Joachim Grevel (1) Garrit Jentsch (1), Rupert Austin (1), John Lettieri (2), Bart Ploeger (3), Ping Ji (2), Gerold Meinhardt (2), and Carol Peña (2)
(1) BAST Inc Ltd., UK, (2) Bayer Healthcare Pharmaceuticals, US, (3) Bayer Pharma AG, Germany
Objectives: To develop a time-to-event (TTE) model for adverse events (AE) and clinical response (progression-free survival, PFS) in relation to sorafenib dose rate (DR) and plasma exposure in DTC. To simulate the benefit and risk of various starting doses and dose reduction schedules.
Methods: PFS, AE and PK data from 206 patients randomised to 400 mg sorafenib BID and 207 placebo treated patients from a Phase 3 DTC trial (DECISION; NCT00984282) were used. Sparse PK data were converted into individual AUCs by an existing popPK model [1]. DR and AUC(0-12) were tested as covariates (average until the event or time-varying) in a TTE analysis (using NONMEM) of PFS (centrally assessed) and 5 AEs (any first event of Grade ≥3, any first event of Grade ≥4, hand-foot skin reaction (HFS) Grade 3, hypertension Grade ≥3, diarrhoea Grade ≥2). The parametric hazard models with covariate influences were used to simulate events for starting doses of 800, 600, or 400 mg/day, and for 2 cycles of 800 mg/day followed by either 600 mg/day or 400 mg/day either maintained or for 2 cycles followed by up-titration.
Results: The average AUC(0-12) had the strongest influence on PFS compared to the other dose/exposure covariates which were less significant. The hazard ratio for 800 mg/day versus 600 mg/day and 400 mg/day was 0.814 and 0.638, respectively. The influence of average DR on the hazard of any Grade ≥3 AEs and on HFS was stronger than that of the time-varying DR. The AUC exposure metrics had no significant influence. For diarrhoea, time-varying DR had the strongest influence. For hypertension and Grade ≥4 AEs, no covariates were significant. Benefit/risk was assessed graphically for DR ranging from 800 mg/day to 0 mg/day (placebo) as the fraction of patients with AEs Grade ≥3 and with tumour progression over a 400 day treatment period.
Conclusions: All simulations emphasised the importance of high DR or exposure early in treatment. All dose reduction schemes starting with 2 cycles of 800 mg/day of sorafenib had similar benefit and risk. Lower starting doses than 800 mg/day could not be justified as less AEs were offset by reduced efficacy.
References:
[1] Prins NH, Grevel J , Mitchell D, Lettieri J,4 Ploeger B, Peña C. Population pharmacokinetics of sorafenib: a meta-analysis of patients with hepatocellular carcinoma, renal cell carcinoma, and differentiated thyroid carcinoma. J Pharmacokin Pharmacodyn (2014) 41:S23.
Reference: PAGE 25 () Abstr 5906 [www.page-meeting.org/?abstract=5906]
Poster: Drug/Disease modeling - Oncology