Tumor-stroma ratio is associated with Miller-Payne score and pathological response to neoadjuvant chemotherapy in HER2-negative early breast cancer.

Abstract

The tumor‐stroma ratio (TSR) has proven to be a strong prognostic factor in breast cancer, demonstrating better survival for patients with stroma‐low tumors. Since the role of the TSR as a predictive marker for neoadjuvant chemotherapy outcome is yet unknown, this association was evaluated for HER2‐negative breast cancer in the prospective DIRECT and NEOZOTAC trials. The TSR was assessed on 375 hematoxylin and eosin‐stained sections of pre‐treatment biopsies. Associations between the TSR and chemotherapy response according to the Miller‐Payne (MP) grading system, and between the TSR and pathological response were examined using Pearson's chi‐square, Cochran‐Armitage test for trend and regression analyses. A stroma‐low tumor prior to neoadjuvant chemotherapy was significantly associated with a higher MP score (P = .005). This relationship remained significant in the estrogen receptor (ER)‐negative subgroup (P = .047). The univariable odds ratio (OR) of a stroma‐low tumor on pathological complete response (pCR) was 2.46 (95% CI 1.34‐4.51, P = .004), which attenuated to 1.90 (95% CI 0.85‐4.25, P = .119) after adjustment for relevant prognostic factors. Subgroup analyses revealed an OR of 5.91 in univariable analyses for ER‐negativity (95% CI 1.19‐29.48, P = .030) and 1.48 for ER‐positivity (95% CI 0.73‐3.01, P = .281). In conclusion, a low amount of stroma on pre‐treatment biopsies is associated with a higher MP score and pCR rate. Therefore, the TSR is a promising biomarker in predicting neoadjuvant treatment outcome. Incorporating this parameter in routine pathological diagnostics could be worthwhile to prevent overtreatment and undertreatment.