The Antitumor Activity of Combinations of Cytotoxic Chemotherapy and Immune Checkpoint Inhibitors Is Model-Dependent

Download pdf:

INTRODUCTION

Immune checkpoint inhibitors (ICI), such as anti-PD-1 (Programmed cell death 1) or anti-PD-L1 (Programmed death-ligand 1) antibodies, are among the most important recent breakthroughs in oncology. As an example, monoclonal anti-PD-1 and anti-PD-L1 antibodies (mAb) showed impressive efficacy in clinical trials for the treatment of unresectable or metastatic melanoma, metastatic non-small cell lung cancer, renal cancer, and more recently for urothelial carcinoma and Hodgkin’s lymphoma (1–7). These monoclonal antibodies block the interaction between PD-1 (Programmed Death 1) molecule, expressed at the surface of T cells and other immune cells, and PD-L1 (Programmed Death-Ligand 1), expressed in multiple types of cancer cells. The PD-1/PD-L1 axis induces an inhibitory signal in T cells, and PD-1/PD-L1 pathway blockade restores T cell function resulting in increased proliferation and cytotoxic activity, subsequently improving anti-tumor immune response (8). Unfortunately, a non-negligible proportion of patients presents with innate resistance to PD-1/PD-L1 axis blockade, particularly because of lack of expression of PD-L1 by tumor cells or to the immunosuppressive effect of the tumor microenvironment. As a consequence some common cancer types have very low response rates, such as breast and prostate cancer (9, 10). More alarmingly, a significant subpopulation of patients treated with ICI who presented an initial response to therapy will develop acquired resistance, with disease progression after some period of time. Zaretsky et al. highlighted mutations in beta-2-microglobulin, resulting in reduced HLA class I surface expression, as a cause for acquired resistance to anti-PD-1 mAb therapy. In an additional study, they described mutations in JAK1 and JAK2, involving the interferon gamma pathway (10–12). Further hypotheses for this resistance phenotype include genetic  alterations (mutations, deletions, epigenetic modifications)  which can lead to altered expression of tumor neo-antigens (10). Expression of alternative checkpoints such as LAG-3, TIGIT,  TIM-3, and ICOS in the tumor microenvironment may also play a key role regarding clinical outcomes and are currently being explored. Evidence of immune checkpoint expression modulation in association with acquired resistance to anti-PD1 has been shown by Koyama et al., with an up-regulation of TIM-3 (13). For this reason, it is crucial to develop new therapeutic approaches to enhance the therapeutic effects of PD-1/PDL1 blockade, and to avoid resistance phenomena. To this end, efforts are currently engaged including the use of conventional chemotherapies to improve the anti-tumor activity of monoclonal antibodies. Some of these combination strategies are currently studied in various cancer types (14–20). One of the aims of combining chemotherapy with ICIs is to trigger antigen release via the cytotoxic cell death activity of chemotherapy leading to immune stimulation and improving the activity of PD-1 / PD-L1 blocking agents. Moreover, the impact of chemotherapies on the leucocyte composition of the tumor infiltrate might be crucial: as an example, cyclophosphamide has been described to decrease the proportion of regulatory T cells (Treg) and gemcitabine has been outlined to reduce MDSC, two immunosuppressive cell populations usually associated with bad prognosis in cancer (21–23). Several  clinical phase I trials combining ICI and chemotherapies are currently ongoing, in particular in NSCLC, with nivolumab in combination with associations of cisplatin and gemcitabine, cisplatin and pemetrexed, or carboplatin and paclitaxel, as well as pembrolizumab in combination with two different  chemotherapies (15). Overall, the cytotoxic role of chemotherapy potentially drives immune activation, supporting the study of combinations between ICI and chemotherapies (20, 24, 25).  To explore the possible interaction between conventional regimens and ICIs, we evaluated four preclinical models in order to determine the impact on the in vivo efficacy of these combinations, and to analyse the consequences on the tumor microenvironment. Grasselly C Frontiers in Immunology 2018