Amikacin and gentamicin accumulation pharmacokinetics and nephrotoxicity in critically ill patients

Abstract

Twenty-five critically ill adults receiving blood level-adjusted doses of amikacin were prospectively studied with serum, urine, and, when possible, tissue amikacin concentrations. These data were fitted to a two-compartment pharmacokinetic model. Prolonged urine collections or postmortem tissues (or both) were used to confirm predicted tissue accumulation. Nephrotoxicity was also investigated. Patients were defined as having renal damage if they showed an increase in serum creatinine of greater than 0.5 mg/100 ml, an increase in urine beta 2-microglobulin of greater than 50 mg/day, and presence of urinary casts of greater than 500/ml. Renal damage was attributed to amikacin if there was, in addition to the above, tissue accumulation of amikacin of greater than 600 mg. These patients were matched with 25 patients treated with gentamicin during the same time period. There were no statistical differences between the gentamicin- and amikacin-treated patients in age, sex, weight, base-line creatinine clearance, concurrent cephalosporins or diuretics, treatment duration, site of infection, normalized (amikacin/gentamicin dosing ratio of 3:1) total dose, mortality, or tissue accumulation. More amikacin-treated patients (19 of 25) than gentamicin-treated patients (9 of 25) received prior aminoglycosides (P less than 0.01). The only pharmacokinetic parameter that differed between amikacin and gentamicin was a greater K21 for gentamicin. Nephrotoxicity was observed in 4 gentamicin was a greater K21 for gentamicin. Nephrotoxicity was observed in 4 gentamicin-treated patients (16%) and 5 amikacin-treated patients (20%). At a 3:1 dosing ratio, there were no significant differences between amikacin and gentamicin two-compartment pharmacokinetics and nephrotoxic potential in matched critically ill patients, but the trend of these data showed greater amikacin tissue accumulation. However, at an amikacin/gentamicin dosing ratio of 4:1, their tissue accumulation potential appeared to be almost identical.