Abstract Introduction: Breast cancer metastases (BCM), which cause most breast cancer (BC)-associated mortality, have increased genetic complexity compared to early-stage disease. However, the contribution of genetic alterations to site of BCM is not well-understood. Different breast cancer subtypes have varying patterns of BCM, e.g., lobular carcinoma more frequently spreads to gynecologic (Gyn) organs and the GI tract, perhaps hinting at selection pressures wherein some organs are hospitable to tumors with certain genetic alterations. Methods: Relationships between BCM site and mutations detected by DNA next-generation sequencing (NGS; NextSeq 592 gene panels or NovaSeq whole exome sequencing) were investigated using 12,464 BC samples sequenced at Caris Life Sciences (sample sizes, Table 1). PD-L1 expression was tested through IHC (Clone SP-142 (cut-off ≥1, 1%)). Tumor mutational burden (TMB) was measured by totaling somatic mutations per tumor (high ≥ 10 mutations per MB). Immune cell fractions were calculated by deconvolution of whole-transcriptome data (NovaSeq) using Quantiseq (reference). Statistical significance was determined using chi-square and Wilcoxon rank sum tests adjusted for multiple comparisons. Results: Compared to primary breast tumors, BCM had increased frequency of TMB-H (10.08% vs. 4.94%), decreased PD-L1 positivity (21.09% vs. 35.82%), and were enriched for PIK3CA (34.62% vs 30.53%) and ESR1 mutations (13.34% vs 2.17%) (all P<0.001). PD-L1 positivity was highest in BCM to lymph nodes (43.06%) and axilla (39.77%). BCM to Gyn organs had more lobular histology, the highest rate of hormone receptor (HR)+ tumors (77.17%), and rarely had high TMB (6.73%) or were PD-L1 positive (11.39%). Double dendrogram hierarchical clustering of BCM site by mutation frequency and pathway alterations revealed BCM to Gyn organs as a simplicifolious clade with a unique mutational pattern. Compared to BC in breast, BCM to Gyn organs had higher rates of mutations of PIK3CA, AKT1, and BRAF; more mutations in DNA repair (0.79% vs 0.06%), transcription factor (4.72% vs 0.93%), and Wnt signaling pathways (2.36% vs 1.47%); but no increase in BRCA mutations. BCM to brain had the most p53 pathway and homologous recombination (HR) pathway mutations (64.71% and 14.01%), while Gyn had the least (19.69% and 7.09%). Quantiseq RNA deconvolution revealed differences in tumor immune cell infiltrate by BCM site. Gyn metastases vs breast tumors had increased B cells (6.20% vs 5.40%), M2 macrophages (5.71% vs 4.07%), and NK cells (3.82% vs 3.18%) (all P<0.01) and a M2/M1 macrophage ratio of 22.8:1 vs 1.3:1. Conclusions: BCM to Gyn organs have a unique mutational and immune suppression profile. Integrating the profiling with clinical outcomes may extend this prognostic signature and set the stage for improved treatment strategies for these patients. Confirmation from matched or sequential specimens could clarify tumor evolution. Our data support repeat biopsy of Gyn site metastases since more targetable mutations might be revealed. Targeting mechanisms of immunosuppression in Gyn BCM could expand therapeutic options. Table 1.Breast Cancer TumorsTumor SiteTotalPredominant Breast Cancer SubtypeAll12464HR+/HER2- (51.6%)Breast5014HR+/HER2- (46.5%)Liver2003HR+/HER2- (63.3%)Bone1132HR+/HER2- (69.4%)Axilla1051HR+/HER2- (47.7%)Lung823HR+/HER2- (46.1%)Lymph Node647HR+/HER2- (43.4%)Chest/Chest Wall375HR+/HER2- (44.8%)Brain359TNBC (38.2%)Other315HR+/HER2- (57.1%)Skin282HR+/HER2- (50.7%)Connective Tissue193HR+/HER2- (51.8%)GI Organs143HR+/HER2- (69.9%)Gynecologic Organs127HR+/HER2- (76.4%) Citation Format: Anna D Louie, Rani Chudasama, Sharon Wu, Marzia Capelletti, Daniel Magee, W. Michael Korn, Virginia Kaklamani, Antoinette R Tan, Pavani Chalasani, Wafik S El-Deiry, Don Dizon, Stephanie L. Graff. Mutational landscape and immune infiltration of breast cancer metastases to gynecologic and other organs [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD6-04.
Read full abstract