Researchers at University College London (UCL) report that an artificial intelligence (AI) tool they developed identified a subgroup of patients with advanced rectal cancer who appear to benefit from adding an irinotecan‑hydrochloride‑trihydrate chemotherapy to standard treatment, according to a statement on EurekAlert.
The team trained the AI specifically to distinguish between patients by tumor cell density in biopsy samples taken at diagnosis. The algorithm analyzed a total of 414 samples.
Among patients whose pre‑treatment biopsies showed high tumor cell density, supplementing standard therapy with the irinotecan‑containing drug was associated with a 43% reduction in the risk of cancer recurrence and a 50% increase in the chance of survival at five years of follow‑up. These comparisons were made against the standard regimen comprising chemotherapy with a capecitabine‑containing agent plus radiotherapy.
In contrast, patients with low tumor cell density did not show a measurable benefit from the irinotecan addition.
Irinotecan‑containing chemotherapy is already commonly used in advanced colorectal cancer, and there has been interest in its role in rectal cancer as an add‑on; however, prior studies did not demonstrate an overall improvement across unselected patient populations. Rectal cancer is the fourth most deadly cancer in the United Kingdom, and once it reaches an advanced stage it carries a high risk of recurrence, motivating investigation of stronger treatment strategies.
The researchers note that the drugs involved can cause significant side effects, such as diarrhea and reductions in white blood cell counts, so combined regimens are prescribed cautiously and only when clearly indicated.
A practical output of the work is Octopath, a free tool the authors developed that allows clinicians to upload biopsy results for AI analysis. The system aims to help clinicians identify which patients are most likely to benefit from irinotecan supplementation and which patients may not gain enough benefit to justify the increased toxicity risk.
The findings are relevant to tailoring treatments more precisely, but further evaluation and clinical validation are required before broad implementation, particularly to balance potential survival gains against treatment‑related harms.
Key points
- The observed benefit was limited to the high tumor cell density subgroup, based on analysis of 414 samples.
- Benefit was measured against standard capecitabine plus radiotherapy treatment.
- Potentially increased toxicity means careful clinical selection is necessary.
Octopath is provided as a decision‑support tool to assist oncologists in classifying biopsy samples by tumor cell density and informing individualized treatment choices.
