The association of amphotericin B with nephrotoxicity is well known, but ri
sk factors for this complication are not well characterized. One hundred an
d seventy-eight patients who received >3 days of intravenous amphotericin B
and a minimal total cumulative dose >100 mg were reviewed retrospectively.
The mean age, average cumulative dose of amphotericin B and duration of th
erapy were 46 +/- 22 years, 536 +/- 547 mg and 16.6 +/- 8.2 days, respectiv
ely. Eighty-six percent of patients received amphotericin B for empirical t
herapy of febrile neutropenia. Various definitions of nephrotoxicity were u
sed; these were as follows (the incidence of nephrotoxicity as determined b
y the given definition is given in parentheses): definition 1, a change in
creatinine of >46 mu mol/L over baseline (50%); definition 2, a doubling of
creatinine over baseline (49%); definition 3, a change in creatinine of >9
2 mu mol/L (29%); definition 4, a doubling and/or a change in creatinine of
>92 mu mol/L (49%); definition 5, an increase in creatinine to >230 mu mol
/L (8%). Multivariate analysis showed that nephrotoxicity was associated wi
th a greater cumulative dose of amphotericin B and receipt of concomitant n
ephrotoxic drugs for all definitions (P < 0.05). In those patients who expe
rienced severe nephrotoxicity (creatinine increased to >230 mu mol/L), cycl
osporin therapy was the most significant risk factor (odds ratio 18.8, P =
0.022). Haemodialysis was necessary in one patient, but multiple concomitan
t risk factors for renal dysfunction were present. No patient experienced i
rreversible nephrotoxicity. These findings allow for stratification of pati
ents at risk for amphotericin B-induced nephrotoxicity and rational use of
alternative agents.