Abstract & Commentary
Source: Graham DJ, et al. JAMA. 2004;292:2585-2590.
A series of reports appeared in a recent issue of JAMA dealing with the withdrawal of Baycol (cerivastatin or ceriva) from the market in 2001. The lead article is an analysis of a large database demonstrating an extremely high rate of rhabdomyolysis (rhabdo) with ceriva and, particularly, the combination of ceriva and gemfibrozil.1 The next article is a detailed commentary on the FDA and the subsequent actions taken by pharmaceutical companies.2 In addition, there are pro and con articles regarding the corporate response of Bayer, regarding the actions taken by the company from 1999 until Baycol was first withdrawn from the US market in 2001.3-6 Fontanarosa et al contribute to this interesting regulatory political story.3 The current crisis regarding Vioxx and other Cox-2 agents echoes the ceriva story, with suspicion by many that the drug companies (Bayer and Merck) had sufficient data regarding serious adverse reactions to warrant discontinuation of marketing well before the decision was made.
The lead article in JAMA is an analysis of 11 large health-plans, regarding the incidence rates of rhabdo per person-years of treatment, and the relative risk of rhabdo with specific statins. This is an extensive survey establishing the incidence of rhabdo in individuals treated with statins, fibrates, or the combination. Rhabdo is defined as severe muscle injury, with a creatine kinase level more than 10 times the upper limit of normal. Severe rhabdo is a subset of individuals with a CK > 10,000 IU/L, or 50 times the upper limit of normal.
Relative risk estimates were obtained for gender, age, and the presence of diabetes. The database included 252,000 patients, most were exposed to statin or fibrate monotherapy and 7300 person-years of combinination therapy. Twenty-four patients had rhabdo; almost all had muscle pain or weakness for a week (range 1-30 days prior to admission), and 18 had severe rhabdo. The incidence of rhabdo for monotherapy with atorvastatin, pravastatin, and simvastatin was comparable. However, the incidence rate for ceriva or gemfibrozil as monotherapy was much greater than that of the other statins. Norhabdo was reported for fenofibrate monotherapy. The incidence rate for rhabdo of a statin-fibrate combination was almost 3-fold vs statins alone. The NNT for 1 year for a case of rhabdo was approximately 1700, with statin-fibrates vs 10-13 patients for combination ceriva-gemfibrozil. Risk with rhabdo was increased in subjects over 65 and diabetics. Gemfibrozil use alone was associated with a 5.5-fold rhabdo increase, ceriva alone with a 10-fold increase in risk, and the combination of non-cerivastatin and gemfibrozil with a 12-fold increased risk vs statin monotherapy. Ceriva had a 1400-fold risk when used with gemfibrozil! Graham et al conclude that use of atorva, prava, or simva has a "relatively low risk of rhabdo." They suggest that higher risk subjects, ie, diabetics, or combination drug therapy, should be counseled to stop the statin immediately if any symptoms suggestive of rhabdo occur. This is particularly important in that there likely will be a substantial increase in the number of patients treated with statins over the next several years.
Another article carefully reviews much of the medical and legal record with respect to the safety of ceriva.2 Many aspects of the regulatory process are examined. Documents from litigation, as well as the published literature are reviewed. Ceriva was launched in the United States in the spring of 1998. Almost immediately, 6 cases of ceriva-associated rhabdo were reported as potential serious adverse drug reactions. In July 1998, rhabdo was added to the warnings in the ceriva packet insert. In 1999, reports of rhabdo with ceriva-gemfibrozil became available. Unpublished clinical trials demonstrated a high incidence of CK elevations and side effects using the 1.6 mg dose, with an estimate that combined ceriva-gembrozil were used in approximately 50% of rhabdo cases. In internal reports from early 2000, Bayer scientists confirmed that ceriva monotherapy was associated with a "substantial elevated risk of rhabdo compared with other statins." Elderly or thin women were thought to be at increased risk with the 0.8 mg dose. Label updates occurred in late 2000 and spring 2001. In August 2001, the drug was withdrawn from the US market. Graham and colleagues describe the history of prescription drug evaluation by the FDA in considerable detail. Federal funding constraints have impacted adversely on the FDA, with respect to pre- and post- marketing safety, and the pressure has increased for "more and faster new approvals", without additional funds for safety surveillance.
The overall experience of rhabdo in long term non-ceriva statin trials is reported, with an estimated 5.3 per 100,000 person years with active treatment vs 3.3 per 100,000 person years with placebo. The ceriva data are much more worrisome, with an increased risk of rhabdo, 65 times more than the other statins combined. Mortality rates with rhabdo were very high compared to other statins. It was subsequently discovered that combined gemfibrozil-ceriva use increased ceriva blood concentrations by 500%. Graham and colleagues focus on the interactions between the pharmaceutical industry and the FDA, and suggest that FDA actions might have been delayed because of inadequate spending for safety monitoring after the release of ceriva. They discuss the conflict of interest that is present in reporting details of drug safety by a pharmaceutical company regarding risks vs benefits of a drug to be FDA approved. Graham and colleagues conclude that serious flaws exist in the current system for severe adverse drug reaction (ADR) reporting and monitoring. They recommend establishment of an independent board for monitoring of drug safety or toxicity (or to provide sufficient support to the FDA for this purpose). Once a drug is approved for marketing, there are no automatic re-reviews. Graham and colleagues state that provision of "post marketing fees and regular re-review of approved drugs might enhance patient safety in the United States."
