A potential immune biomarker could help refine adjuvant treatment decisions for patients with estrogen receptor (ER)-positive, HER2-negative breast cancer, with new research suggesting some patients with an intermediate Oncotype DX Recurrence Score may experience poorer long-term outcomes following chemoendocrine therapy.
The study, published in Nature Communications, found that a high density of stromal tumour-infiltrating CD8+ lymphocytes (sTIL CD8+) was associated with poorer long-term outcomes among patients with an intermediate Oncotype DX Recurrence Score (16–25) who received chemoendocrine therapy.
REGISTER FOR BREAST CANCER TRIAL’S ANNUAL SCIENTIFIC MEETING | JULY 2026
The findings suggest stromal CD8+ T-cell density may represent a candidate predictive biomarker that could complement genomic testing when determining which patients are most likely to benefit from adjuvant chemotherapy.
Current treatment decisions for patients with ER-positive, HER2-negative breast cancer commonly incorporate the Oncotype DX Recurrence Score. However, uncertainty remains for patients with intermediate scores, where the benefit of chemotherapy can be difficult to predict.
To investigate whether the tumour immune microenvironment could improve risk stratification, researchers analysed tumour samples from patients previously enrolled in the TAILORx clinical trial using spatial proteomics (440 patients) and spatial transcriptomics (359 patients).
The analyses revealed that immune activity within the tumour stroma and tumour epithelium was more complex than previously appreciated, with distinct “immune-hot” and “immune-cold” microenvironments that were not fully reflected by genomic recurrence scores.
Importantly, among patients with an intermediate Recurrence Score who were randomised to receive chemoendocrine therapy, increasing stromal CD8+ T-cell density was associated with significantly poorer invasive disease-free survival. The findings were further supported through an orthogonal validation using whole tumour resection specimens, where higher stromal CD8+ T-cell density again identified patients who experienced poorer outcomes following chemoendocrine therapy.
The authors suggest immune profiling could complement existing genomic assays by providing additional biological information not captured by the Oncotype DX Recurrence Score alone.
Based on modelling within the study cohort, the researchers estimated the biomarker could potentially alter adjuvant treatment recommendations for up to half of patients with intermediate Recurrence Scores. Specifically, the modelling suggested some pre- or peri-menopausal women currently recommended chemotherapy might instead avoid treatment, while some postmenopausal women may warrant chemotherapy despite current guideline recommendations. However, the authors emphasise these findings require comprehensive external validation before influencing clinical practice.
Beyond its potential clinical utility, the study also provides new insights into the immune biology of ER-positive breast cancer.
Although these tumours have traditionally been considered immunologically “cold”, the researchers identified diverse immune microenvironments characterised by differing levels of inflammation, immune activation and immune suppression.
Tumours with high stromal CD8+ T-cell infiltration frequently demonstrated molecular features associated with chronic immune stimulation, extracellular matrix remodelling and increased expression of immune checkpoint molecules including CTLA4, TIGIT and CD96.
The researchers hypothesise these immune-rich tumours may represent a constrained or dysfunctional anti-tumour immune response, rather than one capable of effectively eliminating residual cancer cells, although they note the underlying biological mechanisms remain speculative.
They also suggest the findings may help inform future studies exploring immune checkpoint inhibitors in selected patients with ER-positive, HER2-negative breast cancer.
While the results are promising, the investigators stress that stromal CD8+ T-cell density should currently be regarded as a candidate predictive biomarker rather than one ready for clinical implementation.
The authors conclude that comprehensive external validation, ideally in larger randomised cohorts such as the broader TAILORx population, will be required before the biomarker can be incorporated into routine treatment decision-making.
Paper: 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). Access online here.
