Ed Silverman of the Star ledger posted the following alarming story on his blog this morning. The story raises serious issues for Americans depending upon Drug companies to tell the truth about all the risks associated with products they are ingesting.
Merck Hid Evidence Of Vioxx Harm: Study 4 Comments By Ed Silverman // July 10th, 2012 // 8:22 am And so, the Vioxx saga continues. Nearly eight years after Merck withdrew the controversial painkiller over links to heart attacks and strokes, a new paper indicates the drugmaker hid evidence that Vioxx tripled the risk of cardiovascular death for more than three years before taking the pill off the market in 2004. During the same period, the paper in the American Heart Journal notes, Merck had regularly insisted such an increased did not exist. The authors came to this conclusion after running an intent-to-treat analysis of three placebo-controlled studies that Merck conducted to determine whether Vioxx was useful in preventing or treating Alzheimer’s disease. Merck, however, did not publish such an analysis; instead, the drugmaker relied on so-called on-treatment analysis, which is not the preferred method for ferreting out potentially harmful side effects that may occur in a clinical trial (here is the abstract). Simply put, an intent-to-treat analysis will examine patients who were on the study drug, as well as time spent off the drug, but still enrolled in the study. The on-treatment approach used by Merck will not capture as many side effects. If a drug causes people to get sick and drop out of a study, then counting only the results that occur on treatment can bias results to favor the drug being studied, since the placebo arm is unlikely to cause as many patients to drop out. This was an issue with three Alzheimer’s studies, the researchers explain, because cardiovascular events attributed to Vioxx – such as heart attacks and strokes – were likely to occur more than two weeks after treatment was ended. And Merck did not adjudicate, or evaluate, any non-fatal cardiovascular events that were reported by its clinical trial investigators that occurred more than two weeks after patients stopped taking Vioxx. Intention-to-treat “is the gold standard for evaluating clinical trial results,” Jerry Avorn, one of the co-authors, who is chief of the Division of Pharmacoepidemiology and Pharmacoeconomics in the Department of Medicine at Brigham and Women’s Hospital and a professor of medicine at Harvard Medical School, writes us. “This is a fundamental tenet of clinical trial design.” Initially, Merck had sought to conduct intent-to-treat analyses for all three Alzheimer’s studies and, as the paper notes, had collected follow-up data, regardless of whether the patient continued to take Vioxx. However, the information was never presented publicly and there was no oversight because none of the three Alzheimer’s trials had a Data Safety Monitoring Board that have might persuaded the drugmaker to include this information. “The Alzheimers studies were true (intent-to-treat) studies and (intent-to-treat) analysis was pre-specified,” David Madigan, one of the authors and the chair of the statistics department at Columiba University, writes us. “Several Merck employees have acknowledged this in deposition testimony and it is clear from the protocols and data analysis plans. In fact, in the original protocols for (Alzheimer’s studies known as) 078 and 091, (intent-to-treat) was the only specified analysis of safety and efficacy.” The paper notes that the ITT analyses originally planned would have revealed substantially greater risks and statistically significant disparities. As of March 16, 2001, which was the cutoff date for labeling approved the following year and based on a study called VIGOR, the analysis would have revealed 10 confirmed cardiovascular deaths in patients randomized to Vioxx compared with three in the placebo groups in two Alzheimer’s trial. A subsequent intent-to-treat analysis found that as of April 11, 2002, when the FDA approved Vioxx labeling, there were 17 confirmed cardiovascular deaths on Vioxx compared with five on placebo in the same two trials. “In my view, given all the clear evidence of possible cardiovascular harm from Vioxx as early as 2000, Merck should have stuck with its planned analyses for these Alzheimer studies instead of moving to a less-standard approach, and looked at these data as they emerged to make sure such harm wasn’t occurring” among patients, Avorn writes us. “If we could do this from the same data they had, why didn’t they, especially given all the prior concern about possible cardiovascular toxicity? The company certainly should not have gone around and cited these very trials as ‘proof’ that the drug was safe.” Of course, the details surrounding the Vioxx withdrawal have been picked over repeatedly in recent years, amid protracted litigation and sometimes sensational disclosures about behind-the-scenes Merck decisionmaking. And while the episode prompted the FDA to place greater emphasis on safety issues, these latest findings may reinforce ongoing concerns, in some quarters, that the pharmaceutical industry – not just any one drugmaker – remains incapable of making patient safety a high priority. “This is the second disturbing drug safety-related revelation in the last week,” according to Steve Nissen, who chairs the cardiovascular department at the Cleveland Clinic and was on the FDA advisory panel in 2005 that reviewed, where he called it a “mistake” for Merck not to have provided an intent-to-treat analysis that evaluated patients beyond two weeks after having stopped the drug (see page 279 of the transcript). “First, we learned that Glaxo plead guilty to criminal fraud involving multiple products, including Avandia (read here). Now, we learn that the studies used to defend the safety of Vioxx actually showed the opposite,” he wrote us in an e-mail. “There are many fine people within the pharmaceutical industry with ethical standards similar to my own. However, a few ‘bad actors’ have poisoned public perceptions about this important industry. They must be held accountable.” To prevent a repeat of such situations, the authors recommend that regulators require clinical trial designs that use an intent-to-treat analysis and undertake intent-to-treat safety analyses of randomized controlled trial data. “We also recommend that regulators ensure the presence of Data Safety Monitoring Board “whenever appropriate.” This may not a panacea, though, because these boards answer to drugmakers, not independent bodies and, therefore, have prompted concerns that decisions can be skewed by corporate priorities. Such questions were raised after a recent decision to halt a trial for the Zytiga prostate cancer drug sold by Johnson & Johnson (see this). The authors also write that “data on adverse events occurring within clinical trials are sometimes considered proprietary information owned by the company that sponsored the study and not made available by governmental regulatory bodies to the public or to researchers. But in the wake of the Vioxx episode, they suggest that “this policy may require reassessment in the name of public health.” As for Merck, the drugmaker declined to comment about the specifics of the study, but did send this note: “In the studies that served as the basis for the approval of Vioxx, Merck compared Vioxx to placebo and to non-selective NSAIDs (non-steroidal anti-inflammatory drugs) in almost 10,000 patients. Overall, the extensive data from (the) clinical development program did not show a difference in mortality between Vioxx, other pain relievers and placebo. Merck made the decision to withdraw Vioxx from the market within a week of learning the preliminary results of the APPROVe study, which was the first time a clinical trial revealed a difference in thrombotic events between patients taking Vioxx and patients taking placebo.” A paper published in 2009, by the way, did conduct an intent-to-treat analysis of cardiovascular adverse events reported by trial investigators in the Alzheimer’s studies and other placebo-controlled studies, and did show a statistically significant elevated cardiovascular risk as early as June 2001 (read this). Merck, however, charged that the methods used to reach this conclusion were unreliable (see here). Madigan was a co-author of this paper as well. We should note that four of the study authors, including Avorn and Madigan, have served as consultants for plaintiffs in litigation against Merck related to Vioxx, although Avorn did this work pro bono and did not receive any compensation, according to the disclosure statement. In addition, another author, Dan Sigelman, who is currently with the Office of Policy, Office of the Commissioner at the FDA, previously represented plaintiffs in litigation against Merck related to Vioxx
