New AI Tool Could Reduce Unnecessary Chemotherapy in Breast Cancer Care - Video

Could artificial intelligence help some breast cancer patients avoid unnecessary chemotherapy? A new study suggests it might. Researchers from RCSI University of Medicine and Health Sciences and University College Dublin (UCD) have identified immune markers that, when analyzed using AI, may help doctors determine which women with early-stage breast cancer are unlikely to benefit from chemotherapy.

The findings, published in Nature Communications, focus on ER-positive/HER2-negative (ER+HER2-) breast cancer, the most common subtype, accounting for about 70% of breast cancer cases. Although chemotherapy can reduce the risk of cancer returning, it often causes significant side effects, and many patients receive it even when the benefit is uncertain.

Currently, treatment decisions rely heavily on genomic risk scores. However, many patients fall into an intermediate-risk category, making it difficult for doctors to decide whether chemotherapy is necessary.

To improve these decisions, researchers analyzed tumor samples from Irish patients who participated in a clinical trial comparing hormone therapy alone with hormone therapy plus chemotherapy. Using AI, they examined the tumor microenvironment, particularly the immune cells surrounding the cancer.

The study found that a high density of cytotoxic T-cells, immune cells that attack cancer, provided additional information beyond current genomic tests. Surprisingly, patients with higher levels of these T-cells had poorer outcomes after chemotherapy, suggesting this marker could help predict which patients are less likely to benefit from the treatment.

Because the approach uses standard tissue samples already collected during routine care, researchers believe it could be widely adopted if validated in larger studies.

While further research is needed, the findings move doctors closer to more personalized breast cancer treatment, helping some patients avoid unnecessary chemotherapy while ensuring others receive the treatments most likely to benefit them.

REFERENCE: Kinsella, Z., Jahangir, C.A., Nyarko, H.N. et al. Spatial analyses implicate high stromal tumour-infiltrating CD8+ lymphocytes as a negative predictive marker for chemotherapy in estrogen receptor-positive breast cancer. Nat Commun 17, 4863 (2026). https://doi.org/10.1038/s41467-026-73432-2