Abstract P4-10-12: Characterizing the tumor and immune microenvironment through treatment to predict response to neoadjuvant HER2-targeted therapy using the Digital Spatial Profiler

Abstract

Abstract Background: While introduction of HER2-targeted therapies has dramatically improved outcomes for patients with HER2-positive disease, even with the addition of HER2-targeted agents, 40-50% of patients do not achieve a pCR (pathologic complete response) following neoadjuvant therapy implying that clinical or molecular differences may be present in responders versus non-responders. While recent bulk expression studies have identified several biomarkers associated with response to HER2-targeted therapies in the neoadjuvant setting, these studies are limited in their ability to assign observed changes to specific geographic or phenotypic cell populations, such as the malignant tumor core or the surrounding microenvironment. Methods: Here we used the Digital Spatial Profiler (DSP, NanoString Technologies, Inc.) to profile regions-of-interest containing pancytokeratin (panCK)+ tumor cells and infiltrated immune cells that are co-localized with the tumor cells. Using this technology, we assayed archival tissue from 28 patients with HER2-positive breast cancer from the TRIO-B07 (NCT00769470) clinical trial, who were treated with trastuzumab, lapatinib, or both, followed by standard chemotherapy plus HER2-targeted therapy. Tissue specimens were collected from the pre-treatment diagnostic biopsy (Baseline) and after one cycle of targeted therapy (Runin). To study regional heterogeneity, we selected an average of four panCK-enriched tissue regions from each sample. Using DSP, we performed multiplexed quantification of 38 tumor and immune protein markers and 96 RNA markers on the selected tissue regions and compared our findings to bulk mRNA expression data from the same cohort. Results: Within the panCK-enriched regions, DSP revealed significant treatment-associated decreases in HER2 protein levels and the downstream PI3K-Akt signaling pathway in Runin compared to Baseline samples. In tandem, we observed a significant increase in infiltrating leukocytes, with CD45, a pan-leukocyte marker, and CD8, a marker for T cells that mediate tumor cell killing, showing the most dramatic changes. These changes in Runin compared to Baseline were more significant in the subset of cases that achieved a pCR versus those that do not, independent of ER status. Comparison of Runin samples to matched Baseline samples from the same patient enabled improved prediction of patient outcome (pCR) compared with analysis of a single timepoint alone. We also found that the DSP panCK enrichment strategy captures additional signal not observed in bulk expression data. For instance, using bulk expression, a decrease in HER2 RNA levels between Baseline and Runin was evident but there was no difference in the degree of decrease in HER2 mRNA between pCR and no pCR cases. Using DSP, we observed that the significant decrease in HER2 levels at Runin is more pronounced in cases that achieved a pCR. Across both tumor and immune markers, regional heterogeneity increased at Runin compared to Baseline. Conclusions: In this study, we used DSP and a panCK enrichment strategy to retrospectively delineate the changes that occurred in tumor cells and co-localized immune cells during HER2-targeted therapy. In comparison to traditional or multiplexed IHC, DSP allows for simultaneous profiling of a large number of markers, enabling the characterization of multiple cancer signaling pathways and immune markers on a single tissue specimen. This study demonstrates the utility of pancytokeratin-enriched spatial proteomic profiling to characterize treatment-associated changes and identify predictive biomarkers. NanoString’s Digital Spatial Profiler is for Research Use Only. Not to be used for diagnostic procedures. Citation Format: Katherine Lee McNamara, Jennifer L. Caswell-Jin, Zhicheng Ma, Jason J. Zoeller, Michelle Kriner, Zoey Zhou, Jason Reeves, Margaret Hoang, Joseph Beechem, Dennis J. Slamon, Michael F. Press, Joan Brugge, Sara A. Hurvitz, Christina Curtis. Characterizing the tumor and immune microenvironment through treatment to predict response to neoadjuvant HER2-targeted therapy using the Digital Spatial Profiler [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-12.