Effects of liposomal amphotericin B versus an amphotericin B lipid complex on liver histopathology in patients with hematologic malignancies and invasive fungal infections: a retrospective, nonrandomized autopsy study.

BACKGROUND: Abnormal liver function test (LFT) results are common in patients with hematologic abnormalities, making the assessment of drug-related hepatotoxicity difficult. Studies based on elevated LFT levels have found that use of liposomal amphotericin B (L-AMB) was associated with increased hepatotoxicity compared with amphotericin B (AMB)/deoxycholate or amphotericin B lipid complex (ABLC). Because LFT abnormalities are multifactorial in severely immunocompromised patients, uncertainty remains regarding the clinical significance of these laboratory findings.

OBJECTIVE: The aim of this study was to present the hepatic histopathologic findings on autopsy in patients who had hematologic malignancies and fungal infections and had received L-AMB or ABLC.

METHODS: This study was conducted at The University of Texas M.D. Anderson Cancer Center, Houston, Texas. Records from 1995 to 2004 of patients who had received L-AMB or ABLC for > or =7 days, within 30 days before death, were reviewed by 1 investigator. Hepatic autopsy slides were independently reviewed by another investigator (pathologist) in a blinded fashion. Histopathologic evidence of amphotericin-related hepatotoxicity was predetermined based on histopathologic abnormalities reported in animal studies (eg, macrophage vacuolation, multifocal hepatocellular necrosis). Based on data from animal studies and in view of the lack of studies in humans, multifocal necrosis, fatty infiltration, macrophage vacuolation, and/or "foamy macrophage" accumulation were all considered histopathologic abnormalities associated with the use of lipid formulations of AMB.

RESULTS: Data from 64 patients were included (32 patients per group). The demographic characteristics were comparable between the ABLC and L-AMB groups (median ages, 47.5 and 53.0 years, respectively; male, 44% and 53%; white, 75% and 78%; median weight, 67 and 78 kg; active underlying malignancy, 84% and 78%). There were no significant between-group differences in cumulative dose (6 and 7 g), median daily dose (both, 5 mg/kg), or median duration of treatment (19.5 and 19.0 days). Abnormal results (>5 x from baseline) on LFT were found in 12 (38%) and 10 (31%) patients who received ABLC and L-AMB, respectively, but these findings were thought to be associated with concomitant use of triazoles (4/12 [33%] and 1/10 [10%] patients, respectively), hepatotoxic antibiotics (8/12 [67%] and 5/10 [50%]), and/or other hepatotoxic medications (2/12 [17%] and 1/10 [10%]). Nonspecific abnormalities were observed on histopathology in 94% of patients. There was no evidence of histopathologic abnormalities reported in animal toxicity studies of lipid AMB, such as macrophage vacuolation or multifocal hepatocellular necrosis.

CONCLUSIONS: Although abnormal results on LFT and/or histopathologic changes in liver were found in 92% of these debilitated patients with hematologic malignancy, direct histopathologic evidence of toxicity associated with lipid formulations of AMB was not established in our study.