Charting the ecosystem of dormancy and persistence after therapy in cancer

The ability of cells to undergo molecular and physical changes in response to internal and external factors is a key mechanism that ensures healthy tissue development, regeneration and adaptation to stress. In cancer, the same mechanisms are often hijacked to allow tumour cells to proliferate within the tissue in an uncontrolled manner, to hide from the immune system and persist after therapeutic interventions. There is increasing evidence indicating that tumours contain rare subpopulations of cells that divide slowly or entirely stop proliferating, entering a state called "dormancy", and in this state they are able to resist a wide range of anti-cancer therapies, leading to long-term clinical relapse. Identifying and eliminating these rare and dynamic cells within the tumour is critical in order to improve patient outcomes in the clinic, but current therapies are not tailored to specifically target dormant cells. Despite its importance in cancer, dormancy remains poorly understood. We and others have previously described some of the molecular changes that influence dormancy in diverse cancer types, but little is known about how these cells interact with other immune and structural cells in their environment. Understanding these factors could help us predict which patients may present therapeutic resistance and cancer relapse due to dormancy, and how the immune system might be reactivated to detect and kill these cells.

Our previous studies indicate that dormant tumour cells might exploit specific signalling mechanisms to avoid recognition and killing by the immune system. In this proposal, we will employ advanced artificial intelligence (AI) and data science approaches to explore how dormant tumour cells co-opt their environment to persist in the tissue and resist standard-of-care treatments like chemoradiotherapy. We will investigate these complex interactions using sequencing technologies that measure the level of gene activity within single cells or spatially within the cancer tissue, as well as digital slides of cancer tissue that are produced routinely in the clinic and employed by pathologists for cancer diagnosis. The aim is to identify the factors within the immediate tumour environment that enable dormant cells to survive and thrive during the earlier stages of cancer development, before treatment, but also to describe how these factors change upon therapy. For this purpose, we will employ data from gastrointestinal tumours (from the oesophagus and colon) that have been profiled before and after chemoradiotherapy. This research will shed light on some of the fundamental mechanisms enabling therapeutic resistance in cancer, and will generate new AI-based tools that can be widely employed by the scientific community to study similar problems in cancer and other diseases. Ultimately, this research also aims to deliver new AI-informed models that can predict chemoradiotherapy resistance and to identify biomarkers that could be easily measured in the clinic in order to inform the treatment management of cancer patients.