Diabetes Drug Linked to Higher Mortality in Medicare Patients

By Peggy Peck
BOSTON, 01 dez 2008-- Medicare patients started on a thiazolidinedione for diabetes had a higher mortality rate and were more likely to develop congestive heart failure if given rosiglitazone (Avandia) than pioglitazone (Actos), researchers here reported.
Analysis of data from prescription records of 28,361 patients showed that rosiglitazone use was associated with a 7% to 15% increased risk of mortality compared with pioglitazone therapy, Wolfgang C. Winkelmayer, M.D., Sc.D., of Brigham and Women's Hospital, and colleagues reported in the November 24 issue of Archives of Internal Medicine.
The analysis also found an 11% to 13% increased risk of developing congestive heart failure in the rosiglitazone arm, but, importantly, no increase in the rate of myocardial infarction or stroke.
This latest analysis comes on top of negative findings from the APPROACH trial and the decision by the American Diabetes Association to exclude rosiglitazone from its latest treatment guidelines. (See AHA: Rosiglitazone Fails to Slow Plaque Progression http://www.medpagetoday.com/MeetingCoverage/AHA/11761 )
"This study confirms the safety concerns that have been raised for rosiglitazone compared with pioglitazone, which, in turn, also cannot be considered a very safe drug given its well-documented effect on the risk of CHF," the authors wrote.
GlaxoSmithKline, maker of rosiglitazone, continued to defend the safety and efficacy of the drug.
In a prepared statement, the company said the new study was "inconsistent with evidence from randomized clinical trials and has significant limitations.
"The primary outcome in this observational analysis is all cause mortality. The [rosiglitazone] prescribing information includes data from RECORD, an ongoing, long-term randomized clinical trial that has shown no statistically significant differences between the [rosiglitazone] group and the control group regarding death from cardiovascular causes or any cause."
Although the company cited the RECORD study, that trial compared rosiglitazone plus metformin with a sulfonyurea plus metformin--it did not compare rosiglitazone with pioglitazone, as is the case in the Winkelmayer study.
The company added that the results of the study by Winkelmayer et al might be "biased due to imbalances in comorbid conditions (cardiovascular disease, chronic obstructive pulmonary disease and malignancies) between the two treatment groups."
Finally GlaxoSmithKline said, "long-term, randomized clinical trials are considered to be the gold standard for answering safety questions and making clinical decisions about prescription medicines." It urged clinicians to await results of two such trials -- RECORD and BARI2D.
But unlike other analyses that have suggested an increased risk of MI and stroke with rosiglitazone, this study found no such association.
Faced with that apparent dilemma, Dr. Winkelmayer and colleagues devised a guilt-by-supposition rather than a guilt-by-association scenario.
They reasoned that, because cardiovascular disease "represents more than 75% of mortality in patients with diabetes, there must almost certainly be a link."
The explanation, they wrote, was that many of the deaths were from MI or stroke, but those "presumably cardiovascular deaths in our cohort of elderly patients may have occurred suddenly or before the diagnosis was established."
The study findings, therefore, "suggest a higher cardiovascular case fatality rate for rosiglitazone."
Leading Dr. Winkelmayer to conclude that the "difference in all cause mortality may be even more important to consider in elderly patients."
All patients in the study were 65 or older and all participated in state-sponsored prescription drug plans in New Jersey or Pennsylvania. All patients initiated rosiglitazone or pioglitazone therapy from January 1, 2000 through December 31, 2005.
Slightly more patients (50.3%) were initiated on rosiglitazone. The baseline characteristics of the two groups were similar, although rosiglitazone users were more likely to have a history of coronary artery disease and congestive heart failure and were less likely to be on beta blocker and statin therapy prior to the index date.
Rosiglitazone users were also more likely to be nursing home residents and more likely to have been hospitalized in the previous six months.
Among the findings:
There were 1,869 deaths during 29,060 person-years of follow-up
Median time exposed to the study drug was 217 days for pioglitazone and 215 days for rosiglitazone
The on-drug incidence rate ratio for all-cause mortality was 1.15 (95% CI, 1.05-1.26) for rosiglitazone
Constant-exposure incidence rate ratio for all-cause mortality was 1.07 (95% CI 1.01-1.14)
The on-drug incidence rate ratio for CHF hospitalization was 1.13 (95% CI 1.01-1.26) for rosiglitazone
Constant-exposure incidence rate ratio for CHF hospitalization was 1.11 (95% CI 1.03-1.19)
The authors noted a number of limitations of the study ranging from the lack of randomized data to the lack of laboratory data on glycemic control.
Those limitations were, however, balanced by the large database of lower-middle class senior citizens that generated "a large cohort of new TZD users," the researchers said.
Moreover, by "focusing on the specific comparison between two similar drugs that were perceived as equal at the time, we were able to avoid dealing with unobserved confounding that is arguably present in studies comparing TZDs with other diabetes regimens," they wrote.
The study was supported by the American Heart Association, Satellite Healthcare, Amgen, Fresenius Medical Care, and GlaxoSmithKline.
Dr. Winkelmayer disclosed that he served as an unpaid member of advisory boards of Amgen, Roche, Genzyme, and Fresenius.
Primary source: Archives of Internal MedicineSource reference:Winkelmayer WC et al "Comparison of Cardiovascular Outcomes in Elderly Patients with Diabetes Who Initiated Rosiglitazone vs. Pioglitazone Therapy" Arch Intern Med 2008; 168: 2368-2375