A study of nearly 3,500 patients indicates no link between thiazolidinedione use and diabetic macular edema, but given case reports of such an association, the findings still must be interpreted with some caution, researchers say.
The authors of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial Eye Substudy say the findings are reassuring yet inconclusive.
“We cannot rule out the possibility of either a modest protective or deleterious association,” wrote Walter T. Ambrosius, Ph.D., of Wake Forest University, Winston-Salem, N.C., and his colleagues in the ACCORD Study Group, explaining that the cross-sectional analysis found an adjusted odds ratio (OR) of 0.97 with a wide confidence interval (0.67-1.40), and did not account for duration of thiazolidinedione exposure, past exposure, or type of exposure.
“A more definitive answer may be provided from the 4-year follow-up data, which will enable us to examine prospectively the relationship between thiazolidinedione exposure and [diabetic macular edema] incidence.”
The ACCORD Eye Substudy is the largest study to date to examine the association between diabetic macular edema (DME) and thiazolidinedione, the authors noted (Arch. Ophthalmol. 2010;128:312-8).
The Eye Substudy involved 3,473 participants from the larger ACCORD trial. Subjects were eligible only if they had no previous laser photocoagulation or vitrectomy for diabetic retinopathy in either eye. Participants had a mean age of 62 years.
DME was scored separately for each eye on a scale of 0 (none) through 3 (retinal thickness or adjacent hard exudates within 500 mcm of the center of the macula). A score of 1 was considered questionable, while 2 was defined as a zone of retinal thickness 1 disk area or greater and within 1 disk diameter or less from the center of the macula. Eyes graded at level 2 or 3 were considered to have clinically significant macular edema.
Thiazolidinedione use was defined as self-reported current use of rosiglitazone or pioglitazone at baseline. The duration of exposure before baseline was not known, however inclusion criteria for the main ACCORD trial required an antihyperglycemic washout period of 3 months before the start of the study.
Among the 3,473 participants, 695 (20%) had used thiazolidinediones, and 217 (6.2%) had DME. In the adjusted analysis, thiazolidinedione use was not significantly associated with DME (OR 0.97), nor were hemoglobin A1c, duration of diabetes, gender, or ethnicity. Significant association was found between thiazolidinedione and both retinopathy and age.
Former and current smokers had lower prevalence of DME than did those who had never smoked, a finding that “is perhaps attributable to chance or the inclusion process,” wrote the authors, citing a previous study that contradicts this finding (Invest. Ophthalmol. Vis. Sci. 1998;39:233-52).
Thiazolidinedione use was also associated with marginally better visual acuity (P = .009), although “we do not know whether this finding is clinically significant,” they wrote.
The authors pointed out several limitations to the findings, including the possibility that “perhaps longer-term exposure to a thiazolidinedione is necessary for risk to develop.
Disclosures: The study was funded by the National Eye Institute and the National Heart, Lung, and Blood Institute. Dr. Ambrosius disclosed no conflicts. Among his associates, Dr. Hertzel C. Gerstein has received honoraria and grants from GlaxoSmithKline for speaking, consulting, and research related to thiazolidinediones and/or rosiglitazone. Since 2005, the University of North Carolina, Chapel Hill, has contracted with pharmaceutical companies for Dr. John B. Buse's research or consulting on thiazolidinediones and related compounds. Dr. David C. Goff Jr. has received research funding from Merck and Co. for a trial involving the glucose-lowering medication sitagliptin.
Reflectance map shows edema associated with diabetic retinopathy.
Source Image courtesy Heidelberg Engineering