What Vivus Fails To Say About Its Diet Pill Study
11 CommentsBy Ed Silverman // April 12th, 2011 // 8:34 am
Here is a curious oversight. Anxious to see its Qnexa diet pill find its way to medicine cabinets, Vivus yesterday issued a press release touting a study in The Lancet medical journal that the drug demonstrated “significant weight loss and broad improvements in co-morbidities.” The Lancet, in its own press release, suggested “this promising new treatment has additional metabolic benefits - improving blood pressure, lipids, glycaemia, and inflammatory markers.”
The upshot is that the study suggests the Qnexa diet pill could help overweight people lose more pounds than those who were given a placebo and, therefore, possibly reduce their reliance on meds for other risk factors, namely diabetes, high blood pressure or cholesterol. Such a finding is potentially noteworthy, of course, and is the sort of conclusion any drugmaker with a diet pill would like to publicize. This is especially true of Vivus, since the FDA rejected Qnexa last fall.
However, Vivus failed to mention one thing. The study, which is called Conquer, was one of the same two clinical trials the FDA reviewed last year before deciding to deny approval for Qnexa, The Los Angeles Times points out. The combined results prompted the FDA’s Division of Metabolism and Endocrinology Products to conclude the weight-loss difference between people given Qnexa and those given a placebo was “of nominal statistical significance” (see page 3).
As the paper notes, the subtle use of language may leave a different impression, especially among those who do not compare findings for the drug, which combines topiramate and phentermine, the surviving half of the infamous fen-phen weight-loss cocktail. The Lancet authors wrote that, “most importantly, weight loss achieved with phentermine and topiramate was sustained during 56 weeks with improvements in blood pressure, lipids, glycaemia, and inflammatory markers.”
But the FDA reviewers were more matter-of-fact: the groups treated with Qnexa “had the expected improvements in blood pressure, lipids, and glycemia.” In other words, the drug was effective, but The Lancet findings were packaged in a glowing press release that overlooked an FDA review that had already taken the results into consideration. The combined press releases might suggest Qnexa is on the verge of a turnaround.
For those those who may not recall, Qnexa has failed to pass muster with the agency over safety issues. An FDA advisory panel last summer refused to recommend approval over cardiovascular risks and teratogenic concerns. One stipulation before Vivus can win approval involves creating a database that examines the incidence of oral clefts in babies whose moms took topiramate (read here).
As the Times notes, the Conquer study was funded by Vivus and three of the seven authors are Vivus employees. Another author works for the contract research organization that coordinated the study for Vivus and the lead author has served as a Vivus consultant. The second author listed also acknowledged receiving donations, honoraria, consulting fees or grants from Vivus (as well as other dugmakers with interests in diet pills).
UPDATE: We should note that this is not the first time Vivus has practiced what some may call selective disclosure. Last week, the erstwhile drugmaker disclosed new, positive heart-risk data from a long-term Qnexa study in a press release. But as The Street wrote, Vivus omitted negative data from the same study. Interestingly, the press release was issued even as complete data was presented to at the recent American College of Cardiology annual meeting. Perhaps, the Securities and Exchange Commission may want to start reading Vivus press releases.
Hat tip to Cardiobrief
carl
What you failed to understand is that the FDA reviewed 1 year data when it rejected QNEXA . This study is now showing the extension of the study with people that went on to finish 2 years.
Good homework ! Way to try and grab a headline.
1 year VS 2 year is not the same thing.
industry insider
The metabolic side benefits of the Vivus pill are not specific to the pill itself, and are a consequence of weight loss by any means, dietary, pharmacologic, etc. That’s why you can’t turn such benefits into indications. Maybe that’s why management did not make as big a deal out of them as might be expected.
Ed Silverman
Hi Carl,
Thanks for the note, although I have read and re-read both press releases and they state these are results from patients who took the pill for 56 weeks. No extension is mentioned. Nor is any further period of time in which the drug was taken or studied or followed up.
For instance, the Vivus press release contains this statement before detailing the findings: “Specific results for all patients through 56 weeks as published in The Lancet are as follows…”
If you can be more specific, I would like to know.
Regards
ed
exMBB
Also, the article interpreted the phrase “of nominal statistical significance” out of context. In statistical terms, reaching nominal statistical significance means p < 0.05.
