Further Evidence of CV Harm With Rosiglitazone

Sue Hughes
December 13, 2007 — More evidence supporting the idea that rosiglitazone (Avandia, GlaxoSmithKline) does increase the risk of cardiovascular events has come from a new population-based study [1].
The retrospective case-control study, published in the December 12, 2007 issue of the Journal of the American Medical Association, was conducted in older patients with diabetes and showed that thiazolidinedione (TZD) treatment, primarily with rosiglitazone, was associated with an increased risk of congestive heart failure (CHF), myocardial infarction (MI), and mortality when compared with other combination oral hypoglycemic agent treatments.
The study has reignited the arguments surrounding the safety of rosiglitazone, with Dr Steven Nissen (Cleveland Clinic) issuing new calls for more forceful action on the drug from the Food and Drug Administration (FDA), while GlaxoSmithKline highlights limitations of the new study and continues to defend the cardiovascular profile of its product.
In the paper, the authors, led by Dr Lorraine Lipscombe (Institute for Clinical Evaluative Sciences [ICES], Toronto, ON), note that most studies to date have examined adverse cardiovascular outcomes associated with TZDs among clinical-trial samples, but the extent to which these adverse effects apply to real-world populations is less clear. In addition, the majority of clinical trials were limited to patients younger than 65 years, even though those aged 65 years or older have the highest prevalence of diabetes and represent more than 40% of the population with diabetes.
They therefore investigated the risk of heart failure, MI, or death in diabetic patients over the age of 65 taking a TZD compared with those taking other oral hypoglycemic agents in a nested case-control analysis of a retrospective cohort study using healthcare databases in Ontario. The study population consisted of 159,026 diabetes patients (mean age 74.7 years) being treated with an oral hypoglycemic agent. During the median 3.8-year follow-up, 7.9% of the patients had a hospital visit for CHF, 7.9% had a visit for MI, and 19% died. The risks of these events were compared between persons treated with TZDs (rosiglitazone and pioglitazone [Actos, Takeda Pharmaceuticals]) and other oral hypoglycemic agent combinations, after matching and adjustment for prognostic factors.
Results showed that treatment with TZD monotherapy was associated with a significantly increased risk of congestive heart failure, MI, and death and that this increased risk appeared limited to rosiglitazone.
Lipscombe et al report that treatment with TZDs was also associated with a higher risk of CHF and death regardless of whether it was used as monotherapy or in combination with other oral agents, which they say further enhances the argument for a causal relationship with these outcomes.
They write: "Our findings argue against the current labeling of TZDs, which warns against use only in persons at high risk of CHF, as we did not identify any subgroup of older diabetes patients who may be protected from the adverse effects of TZDs." They add: "Our results suggest that the numbers needed to harm over four years were approximately 26 for acute myocardial infarction (AMI), 34 for CHF, and 22 for death. The magnitude of harm observed in our study may therefore be sufficient to outweigh any potential benefits associated with TZDs in an older, higher-risk population."
Noting that the US FDA has recently decided that the available data are thus far inconclusive to warrant withdrawing rosiglitazone from the market, they say that this "well-designed population-based study provides more convincing evidence that rosiglitazone is associated with an increased risk of cardiac events and deaths among elderly patients with diabetes."
Nissen: Further action now required
Commenting on the study for heartwire, Nissen said, "These are pretty striking data. It has the weakness of being an observational study, which is not the same as randomized data, but it has the advantages of being very large, with 159,000 patients, and it is not sponsored. The results are quite remarkable in that they show a very similar increase in risk of MI as we did in our meta-analysis. And the numbers needed to harm one patient are very low in this study. It suggests that for every 22 patients treated with rosiglitazone, one will die. I don't remember seeing numbers that low before in terms of benefit or harm. Even if these data are retrospective, it must be considered within the context of what we already know. And we have already had four meta-analyses, all of which have shown increased cardiovascular risk with rosiglitazone. These new data agree completely with what is already out there and have shown again that everything is going in the wrong direction with this drug. These data really increase the pressure on the FDA to take further action on rosiglitazone. The FDA's recent warning was very weak, and now there is even more evidence to support more forceful action."
