Abstract
Prognostic and predictive value of immune infiltrates in the context of regional nodal radiation (RNI) for breast cancer has not been assessed. Stromal tumor infiltrating lymphocytes (sTILs) were assessed on scanned images of hematoxylin and eosin (H&E) stained sections and by CD8 immunohistochemistry on tissue microarrays available from the MA.20 trial. Cox proportional modelling was used, and hazard ratios (HR) with 95% confidence intervals (CI) are reported for primary and secondary endpoints. Predictive value was assessed by an interaction test. H&E sTILs (continuous parameter) were prognostic for distant-DFS (HR 0.99, 95% CI 0.98-1.00, P = 0.04). CD8+sTILs were associated with significantly improved disease-free survival (DFS) (HR 0.99, 95% CI 0.98-1.00, P = 0.02) and distant-DFS (HR 0.98, 95% CI 0.97-0.99, P = 0.0002). CD8+sTILs was predictive of benefit from RNI for distant-DFS (continuous variable: HR 0.98, 95% CI 0.96-1.00, P((interaction)) = 0.04; exploratory categorical variable: CD8+ sTILs < 44, HR = 0.83; 95% CI 0.57-1.21, and CD8+ sTILs ≥ 44; HR 0.09; 95% CI 0.01-0.74, P((interaction)) = 0.04). In MA.20 breast cancer patients, pre-treatment sTILs were prognostic for DFS (CD8+sTILs) and distant-DFS. CD8+sTILs also appeared to be predictive for the effectiveness of RNI on distant-DFS, suggesting that immune mechanisms may in part be responsible and merits further investigation.