Comparative nephrotoxicity of gentamicin and tobramycin: pharmacokinetic and clinical studies in 201 patients

Abstract

A total of 201 critically ill patients were studied during 267 courses of gentamicin or tobramycin treatment (139 gentamicin courses and 128 tobramycin courses). Of these 267 courses, pharmacokinetic and clinical data were obtained for 240 (120 gentamicin and 120 tobramycin). The data collected for pharmacokinetic analysis included measurements of serial blood and urine levels, urinary excretion of beta 2-microglobulin, protein levels, and granular casts. A two-compartment model was used to assess tissue accumulation, and in 89 courses the predicted accumulation was confirmed by cumulative urine collection or postmortem tissue analysis. As groups, the patients given gentamicin and tobramycin did not differ in age, weight, creatine clearance, total dose given, duration of treatment, initial aminoglycoside through serum levels, number of dosage adjustments, concurrent use of furosemide, or concurrent cephalosporins. Previous aminoglycoside treatment (usually gentamicin) had occurred more frequently in the tobramycin treated patients (P less than 0.01), and more males than females received tobramycin (P less than 0.05). Pharmacokinetic assessments of renal damage were based on both changes in glomerular filtration rate (serum creatinine levels, creatinine clearance) and renal tubular damage (beta 2-microglobin, casts), but only patients with elevated aminoglycoside tissue levels leading to renal tubular damage and subsequent creatinine clearance decreases were considered to have experienced aminoglycoside nephrotoxicity. In the pharmacokinetic analysis of nephrotoxicity, 29 gentamicin courses (24%) and 12 tobramycin courses (10%) were complicated by nephrotoxicity (P less than 0.01). The 201 study patients were also evaluated independently for clinical nephrotoxicity (defined as a serum creatinine level increase of 0.5 mg/dl or more). Clinical nephrotoxicity occurred at rates of 37% in the gentamicin-treated group and 22% in the tobramycin-treated group (P less than 0.02). In these similar groups of critically ill patients, tobramycin was less nephrotic than gentamicin.