Extrapolation and Justification of Nplate Dosing to Improve Overall Survival in Acute Radiation Syndrome

Abstract

Background: Acute Radiation Syndrome (ARS) is an acute illness caused by exposure to a high dose of penetrating radiation over a short period of time. Hematopoietic subsyndrome of ARS (HS-ARS) is characterized by dose dependent bone marrow depression leading to lymphopenia, neutropenia, thrombocytopenia, and anemia. Death due to HS-ARS from infection or excessive bleeding usually occurs within 2 to 3 weeks. Duration of thrombocytopenia is a predictor of overall survival (OS) in irradiated animal models, suggesting that a treatment for thrombocytopenia may increase survival in humans with HS-ARS. Romiplostim, a thrombopoietin receptor agonist, treatment resulted in prevention of severe thrombocytopenia and increased OS in irradiated animals. As human clinical trials for HS-ARS are not feasible or ethical, a romiplostim pharmacokinetic/pharmacodynamic (PKPD)-OS model for irradiated humans was developed. The model, informed by PKPD data in healthy/irradiated rhesus monkeys (RM) and healthy volunteers (HV), was subsequently used to predict the survival benefit of romiplostim relative to placebo in humans with HS-ARS. Methods: A PKPD model of romiplostim exposure-platelet response in healthy RM was developed and updated with radiation parameters to estimate the radiation effects on thrombopoiesis, on PKPD of romiplostim, and on differences in platelet response due to sex and body weight (Pritchard-Bell, ACoP11, 2020). A parametric time-to-event model relating platelet time course to OS in irradiated RM with/without romiplostim treatment was developed to quantify the impact of platelet response on OS (Jones, ACoP11, 2020). To extrapolate radiation effects to humans, romiplostim PKPD radiation parameters and radiation effects estimated from irradiated RM were applied to a romiplostim HV PKPD model. The extrapolated irradiated human PKPD model and irradiated RM OS model were combined and calibrated against published historical mortality data of humans exposed to acute radiation (Scott BR 1990). Following calibration, the extrapolated irradiated human PKPD and OS models were used to conduct simulations of OS in irradiated humans receiving romiplostim or placebo treatment. Relative survival benefit (RSB) of romiplostim, proportion of romiplostim treated humans surviving relative to placebo treated humans post radiation exposure, was summarized for various romiplostim doses (1 [initiation dose per label], 3, or 10 µg/kg [maximum dose per label]) and treatment scenarios (time after irradiation, adult vs pediatric) based on simulations of 10,000 humans/scenario randomized 1:1 to receive romiplostim or placebo. Results: An HV PKPD model of thrombopoiesis was updated to incorporate radiation specific scaling factors estimated from irradiated RM on romiplostim pharmacokinetic parameters and platelet lifespan, nonlinear inhibitory effect of radiation dose on megakaryocyte production, treatment effect of romiplostim on platelet lifespan, radiation sensitivity and radiation intensity. The extrapolated PKPD model for irradiated humans combined with the OS model from irradiated RM was calibrated with a scaling factor of 1.24 on radiation parameters to result in 50% survival 60 days post radiation exposure for humans exposed to a 3.07 Gy radiation dose (1 Gy/hr over 3.07 hr) consistent with published data. The calibrated OS model was used to simulate OS following romiplostim treatment 24 hours post irradiation and predicted 75% (RSB: 1.5), 80% (RSB: 1.6), and 87% (RSB: 1.7) survival on day 60, for romiplostim doses of at 1, 3, and 10 µg/kg, respectively compared to 50% survival for placebo. Percent survival was >75% when 10 µg/kg romiplostim was administered 24, 48, or 72 hours post irradiation and when administered to pediatric sub-groups (0-2, >2-6, >6-12, >12-18). Conclusions: A PKPD model of romiplostim exposure-response in irradiated humans was used to predict platelet response in humans with HS-ARS with/without treatment with romiplostim and was combined with an OS model to simulate 60-day survival in humans with HS-ARS. Simulations demonstrated the robust survival benefit of a single 10 µg/kg romiplostim dose compared with placebo in both adult and pediatric humans with HS-ARS following an acute radiation event. Selection of a romiplostim dose for treatment of HS-ARS is pending review of the irradiated RM studies and subsequent modeling and simulation analyses by the FDA. Disclosures Doshi: Amgen Inc: Current Employment. Jones:Amgen Inc: Current Employment. Pritchard-Bell:Amgen Inc: Current Employment. Park:Amgen Inc: Current Employment. Olsson Gisleskog:POG Pharmacometrics Ltd: Current Employment.