Population pharmacokinetics of cisplatin in adult cancer patients

Purpose To characterize the pharmacokinetics of the anticancer agent cisplatin, and explore the influence of patient covariates and interoccasion variability on drug disposition.Methods Data were obtained from 285 patients (519 complete curves; 3483 plasma samples) who received the drug as a 3-h intravenous infusion at a mean dose of 144 mg (range 75–210 mg). The population model was built with the use of NONMEM, performing generalized-additive modeling to identify candidate covariates including body-surface area (BSA), age, sex, height, weight, hematocrit, total protein, albumin, serum creatinine, and creatinine clearance, and using a backward deletion protocol to obtain the final models for clearance (CL) and volume of distribution (V).Results The final model was a one-compartment linear model with BSA (in meters squared) as the only significant covariate that impacted on both CL and V: TVCL (in liters per hour)=51.7+26.3×(BSA–1.855) and TVV (in liters)=41.1+24.6×(BSA–1.855), where TVCL and TVV are referred to as typical values that could be used a priori in dosage regimen design. The interindividual and interoccasion variability estimates for CL and V were 16.82 and 20.35%, and 13.93 and 22.91%, respectively.Conclusion A population pharmacokinetic model for cisplatin has been developed that incorporates measures of body size to predict clearance. In this patient population, cisplatin pharmacokinetics were not associated with age, sex, or measures of renal dysfunction.Presented previously in part at the 92nd Annual Meeting of the American Association for Cancer Research, 24–28 March 2001, in New Orleans, LA, USA.