Liposomal nystatin (LNY) vs. amphotericin B (AMB) for Candidemia in non-neutropenic patients: A historical comparison

Candidemic, non-neutropenic pts. (N=43)were treated with open-label LNY @ 2 mg/kg/day and the results compared to published data on 103 non-neutropenic pts. treated with AMB @ 0.5 mg/kg/day (NEJM 331:1325,1994). The groups were similar: mean ± SE age 60 ± 2 vs. 60 ± 2 (LNY vs. AMB), cancer 23 vs. 31%, diabetes meflitus 26 vs. 22%, corticosteroids 26 vs. 23%, APACHE II (AP2) score at enrollment 13 vs. 16 (p for AP2 comparison = 0.02, p for all others > 0.05). Both studies required candidemia within 96 h of enrollment, neutrophil counts ≥ 1000/mm³, and fever or other signs of sepsis. To be considered a success, resolution of entry signs of fungal infection and negative fungal blood cultures were required. To increase comparability, pts. who died within 14 days post therapy were considered failures and pts. receiving a concomitant antifungal (AF) medication for other than candidemia were considered not evaluable. Results at 6 weeks. LNY pts.: 20 success, 10 failure, 13 unevaluable. AMB pts.: 68 success, 34 failure and 16 unevaluable. For LNY, 12 of 13 unevaluable pts. had no cultures taken, were missing temp, information, or were taking a concomitant AF medication. Success rates were 20/30 (67%) for LNY and 68/102 (67%) for AMB, for a difference of 0%, (95% CI=20 to 20%). During followup, 10 (23%) LNY pts. vs. 41 (40%) of AMB pts. died (p=0.056). Significant renal toxicity was judged by a review panel to have occurred in 37% of AMB pts. Among patients treated with LNY, 6/43 (14%) experienced a doubling in creatinine. Conclusions: (1) These studies used similar eligibility and could be scored using similar outcome criteria. Although limited by the inherent biases of all open-label studies, a meaningful historical comparison is thus possible. (2) LNY and AMB appear to have similar efficacy for candidemia in non-neutropenic pts. (3) LNY may have less nephrotoxicity than AMB.