NEW ORLEANS — Findings from the RECORD study confirm the growing consensus that rosiglitazone should not be used in patients with a history of heart failure, with previous problems that might have led to myocardial damage, or in women who are at increased risk for fractures.
However, “Although our evidence is insufficient to rule out a small increased risk of myocardial infarction caused by rosiglitazone when compared with other glucose-lowering agents, it appears that for people with type 2 diabetes, rosiglitazone can be used without concern that there is increased overall cardiovascular morbidity and mortality, or for that matter, all-cause mortality,” study chair Philip D. Home said at the annual scientific sessions of the American Diabetes Association.
“We recommend monitoring for fluid retention,” added Dr. Home, professor of diabetes medicine at Newcastle University, England.
In the prospective, multicenter, open-label Rosiglitazone Evaluated for Cardiovascular Outcomes in Oral Agent Combination Therapy for Type 2 Diabetes (RECORD) trial, funded by GlaxoSmithKline, addition of rosiglitazone (Avandia) to either sulfonylurea or metformin therapy in people with type 2 diabetes increased their risk for heart failure but did not increase the risk of overall cardiovascular morbidity or mortality in a mean follow-up of 5.5 years (Lancet 2009 June 5 [doi:10.1016/S0140-6736(09)60953-3]).
The study was originally designed to evaluate noninferiority of rosiglitazone on cardiovascular outcomes and glucose control, compared with sulfonyl-urea, over a 6-year period. It included a total of 4,447 patients with type 2 diabetes whose glucose levels were inadequately controlled with either sulfonylurea or metformin alone.
Patients in the trial—seen in 364 centers in 25 countries in Europe and Australasia—were aged 40-75 years, had a body mass index greater than 25 kg/m2, and had hemoglobin A1c levels between 7.0% and 9.0% (mean 7.9%) despite maximal doses of either sulfonyl-urea or metformin monotherapy. Patients who were taking sulfonylurea initially were randomized to add-on treatment of either rosiglitazone or metformin (as “active comparator”) while those already on metformin were randomized to the addition of either rosiglitazone or sulfonylurea.
Rosiglitazone was added at a starting doses of 4 mg/day and titrated up as needed to achieve a target HbA1c of 7.0% or less. Any patient who had an HbA1c of 8.5% or more with the two agents was “rescued” with either a third oral agent (if taking rosiglitazone) or transferred to insulin if on metformin/sulfonylurea. A total of 2,220 were randomized to receive rosiglitazone, and 2,227 to the active control group (sulfonyl-urea plus metformin).
The overall rate of the primary end point, cardiovascular hospitalization or cardiovascular death, was 28 per 1,000 person-years, occurring in 321 rosiglitazone patients and 323 active control patients. The hazard ratio of 0.99 met the prespecified noninferiority criteria. Excluding cardiovascular events not of atherosclerotic origin gave similar results (HR 0.97).
The predefined composite secondary end point of cardiovascular death, myocardial infarction, and stroke gave a hazard ratio of 0.93 for rosiglitazone vs. active comparator, suggesting a slight but not statistically significant benefit for rosiglitazone. Cardiovascular death and all-cause death were also slightly in favor of rosiglitazone but not statistically significant (HR 0.84 and 0.86, respectively). The hazard ratio for MI was 1.14—there were eight excess cases with rosiglitazone—but this was also statistically nonsignificant.
Heart failure was highly significantly increased with rosiglitazone, with a hazard ratio of 2.10. Increased heart failure with rosiglitazone has been reported previously in other studies, and in an earlier interim analysis of RECORD (N. Engl. J. Med. 2007;357:28-38).
The overall incidence of participant-reported bone fractures was higher in the rosiglitazone group than in the active control group, with a risk ratio of 1.57, higher for women than men (1.82 vs. 1.23). Fractures of the upper limb and distal lower limb were increased, but hip and femur fractures were not.
In the final results, the statistical power for RECORD was less than initially planned because the overall primary event rate in the 4,447 patients followed for a mean of 5.5 years was substantially lower than anticipated. Reasons for this aren't entirely clear, but an editorial that accompanied the interim analysis of the study suggested that incomplete analysis due to a high loss to follow-up (10%) and the differences in medical care in the various countries where the trial was conducted were possible factors.
Dr. Home disclosed that he is on an advisory panel and speakers bureau for, and receives research support from, GlaxoSmithKline. He also has financial relationships with a variety of other companies that make diabetes-related products.
Hazard ratios for rosiglitazone patients were 2.10 for heart failure and 1.57 for bone fractures. DR. HOME