Randomised controlled trials (RCTs) with an intervention arm that received oral pentoxifylline and in adults with diabetic kidney disease were eligible for inclusion in the review. Diabetic kidney disease was defined as albuminuria with greater than 30mg/day of albumin, or glomerular filtration rate of less than 60mL/min/1.73m2. Trials that included patients with renal transplants or acute renal failure were excluded.
The primary outcome was change in proteinuria, stratified by whether pentoxifylline comparison was with rennin-angiotensin system blockade. Secondary outcome measures were changes in systolic or diastolic blood pressure, and glomerular filtration rate (defined as change in creatinine clearance, measured or calculated).
Included trials compared pentoxifylline (with or without angiotensin-converting enzyme inhibitors or angiotensin receptor blockers ) with placebo, standard care, angiotensin-converting enzyme inhibitors (including captopril, ramipril, and lisinopril), or angiotensin receptor blockers. Pentoxifylline doses ranged from 400mg/day to 1,200mg/day; duration of therapy ranged from two to 12 months (medium six months). Most trial populations included patients with diabetes with overt proteinuria and/or microalbuminuria; one trial included patients with diabetes with proteinuria and chronic renal failure. The mean age of included participants ranged from 25 to 65 years. Trials were conducted in Europe, USA, Mexico, Brazil and Iran.
Two reviewers independently assessed studies for inclusion. Disagreements were resolved through discussion.