Incidence and mortality of colorectal cancer (CRC) is the second leading cause of cancer death in Europe. With current screening programs in most European countries, many CRCs are detected in an early and curable stage with excellent prognosis. The current standard treatment for early CRC is surgical tumour removal. This is considered as overly invasive especially for the growing number of elderly and frail patients. Expanding technical possibilities of local tumour removal by flexible endoscopy along with initial data supporting its oncological safety open the way to organ preserving therapy. This may result in avoiding surgery and improved patients´ quality of life (QoL). However, high-quality, long-term oncological outcome data of endoscopic removal as compared to surgery are lacking, and precise oncological risk criteria prior to the start of the therapy are urgently needed. The current project has two major aims: a) to assess oncological efficacy, safety, and patient QoL of local endoscopic removal of early CRC as compared to surgery in 3 randomized trials, and b) to develop and test criteria in this patient population to primarily identify high-risk lesions based on novel biomarkers, endoscopic imaging, and other clinical criteria, also within an artificial intelligence-guided treatment algorithm. We propose to provide state-of-the-art assessment of new, sustainable, less invasive treatment modalities for early-stage CRC and determine patient preferences around these therapies. We will perform large, international head-to-head randomized clinical trials. These trials will allow carbon cost effectiveness analyses, biomarker research and development of clinical practice guidelines to carve out the next-generation best clinical practice for patients with early-stage CRC. Our proposal aims at less invasive, sustainable organ-preserving therapies maintaining high oncological efficacy and helping physicians and patients to perform individual shared decision-making.