http://www.spacedoc.net/board

I’m not saying you’re in denial about the power of meta-analyses or the overall hazards of statins. And I’m also not saying people shouldn’t be warned about the potential hazards. What I am saying is you are completely off-base in terms of the relative risk-benefit ratio as concerns the study at hand if you really think these are equivalent endpoints on a percentage-to-percentage basis.

And you’ve got to be kidding me if think you accurately represented the entirety of the safety discussion with that little blurb.

How men would rather die? What? True or not, the irrelevance of the point is staggering, and makes clear to me that any further serious discussion of this study and your (mis)representation of it is a colossal waste of my time.

Walsh JM, Pignone M. Drug treatment of hyperlipidemia in women. J.A.M.A. 2004;291:2243-2252.

And the meta-analysis I cited is the best way to look at a question like whether statins cause cancer, since it puts all the studies together, and therefore has more power.

If I disagree with the points above, does that mean I am in denial of Reality?

And just to be complete, here’s how you summarized the authors’ take on prostate cancer:

“The authors stated that the increase was “probably due to
chance and not causally related. A similar ruse was given for the finding from PROSPER of a 25% increase in cancer, where the authors said that it was due to “the play of chance”.”

I included the second sentence to show how you injected your own clear bias.

For people who actually care, here’s what the authors ACTUALLY wrote in the discussion:

“The only cancer outcome that occurred with significantly greater frequency in the pravastatin group than in the placebo group was prostate cancer. Taking into account the multiple statistical tests performed, it is arguable that this result was due to chance rather than to a causal relationship. In the meta-analysis of statin trials,(17) data were not reported for prostate cancer as a separate site of cancer. However, there was no excess in the number of genitourinary cancers in that analysis, and the results of an observational case–control study suggest a possible reduction in prostate cancer (18) with statin use. A meta-analysis of the incidence (19) of cancer in statin trials showed no significant difference in the incidence of prostate cancer between the statin and control groups in trials involving a total of 20,063 participants. However, reports on these other studies did not include data on long-term follow-up, and some cancers may have a long latency period.

“Our results should be compared with those observed in the long-term follow-up of the Scandinavian Simvastatin Survival Study (5) and the LIPID trial.(7) Both were secondary prevention studies, and in both studies most participants were treated with open-label cholesterol-lowering drugs at the start of the post-trial follow-up period (84% in the Scandinavian study and 90% in the LIPID study). The Scandinavian group reported on an extended 5-year follow-up of deaths and incident cancers (5). They found no ongoing benefit with respect to a reduction in the rate of death from any cause or from cardiovascular disease in the follow-up period and no evidence of an increase in the incidence of cancer. There were more cases of prostate cancer in the placebo group than in the statin group. The LIPID Study Group reported results (7) for a 2-year extension. During this period, there was ongoing evidence of a benefit with respect to a reduction in deaths from coronary cause and deaths from any cause. There was no evidence of an excess incidence of cancers of the prostate and testes combined in the group randomly assigned to receive pravastatin. There is no obvious explanation for the apparent disparity between the two studies with respect to the extended benefit achieved. It is possible that the extended benefit of a period of treatment with a statin will depend on the severity of coronary disease at baseline. In that respect, our study population would rank as having the least severe coronary disease at baseline, and the Scandinavian population as having the most severe disease. Hence, it is not clear that the results achieved
during long-term follow-up in our study can be extrapolated to the treatment of patients with established coronary heart disease. In addition, all study participants were middle-aged men, which means that our findings cannot necessarily be extrapolated to any other group.”

When you give the whole context, it kind of takes away from your “gotcha” moment, don’t you think? It’s worth noting, too, that this is about 1/4 of the discussion by eyeballing it. And this is NEJM, where space is at a premium.

I hope that folks like John Mack see fit to either amend or correct their posts regarding Doug’s assertion, selective quotation, and misleading information.

I’ll close statement from your post:

“…I think it is time for these researchers to start playing by the rules of statistics, i.e. don’t use statistics only when it goes in your favor.”

Noted without comment.

Your post on your blog implies that I’m somehow dodging you, which is preposterous. You can post all the meta-analyses about statin safety you want. That was never the point, and it was something I acknowledged, and I think my posts bear that out. The points were that:

1. You make an overly simplistic 1-to-1 comparison of a 25% percentage decrease in coronary events to a 51% increase in prostate cancer as thought they’re on an equal scale. If one of your medical students or residents made this type of comparison, I’m sure you’d point out that it is a false comparison.

If a new drug has been shown to decrease the incidence of prostate cancer by 51%, but it has this tiny AE of 20% increased mortality. How do you feel about Pharma Co. X getting FDA approval for this new drug?

Let’s take it a little further: I am not a statistician by any stretch of the imagination, but as I calculate it, the NNT for the combined endpoint is 27, and 55 for all-cause mortality. The NNH for prostate cancer incidence is 111. Given that I’m much more likely to die from CVD than prostate cancer (and we’re not even talking about death from prostate CA, just incidence), I’ll take my chances with a statin.

2. You grossly mischaracterized the authors’ treatment of the incidence of prostate cancer, which you still haven’t acknowledged.

This is actually somewhat analogous to what we were talking about in the Gardasil post. Given what we know about the natural histories of the diseases, the overall effect on mortality by the 25% decrease in CV mortality almost certainly outweighs the effect by a 51% increase in prostate cancer incidence. This supported, at least in part, by the similar rates in non-CV mortality, including cancer mortality, both of which have HRs close to 1, and pretty narrow CIs. What would’ve been helpful would be a mortality rate for prostate cancer alone.

I think you give short shrift to the authors’ discussion of the increased incidence. They go well beyond merely saying it’s probably due to chance–they spend a decent portion of 3 paragraphs discussing it and reviewing earlier studies showing no increased incidence with statins.

And I wouldn’t dismiss out-of-hand the issue of multiplicity–it’s a legitimate statistical concern.

I do think, however, that such a large increase is worthy of further investigation.