In an article by Strom,4 an epidemiologist at the University of Pennsylvania, the FDA post-marketing surveillance system is described.4 Identification of possible ADRs after marketing "relies on spontaneous voluntary reporting to industry regulators of reactions observed in clinical practice, fundamentally a 1950’s era approach." The large majority of such reports are received by the manufacturers and not the FDA; (almost 300,000 annually). Strom agrees with Psaty et al2 in calling for complete public disclosure of the results of all clinical trials, and suggests that pharmaco-epidemiology studies be included as well. He charges that the ceriva matter indicates that "critical gaps" remain in the FDA safety system. "The resources expended to ensure drug safety are extremely limited." He questions oversight by pharmaceutical companies because of the "inherent conflict of interest" related to possible adverse effects of newly marketed drugs. "No organization is charged with developing and testing new methods for improving physician’s use of drugs in clinical practice." A 1990 GAO study found that half of approved drugs have serious adverse effects undetected before approval. The ability of the FDA and others to insure safety "are extraordinarily limited." Furthermore, "there is indeed a conflict of interest in asking an industry to monitor its own drugs." Strom also calls for bolstering the FDAs efforts to monitor drug safety and establish a national center for drug surveillance to "complement the regulatory mission of the FDA." Another article weighs in on this theme, placing much of the blame on industry, "yet the major problem with the current system for ensuring the safety of medications is that drug manufacturers are largely responsible for collecting, evaluating, and reporting data from post-marketing studies of their own products. This approach has many inherent problems."3 Fontarosa and colleagues mention the SSRI and Cox-2 problems, focusing on the recent withdrawal of Vioxx from the market. They echo Psaty et al, and state that unpublished data obtained from company documents indicate that "the company was well aware of the risks of rhabdo and the interaction of ceriva with gemfibrozil even as early as 4 months after the launch of ceriva." They state that "companies also have financial incentives and economic pressures that may influence the interpretation of adverse event data and may delay full reporting to the FDA." Fonterosa and colleagues are pessimistic as to whether the FDA can overcome the loss of trust and confidence of healthcare workers and the public regarding "its first absolute priority to protect public health." The approval process must uncouple post-marketing surveillance from the drug approval process. They concur with demands for an independent drug safety board or agency that would specifically oversee post-marketing surveillance of drugs and devices. In yet another article, a Bayer lawyer states, "Bayer’s conduct in the marketing of ceriva from 1997 until its voluntary withdrawal. . . was responsible, appropriate, and consistently motivated by the concern for the safety and welfare of patients."5 Psaty et al robustly respond to this article.6
Comments by Jonathan Abrams, MD
The 6 articles in the December 1, 2004, issue of JAMA represent a detailed overview of drug safety in the United States. Not only does the withdrawal of cerivastatin reflect the current situation, but problems with antidepressants in youngsters and, more recently, the identification of adverse reactions with the 3 Cox-2 inhibitors, raise important questions regarding the ability of the FDA to effectively police ADRs after release of a drug. The pressures on the FDA from Big Pharma companies and the US Congress to speed up the drug approval process, as well as in funding changes, leaving the FDA surveillance system without sufficient support, are critical matters. Whether or not a new drug safety Czar is the answer is not clear. It would appear that specific legislation enhancing the FDA’s surveillance and safety abilities, independent from the drug approval process, might suffice, as opposed to setting up an entirely new federal commission or agency. Nevertheless, there is unfortunately too much evidence over the past several years that "Something is Wrong in Denmark." The problems in part are complicated by the enormous popularity of the SSRIs, statins, and Cox-2 inhibitors, resulting in billions of dollars in sales in the United States, and even more worldwide. Multiple recent newspaper and television reports dealing with the current Cox-2 inhibitor story accentuate the problem. Some of these drugs are like a fancy car; fine if you can afford it, but not a truly better means of transportation than a standard automobile. Regarding Baycol, the unexpected hazard was identified very early after release was either unrecognized or not acted upon, or both, until ultimate withdrawal from the market, arguably later than it should have been. This is, alas, very similar to the problem with the Cox-2 inhibitors, which have already achieved remarkable sales, when compared to Baycol. In the case of both ceriva and Cox-2 adverse events, it is regrettably true that these drugs are not essential, and that many alternatives are available. This may not be the case for SSRI therapy in teenagers and children, where depression can lead to suicide and severe psychological disturbances, for which an effective dose would be highly valuable. No matter which side of the safety argument one is on, it is clear that there is enough blame to go around, including the FDA and big pharmaceutical companies. Vigilance and hyper-alertness may not be sufficient after a new drug has been approved, which could result in hundreds of thousands, if not millions, of new users. It is now clear that such patients may serve as an unwanted sensor to detect rare but extremely harmful adverse effects. Let us hope that there will be more light than heat arising from these unhappy experiences in the years ahead.
Dr. Abrams, Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque, is on the Editorial Board of Clinical Cardiology Alert.
1. Graham DJ, et al. JAMA. 2004;292:2585-2590.
2. Psaty BM, et al. JAMA. 2004;292:2622-2631.
3. Fontanarosa PB, et al. JAMA. 2004;292:2647-2650.
4. Strom BL. JAMA. 2004;292:2643-2646.
5. Piorkowski JD Jr. JAMA. 2004;292:2655-2657.
6. Psaty BM, et al. JAMA. 2004;292:2658-2659.