Pentoxifylline for heart failure: a systematic review.

BACKGROUND Recent trials have indicated a beneficial effect of pentoxifylline on measures of inflammation and markers of cardiac dysfunction in people with heart failure. However, it is uncertain whether pentoxifylline should be used routinely in the management of heart failure. OBJECTIVE To determine the effectiveness of pentoxifylline in heart failure. DESIGN Systematic review of randomised controlled trials. METHODS We searched MEDLINE (1 January 1966 - 20 November 2004), the Cochrane Controlled Trials Register (issue 4, 2004), and reference lists of related papers, for randomised controlled trials of pentoxifylline in the treatment of heart failure. Prospective, randomised, double-blind controlled trials were sought for inclusion in the study. The two reviewers independently assessed trial quality and extracted data, which were analysed using RevMan statistical software. The following outcome measures were evaluated: (i) New York Heart Association (NYHA) functional class; (ii) left ventricular ejection fraction (LVEF); (iii) frequency of hospitalisation; and (iv) death from all causes. RESULTS Four studies with a total of 144 participants met the inclusion criteria. Statistical pooling (or meta-analysis) was not performed owing to the significant clinical heterogeneity and differences in reporting of the outcomes in the included studies; instead, the trials were analysed separately for the outcomes of interest. The four studies tested the use of pentoxifylline versus placebo in patients with heart failure of varying aetiology (idiopathic dilated cardiomyopathy, 3 studies; ischaemic cardiomyopathy, 1 study). In 2 of the idiopathic dilated cardiomyopathy studies, patients were classified as NYHA class II or III, while the study population in another idiopathic cardiomyopathy study was in NYHA class IV. The trial of patients with ischaemic cardiomyopathy included patients in NYHA functional classes I - IV. The use of pentoxifylline was associated with significant improvement in symptoms (i.e. NYHA functional class) and cardiac function (i.e. LVEF) in 3 out of 4 studies. The beneficial effect on symptoms of heart failure and cardiac function was seen in all grades of severity of heart failure and in patients with ischaemic and idiopathic dilated cardiomyopathy. All 4 studies showed a trend towards reduction of mortality, but this effect was not statistically significant. The effect of pentoxifylline on the frequency of hospitalisation has not been tested in randomised controlled trials. INTERPRETATION Pentoxifylline may have a beneficial effect on NYHA functional class, ejection fraction and mortality in heart failure, but published trials are too small to provide conclusive evidence. There is a need for large, placebo-controlled trials of pentoxifylline in heart failure, involving a diverse group of patients with regard to cause and severity of heart failure.