Table 1—

RTCs of TZDs that evaluated heart failure outcomes

Study (ref.)Heart failure rates in the treatment armHeart failure rates in the control armParticipantsDrug and daily dosageDuration of treatmentCase ascertainment and definition
DREAM 2006 (10)14 of 2,6352 of 2,634Impaired glucose tolerance and/ or fasting glucose, not on hypoglycemic medication. Patients with heart failure excluded.8 mg rosiglitazone3 years (median)Active surveillance. Confirmed heart failure was secondary outcome adjudicated by blinded independent monitoring committee. Prespecified definition was patient requiring acute treatment plus 2 of 3 criteria from: 1) typical signs/symptoms; 2) radiologic findings; 3) use of diuretics, vasodilators, or inotrope.
ProActive 05 (18)281 of 2,605198 of 2,633Type 2 diabetic patients with vascular disease. Less than one-third on insulin. Excluded if greater than NYHA Class II.15–45 mg titrated pioglitazone34.5 monthsActive surveillance but no independent monitoring. We extracted data on all reported heart failure events. Post hoc independent blinded adjudication using clinical, radiologic, and laboratory criteria confirmed accuracy of original diagnoses.
GlaxoSmithKline 211 (17)19 of 11010 of 114Type 2 diabetic patients with NYHA Class I or II; single blinded. No information on concomitant medications.4–8 mg titrated rosiglitazone12 monthsActive surveillance and worsening heart failure a prespecified secondary outcome. We extracted data on all investigator-reported cardiac failure events. No information on diagnostic criteria or adjudication.