AI helps predict bowel cancer recurrence

Colorectal cancer, or bowel cancer, is one of the most common cancers worldwide, and being able to predict whether the cancer will come back is crucial for guiding further treatment.

The role of AI in predicting recurrence

The study, part of the larger QUASAR trial, examined tumour samples from 868 patients with stage II and III colorectal cancer. Researchers focused on two types of immune cells – CD3 and CD8 T-cells. These cells are part of the body's natural defence system, attacking cancer cells and helping to control tumour growth. The team used advanced AI algorithms to count the number of these immune cells in different parts of the tumours, making the process much quicker and more precise than manual counting by pathologists.

Using AI, the study found that patients with a high density of CD3 and CD8 T-cells in their tumours were less likely to experience cancer recurrence. Specifically, patients with higher numbers of these immune cells had half the risk of cancer returning compared to those with lower numbers. This was consistent for both stage II and III cancers, and across colon and rectal cancers.

Prognosis vs chemotherapy benefit

The AI test helped predict who was more likely to see their cancer return after surgery. However, while CD3 and CD8 T-cells were good markers for prognosis, they did not predict which patients would benefit most from chemotherapy. Both high-risk and low-risk patients who received chemotherapy saw similar reductions in recurrence rates. This means that, although AI can identify who is at greater risk of cancer coming back, it doesn’t yet help doctors decide whether chemotherapy will be more effective for these patients.

These findings are important because they offer a clearer way to predict which patients are at higher risk of cancer recurrence. For patients with a high density of CD3 and CD8 T-cells, the risk of recurrence is lower, potentially meaning fewer aggressive treatments might be needed after surgery. On the other hand, those with lower levels of these immune cells may require closer monitoring and potentially more intensive follow-up care.

However, because the immune markers did not indicate whether chemotherapy would be more effective, doctors will still need to rely on other clinical factors when deciding whether chemotherapy is necessary after surgery. The AI-driven immune cell test may eventually be a helpful tool in guiding conversations about post-surgery treatments, but more research is needed before it can reliably inform chemotherapy decisions.

Looking forward

This AI-driven test holds promise for making cancer treatment more personalised and less invasive. By accurately predicting which patients are at higher risk of recurrence, it could help avoid unnecessary treatments, such as chemotherapy, in patients who are less likely to benefit.

While AI is helping to open new doors in cancer care by providing more detailed and precise ways to assess the risk of cancer returning, there’s still more to understand about how these findings will affect treatment decisions like chemotherapy.