Predictive Biomarkers for Immune Checkpoint Inhibitors in Metastatic Breast Cancer

Abstract

We examined a large dataset of female metastatic breast cancers (MBCs) profiled with comprehensive genomic profiling (CGP) to identify the prevalence and distribution of immunotherapy responsiveness-associated biomarkers. DNA was extracted from 3831 consecutive MBCs: 1237 (ER^pos /HER2^neg ), 1953 ER^neg /HER2^amp , and 641 triple-negative breast cancer (TNBC). CGP was performed using the FoundationOne^® or FoundationOne^® CDx NGS assay. Tumor mutational burden (TMB) and microsatellite instability (MSI) were determined in a subset of cases. PD-L1 expression in immunocytes in a subset of cases was determined by immunohistochemistry using the companion diagnostic VENTANA PD-L1 SP142 Assay. The median age of the cohort was 54 years (range 20-89). Genomic alterations (GAs)/tumor were similar (range: 5.9-7.3). Markers of potential immune checkpoint inhibitor (ICPI) benefit included: CD274 (PD-L1) amplification (1%-3%), BRAF GA (1%-4%), TMB of ≥10 mutations/Mb (8%-12%), MSI-high (0.1%-0.4%), PBRM1 GA (1%), and positive PD-L1 staining of immunocytes ranging from 13% in ER^pos /HER2^neg and 33% in ER^neg /HER2^amp to 47% in the TNBC group. Potential markers of ICPI resistance included inactivating STK11 GA (1%-2%) and MDM2 amplification (3%-6%). MTOR pathway targets were common with lowest frequency in TNBC. ERBB2 short variant mutations were most frequent ER^pos /HER2^neg and absent in TNBC. BRCA1/2 GA were least frequent in ER^neg /HER2^amp . The demonstrations of clinical benefit of immunotherapy in MBC support the need for development and utilization of biomarkers to guide the use of ICPIs for these patients. In addition to guiding therapy selection, CGP shows potential to identify GA linked to response and resistance to ICPI in MBC.

Keywords: PD-L1; biomarkers; comprehensive genomic profiling; immunotherapy; metastatic breast cancer; microsatellite instability; tumor mutational burden.

FIGURE 1

Most prevalent genomic alterations by gene and alteration type across breast cancer subtypes. (A) Genomic alterations in 1237 ER^pos/HER2 ^neg metastatic breast cancers. (B). Genomic alterations in 1953 ER^neg/HER2 ^amp metastatic breast cancers. (C) Genomic alterations in 641 triple‐negative breast cancers

FIGURE 2

Case of stage IV triple‐negative breast cancer in a 72‐year‐old woman with a CD274 amplification. The genome‐wide copy number plot demonstrates an amplification of the CD274 gene at 12 copies per cell

FIGURE 3

Case of stage IV triple‐negative breast cancer in a 57‐year‐old patient with MDM2 amplification. Comprehensive genomic profiling revealed multiple potentially targetable gene amplifications in FGFR1 and PIK3CA, but also featured amplification of MDM2