Colorectal cancer (CRC) is the third most common cancer worldwide and the second most lethal. Nearly 30% of CRC are in the rectum (RC). The standard treatment of locally advanced rectal cancers (LARC) is preoperative chemo-radiotherapy (CRT) followed by invasive surgery and CAPOX adjuvant chemotherapy. Intensification of the neoadjuvant treatment by anticipating CAPOX in addition to CRT prior to surgery, is well tolerated and associated with a higher pathological complete response (pCR) rate. This approach is known as total neoadjuvant therapy (TNT). Achieving a pCR can translate in a No Operative Management (NOM) of the patient, with full organ function preservation and a vast improvement in quality of life. The feasibility of NOM however depends on defining a response marker with optimal predictive power in each patient.

 Artificial intelligence (AI) applied to radiological images (radiomics) or whole-slide images (WSI, pathomics) is effective in delivering outcomes predictive signatures for many cancer types, including LARC. Tumor molecular profiling also provides valuable prognostic and predictive information guiding treatment choices. Each of these omics has been used in RC but with a single modal approach. We posit that deploying the three techniques concurrently could provide a highly informative snapshot of the tumor response to treatment and deliver an integrated marker with high predictive power for choosing among vastly different treatment options.