---

Radiotherapy (RT) is the most common form of cancer treatment, with more than 60% of patients receiving radiation (600K+ patients per year in the US alone), and RT is responsible for over 40% of all cancer cures. Yet RT dose is not personalized based on biomarkers, and advances in radiation oncology have primarily focused on beam properties. RT is still prescribed using a one-size-fits all approach, where patients receive uniform doses that are based on clinicopathological features and decades of empirically derived tolerance of normal tissue. An extensive body of literature has emerged relating to the immune-activating ability of RT. In fact, RT efficacy may be a combination of the direct cytotoxic effect of radiation and, possibly more importantly, the subsequent indirect effect of stimulating a successful antitumor immune response. However, current RT fractionation has not specifically focused on enhancing immune responses despite evidence that fewer, larger doses induce significantly stronger antitumor immunity. Daily RT over many weeks may even be detrimental to antitumor immunity, as CD8+ T cells are generally very radiation sensitive. Therefore, understanding the complex, non-linear cytotoxic and immunologic consequences of RT dose and fractionation is of high biological interest and clinical value. Innovative models based upon interactions of the tumor with its immune-environment could elucidate new treatments that improve tumor control and protect normal tissues.

We profiled 10,469 primary tumors for their metrics of radiosensitivity and immune cell infiltrate (ICI) and developed a novel in silico model that mimics the dynamic relationships between tumor growth, ICI flux and the response to radiation therapy. Our data presents the most extensive portrait of the interaction between the tumor immune ecosystem (TIES) and radiosensitivity in human cancer. Furthermore, the in silico model, represents the TIES as a juxtaposition of two balancing phenotypes: anti-tumor vs. pro-tumor that is based on the presence of ICI (effector vs suppressor). We validated this model in the 10,469 primary tumors as well as a separate cohort of 59 lung cancer patients treated with radiotherapy. These results now mean we can use pre-treatment molecular and tissue-pathological features to identify patient candidates for: (i) safe radiation dose de-escalation; (ii) radiation dose escalation; (iii) combination therapy with immunotherapeutic agents. This work offers the first radiation therapy personalization platform in the context of cancer immunology. These analyses explain radiation response based on its effect on the TIES and quantify the likelihood that radiation can promote a shift to anti-tumor immunity. All parameters that define TIES and radiosensitivity are directly measurable from tumor biopsies and can be readily tested in prospective clinical trials combining RT and immune therapies.

This work is recently published as a bioRxiv preprint at: https://www.biorxiv.org/content/10.1101/2020.02.11.944512v1 .