Population pharmacokinetics and pharmacodynamics of oral etoposide

Aims To study the population pharmacokinetics and pharmacodynamics of oral etoposide in patients with solid tumours. Methods A prospective, open label, cross-over, bioavailability study was pe rformed in 50 adult patients with miscellaneous, advanced stage solid tumou rs, who were receiving oral (100 mg capsules) etoposide for 14 days and i.v . (50 mg) etoposide on day 1 or day 7 in randomised order during the first cycle treatment. Total and unbound etoposide concentration were assayed by h.p.l.c. Population PK parameters estimation was done by using the P-Pharm software (Simed). Haematological toxicity and tumour response were the main pharmacodynamic endpoints. Results Mean clearance was 1. 14 1 h(-1) (CV 25%). Creatinine clearance was the only covariable to significantly reduce clearance variability (residua l CV 18%). (CL=0.74+0.0057 CLCR; r(2)=0.32). Mean bioavailability was 45% ( CV 22%) and mean protein binding 91.5% (CV 5%). Exposure to free, pharmacol ogically active etoposide (free AUC p.o.) was highly variable (mean value 2 .8 mg l(-1) h; CV 64% range 0.4-9.5), It decreased with increased creatinin e clearance and increased with age which accounted for 9% of the CV. Mean f ree AUC p.o. was the best predictor of neutropenia. Free AUC(50) (exposure producing a 50%, reduction in absolute neutrophil count) was 1.80 mg l(-1) h. In patients with lung cancer, the free AUC p.o. was higher in the two pa tients with responsive tumour (5.9 mg l(-1) h) than in patients with stable (2.1 mg l(-1) h) or progressive disease (2.3 mg l(-1) h) (P=0.01). Conclusions Exposure to free etoposide during prolonged oral treatment is h ighly variable and is the main determinant of pharmacodynamic effects. The population PK model based on creatinine clearance is poorly predictive of e xposure. Therapeutic drug monitoring would be necessary for dose individual ization or to study the relationship between exposure and antitumour effect .