Immunotherapy revolutionized the treatment of many malignancies. However, many patients fail to respond to immuno-oncology (IO) drugs, although satisfying the accepted clinical criteria for IO treatment. This stems partially from the very incomplete understanding of the network-level interactions between various cells within the tumor itself, which together determine the fate of the anti-cancer immune responses. Using a novel, high-resolution method to resolve all the cells that together comprise the tumour micro-environment, we are able to infer important inter-cellular connections that may protect the tumor from an immune assault. By combining many patient tumor-immune networks, we at Pyxis, can predict IO drug response for many cancer types. Our platform can also suggest alternative treatment routes in cases where the predicted drug response is low, improving the quality of life of oncological patients.
