ridker3 supports
aspirin use to lessen myocardial
infarction risk through reduction of
high-sensitivity C-reactive protein
(hsCRP). Meanwhile we must consider
that if we do treat patients on the basis
of their hsCRP concentration, most
25
 Correspondence
western populations would require
statins (which might be excessive
without more long-term data).
A perhaps more subtle point that
readers should consider is that statins
are not only anti-inflammatory and
cholesterol-lowering, but also partly
antithrombotic themselves4 and able
to reduce venous thromboembolism.5
If the benefits of rosuvastatin in the
JUPITER study are corroborated by
another similar study in which aspirin
is adequately used, then we should
consider broadening statin therapy.
I declare that I have no conflicts of interest.
Jeffrey M Bloom
jmb4u2c@charter.net
1334 Marsh Street, San Luis Obispo, CA 93401, USA
1 Ridker PM, Danielson E, Fonseca F, et al.
Reduction in C-reactive protein and LDL
cholesterol and cardiovascular event rates
after initiation of rosuvastatin: a prospective
study of the JUPITER trial. Lancet 2009;
373: 1175–83.
2 Hayden M, Pignone M, Phillips C, Mulrow C.
Aspirin for the primary prevention of
cardiovascular events: a summary of the
evidence for the US Preventive Services Task
Force. Ann Intern Med 2002; 136: 161–72.
3 Ridker PM, Cushman M, Stampfer MJ,
Tracy RP, Hennekens CH. Inflammation,
aspirin, and the risk of cardiovascular disease in
apparently healthy men. N Engl J Med 1997;
336: 973–79.
4 Rosenson RS, Tangney CC.
Antiatherothrombotic properties of statins:
implications for cardiovascular event
reduction. JAMA 1998; 279: 1643–50.
5 Glynn RJ, Danielson E, Fonseca FA, Genest J,
Gotto AM. A randomized trial of rosuvastatin
in the prevention of venous
thromboembolism. N Engl J Med 2009;
360: 1851–61.
Paul Ridker and colleagues1 conclude
that, for people who choose to
start pharmacological prophylaxis,
reductions in both LDL cholesterol and
high-sensitivity C-reactive protein are
indicators of the success of treatment
with statin therapy.
In the initial description of the
JUPITER study,2 table 1 contains
baseline estimated glomerular
filtration rates (eGFR) of 73·3 mL/
min/1·73 m² for the rosuvastatin
group and 73·6 mL/min/1·73 m² for
the placebo group. 1 year later, as
shown in table 4, median eGFRs were
66·8 and 66·6 mL/min/1·73 m²,
26
respectively.2 According to these data,
a relevant proportion of patients
included in JUPITER could poten-
tially have established chronic
kidney disease (defined as eGFR
<60 mL/min/1·73 m²).
Chronic kidney disease is character-
ised by an increased cardiovascular
risk and by a positive response to
statins.3,4 In our opinion, the frequency
of chronic kidney disease should be
established in this population, as
should whether the positive response
to rosuvastatin was affected by the
presence of chronic kidney disease.
We declare that we have no conflicts of interest.
*J Segura, L M Ruilope
hta@juliansegura.com
Hypertension Unit, Nephrology Department,
Hospital 12 de Octubre, 28041 Madrid, Spain
1 Ridker PM, Danielson E, Fonseca FAH, et al.
Reduction in C-reactive protein and LDL
cholesterol and cardiovascular event rates after
initiation of rosuvastatin: a prospective study
of the JUPITER trial. Lancet 2009; 373: 1175–82.
2 Ridker PM, Danielson E, Fonseca FAH, et al.
Rosuvastatin to prevent vascular events in
men and women with elevated C-reactive
protein. N Engl J Med 2008; 359: 2195–207.
3 Tonelli M, Isles C, Curhan GC, et al. Effect of
pravastatin on cardiovascular events in people
with chronic kidney disease. Circulation 2004;
110: 1557–63.
4 Navaneethan SD, Pansini F, Perkovic V, et al.
HMG CoA reductase inhibitors (statins) for
people with chronic kidney disease not
requiring dialysis. Cochrane Database Syst Rev
2009; 2: CD007784.
Authors’ reply
We agree with Jaime García de Tena
that smoking cessation, blood
pressure control, and lifestyle modifi-
cation remain important for primary
prevention. However, half of all vascular
events occur among those with average
or low concentrations of cholesterol.
JUPITER showed that targeting statin
therapy to those with elevated high-
sensitivity C-reactive protein (hsCRP)
and low LDL cholesterol reduces heart
attack and stroke by 50% and total
mortality by 20%. Similar benefits were
seen among those with elevated hsCRP
and no other major risk factors.
In our paper and webappendix, we
show that the effects of hsCRP are
independent of LDL cholesterol and
the ratios of apolipoprotein B to apo-
lipoprotein A and of LDL cholesterol
to HDL cholesterol. We also present
in paragraph 4 of the Results data on
outcomes stratified by baseline LDL
cholesterol and hsCRP. Thus, the infor-
mation sought by both W E Feeman
and Nicolas Danchin are already
provided. JUPITER did not enrol
patients with low LDL cholesterol and
low hsCRP because we previously saw
within the AFCAPS/TexCAPS trial1
that such individuals have a very low
absolute risk and did not benefit from
statin therapy.
Whereas Robert Rosenstein and
David Parra focus on p values within
specific subgroups, our prespecified
analyses assessed trends across levels
of achieved LDL cholesterol and
achieved hsCRP where effects are
highly significant (p<0·0001). The
independent data and safety moni-
toring board was highly experienced
and followed rigorous principles
in their prespecification that early
termination would require proof
beyond a reasonable doubt. The
board elected to continue the trial
for an additional 6 months after the
conservative stopping boundary was
crossed. Data that accrued thereafter
independently confirmed both the
magnitude and statistical significance
of the main effects. We thus believe the
board protected the interests of society
and the trial participants and provided
a valid estimate of treatment effect.2
Cut-points for both LDL cholesterol
and hsCRP were not arbitrary but
prespecified on the basis of previous
work.3 In continuous rather than
categorical analyses, we obtain largely
similar results, as suggested by our use
of percentage reductions in addition to
absolute reductions. Allan Sniderman is
correct in that achieved levels depend
in part on baseline characteristics, and
our analyses therefore adjusted for
a wide range of baseline covariates
including LDL cholesterol, hsCRP,
and HDL cholesterol. Only 1% of the
placebo-allocated group achieved LDL
cholesterol concentrations of less than
www.thelancet.com Vol 374 July 4, 2009