Exploring the synergistic effects of cabozantinib (cabo) and a programmed cell death protein 1 (PD1) inhibitor in metastatic renal cell carcinoma (mRCC) with artificial intelligence (AI).

Abstract

e16555 Background: Nonclinical and clinical data suggest that cabo with a PD1 inhibitor provides synergistic antitumor activity in patients with mRCC, possibly by a cabo-induced switch to an immunopermissive tumor microenvironment. We used a complementary, unbiased, AI approach to gain a holistic view of the complex interplay between multiple pathways, cells and molecules and identify the mechanisms that may underpin this synergism. Methods: Biological targets associated with mRCC pathophysiology or drug actions were identified from proteomic, genomic and transcriptomic databases and literature. Using systems- and AI-based technology, the data were integrated using machine learning into mathematical models of the human mRCC protein network topology. The combined effects of cabo and a PD1 inhibitor on biological targets were simulated assuming target receptors were fully activated or fully inhibited. Relevant effects on known cancer processes (e.g. angiogenesis, metastasis, cell proliferation, immune evasion) were identified using artificial neural networks. Biologically plausible synergistic mechanisms were described with sampling methods. Results: Inhibition of VEGF/VEGFR and GAS6/TAMR axes by cabo enhanced the known effects of PD1 inhibitors on immune evasion mechanisms by modulating multiple humoral and cellular components of the innate and adaptive immune responses (Table). PD1 inhibitors further enhanced the anti-angiogenic and tumor pro-apoptotic effects of cabo by modulating pro- and anti-angiogenic factors and T cell cytotoxicity. Conclusions: These data provide a mechanistic rationale and further support for the beneficial combination of cabo and a PD1 inhibitor and may guide future nonclinical and clinical research.[Table: see text]