Effects of short-course preoperative endocrine therapy on tumour morphology and immunohistochemical profile in oestrogen receptor-positive, HER2-negative breast cancer.
AIMS: Short-term preoperative endocrine therapy (ET) is increasingly used in oestrogen receptor (ER)-positive, HER2-negative breast cancer as a functional test of endocrine sensitivity. We aimed to characterise histomorphological and immunophenotypic changes following preoperative ET and to identify predictors of endocrine response, defined as post-treatment Ki67 ≤ 10%. METHODS AND RESULTS: In this retrospective single-centre study, 180 patients treated with short-course preoperative ET (median duration 29 days) were compared with 151 patients undergoing primary surgery without ET. Paired biopsy and resection specimens were assessed for histological features, stromal proportion, stromal tumour-infiltrating lymphocytes (strTILs) and expression of ER, progesterone receptor (PR), HER2 and Ki67. Genomic risk was determined using the MammaPrint assay. Preoperative ET was associated with a significant reduction in tumour proliferation, with 73.9% of cases showing post-treatment Ki67 ≤ 10% compared with none in controls (P < 0.001). Histological grade decreased in 36.7% of ET-treated tumours versus 7.9% of controls (P < 0.001), predominantly reflecting reduced mitotic activity. ER expression remained stable, whereas PR expression decreased more frequently following ET (P < 0.001) and was independently associated with Ki67-defined response. HER2-low status was more frequently observed after ET (P < 0.001), but HER2 expression and microenvironmental parameters, including strTILs, were not associated with response. High genomic risk was independently associated with a lower likelihood of achieving post-treatment Ki67 ≤ 10% (P < 0.001). CONCLUSIONS: Short-course preoperative ET induces rapid and reproducible morphological and immunophenotypic changes in ER-positive, HER2-negative breast cancer. Ki67-defined response is associated with genomic risk and PR expression, whereas microenvironmental features appear to have limited predictive value.