Genomic landscape of cell-free DNA in a diverse cohort of metastatic breast cancer patients.

Abstract

e13066 Background: Breast cancer is the most prevalent cancer and the second leading cause of cancer deaths among women in the United States. Once the primary breast cancer has metastasized to other parts of the body, despite treatment, patients have an incurable disease. Compared to the primary tumor, often the metastasized tumor sites have different mutational profiles that are not represented in the initial biopsy. Liquid biopsy is becoming a more commonly accepted method in clinical practice to determine the heterogeneity of metastatic tumors by analyzing cell-free DNA (cfDNA). Several studies have demonstrated that cfDNA can be utilized as an effective prognostic tool and can reveal multiple clinically targetable alterations in metastatic breast cancer (mBC). Methods: We characterized the genomic landscape of cfDNA in a diverse patient population and across different subtypes of mBC. We performed a retrospective chart review in patients (n = 41) with metastatic breast cancer from a medical oncology clinic within the Hawaii Pacific Health system. Patients who had at least one cfDNA testing by Guardant360 were evaluated. Patient demographics, age at diagnosis, race, subtype and mutations were reviewed along with treatment information. Results: We found no statistically significant differences in genomic landscape according to race, however there were differences seen in tumor genomics according to age and subtype. Older patients were less likely to have detectable disease on liquid biopsy compared to younger patients (age 70+ and 60-69 vs 30-39; xp 0.02). Consistent with subtype, mutations in ESR1 (n = 12) and PIK3CA (n = 14) were more commonly seen in hormone positive (HR+) mBC where known tailored treatment options are available (ie. fulvestrant and alpelisib respectively). Postmenopausal patients were more likely to have ESR1 mutations than premenopausal patients (50-59 and 70+ vs 30-39 and 40-49, xp 0.009). About half (n = 20) had TP53 mutations, followed by PIK3CA (n = 14), ESR1 (n = 12), CCND1 (n = 7) and MYC (n = 6). Conclusions: We evaluated 41 unique cases of mBC and correlated them with their genomic profiles. We plan to further investigate genomic changes according to treatment with subsequent repeat liquid biopsies. We hope to utilize these findings to help prognosticate and guide future treatment options.