Abstract

Investigating the utility of breast cancer (BC) molecular signatures and immune signature profiling (ISP) as predictors of pathological complete response (pCR) to neoadjuvant chemotherapy + trastuzumab (NAC+T) in HER2+ BC.

Author
person Salomon M. Stemmer Davidoff Cancer Center, Rabin Medical Center-Beilinson Hospital, Petah Tikva, Israel info_outline Salomon M. Stemmer, Yulia Shvartser, Itay Israel Shemesh, Idit Peretz, Rinat Yerushalmi, Daniel Hendler, Izhak Haviv, Heather Ann Brauer, Michael Castro
Full text
Authors person Salomon M. Stemmer Davidoff Cancer Center, Rabin Medical Center-Beilinson Hospital, Petah Tikva, Israel info_outline Salomon M. Stemmer, Yulia Shvartser, Itay Israel Shemesh, Idit Peretz, Rinat Yerushalmi, Daniel Hendler, Izhak Haviv, Heather Ann Brauer, Michael Castro Organizations Davidoff Cancer Center, Rabin Medical Center-Beilinson Hospital, Petah Tikva, Israel, Inst of Pathology, Rabin Medical Center, Petah Tikva, Israel, NanoString Technologies, Inc., Seattle, WA, Davidoff Center, Rabin Medical Center, Petah Tikva, Israel, Davidoff Cancer Center, Rabin Medical Center, Petah Tikva, Israel, TyrNovo, Tel Aviv, Israel, Personalized Cancer Medicine PLLC, Los Angeles, CA Abstract Disclosures Research Funding Other Background: Trastuzumab targets HER2 directly as well as contributes to activation of the immune system against the tumor. We investigated the utility of BC molecular signatures and ISP as predictors of pCR to NAC+T in HER2+ BC pts. Methods: The analysis included 73 consecutive HER2+ BC pts with pathological samples at diagnosis, who received NAC+T. RNA was extracted from 73/22 biopsy/surgical specimens (FFPE). RNA was analyzed using NanoString PanCancer Immunoncology 360 (IO360) panel with PAM50 spike-in. Raw gene counts were log2-transformed and normalized (using housekeeping genes). The analysis assessed 47 gene signatures including immune related and PAM50 signatures. Signatures were computed after removing samples with low expression of the housekeeping genes and normalization. Signature scores range from ~0 to 10 and have an average value of 5 in tumor samples from The Cancer Genome Atlas. Results: Overall, 46% of the pts achieved pCR; of 13 pts who were IHC 2+ and FISH positive, one achieved pCR. Significant associations were found between response to NAC+T (pCR, no response) and molecular signatures/ISP from the initial biopsy (Table). For pts who did not experience pCR and who underwent surgery, paired sample analysis was performed assessing molecular signatures/ISP in the initial biopsy sample vs the surgical sample. This analysis demonstrated a significant increase post-treatment (compared to the initial biopsy sample) in Risk of Recurrence (ROR) score (p=.036), interferon (IFN) downstream signature (p=.038), and antigen‐processing machinery (APM) loss (p=.002). Conclusions: Combining molecular BC signatures and ISP on biopsy samples may have a predictive value in HER2+BC pts for whom NAC+T is considered. Luminal A pts and IHC 2+ (FISH positive) are unlikely to achieve pCR with NAC+T. Parameter Log 2 fold-change (pCR vs no response) and 95% CI P value ROR 0.67 (0.16, 1.17) .01 HER2-enriched score 1.18 (0.63, 1.74) <.01 Luminal A score -0.81 (-1.42, -0.20) .01 CD8 0.47 (0.01, 0.94) .04 PD-L2 0.68 (0.15, 1.20) .01 Macrophages 0.35 (0.05, 0.66) .02