HOPE-3: Statin mildly beneficial (but in what context)
Remarks on the HOPE-3 Trial
The fulltext is available for open access on the NEJM:
and Cholesterol Lowering in Persons without Cardiovascular
PDF is here. The Protocol
PDF is here.
Subsequent to my authoring the article following, a formal paper was published
Flaws in Evidence-Based Medicine: Statin Effects in Primary
Prevention and Consequences of Suspending the Treatment
Even the deminimus touted news from HOPE-3 may be no such thing.
This was a 7-year randomized trial of 12705
“intermediate risk” subjects assigned
to one of four arms, shown in the table below.
The results are being touted as:
“The decrease in the LDL cholesterol level
was 33.7 mg per deciliter
(0.87 mmol per liter) greater in
the combined-therapy group than in the dual-placebo
group, and the decrease in systolic blood pressure
was 6.2 mm Hg greater with combined
therapy than with dual placebo. … The
combination of rosuvastatin …, candesartan …,
and hydrochlorothiazide … was associated
with a significantly lower rate of cardiovascular
events than dual placebo among persons at intermediate
risk who did not have cardiovascular disease.”.
Yes, they drove LDL (presumably the mythical LDL-C)
down by a striking ~30%.
You may see headlines about a “24%”
reduction in heart attack and stroke, and
“5%” reduction in all-cause mortality,
but those are the notorious “relative risk”
numbers favored by Marketing. The absolute MI/stroke
reduction was a less impressive 1.1%, and deaths from
other causes went up. What really matters is all-cause mortality.
|placebo + placebo
|statin + placebo
|placebo + BP meds
|statin + BP meds
For driving LDL down by a whopping 30%, total all-cause
mortality only went down by 0.24% or 0.41%
(yep, ¼ or ½ of 1%).
You’d think that between this and the CETP
results, they might figure
out that there’s not very much leverage in shoving
LDL-C around. Don’t
count on it (video from the early 2016 conference): ’I
Have a Patient Who Won’t Take a Statin...’
leads off with Captain Statin.
And due to confounders not recorded or reported in
HOPE-3, we don’t
even know if the ¼% or ½% are real.
Apart from the normal suspicion of bias due to being
funded by the product producer, suppose, for example, that the placebos
were calcium pills (or capsules containing sugar or
soybean oil). In real life, people declining statin
treatment don’t take placebos that
might actually be antagonist agents. Had they included
an untreated arm, they might have been able to shed
some light on that.
The statin was rosuvastatin (Crestor, 10 mg per day).
The BP meds were candesartan (16 mg per day) plus
hydrochlorothiazide (12.5 mg per day). The placebo
was not specified (and so may have been no such thing).
You can check Table 1 (PDF page 5) for a
list of other meds that the arms were on. It may or may
not be comprehensive, excluding perhaps things like
CoQ10 and K vitamins.
Persons with [diagnosed] cardiovascular
disease and those with an indication for or contraindication
to statins, angiotensin-receptor blockers,
angiotensin-converting–enzyme inhibitors, or
thiazide diuretics were excluded.
But these were not healthy people, despite lacking a CVD diagnosis:
||Average Baseline Reading
HbA1c and actual lipoproteins were not reported.
There’s no indication that any attention whatever
was paid to diet, so the volunteers may well have
continued to eat in the same manner that made them
unhealthy (or followed their doctors’ advice on
how to eat, which might amount to the same thing).
This trial is already being spun as a justification
for broader prophylactic use of statins. As a commenter
on a statin skeptic site put it:
that’s not a study - it’s an advertisement.