Comments on: The Torcetrapib Mystery Unfolds, Just A Little

By: Pharma News 11_9 « Impactiviti blog

Pharma News 11_9 « Impactiviti blog — Fri, 09 Nov 2007 13:05:42 +0000

[...] Why did Pfizer’s torcetrapib fail? Some reasons may be surfacing. [...]

By: Steven Levy-correction

Steven Levy-correction — Wed, 07 Nov 2007 19:38:39 +0000

My apologies for an error yesterday. P value significance is somewhat arbitrary but in drug work statistical significance is generally accepted as p less than or equal to .01, not .001. This does not change my concerns. By use of .01 p values in table 3, ‘death from any cause’, ‘peripheral vascular disease(diagnosis)’, ‘hospitalization w primary diagnosis of CHF’ and ‘major CV event coronary revascularization procedure and PVD’ become statistically significant. With the possible exception of ‘hospitalization w primary diagnosis of CHF’ I would still submit the relatively modest (4 to 4.5 mmHG systolic difference between the trial legs)of aldosterone induced HTN may not be the primary cause of the increased morbidity and mortality observed within the torcetrapib leg. The authors certainly conclude the mechanism of increase morbidity and mortality is unknown. I simply wish to amplify that although there is an off-target HTN effect more investigation is warrented.
Steven Levy MD
stevenlevymd@gmail.com

By: Steven Levy

Steven Levy — Tue, 06 Nov 2007 22:11:54 +0000

Re: NEJM article- I respectfully disagree that aldosterone and HTN are the cause of this negative outcome based on this report. It appears the major CV events between the torcetrapib/non-torcetrapib legs noted on the Kaplan-Meier curves (figure 2) begin to separate at about 45 days. Statistical significance between the legs in CV events (table 3) is only present (p less than or equal to .001) for the primary composite outcome of time to first major CV event (torcetrapib 464 days vs non-torcetrapib 373 days), hospitalization for unstable angina and coronary revascularization procedure without major CV event (torcetrapib 505 days vs non-torcetrapib 403 days) or with major event (torcetrapib 738 vs non-torcetrapib 589). No other CV events reached statistical significance. It’s not clear what the cut offs were to remove patients from the study or when they were removed for non-fatal side effects including HTN, GI side effects, HA, although since more patients were removed from the torcetrapib leg- it’s likely there were even more patients with HTN from torcetrapib earlier in the study. But if the negative outcome with torcetrapib was directly related to HTN it would seem the CV events should have occured earlier with torcetrapib, not later, and that those relating directly to HTN such as hemorrhagic stroke would reach statistical significance-which they did not. Just speculation here but perhaps there is some instability in the ‘at risk’ plaques being created by elevation of the HDL’s that manifests late in the torcetrapib leg. Perhaps the increase in HDL’s is too rapid or the larger size of the HDL particles created with torcetrapib is creating an issue. Also of interest is the separation of the Kaplan-Meier curves (figure 2) for ‘deaths from any cause’ which appears to begin at 180 days. Therefore deaths from non-CV causes are separating out even later. Those deaths include (table 4) cancer (torcetrapib 24 vs non-torcetrapib 14) and infections (9 vs 0). There are no statistics regarding p values so significance is not clear, but it is interesting to speculate whether the drug itself or the change in HDL’s is affecting the immune system adversel- perhaps through elevation of glucocorticoids/cortisol. My point is that the the negative outcomes reported may not be due solely to elevated aldosterone and subsequent HTN that would seem addressable with proper patient follow-up.
Steven Levy MD
Search Consultant. Head Healthcare & Life Sciences.
The Landstone Group. Office 212-972-7300 CP 203-246-4742
stevenlevymd@gmail.com

By: www.topcholesteroladvice.info » The Torcetrapib Mystery Unfolds, Just A Little

Tue, 06 Nov 2007 05:39:08 +0000

[...] Ed Silverman added an interesting post on The Torcetrapib Mystery Unfolds, Just A Little.Here’s a small excerpt:Since then, the study has been picked over and this morning the results were finally disclosed at the annual American Heart Association meeting in Orlando, Florida (you can also read them in The New England Journal of Medicine). … [...]

By: Reality

Reality — Mon, 05 Nov 2007 15:28:44 +0000

Although Stein can’t say it without getting kicked out of the Cardiology fraternity, the question isn’t why Pfizer didn’t know about the aldosterone link (despite the fact that they looked), but why the folks like Kastelein and Nissen (or Stein himself for that matter) didn’t know. After all, Nissen has made a career of attacking other companies for not anticipating side effects given known physiology (e.g., Merck and Vioxx, GSK and Avandia) or seeing signals of such problems.