Efficacy and safety of amphotericin B formulations: a network meta-analysis and a multicriteria decision analysis

Abstract Objectives Despite its broad spectrum, conventional amphotericin B (AB) is associated with serious adverse events. Lipid-based formulations may offer safer options. We aimed to synthesize the evidence of efficacy and safety of AB formulations. Methods We performed a systematic review and network meta-analysis (NMA) to compare all available formulations: conventional AB; lipid complex or ABLC; colloidal dispersion or ABCD; liposomal or LAB; AB in Intralipid. Randomized controlled trials were searched in four databases. Cure, fever, chills, nephrotoxicity, death and drug discontinuation were assessed. NMA was based on Bayesian methods accounting for direct and indirect comparisons. Probability ranks estimating the best formulation were built for each outcome. The relative benefit–risk of formulations was assessed with stochastic multicriteria acceptability analyses (SMAA). Key findings We identified 25 trials (n = 2996). No significant differences among drugs were observed for cure or death. All lipid-based formulations were safer than conventional AB for nephrotoxicity. AB-Intralipid was more tolerable than conventional AB and caused less chills than ABCD. AB-Intralipid was the best therapy (>60%) regarding nephrotoxicity, fever, chills and discontinuation. The scenario from SMAA favoured AB-Intralipid (81% acceptability). Conventional AB was secondary to all lipid-based formulations. Conclusions Amphotericin B-Intralipid was identified as safer, cost-saving treatment in comparison with other formulations.