PII-026 - AN INTEGRATED PK/PD MODELING FRAMEWORK OF NONCLINICAL EFFICACY AND CLINICAL HEMATOLOGICAL SAFETY TO INFORM DOSE SELECTION OF ATRI TUVUSERTIB IN COMBINATION WITH NIRAPARIB IN PATIENTS WITH PARPI-RESISTANT OVARIAN CANCER

J. Mukker¹, N. Dupuy², P. Diderichsen³, T. Yap⁴, R. Plummer⁵, A. Tolcher⁶, J. de Bono⁷, Z. Szucs⁸, A. Zimmermann⁹, F. Hellmann³, C. Villette², R. Strotmann⁹, A. Zutshi¹, I. Gounaris⁸, K. Venkatakrishnan¹; ¹EMD Serono, Billerica, MA, USA, ²Physiomics Plc, Oxfordshire, UK, ³Certara USA, Radnor, PA, USA, ⁴Department of Investigational Cancer Therapeutics, University of Texas MD Anderson Cancer Center, Houston, TX, USA, ⁵Newcastle University and Northern Centre for Cancer Care, Newcastle Hospitals NHS Trust, Newcastle Upon Tyne, UK, ⁶Clinical Research, New Experimental Therapeutics (NEXT), San Antonio, TX, USA, ⁷Division of Clinical Studies, Royal Marsden Hospital, Sutton, UK, ⁸Merck Serono Ltd., Feltham, UK, an affiliate of Merck KGaA, Darmstadt, Germany, Feltham, UK, ⁹the healthcare business of Merck KGaA, Darmstadt, Hessen, Germany.

Background: The combination of ATR inhibitor (ATRi) tuvusertib and PARP inhibitor (PARPi) niraparib is believed to induce synergistic cancer cell death. Part B1 of the DDRiver solid tumors 301 study (NCT04170153) assessed the safety, pharmacokinetics (PK), pharmacodynamics (PD), and preliminary efficacy of this combination. Semi-mechanistic PK/PD analyses of nonclinical tumor growth inhibition (TGI) and clinical hematological safety were conducted to inform the clinical dosage for this combination.

Methods: Nonclinical PK-TGI relationships for tuvusertib, niraparib, and other ATRis and PARPis were assessed using published and in-house xenograft models. A semi-mechanistic ‘virtual tumor’ efficacy model was developed incorporating cell cycle, drug, and tumor parameters, key features of DNA damage repair, and diverse genetic backgrounds. A non-linear mixed effects modeling approach was used to describe clinical PK and PK/PD for the time course of reticulocytes, red blood cells, and hemoglobin (HGB), as applicable. Tuvusertib and niraparib PK were described using a previously developed and a literature population PK model, respectively.

Results: The virtual tumor model adequately described the nonclinical TGI data, including patient and cell line derived HBCx, FaDu, and MDA-MB-436 xenograft models. Clinical translation was qualified to account for inter-species differences in plasma free fraction, PK, and clinical efficacy data to predict the relative reduction in tumor volume. An Emax model of tuvusertib and a linear model of niraparib implementing additive concentration-dependent inhibitory effects on the production of progenitor cells adequately described the preliminary time course of HGB and predicted the rate of grade 3 anemia after 4 treatment cycles. By rank-ordering predicted benefit (relative efficacy) versus risk (% grade 3 anemia) of clinically tolerable intermittent and continuous dosages, tuvusertib 180 mg QD 1 week (w) on/1w off + niraparib 100 mg QD 1w on/1w off was selected for further clinical evaluation in patients with PARPi-resistant ovarian cancer.

Conclusion: The developed framework represents an integrated PK/PD modeling approach to describe nonclinical efficacy and clinical safety data and inform dosage selection in early phase combination treatment development.