This article is more than 5 years old. Information may no longer be current.
Meta-analysis provides new insights on CV risk with rosiglitazone
Rosiglitazone was linked to an increased CV risk, particularly for HF events, in trials with available individual patient-level data, according to a systematic review and meta-analysis published in BMJ.
“Avandia (GlaxoSmithKline) was first approved approximately 20 years ago, and since 2007, there have been conflicting findings about whether Avandia increases the risk of myocardial infarction,” Joshua D. Wallach, PhD, MS, assistant professor of epidemiology in the department of environmental health sciences at Yale School of Public Health, told Healio. “Similar to previous meta-analyses, we did observe an increased risk of heart failure and myocardial infarction. However, the increased risk for myocardial infarction was not as strong as what had been previously reported based on summary-level data sources. This finding demonstrates the potential for different results from different data sources and supports efforts to promote clinical trial data sharing.”
Trials comparing rosiglitazone with control
Researchers analyzed data from 21,156 patients from 33 randomized controlled trials with available individual patient-level data that compared rosiglitazone with a control in adults for at least 24 weeks. The meta-analysis for MI included 103 trials without individual patient-level data for 23,683 patients, in addition to another 103 trials with 22,772 patients for the meta-analysis on CV-related death.
The primary outcome for trials with available individual patient-level data was HF events, MI events, non-CV-related deaths and CV-related deaths. These outcomes were also assessed individually in secondary analyses. In trials without available individual patient-level data, outcomes of interest were MI and CV-related deaths.
Individual patient-level data helped identify more MIs compared with summary level data.
Compared with controls, patients assigned rosiglitazone had a 33% increased risk for a composite event when analyses were limited to trials with individual patient-level data and when a random-effects model and a constant continuity correction of 0.5 were used (274 events in the rosiglitazone group vs. 219 events in the control group; OR = 1.33; 95% CI, 1.09-1.61).
When examined independently, the OR for HF was 1.54 (95% CI, 1.14-2.09), 1.17 for MI (95% CI, 0.92-1.51), 1.18 for non-CV-related death (95% CI, 0.6-2.3) and 1.15 for CV death (95% CI, 0.55-2.41). For trials without individual patient-level data, rosiglitazone increased the odds for MI by 9% (OR = 1.09; 95% CI, 0.88-1.35) and the odds for CV death by 12% (OR = 1.12; 95% CI, 0.72-1.74).
Results were consistent when analyses were repeated with either of the two continuity corrections and with trials with zero events in both arms.
Clinical implications
“Avandia is no longer available in Europe and is rarely used in the United States,” Wallach said in an interview. “Therefore, these findings may not have any immediate clinical implications. However, our study suggests that we could have had better estimates if the trial data were made available years ago. Future meta-analyses focused on drug safety should locate, request and utilize raw data to accurately assess adverse events. These efforts can allow patients, clinicians and policymakers to make more informed decisions.” – by Darlene Dobkowski
For more information:
Joshua D. Wallach, PhD, MS, can be reached at joshua.wallach@yale.edu; Twitter: @joshuadwallach.
Disclosures: Wallach reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.
Perspective
Back to Top
Clinically speaking, the findings do not add that much to the knowledge base. The CV risks associated with rosiglitazone are well established, as are the limitations of the RECORD trial (Home PD, et al. Lancet. 2009;doi:10.1016/S0140-6736(09)60953-3) such as low event rate, which suggested there was no increased risk of MI with rosiglitazone. I do commend the authors, however, on using individual patient-level data and calling for increased clinical trial transparency and data sharing. I believe this aspect of the article was timely and relevant.
Other than reinforcing what we’ve known for many years, these findings don’t really have any implications for clinical practice. The use of thiazolidinediones has declined significantly over the past decade, and with the robust CV and renal data we now have with the SGLT2 inhibitors and GLP-1 receptor agonists, we just have better tools in the toolbox. Only when patients cannot tolerate other therapies (eg, metformin) and newer therapies are unaffordable, should thiazolidinediones even be considered at all.
There have been numerous articles evaluating the rosiglitazone controversy and how the FDA responded, etc. Right or wrong, it ended up being better for patients because we discovered that select antihyperglycemic therapies (ie, SGLT2 inhibitors and GLP-1 receptor agonists) actually reduce CV and renal complications from diabetes. I just feel like it’s time to move on and for the clinical community to focus on implementation of contemporary diabetes management, which doesn’t include thiazolidinediones.
Read more about