Immunotherapy predictive biomarkers in metastatic breast cancer (MBC).
1023 Background: We queried whether comprehensive genomic profiling (CGP) of mBC subtypes could identify biomarkers that have been linked to responsiveness to immunotherapy (IO) treatments. Methods: DNA was extracted from 3,871mBC: 1,388 ER+/HER2-, 1,969 HER2 amplified (amp) and 643 TNBC. CGP was performed using a hybrid-capture assay. Tumor mutational burden (TMB) was determined on 1.1 Mbp of sequenced DNA and microsatellite instability (MSI) was determined on 114 loci in 564 cases. PD-L1 was determined by IHC (Dako 22C3). Results: Patient ages were similar. Genomic alterations (GA)/tumor were similar ranging from 5.9 to 7.3. MTOR pathway targets were common; lowest in TNBC. CDH1 and ESR1 GA most frequent in ER+/HER2- cases. ERBB2 short variant (SV) mut were most frequent in ER+/HER2- and HER2 amp and not seen in TNBC. Other kinase targets were uncommon except for FGFR1 GA in ER+/HER2-. BRCA1/ 2 GA least frequent in HER2 amp. AR was amplified in 1% of all mBC. Markers of potential IO benefit: CD274 ( PD-L1) amp (1-3%), BRAF GA (1-4%), TMB of ≥10 mut/Mb (8-12%), TMB of ≥20 mut/Mb (2-3%), MSI-High (0.1-0.4%), PBRM1 GA (1%) and low (1-10%) or high (0-3%) PD-L1 staining. Potential markers of resistance: inactivating GA in STK11 (1-2%) and MDM2 amplification (3-6%). Examples of mBC patients responding to IO therapies will be presented. Conclusions: In addition to guiding targeted therapy selection, CGP shows potential to identify GA linked to response and resistance to IO in mBC. The demonstrations of clinical benefit of immunotherapy in mBC supports the need for the development of biomarkers used to guide the use of ICPI drugs for these patients. [Table: see text]