Background: Apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like (APOBEC) enzymes drive genomic instability in breast cancer (BC) characterized by specific single base substitution mutation signatures (COSMIC S2 and S13). Computational algorithms have the potential to derive signatures from targeted sequencing assays with relatively small genomic footprints. We sought to develop a computational approach to identify APOBEC mutagenesis in samples subjected to targeted sequencing and to characterize the genomic and clinical features of APOBEC-positive BC. Methods: We retrieved clinical and genomic data from 4,595 consecutive BCs that had been subjected to targeted sequencing using the FDA-approved MSK-IMPACT assay. Additional whole exome sequencing (WES) data were retrieved from publicly available datasets, including TCGA (primary BC, n=1019) and Bertucci et al. cohorts (metastatic BC, n=617). APOBEC-related mutational signatures were computed by using the SigMA algorithm (Gulhan, Nat Genet 2019) for MSK-IMPACT samples with at least 5 single nucleotide variants (SNV) (n=2,983). Survival analyses were performed by using Kaplan Meier method with log-rank test and Cox proportional-hazards models. Results: A significant positive correlation between APOBEC exposure calculated by SigMA vs other computational tools (e.g., DeconstructSig, SigProfiler, MutationalPatterns) was found (r=0.99). After down-sampling the WES data from the TCGA BC dataset to the genomic footprint of MSK-IMPACT, SigMA showed high concordance in determining APOBEC exposures (r=0.78). BCs with high APOBEC exposure were enriched for pathogenic somatic mutations affecting several genes, including PIK3CA, CDH1 and GATA3 (q<0.05), as compared to tumors with no/low APOBEC. In the MSK-IMPACT cohort, median APOBEC mutational exposure was significantly higher in metastatic than primary samples (0.27 vs 0.19, p<0.001) and was the dominant mutational signature in 26% of primary and 34% of metastatic BC. A significant enrichment for APOBEC exposure in the metastatic setting was observed in both hormone receptor-positive (HR+) (p<0.001) and triple-negative (TN) (p<0.001) subtypes. Invasive lobular BCs displayed a higher proportion of cases with high APOBEC mutagenesis when compared to invasive ductal BCs in both primary and metastatic setting (p<0.001). Enrichment of APOBEC in metastatic samples could be due to signature acquisition or enrichment in poor prognosis tumors, to address this, we assessed the presence of APOBEC mutational exposure in patients with paired samples, which included 352 primary/metastasis (P/M) cases and 514 metastasis/metastasis (M/M) cases. Strikingly, no statistically significant increment of APOBEC exposure was found in the P/M nor in the M/M cohort (Wilcoxon signed-rank test p>0.05), overall and in different BC subtypes. We also assessed the association between APOBEC and response to endocrine therapy (ET). Patients with metastatic HR+ BC and APOBEC as dominant signature had a statistically significant lower progression-free survival (PFS) on first-line single-agent ET (aromatase inhibitors or SERDs; median PFS of 9.7 versus 14.7 months, p<0.001). This difference was more pronounced in patients treated with SERD with a median PFS of 4.7 and 14.3 months for APOBEC-dominant and APOBEC non-dominant tumors (p<0.001), respectively. Conclusions: Evidence for APOBEC mutagenesis is present in >30% of metastatic breast cancers and associated with resistance to ET. Unlike ESR1 mutations, which are acquired as a late event in metastatic HR+ BCs, APOBEC mutagenesis is commonly present in paired primary and metastatic BCs, indicating its relatively early onset and potential role in driving the poor prognosis of these cancers.

Citation Format: Antonio Marra, Andrea Gazzo, Pier Selenica, Xin Pei, Avantika Gupta, Fresia Pareja, Giuseppe Curigliano, Reuben Harris, Nadeem Riaz, Jorge S. Reis-Filho, Sarat Chandarlapaty. Apobec mutagenesis is a pervasive feature of poor prognosis breast cancer associating with ESR1 wild type, endocrine resistant disease [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 PD1-06.