Impact of HER2 low status on genomic signatures in triple negative breast cancer (TNBC).

Melissa Taylor Yale New Haven Hospital, New Haven, CT info_outline Melissa Taylor, Shilpa Reddy, Prashanth Ashok Kumar, Dana Hariri, Ethan Sokol, Smruthy Sivakumar, Julia Quintanilha, Dean Pavlick, Mia Alyce Levy, Jeffrey S. Ross, Maryam B. Lustberg

Authors Melissa Taylor Yale New Haven Hospital, New Haven, CT info_outline Melissa Taylor, Shilpa Reddy, Prashanth Ashok Kumar, Dana Hariri, Ethan Sokol, Smruthy Sivakumar, Julia Quintanilha, Dean Pavlick, Mia Alyce Levy, Jeffrey S. Ross, Maryam B. Lustberg Organizations Yale New Haven Hospital, New Haven, CT, Upstate University Hospital, Syracuse, NY, SUNY Upstate University Hospital, Syracuse, NY, Foundation Medicine, Inc., Cambridge, MA, Foundation Medicine, Inc., Boston, MA, Foundation Medicine, Cambridge, MA, Smilow Cancer Hospital, Yale Cancer Center, New Haven, CT Abstract Disclosures Research Funding No funding sources reported Background: HER2 status has become more important with new treatment regimens subcategorizing HER2 to into low (1+or 2+ FISH-) or absent HER2 expression. Through comprehensive genomic profiling (CGP), we aim to identify distinctive molecular features associated with various immunohistochemistry (IHC) categories of HER2 low tumors. Methods: Retrospective review of pathology specimens submitted to Foundation Medicine were done and 514 TNBC (ER-/PR-/HER20,1+,2+FISH-) and 191 3+/HER2 amplified clinically advanced BC underwent hybrid capture-based CGP to evaluate genomic alterations (GA) in 324 genes, MSI status and tumor mutational burden (TMB, mutations/Mb). PD-L1 expression was determined by IHC (Dako 22C3 TPS). Chi-square test and Mann Whitney U test were used in the statistical comparisons of the 2 groups. Statistical significance was defined as p < 0.05. Results: 53.5% were HER2 IHC 0+, 31.5% were HER2 IHC 1+ and 15.0% were IHC 2+/FISH negative. The HER2 3+/amplified BC were 59% ER+ and 41% ER-. Notable differences in GA frequencies included a significant increase in PIK3CA GA in the HER2 2+ vs the HER 0+ and 1+ groups (p=.0003). A lower frequency in BRCA1 GA in the HER2 2+ (5.2%) vs the HER2 0+ (9.8%) was significant when compared to the 1.0% in the HER2 3+/amplified cases (p<.0001). BRCA2 GA frequencies did not show statistically significant differences among the TNBC HER2 IHC score groups. TP53 and RB1 had the highest frequency in HER2 0+ samples, with decreasing prevalence in the HER2 1+, 2+ and 3+/amplified cases (p<.0001 for both). There was also a trend towards higher frequency of CDH1 GA as the HER2 IHC score increased from 0 to 2+ among TNBC. There were no significant differences in immunotherapy biomarkers (MSI, TMB and PD-L1) based on HER2 status among the TNBC groups. TMB >= 10 mutations/Mb was significantly higher in the HER2 3+/amplified group (p=.004). Conclusions: We found that TNBC HER2 2+ positive FISH show key molecular differences from their HER2 low (1+) or absent (0+) counterparts. The most substantial difference was an increase in GA in PIK3CA. While this is targetable in the hormone positive setting, its potential role as a target in the hormone negative setting is uncertain. Also noted but did not reach significance was a decrease in GA within BRCA1/2 and a positive linear relationship with CDH1 and HER2 IHC expression. These GA differences may indicate potential treatment options for different HER2 subgroups and warrant further investigation. Limitations of our study include the retrospective nature and the lack of clinical data. TNBC HER2 0+ TNBC HER2 1+ TNBC HER2 2+ (FISH-) HER2 3+ (FISH+) P-Value # of Cases 275 162 77 199 (59% ER+) N/A TP53 90.9% 85.8% 84.4% 72.4% <0.0001 RAD21 22.2% 25.9% 32.5% 33.2% 0.070 PIK3CA 18.5% 18.5% 35.1% 34.2% 0.0003 RB1 17.1% 16.1% 11.7% 2.5% <0.0001 BRCA1 9.8% 7.4% 5.2% 1.0% <0.0001 BRCA2 5.5% 3.1% 1.3% 3.5% 0.21 CDH1 3.3% 6.8% 10.4% 4.5% 0.016 TMB >= 10 mut/Mb 4.7% 9.9% 6.5% 11.6% 0.004