Supplementary file 2_A pathology-based surrogate model for chemotherapy decision-making in intermediate-risk luminal breast cancer: validation of histologic grade and Ki67 in a Chinese population.docx

IntroductionBreast cancer requires precise decisions regarding postoperative adjuvant chemotherapy, as these choices critically influence both survival rates and quality of life. Although genetic testing has enhanced risk stratification, its limited accessibility in grassroots healthcare settings stems from high costs and ongoing debates regarding Western-derived models’ suitability for Chinese populations. This is particularly evident in intermediate-risk luminal-subtype patients without lymph node metastasis, where therapeutic controversies persist. Existing prediction models based on routine immunohistochemical markers face limitations due to data bias from Western cohorts, insufficient standardization of parameters, and interference from heterogeneous high-risk factors. Our study aims to validate clinicopathological predictors of chemotherapy responsiveness (including histological grade and a Ki67 index ≥20%), with the ultimate goal of developing a cost-effective predictive model specifically designed for Chinese patients.
MethodsThis retrospective study included surgically treated intermediate-risk breast cancer patients from Huizhou Third Municipal Hospital (2015–2022). It compared survival outcomes between chemotherapy-treated and chemotherapy-naïve cohorts. Propensity score matching (PSM) was applied to balance baseline disparities. The Kaplan–Meier method and Cox regression analyses were employed to assess disease-free survival (DFS) and overall survival (OS).
ResultsThis study enrolled 476 intermediate-risk breast cancer patients (chemotherapy group: 288 cases; non-chemotherapy group: 186 cases), with a median follow-up of 57.5 months. After propensity score matching (196 cases), multivariate analysis identified histological grade 2–3 (HR = 5.398, p = 0.028), a Ki67 index ≥20% (HR = 5.020, p = 0.012), and absence of chemotherapy (HR = 3.998, p = 0.014) as independent risk factors for DFS. Chemotherapy significantly reduced the risk of recurrence in the Ki67 ≥ 20% subgroup (6.5% vs. 34.9%, p = 0.003) and the histological grade 2–3 subgroup (7.1% vs. 35.6%, p = 0.007), but it showed no significant benefit for the patients with a Ki67 index <20% or grade 1 (p > 0.05). In OS analysis, only a Ki67 index ≥20% emerged as an independent prognostic factor (HR = 9.301, p = 0.035), with no statistically significant difference observed for chemotherapy (p = 0.085).
ConclusionOur findings validate the histological grade and a Ki67 index ≥ 20% as predictive biomarkers of chemotherapy responsiveness in intermediate-risk luminal-subtype breast cancer. This evidence-based stratification offers clinical guidance for resource-constrained settings, while further prospective validation and integration with emerging biomarkers are warranted to refine precision therapeutic strategies.