27 July 2007
Diabetes Care 2007; 2007: Advance online publication
Thiazolidinediones used to lower blood glucose levels in Type 2 diabetes patients may double the risk of heart failure (HF), review findings indicate.
Based on the review of studies and case reports, researchers estimate that one additional patient with Type 2 diabetes would develop HF for every 50 patients taking one of the drugs over a 26-month period.
"These drugs are currently used by more than 3 million diabetic patients in the USA alone, suggesting that several thousand could be harmed," said lead author Sonal Singh, from Wake Forest University in Winston-Salem, North Carolina, USA.
Singh and colleagues analyzed evidence from randomized controlled trials (RCTs), controlled observational studies, anecdotal case reports, case-series, and spontaneous reports in the Canadian Adverse Events Database (CADRMP).
They identified three RCTs including 10,731 patients who provided numerical information on HF events. A meta-analysis of these trials revealed an odds ratio for HF of 2.1 (p=0.03) in patients receiving a thiazolidinedione compared with those on placebo.
Based on this odds ratio, the estimated number-needed-to-harm with thiazolidinediones would be 50 over a 2.2 year follow-up period, the authors note in the journal Diabetes Care.
Meanwhile, results of four observational studies including 67,382 patients gave a pooled odds ratio for HF of 1.5 (p<0.00001) for those taking thiazolidinediones.
A Dose-Time-Susceptibility analysis of 28 published reports and 214 spontaneous reports from CADRMP showed that HF was more likely to appear after several months, with a median duration for HF onset of 24 weeks (range 1 to 260 weeks). Patients on low doses were also at risk of developing HF soon after treatment.
One patient on low-dose (=15 mg) pioglitazone and nine patients on low-dose (=4 mg) rosiglitazone developed HF within the first 4 weeks of therapy. And HF did not develop much earlier overall in patients receiving high doses of the drugs compared with those on low doses.
"The occurrence of HF several months after initiation of treatment suggests a long-term effect of the drugs, which may not be avoided by beginning with low doses," Singh commented.
Furthermore, the increased risk was not limited to the elderly, with one-quarter of cases of HF occurring in people younger than 60 years old. HF occurred with equal frequency in men and women.
Package inserts for the drugs warn against their use in patients with severe heart failure, and of an increased risk of HF in patients on insulin. But the current analysis indicates that the risk is not confined to patients with HF risk factors or those who are on insulin, the authors report.
"Our findings support current efforts by the [US] Food and Drug Administration to add a black box warning to the labeling of those agents," commented co-investigator Curt Furberg, also of Wake Forest University.