Abstract B73: Computational model of combined cancer treatment with radiotherapy and anti-PD-1 immunotherapy

Abstract

Abstract Cancer treatment with combination of radiotherapy (RT) and immunotherapy (IT) (immune check-point inhibitors) has gained promising results in preclinical and clinical studies. Accumulating evidence suggests that RT is beneficial not only because of its direct cytocidal effect but it also acts as an immunogenic hub, turning the irradiated tumor into an in situ cancer vaccine. In around 25% of patients such combined treatment results not only in shrinkage of the irradiated tumor but also in shrinkage of distant metastases (abscopal effect). However, little is known about the optimal RT dose and fractionation scheme, scheduling of IT and about which patients are candidates for responders. The results of preclinical and clinical studies addressing those questions are sparse and often contradictory. To help understanding the mechanisms of such therapy, we developed a computational model capable of simulating tumor response to treatment with RT and anti-programmed death 1 (anti-PD-1) antibodies. The model describes interplay between tumor cells and cytotoxic T lymphocytes (CTLs) with a set of ordinary differential equations. It incorporates intrinsic tumor and CTLs characteristics, such as radiosensitivity coefficients, PD-1 expression on CTLs, PD-1 ligand (PD-L1) expression on tumor cells, RT dose-dependent major histocompatibility complex class I (MHC-I) expression on tumor cells, etc. Additionally, we incorporated RT dose-dependent increase of damage-associated molecular patterns that play a crucial role in immunogenic cell death, such as calreticulin, ATP and high mobility group box 1. Finally, we studied pharmacokinetic and pharmacodynamic properties of a novel anti-PD-1 antibody and included it in our model. With tuning of some free parameters we successfully reproduced experimental results from literature (mice tumor model), where tumor response to 3 different therapies (anti-PD-1, stereotactic ablative radiotherapy (SABR) 1 x 15 Gy, SABR + anti-PD-1) was studied. The focus of our simulations was on primary irradiated tumor. Once we confirmed ability of the model to reproduce experimental results, we performed a sensitivity analysis of free parameters and studied their impact on tumor response. First we analyzed the impact of MHC-I and PD-L1 expression on tumor response to combined therapy. If the fraction of tumor cells expressing MHC-I is low, the therapy with anti-PD-1 has poor effect on tumor regardless of the fraction of tumor cells expressing PD-L1. SABR, as well as SABR + anti-PD-1, result in a temporary delay of tumor growth but the tumor volume is similar to the untreated tumor 2 weeks after irradiation. Secondly, if the fraction of tumor cells expressing MHC-I is high (fixed in this set of simulations) and the fraction of tumor cells expressing PD-L1 is low, even the untreated tumor does not form. If PD-L1 expression is moderate, the untreated tumor does form, whereas all three therapies result in a complete tumor eradication. If PD-L1 expression is high, SABR, as well as anti-PD-1 alone, result only in a delay of tumor growth. On the other hand, SABR + anti-PD-1 therapy results in a great tumor control even if all tumor cells express PD-L1. Finally, we simulated the combination of anti-PD-1 and different RT regimens (1 × 15 Gy, 3 × 5 Gy, 5 × 3 Gy, 7 x 2.14 Gy) for the case of poorly immunogenic tumor cells (low MHC-I, high PD-L1). In our simulations all three fractionation regimens outperform SABR, resulting in a 50% smaller tumor volume 3 weeks after irradiation. To conclude, we showed that the model of combined cancer therapy was able to reproduce preclinical results (mice tumor model). According to our model, MHC-I expression might play an important role in treatment with anti-PD-1 + RT and deserves further attention. Finally, the model predicts that fractionation should outperform SABR when combining RT with anti-PD-1. Citation Format: Damijan Valentinuzzi, Katja Ursic, Martina Vrankar, Urban Simoncic, Urban Simoncic, Robert Jeraj. Computational model of combined cancer treatment with radiotherapy and anti-PD-1 immunotherapy. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2016 Oct 20-23; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2017;5(3 Suppl):Abstract nr B73.