But the way it was reported in the LA Times article, the author made it sound like Qnexa’s efficacy was minimally statistically significant. That is not the case. the primary point of efficacy for all 3 doses were <0.0001.
exMBB
Up on further reading the LA times article, and by citation, this article, should be retracted.
There was never 2 trials. It was 1 trial that measured 2 primary endpoints, the FDA’s 2007 criteria for evaluating weight-loss drugs
1. the average weight loss per patient
2. the % of patients who lost >/=5% of body weight.
THe FDA stated in the same document cited by LA TImes that Qnexa met the efficacy criteria by both measures.
However, at the time, there was concern about the drug’s chronic usage. THus the panel requested additional data, in other words, lengthen the trial to 2 years. Vivus going to collect that anyways.
Basically the 2 sensational charges in both the LA TImes article and in this article are wrong:
1. The new study is just an old study - No sh8t sherlock! It’s a continuation of the 1-year trial. Why would you start a new 2 year study???
2. Qnexa’s efficacy was of nominal statistically significance - this is clearly a misinterpretation of the FDA opinion. of nominal statistical significance means meeting the minimal bar of statistical significance. But it does NOT mean it is AT the minimal bar of statistical significance. p<.000000000001 still met nominal statistical significance.
Again, there were no 2 separate trials. Melissa Healy’s interpretation was just plain ignorant
Salient point
Lancet would probably be more concerned about EMEA than FDA (inasmuch as it concerns itself with regulatory bodies at all). Not sure what the submission history of this product in Europe is.
Ed Silverman
Hi exMBB,
Thanks for your note. Please click on the link to the FDA review from last summer and you will see the following statement…
“The long-term efficacy of PHEN/TPM was assessed in two randomized, double-blind, placebo-controlled 56-week trials, OB-302 and OB-303.”
Also, I noted in my post that efficacy was met.
I hope this helps,
ed
exMBB
Okay, maybe the trials were initiated with 2 slightly different protocols. One with low and high dose, another with mid and high dose. They probably started the latter trial because they afraid that the low dose would be have the requisite efficacy and the high dose may have bad side-effects.
But all the analyses were pooled. So in essence, it really was just 1 pooled trial.
ED and Diabetes
@ Carl: Sure, 1 year is not the same as 2 years, but when how long should the FDA consider for proper testing. If anything, drug companies need to spend longer testing their products before they submit them for approval.
Aside from that, the article is great insight.
y
Check out if there are any past, current, or future SEC actions against the company. By issuing misleading info in press releases that then influence the price of its stock is a no no and the SEC will come knocking. Also, Wall Street is prety smart. See what they do to the stock. If these guys are lying their stock will get hammered. (After Wall Street sucks the moneu out of the idiots who bought on this press release.)
Brian, New York
Ed, I am finding this late, having searched based on recent New York Times articles. I read the FDA documents that are referenced in your post, and in the Los Angeles Times article. The two studies that you reference — 302 and 303, had slightly different inclusion criteria. The first was for BMI greater than 35. The second was for BMI greater than 27, but with comorbidities like diabetes and high blood pressure.
The surprising thing about the FDA document is the disconnect between the Eric Gordon introduction memo, and the actual evaluation by Mary Dunn Roberts, M.D. Gordon went out of his way to minimize benefits and emphasize potential side effects/adverse events.
The weight loss numbers are striking. The “nominal statistical significance”, which in the LA Times article appears to be meant to minimize the findings, actually means that there was less than a one in 1000 chance that the weight loss was not related to the therapy. In other words, it works!
Gordon’s dismissal of lower LDL, higher LDL, and lower blood pressure in those taking the therapy is disingenuous at best; Roberts reported them as also statistically significant.
The only valid concern that I see in the product is the potential for birth defects. The anticonvulsant topirimate has been shown in animal studies to produce birth defects in mice, rats and rabbits.
Interestingly, there were 34 pregnancies among study participants, even though all had to agree to two forms of birth control, either double barrier, or barrier and oral contraceptive. Yet they were still 34 pregnancies. Of the 34, 18 month term, and no birth defects were observed.
The other thing you did not mention is that the Vivus pill is actually a combination of existing approved therapies. Doctors are free to prescribe them individually for weight loss purposes. Many already are.