Background: In vitro data indicate that the sorafenib is a moderate inhibitor of cytochrome P450 (CYP) enzymes, including CYP3A4, CYP2C19, and CYP2D6. This phase I/II study in patients with advanced melanoma evaluated the potential effect of sorafenib on the pharmacokinetics of midazolam, omeprazole, and dextromethorphan, specific substrates of CYP3A4, CYP2C19, and CYP2D6, respectively.
Methods: Twenty-one patients received sorafenib 400 mg twice daily for 28 consecutive days. On days 1 and 28, a cocktail containing midazolam 2 mg, omeprazole 20 mg, and dextromethorphan 30 mg was administered. Pharmacokinetic analyses were performed on day 1 without sorafenib and day 28 after steady-state sorafenib exposure; sorafenib pharmacokinetics were evaluated on day 28. We defined an interaction to be excluded if the 90% confidence interval of the ratio of all day 28:day 1 analyses fell within a range from 0.80 to 1.25.
Results: In all, 18 patients were evaluable. On day 28, area under the plasma concentration-time curve from time 0 to 12 h (AUC(0-12)) and maximum plasma concentration (C(max)) for sorafenib were 38.1 mg h/l and 4.9 mg/l, respectively. Day 28:day 1 ratios for AUC from time 0 extrapolated to infinity (AUC(0-inf)) and C(max) for midazolam were 0.85 and 0.98, respectively. Day 28:day 1 ratio for 5-OH-omeprazole:omeprazole plasma concentration at 3 h postdose was 1.26, slightly outside of the 0.80-1.25 range. Thus, an interaction could not be excluded, but is considered unlikely to be clinically significant. Day 28:day 1 ratio for dextromethorphan:dextrorphan concentration in urine was 0.94. Sorafenib had an acceptable safety profile. The most frequently observed grade 3-4 toxicities in cycle 1 included elevated lipase (19%) and hypertension (10%).
Conclusions: In this patient population, our results demonstrate that exposures of probes of CYP3A4, CYP2D6, or CYP2C19 activity are potentially altered by administration of sorafenib at 400 mg twice daily. However, these differences are sufficiently small that a clinically significant inhibition or induction of these important drug metabolizing P450 isoenzymes is unlikely. Clinical and, where possible, drug level monitoring may still be appropriate for drugs of narrow therapeutic range co-administered with sorafenib.