Background and objectives: The recent significant increase in the number of immigrants entering the European Union from South and Central America means that chronic Chagas' disease is an increasingly frequent diagnosis among immigrants in Europe. Our objectives were to evaluate published evidence on the treatment of chronic Chagas' disease with benznidazole and on the potential benefits of this drug in the chronic phase of the disease.
Methods: We performed a systematic review and meta-analysis by means of an electronic search of the published literature, with no language restrictions, until October 2008. We included studies on chronically infected patients of any age who were in the indeterminate phase or had visceral involvement and for whom treatment with benznidazole was compared with placebo or no treatment. The primary endpoint was response to therapy (whether serological, parasitological or clinical), as it was measured in each of the studies included. Clinical response to therapy was also analysed.
Results: We identified 696 studies, from which we chose 9: 3 clinical trials and 6 observational studies. Compared with placebo or no treatment, benznidazole increases 18-fold the probability of a response to therapy [global odds ratio (OR), 18.8; 95% confidence interval (CI), 5.2-68.3]. This effect was mainly observed in clinical trials (OR, 70.8; 95% CI, 16-314), whereas in observational studies it was much less marked (OR, 7.8; 95% CI, 2.1-28.9), and even less so when only observational studies in adults were considered (OR, 6.3; 95% CI, 1.6-24.7). Patients treated with benznidazole had a significantly lower risk of clinical events (OR, 0.29; 95% CI, 0.16-0.53). Up to 18% of patients discontinued treatment due to toxicity (cutaneous reactions followed by gastrointestinal disturbances); this was less common in children than in adults.
Conclusions: Analysis of available information reveals that the efficacy of treatment in late chronic infection is doubtful. Although data generally point to a beneficial effect, this could be marginal. This uncertainty is largely the result of differences in the study populations, endpoints and follow-up periods, and the fact that almost all of the information on treatment in the late chronic phase comes from non-randomized studies.