Advances in artificial intelligence are reshaping how clinicians decide who needs chemotherapy after colorectal cancer surgery. A Norwegian start-up, DoMore Diagnostics, has developed an AI prognostic tool that aims to cut the number of patients exposed to toxic chemotherapy without benefit.
Targeting Over-treatment in Cancer Care
Standard pathology can miss subtle patterns that predict recurrence. As a result, many patients with stage two and stage three colorectal cancer receive adjuvant chemotherapy despite having low risk of relapse. Some analyses suggest up to 96 to 98 percent of stage two patients and roughly 80 percent of stage three patients may be treated without clear survival benefit, exposing them to side effects, functional decline, and higher costs.
How AI Delivers Precise Prognosis
DoMore Diagnostics, a spin-out from research collaborations that include Oxford and Oslo University Hospital, uses deep learning to scan digitized tissue slides. The model was trained on thousands of images linked directly to long-term patient outcomes rather than only to pathologist labels. This outcome-driven training lets the AI detect morphological and spatial features that correlate with recurrence risk.
In validation studies the tool identified high-risk cases that standard review missed and stratified patients more accurately than routine pathology alone. That increased prognostic precision supports more confident decisions about who truly needs adjuvant chemotherapy.
A Future of Personalized Cancer Treatment
When applied in clinical workflows across Europe, the United States, Japan, and Mexico, the technology has the potential to spare many patients from unnecessary chemotherapy and its harms while focusing treatment on those who stand to benefit. For clinicians and patients, the immediate value is clearer risk information to guide shared decision making.
As external validation continues and regulatory approvals advance, outcome-linked digital pathology tools like DoMore’s promise to shift colorectal cancer care toward less toxic, more personalized treatment plans.
