Modeling the dynamics of anti-drug antibodies in cancer patients treated with oncolytic virus in monotherapy or in combination with immune check- point inhibitors
Zinnia P Parra-Guillen1,2, Iñaki F Trocóniz1,2,3, Tomoko Freshwater4
1Pharmacometrics & Systems Pharmacology, Department of Pharmaceutical Technology and Chemistry, School of Pharmacy and Nutrition, University of Navarra, Pamplona, Spain; 2IdiSNA, Navarra Institute for Health Research, Spain; 3 Institute of Data Science and Artificial Intelligence, DATAI, University of Navarra, Pamplona, Spain, 4Quantitative Pharmacology and Pharmacometrics Immune/Oncology (QP2-I/O) Merck & Co., Inc., Rahway, New Jersey, USA.
Introduction and Objectives: Immuno-oncology (IO) therapeutic strategies, such as oncolytic viruses (OVs), are based on the activation of the endogenous anti-tumor immune response to control or eradicate tumor cells [1,2]. Given the mechanisms of action of these therapeutics, development of antidrug antibodies (ADA) represents an important clinical aspect to evaluate due to their potential to alter drug pharmacokinetics and/or pharmacodynamics which could results in reductions or even loss of efficacy [3]. Our work aims to develop a semi-mechanistic pharmacokinetic/pharmacodynamic (PK/PD) model characterizing the dynamics of ADA in cancer patients receiving the V937 OV in monotherapy or in combination with pembrolizumab. following intratumoral (IT) or intravascular (IV) administration. The impact on the OV PK and tumor response was also evaluated.
Methods: Model building was performed following a sequential and integrative approach using NONMEM 7.4. First, V937 pharmacokinetics was characterized using OV measurements in the absence of pembrolizumab and up to day 8 (to avoid the impact of ADA). The rest of the OV measurements were used for model validation. Then, the time course of ADA was modelled using the PPP&D approach, where population PK parameters are fixed, but individual PK parameters are estimated simultaneously with the ADA model based on both PK and ADA data [4]. Treatment (monotherapy or combination), administration route and pre-existence of ADA were evaluated as covariates at this stage. Finally, the impact of ADA on V937 clearance and tumor response was ultimately explored.
Results: Model results indicate that OV undergoes extensive distribution beyond systemic circulation and tumor. After IT administration systemic bioavailability was estimated in 20%. The high serum viral clearance (111 L/h) and a large apparent volume of distribution (996 L) estimates are in agreement with a previous mechanistic model [5]. Serum V937 initiates the underlying mechanisms of (saturable) ADA expression (e.g. 4.5 fold increase in maximum ADA levels when increasing the IV dose from 1x108 to 10x108 TCID50 ) achieving mature immune response after three weeks from treatment initiation. Despite of their statistical significance the impact of pre-existing ADAs and the route and mode of administration on the turn-over parameters of the model was deemed as clinically irrelevant given the reduced improvement in model performance, and increased model instability and parameter imprecision. Similarly, co-administration with pembrolizumab did not show an effect on ADA dynamics, aligned with the low immunogenicity described for this monoclonal antibody [6]. The current evaluation did not find a significant impact on V937 clearance or the best response in this cancer patient population.
Conclusions: A pharmacokinetic and pharmacodynamic model to simultaneously describe the time course of V937 levels and their triggered ADA development was successfully developed. To the best of our knowledge, this is the first model characterizing ADA development after OV administration. Despite that this evaluation suggests that ADAs do not have a relevant impact on the pharmacokinetics or efficacy of V937 oncolytic virus, this quantitative approach can be adopted in other clinical scenarios were, there is indeed a relationship between neutralizing antibodies and efficacy (i.e, SARS-CO-19).
References:
[1] Finn. Ann Oncol 23:viii6-9 (2012)
[2] Kaufman et al. Nat Rev Drug Discov 14:642-662 (2015)
[3] Davda et al. J Immunother Cance 7: 105 (2019).
[4] Zhang et al. J Pharmacokinet Pharmacodyn 30: 387-404 (2003).
[5] Parra-Guillen et al. PAGE 30 (2022) Abstr 10070 [www.page-meeting.org/?abstract=10070].
[6] van Vugt et al. J Immunother Cancer 7: 212 (2019).