A 51-year-old African-American man with type 2 diabetes (anti-GAD antibody negative and C-peptide positive) of 7 years duration was controlled (A1c 7.0%) on a combination of 1 g metformin, 2 mg glimeperide, and 4 mg rosiglitazone all twice daily and 40 units Glargine insulin at bedtime. He had been taking rosiglitazone for 26 months and troglitazone for 13 months before changing to rosiglitazone.
The patient suddenly developed painful swelling of the right leg. On examination there was no edema in the left leg and pitting edema to just below the knee in the right leg, which had an increased temperature. However, no tenderness was found on deep palpation of the calf, popliteal, and femoral areas or in the adductor canal. Two venous ultrasounds 5 days apart showed no evidence of a deep vein thrombosis, and furosemide self-administered by the patient had no effect on the edema. In addition, the neck veins were not distended, hepatojugular reflux was negative, chest was clinically clear, and there was no hepatomegaly or added heart sounds. Serum albumin and liver profile were normal, and there was no albuminuria.
The patient’s wife, having consulted the Physicians Desk Reference, felt that the edema was due to rosiglitazone and insisted that her husband discontinue the medication. Within 3 days of discontinuing rosiglitazone the edema disappeared. On rechallenge the edema reappeared within 5 days and the thiazolidinedione was permanently discontinued with resolution and no recurrence of edema.
Dependent edema is a side effect of thiazolidinediones due to increased plasma volume, insulin-induced vasodilatation, and increased production of vascular endothelial growth factor (1–4). However, the distribution of the edema is almost invariably bilateral. Because of unilateral edema in this case, the diagnosis of thiazolidinedione-induced dependent edema was not considered. Since the edema resolved with withdrawal of the thiazolidinedione and reappeared on rechallenge and disappeared after withdrawal, never to return, it was the obvious cause of this unilateral edema. There was no anatomical reason, such as venous insufficiency from varicose veins, or a neurological cause, such as a previous stroke, childhood poliomyelitis, or previous trauma to the leg to explain this unilateral edema. Why an edema with systemic involvement should present unilaterally is not known or gleaned from this case.
In conclusion, practitioners need to be aware that dependent edema due to thiazolidinediones can present with unilateral edema, which will disappear with discontinuation of the thiazolidinedione.
D.S.H.B. is a member of the National Speakers’ Bureau and Advisory Panel of and is a consultant for GlaxoSmithKline Pharmaceutical, and has received honoraria and consultant fees from Eli Lilly.
