Background: Patients (pts) with early stage HER2+ breast cancer (HER2+ EBC) are often treated with neoadjuvant chemotherapy (NACT). Recent studies have demonstrated that increased densities of tumor infiltrating lymphocytes (TIL) are associated with increased rates of pathologic complete response (pCR) in HER2+ EBC. However, outside of standard TIL analysis, a more comprehensive description of the tumor microenvironment (TME) and its impact on clinical outcomes in HER+ EBC is not well defined. Using novel methods, we sought to characterize the TME of patients with hormone receptor positive (HR+), HER2+ EBC treated with NACT. In addition to standard H&E based TIL analysis, we utilized quantitative multiplex immunofluorescence (qmIF), spatial analysis and analysis of RNA-based gene signatures to assess for pretreatment TME characteristics associated with pCR to NACT in HER2+/HR+ EBC. Methods: We identified 29 pts with HER2+/HR+ EBC treated with NACT between 2005 and 2015 for which pretreatment biopsy tissue was available. 13 pts (45%) were treated with an anthracycline, 26 (90%) were treated with trastuzumab in addition to chemotherapy and 14 (48%) were treated with trastuzumab and pertuzumab. At surgery, 13 pts (45%) achieved a pCR. A breast cancer pathologist performed TIL analysis according to international guidelines on pretreatment H&E slides. qmIF was performed on 28 pretreatment specimens using an antibody panel targeting CD3, CD8, CD68, FOXP3, and pancytokeratin. We analyzed multispectral images using HALO software to perform tissue segmentation (tumor vs stroma) and phenotypic analysis based on antibody staining patterns. Upon spatially resolving these components, we then performed proximity analysis. We obtained mRNA-based gene signatures via the Nanostring nCounter Breast Cancer 360 Panel for 28 patients. T-tests were performed to evaluate for associations of TME features with pCR. Results: We evaluated the association of pCR with: manual H&E-based TIL assessments, lymphocyte densities and interactions (assessed by spatial proximity) from qmIF, and RNA signatures assessed from the NanoString BC360 panel. Univariate analysis revealed strong associations of inflammation signatures derived from qmIF with pCR, including higher stromal densities of CD3+ (p<0.01), CD8+ (p=0.02) and CD3+CD8-FOXP3- (p=0.02) cells. pCR was also associated with higher ratios of CD3+:CD68+ (p<0.01). This finding remained significant in T-cell subsets CD8+:CD68+ (p=0.04) and was driven by CD3+CD8-FOXP3-:CD68+ (p=0.02). High numbers of lymphocyte interactions, assessed by counting pairs of cells within 30 microns of each other, were also associated with pCR, including CD3+:FOXP3+ (p<0.01) and CD3+CD8-FOXP3-:FOXP3+ (p<0.01). We found no significant associations between pCR and manual H&E based TIL assessments or RNA signatures. Conclusion: In this population of HER2+/HR+ EBC treated with NACT, stromal and intratumoral TIL assessment by standard H&E was not associated with pCR. Higher stromal densities of CD3+, CD8+, and CD3+CD8-FOXP3- TIL subsets were associated with pCR by qmIF and further analysis revealed the presence and proximity of lymphocyte subpopulations correlated with pCR, suggesting that spatial relationships between these populations within the TME contributes to sensitivity to NACT. The lack of association between TILs and pCR in our cohort by standard H&E is likely due to small numbers; however, significance was seen by qmIF, indicating this method as a potentially more sensitive approach to this analysis. Analysis of RNA signatures by NanoString BC360 panel did not reveal signatures associated with pCR . While our study demonstrates the feasibility of this approach to TME evaluation, further studies of larger cohorts are needed to validate markers predictive of response to NACT.
Citation Format: Matthew Ryan Kearney, Rami Vanguri, Qi Wang, Kathleen Fenn, Hua Guo, Douglas Marks, Hanina Hibshoosh, Kevin M Kalinsky, Eileen Connolly. Characterization of the tumor microenvironment in patients with hormone receptor positive, HER2 negative early breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P2-07-09.