Gary Mo (1), Robin L. Jones (2), John R. Baldwin (1), Randy D. Nichols (1), Debra R. Luffer-Atlas (1), Robert L. Ilaria (1), Ilaria Conti (1), William D. Tap (3), Damien M. Cronier (4)
(1) Eli Lilly and Company, Indianapolis, USA, (2) Fred Hutchinson Cancer Research Center, Seattle, USA, Memorial Sloan Kettering Cancer Center, New York City, USA (3) Eli Lilly and Company, Windlesham, UK (4)
Objectives: Olaratumab (Olara), a recombinant human IgG1 monoclonal antibody, is a PGDFRα antagonist. Olara plus doxorubicin (Dox) improved survival vs Dox in an open-label, randomized Phase 2 soft tissue sarcoma (STS) trial (NCT01185964) [1]. We characterized the exposure-response relationship of Olara for progression free survival (PFS) and overall survival (OS).
Methods: Patients received Dox (75 mg/m2 Day 1) with (Arm A) or without (Arm B) Olara (15 mg/kg Days 1 and 8) for up to eight 21-day cycles. In Arm A, Olara monotherapy continued after Dox until disease progression, and Arm B patients could receive Olara after progressing with Dox. Data from patients in both Arm A (n=67) and Arm B (n=66) were used to develop the PFS and OS time-to-event models. Various hazard models were evaluated (exponential, Weibull, Gompertz, combined Weibull/Gompertz). The effect of Olara was explored using the average Olara serum concentration (Cavg) and trough Olara serum concentration in cycle 1 (Cmin1), both simulated using individual patient post-hoc PK parameters estimated by PopPK modelling. Additional covariates were evaluated for potential effects on PFS/OS.
Results: OS and PFS were both best described by survival models with an exponential hazard function. The effect of Olara on PFS/OS was captured by an inhibitory Emax function with Hill coefficient, regardless of the PK endpoint (Cavg or Cmin1) considered. The Olara EC50s for PFS (ECmin150 = 82 μg/mL, ECavg50 = 179 μg/mL) and those for OS (ECmin150 = 66.1 μg/mL, ECavg50 = 134 μg/mL) corresponded to the median and 25th percentile of Cmin1/Cavg in the study, respectively. The maximum predicted improvement in the hazard ratio for OS and PFS was 75% and 65%, respectively, and was predicted to be achieved within the range of Olara serum levels observed in the study. Baseline ECOG status and prior treatments were found to significant covariates for OS. No significant covariates were identified for PFS.
Conclusions: OS and PFS in STS patients receiving Dox alone or in combination with Olara were successfully described by the survival models developed in this study. Olara Cmin1 and Cavg showed a similar predictive role for Olara effect on PFS/OS. The survival models are consistent with a greater effect of Olara on OS than PFS. Maximum improvement in the OS HR was predicted to occur within the range of Olara serum levels achieved in the study.
References:
[1] Tap et al., J Clin Oncol 33, 2015 (suppl; abstr 10501).
Reference: PAGE 25 () Abstr 5779 [www.page-meeting.org/?abstract=5779]
Poster: Drug/Disease modeling - Oncology