Androgen receptor expression and immune characteristics of HER2-low metastatic triple-negative breast cancer.

Methods

Study sample
This study utilized tumor samples and clinical data from the pre-screening phase of 17-024, a phase II trial evaluating abemaciclib in retinoblastoma-positive (Rb-positive) mTNBC (NCT03130439)12. Patients were required to undergo IHC staining for Rb (clone clone G3-245–RRID: AB_395259, BD Bio, catalog No. 554136), with enrollment restricted to those whose invasive tumors stained ≥50% positive for Rb.
Although patients with Rb-negative disease were excluded from the trial intervention, their tumor samples were retained and the consent forms included permission for further research analyses. In total, 196 patients with mTNBC (regardless of Rb status) completed pre-screening and were included in this present study, with available tissue and OS data. Tumor samples were collected from heterogenous locations (i.e., primary tumors or metastatic biopsies were both eligible for the study). All research was performed in accordance with the Declaration of Helsinki. Institutional review board approval for the conduction of this study was obtained from the Dana-Farber/Harvard Cancer Center and from the Duke Cancer Institute. Written informed consent was obtained from all participants.

Biomarker assessment
Centralized IHC staining was performed for AR (clone clone AR441, Dako, catalog No. M3562–RRID: AB_2060174), and PD-L1 (clone 405.9A11, Cell Signaling Technology, catalog No. 29122–RRID: AB_2798970) by experienced pathologists at Brigham and Women’s Hospital. Central quantification of TILs was performed by histopathologic hematoxylin and eosin-stained tissue slides according to the International TILs Working Group guidelines13. HER2 statuses were retrieved from pathology reports. Within the TNBC population, HER2 expression was categorized as HER2-low if IHC 1+ or 2 + /not-amplified, or HER2-0 if IHC 0. AR expression was assessed both continuously and in discrete categories: negative (0%), intermediate (1–50%), and high (>50%). PD-L1 expression was defined by CPS as is standard, with a threshold of CPS ≥ 10 indicating PD-L1 positivity. TILs were measured both continuously and in discrete categories: absent (0%), intermediate (1–5%) and high (>5%), with TIL positivity defined by ≥1%. <1% TILs were considered as absent.

Statistical analysis
The primary aim was to evaluate the association between HER2 and AR expression in mTNBC. Secondary analyses explored the relationship between HER2-low status, PD-L1 expression, TIL levels, and OS.
OS was defined as time from metastatic diagnosis to death due to any cause, or censored at date last known alive. OS was analyzed in the full cohort (n = 196) and compared between HER2-low and HER2-0 subgroups. Median OS with 95% CIs was estimated via the Kaplan-Meier method for each group. Log-rank tests were used to assess differences.
Expression levels of AR and PD-L1 were evaluated as both continuous and categorical variables. Association between the biomarkers and HER2 status was compared using Wilcoxon rank-sum test (continuous biomarker) and Chi-square test (categorical biomarker). Box plots and histograms were generated to visualize AR expression by HER2 status.
OS was further stratified by HER2 status and biomarker positivity (AR or PD-L1), with median OS (95% CIs) reported for each subgroup using the Kaplan-Meier estimation method. All analyses were conducted using R version 4.4.1, with two-sided p-values < 0.05 considered statistically significant.

Supplementary information