Integrating genomic mutations and tumor-infiltrating lymphocytes improves prediction of response to trastuzumab-based adjuvant therapy in patients with HER2-positive breast cancer.

Abstract

Aim: Resistance to trastuzumab remains a major barrier to cure in early-stage HER2-positive breast cancer (HER2+ BC). We investigated the impact of genomic alterations and tumor-infiltrating lymphocyte (TIL) density on treatment resistance and survival outcomes. Methods: We retrospectively analyzed 315 patients with HER2+ BC who received adjuvant trastuzumab at Ruijin Hospital (2009-2019). Whole-exome sequencing and TIL scoring were performed on surgical specimens, and clinical and pathological data were collected. The Cancer Genome Atlas (TCGA) cohort was used for external validation. Genomic alterations and TIL density were compared between trastuzumab-sensitive and -resistant tumors. Survival analyses were conducted to identify prognostic biomarkers. Results: After a median follow-up of 109.3 months, 67 tumors (21.3%) were trastuzumab-resistant, exhibiting lower TIL density (mean 19.8% vs. 26.3%, P = 0.001), higher mutation frequencies in FLG, MAP1A, BRCA1, PTPRD, PAPPA2, NCOR2, FBXW7, MYH7, and VCAN, and more frequent alterations in the TP53/NOTCH pathways compared with sensitive tumors (all P < 0.05). A 15-gene trastuzumab response-associated gene (TRAG) signature independently predicted poorer disease-free survival (DFS) in both our cohort (HR, 3.57, P < 0.001) and the TCGA cohort (HR, 4.99, P = 0.037). A high copy number alteration burden was associated with worse overall survival (HR, 2.49, P = 0.043), whereas TIL density > 10% was associated with improved DFS (HR, 2.44, P = 0.003). A prognostic model integrating tumor size, nodal status, estrogen receptor status, TILs, and the TRAG signature showed strong discriminatory power (c-index 0.743 in the training set; 0.915 in the validation set). Conclusion: Genomic alterations and reduced TIL density underpin trastuzumab resistance. The novel TRAG signature and integrated prognostic model enhance risk stratification and may guide personalized adjuvant therapy in early-stage HER2+ BC.