GlaxoSmithKline response
GlaxoSmithKline issued a statement on the ICES study, saying it believes the retrospective analysis has significant limitations and generates misleading conclusions regarding MI and death. It states: "These conclusions are inconsistent with a more robust body of evidence from large, long-term, prospective, well-designed clinical studies, including A Diabetes Outcome Progression Trial (ADOPT) and the Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes (RECORD). These long-term trials in diabetic patients comparing rosiglitazone with other oral antidiabetic medicines show no increased risk for cardiovascular events compared with other commonly used medications, other than the well-known risk of congestive heart failure with thiazolidinediones."
The company adds that the current study is also inconsistent with the findings of large epidemiological studies presented during the recent FDA advisory committee meeting, the majority of which show that rosiglitazone is not associated with an increased risk of MI compared with other antidiabetic agents.
GlaxoSmithKline suggests that the inconsistency could be due to the fact that the rosiglitazone patients were older and sicker and had more chronic diseases — and therefore higher baseline risk for cardiovascular (CV) events — than patients in other medicine groups. "The problems with the study become obvious with regard to the findings for congestive heart failure — we know members in this drug class (TZDs) have well-documented and similar CHF events, yet this study somehow finds an increase in these events with Avandia vs Actos, a finding that is inconsistent with known science," it says.
The company further notes that the ICES analysis does not correct for the fact that patients prescribed rosiglitazone alone suffered from more chronic diseases compared with those prescribed pioglitazone alone or for the fourfold higher rate of kidney impairment in the TZD patient group. In addition, it notes that the ICES analysis included insulin therapy within the TZD combination group and excluded insulin combinations within the comparison group, which biases the TZD combination group toward increased cardiovascular risk relative to the comparison group.
Nissen described the company's statement as "predictable." He said it was "impossible" to ignore the evidence of increased risk of MI with rosiglitazone, which has been shown in four meta-analyses and now in this independent observational study. "The conclusion that rosiglitazone increases the risk of MI is simply inescapable. The contrary 'evidence' cited by GlaxoSmithKline is not credible. The RECORD trial is an interim report of an ongoing study and the hazard ratio (HR) for MI was 1.23, but with too few events to reach significance. The 'large' GlaxoSmithKline-sponsored observational studies are unpublished and not credible. Finally, the Ontario study is fully consistent with the findings of the meta-analysis of pioglitazone [2] that show no increased risk of MI with pioglitazone," he told heartwire.
Also commenting on the latest results for heartwire, Dr Sanjay Kaul (Cedars Sinai Medical Center, Los Angeles, CA), who coauthored a paper criticizing Nissen's original meta-analysis, said the ICES study was interesting but that the data are not conclusive enough to guide clinical practice. He said: "Case-control observational studies are notorious for introducing biases. Among the methodologic tools that have been developed to minimize bias (confounding) in these studies, propensity-adjusted model with time-varying covariates and adjustment for multiple comparisons is generally considered to be the most robust. I did not see these methods applied to the analysis in this study. The investigators report no statistically significant increase in CHF for pioglitazone, but we know from many other data sources that this agent does increase risk of CHF. The absence of such a finding in their data suggests the likelihood of methodological errors associated with biases that are inherent to analyses of observational data."
The Ontario Ministry of Health and Long-Term Care funded this study. Dr. Lipscombe has received support from the Canadian Diabetes Association, the Canadian Institutes of Health Research, and the New Emerging Team Canadian Institutes of Health Research. Two other study authors have received support from the Canadian Institutes of Health Research and from the Heart and Stroke Foundation of Ontario. One study author has disclosed various financial relationships with INTERxVENT Canada and PrevCan. Another study author has obtained funding.

Sources
Lipscombe LL, Gomes T, Lévesque LE, et al. Thiazolidinediones and cardiovascular outcomes in older patients with diabetes. JAMA. 2007;298:2634-2643.
Lincoff AM, Wolski K, Nicholls SJ, Nissen SE. Pioglitazone and risk of cardiovascular events in patients with type 2 diabetes mellitus: A meta-analysis of randomized trials. JAMA. 2007;298:1180-1188.