|
Comparative cardiovascular effects of thiazolidinediones: systematic review and meta-analysis of observational studies |
Loke YK, Kwok CS, Singh S |
|
|
CRD summary This review concluded that in patients with type 2 diabetes, rosiglitazone was associated with significantly higher odds of congestive heart failure, myocardial infarction, and death, than pioglitazone, in real-world settings. This was a generally well-conducted review. The authors acknowledged the methodological limitations of the evidence available and the conclusions are likely to be reliable. Authors' objectives To compare the effects of rosiglitazone and pioglitazone on myocardial infarction, congestive heart failure, and mortality in patients with type 2 diabetes. Searching MEDLINE and EMBASE were searched, without language restrictions, for articles included from their inception to September 2010; search terms were reported. The websites of the US Food and Drug Administration and European Medicines Agency and study registers were searched and the drug manufacturers were contacted. Bibliographies of the included studies and recent review articles were screened for additional articles. Study selection Controlled observational studies of cohort or case-control design were eligible for inclusion if they reported the cardiac outcomes, for patients with type 2 diabetes mellitus, who were receiving rosiglitazone compared with those receiving pioglitazone. The primary outcome was myocardial infarction; secondary outcome measures were congestive heart failure and overall mortality.
Where reported, the mean age of participants ranged from 54 to 76 years (most participants were over 60 years old) and 26% to 74% of them were male. Thiazolidinedione treatment lasted between 215 and 450 days. Most studies excluded patients who were on insulin and those with a range of comorbid conditions. Drug use was primarily from pharmacy claims databases and the outcomes were assessed by international classification of diseases (ICD) code data.
Two reviewers independently screened studies for inclusion; discrepancies were resolved by checking and consensus or by consultation with a third reviewer. Assessment of study quality Study quality was assessed for participant selection, baseline characteristics, adjustment for confounders, nature of follow-up, ascertainment of drug use, and definition and monitoring of adverse outcomes. The authors did not report the number of reviewers who assessed quality. Data extraction Odds ratios and 95% confidence intervals, or the data required to calculate these, were extracted for the outcomes of interest. Where the relevant cardiovascular outcomes were potentially measured but not reported or were stated to be non-significant and where clarification was needed, the authors were contacted. Where different regimens of thiazolidinedione were reported, the data from participants who were more recently treated were preferred. The risk estimates for the entire cohort rather than specific subgroups were extracted.
Two reviewers independently extracted the data; discrepancies were resolved by checking and consensus or consultation with a third reviewer. If necessary, authors were contacted for clarification. Methods of synthesis Pooled odds ratios and 95% confidence intervals were calculated using a random-effects model, weighted by inverse variance; a fixed-effect model was used as a sensitivity analysis. The analyses were subgrouped by whether studies reported the adjusted odds ratios or not. The number-needed-to-harm was calculated. Heterogeneity was investigated using the I2 statistic, with values between 30% and 60% considered to be moderate. Sensitivity analyses were conducted by excluding two specific studies. Publication bias was investigated using a funnel plot. Results of the review Sixteen studies met the inclusion criteria (n=810,000; range 697 to 227,571); 12 were retrospective cohort and four were case-control studies. The ascertainment of drug use and outcome assessment were similar across studies. Most studies adjusted for confounding factors, but these factors varied widely across the studies. None of the studies reported the severity and consequences of adverse events. Follow-up ranged from 105 days to 7.1 years.
Compared with pioglitazone, rosiglitazone resulted in a significant increase in the incidence of myocardial infarction (OR 1.16, 95% CI 1.07 to 1.24; 15 studies), and congestive heart failure (OR 1.22, 95% CI 1.14 to 1.31; eight studies), and the overall mortality (OR 1.14, 95% CI 1.09 to 1.20; eight studies). Moderate heterogeneity was observed for the analyses of myocardial infarction and congestive heart failure.
None of the sensitivity analyses significantly altered the results and the funnel plots were reported to show no evidence of publication bias. Authors' conclusions In patients with type 2 diabetes, rosiglitazone was associated with significantly higher odds of congestive heart failure, myocardial infarction, and death than pioglitazone, in real-world settings. CRD commentary This review addressed a clear research question, which was supported by appropriate inclusion criteria. The search for published studies was limited to two databases, but language restrictions were not applied and unpublished data were sought. Study selection and data extraction were conducted in duplicate, but it was unclear whether similar methods to reduce error and bias were used for quality assessment. Statistical heterogeneity was moderate for two of the analyses and clinical heterogeneity was evident in the reported study details, but the direction of the effect was fairly consistent across studies. The authors stated that the number-needed-to-harm was calculated, but this was done using incidence data from studies that were not included in the review.
This was a generally well-conducted review. The authors acknowledged the methodological limitations of the evidence available and the conclusions are likely to be reliable. Implications of the review for practice and research Practice: The authors stated that clinicians, patients, and regulatory authorities should carefully consider these results in the context of the available information on the benefits of thiazolidinediones for glycaemic control and the harm of various outcomes.
Research: The authors stated that studies were needed to investigate the additional potential adverse events, such as bladder cancer and fractures. Funding National Centre for Research Resources, grant number 1KL2RR025006-03. Bibliographic details Loke YK, Kwok CS, Singh S. Comparative cardiovascular effects of thiazolidinediones: systematic review and meta-analysis of observational studies. BMJ 2011; 342:d1309 Other publications of related interest Singh S, Loke YK, Furberg CD. Long-term risk of cardiovascular events with rosiglitazone: a meta-analysis. JAMA 2007; 298(10): 1189-1195.
Singh S, Loke YK, Furberg CD. Thiazolidinediones and heart failure: a teleo-analysis. Diabetes Care 2007; 30(8): 2148-2153.
Loke YK, Singh S, Furberg CD. Long-term use of thiazolidinediones and fractures in type 2 diabetes: a meta-analysis. CMAJ: Canadian Medical Association Journal 2009; 180(1): 32-39. Indexing Status Subject indexing assigned by NLM MeSH Diabetes Mellitus, Type 2 /drug therapy /mortality; Heart Failure /chemically induced /mortality; Humans; Hypoglycemic Agents /adverse effects; Myocardial Infarction /chemically induced /mortality; Prognosis; Publication Bias; Risk Factors; Thiazolidinediones /adverse effects AccessionNumber 12011001823 Date abstract record published 30/03/2011 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|