Background: Methylxanthine derivatives are vasodilators. They also inhibit platelet aggregation and thromboxane A2 synthesis, decrease the release of free radicals and may be neuroprotective. NOTE: This review covers an area where no active research is taking place. It will be updated if relevant information becomes available, e.g. on completion of an appropriate study.
Objectives: To assess the effect of intravenous or oral methylxanthines (pentoxifylline, propentofylline, or pentifylline) in patients with acute ischaemic stroke.
Search strategy: We searched the Cochrane Stroke Group trials register (last searched November 2003). For the first version, we also searched EMBASE (1980 to 1999), MEDLINE (1966 to 1999), Science Citation Index (1981 to 1999) and the Ottawa Stroke Trials Registry. We also contacted the manufacturers of methylxanthines and the principal investigators of the identified trials.
Selection criteria: Randomised trials comparing pentoxifylline, propentofylline or pentifylline with placebo or control in patients with definite or presumed acute ischaemic stroke. Trials were included if treatment was started within one week of stroke onset.
Data collection and analysis: Two reviewers independently applied the inclusion criteria. Trial quality was assessed.
Main results: Five trials were included. Four trials tested pentoxifylline in 763 people, and one tested propentofylline in 30 people. No trials of pentifylline were found. The odds of early death (within four weeks) was non-significantly reduced in patients given a methylxanthine drug as compared with those given placebo (odds ratio (OR) 0.64, 95% confidence interval (CI) 0.41 to 1.02). This non-significant trend to less deaths was due mainly to one pentoxifylline trial that found a highly significant reduction in early deaths. Two trials reported early death or disability and found a non-significant reduction (OR 0.49, 95% CI 0.20 to 1.20). There was no significant difference in late death (beyond four weeks), as reported in the propentofylline trial involving 30 patients, although the confidence interval was wide (OR 0.70, 95% CI 0.13 to 3.68). Data for neurological impairment and disability were not in a form suitable for analysis. Data on quality of life, stroke recurrence, thromboembolism and bleeding were not reported.
Reviewers' conclusions: There is not enough evidence to assess adequately the effectiveness and safety of methylxanthines after acute ischaemic stroke.