Comparison of nephrotoxicity associated to different lipid formulations of amphotericin B: a real-life study

Amphotericin B (AmB) use is limited by the occurrence of kidney toxicity. Here, we evaluated the incidence and impact of nephrotoxicity in a large series of patients receiving therapy with amphotericin B deoxycholate (d-AmB), liposomal AmB (L-AmB), or AmB lipid complex (ABLC), in a clinical practice scenario. In a retrospective cohort study, patients treated with different AmB formulations between 2003 and 2012 were evaluated. Medical records and laboratory data were reviewed. Nephrotoxicity was determined according to modified RIFLE criteria. Predictors of nephrotoxicity and mortality were determined and treatment groups were compared. About 431 patients were studied (d-AmB, n = 236; L-AmB, n = 105; ABLC, n = 90). Frequency of severe nephrotoxicity (RIFLE 'Failure') was 11.5%, 2.4% and 7.2% for d-AmB, L-AmB and ABLC, respectively (P = 0.046). Use of L-AmB was found to be an independent protective factor (OR: 0.18; 95% CI: 0.03-0.64; P = 0.006) for severe nephrotoxicity, considering d-AmB as a reference. L-AmB was also a protective factor for mortality (OR: 0.56; 95% CI: 0.32-0.99; P = 0.046). In addition, in-hospital overall mortality was associated with cancer, previous dialysis, evolution to dialysis, and stay in the intensive care unit. Patients treated with ABLC showed similar frequency of severe kidney toxicity than those treated with d-AmB. L-AmB was associated with better outcomes than other formulations, including severe nephrotoxicity and overall mortality.