Retrospective Analysis of Rosiglitazone and Macular Oedema in Patients with Type 2 Diabetes Mellitus

Background: Macular oedema tends to be a more rapid complication of diabetic retinopathy and represents the major cause of blindness. Among subjects with type 2 diabetes mellitus, it can be found in 15% of those who use insulin and 4% of those who do not. Use of thiazolidinediones (glitazones) has recently been associated with some cases of macular oedema.

Methods: We recalled 102 diabetic subjects treated with rosiglitazone to our diabetes centre in order to evaluate a possible association of this drug with macular oedema. Of these 102 subjects, we evaluated all 76 who provided written informed consent to participate in the analysis. All of these underwent a battery of four diagnostic tests: (1) visual acuity, (2) Amsler visual field test, (3) Ishihara colour recognition test, and (4) retinal fundus photography. All retinal photographs were examined by two experienced ophthalmologists.

Results: The most noticeable result was that most subjects (80%) had satisfactory visual acuity. The Ishihara test chart showed that three subjects were colour blind, but this abnormality was already known. On the Amsler test, one subject had a positive result consisting of visual distortion of a series of straight lines. In the retinal photos, two expert ophthalmologists independently identified one case of 'paramacular oedema' in a subject with diabetes of long duration with a proliferative retinopathy. The patient developed bilateral macular oedema during treatment with rosiglitazone 8 mg/day. The patient had been diagnosed with diabetes at the age of 45 years and after a period of 6 years taking oral antihyperglycaemic agents had been switched to insulin, up to four injections per day (total 60-70 IU/day), for the next 15 years. In 2000 a routine examination demonstrated the presence of sustained hypertension and the patient was started on an ACE inhibitor. A computerized test for autonomic neuropathy demonstrated abnormal deep breathing and lying-to-standing responses. Treatment with rosiglitazone was interrupted and the subject underwent a series of retinal photocoagulations for proliferative retinopathy. Two months after rosiglitazone therapy had been discontinued, the visual acuity of the patient reversed to baseline values.

Conclusion: The study shows that rosiglitazone was not linked to formation of macular oedema, with the exception of one case of bilateral and clinically reversible paramacular oedema, where rosiglitazone was given in co-administration with a long-term insulin treatment regimen in a subject with pre-existing diabetic retinopathy. This patient had a long duration of diabetes and had also been hypertensive since 2000.