Abstract P2-02-06: Biomarker analysis by next generation circulating tumor DNA (ctDNA) sequencing in patients with advanced breast cancer
H Li,J Wang,HS Rugo,Y Zhang,L Yang,X Liu,B Shao,Y Xu,L Yang,R Zhang,R Ran,L Chang
DOI: https://doi.org/10.1158/1538-7445.sabcs17-p2-02-06
IF: 11.2
2018-01-01
Cancer Research
Abstract:Abstract Background: Next-generation sequencing is of increasing interest to identify specific targets for both drug development and treatment. The study of metastatic cancer is complicated by lack of tissue and the potential for change in biology over treatment. We evaluated ctDNA in patients with advanced breast cancer to explore the relationship between specific DNA mutations and prognosis as well as therapeutic decision making. Methods:Peripheral blood was collected in EDTA at the time of diagnosis of advanced disease. Samples were sent to Geneplus-Beijing for sequencing. Indexed Illumina libraries were prepared from germline and circulating DNA using the KAPA Library Preparation Kit; the capture probe was designed based on genomic regions selected with 1021 genes, covering the most frequently mutated genes and exons in solid tumors. Clinical characteristics, treatment and outcome data were collected. We analyzed progression free survival (PFS) from first-line therapy and overall survival (OS), endpoints were correlated with observed gene mutations. Results: 54 patients were enrolled; 27 (50%) HER2+, 22 (41%) hormone receptor + (HR+)/HER2-, and 5 (9%) triple negative (TNBC). Median age was 48 (range 26-74). The median follow-up was 8 years (range 12-180 months). First-line therapy included chemotherapy with trastuzumab for HER2+ disease, chemotherapy with endocrine maintenance (17) or endocrine therapy alone (5) for HR+/HER2- disease, and chemotherapy for TNBC. Mutations were found in TP53, PIK3CA, PIK3CA 3140 A>G(p.H1047R) and ERBB (including ERBB1-4), at 40.7%, 35.2%, 20.4% and 25.9%, respectively. In univariate analysis, patients with tumor mutations in TP53 had a shorter OS (median 64 vs 121 months, p=0.006). The PIK3CA 3140 A>G mutation was more frequent in HER2+ (7/27, 25.9%) than HR+/HER2- (4/22 (18.2%) or TNBC (0/5), and was associated with shorter median PFS in HER2+ disease (mutant vs. wild type: 4 (range 2-9) vs. 8 (range 2-22) months, p=0.006). The frequency of ERBB mutation was similar in HER2- 7/27(25.9%) (p=0.707) or HER2+ 7/27(25.9%) disease (p=0.066); there was no significant impact on PFS in any subset. Multivariate analysis for HER2+ disease including age, ER, Ki67, TP53, PIK3CA, PIK3CA 3140 A>G and ERBB), demonstrated that the PIK3CA 3140 A>G mutation was the only factor associated with shorter PFS (p=0.025); further analysis by receiver operating characteristic (ROC) curve showed that the PIK3CA 3140 A>G mutation and the mutation in PIK3CA 3140 A>G and ERBB combination pathway had a large area under the curve (AUC), with AUC of 0.789, and 0.734 respectively. Conclusions: Using NGS in ctDNA, we found that the PIK3CA 3140A>G mutation was more frequent in HER2+ disease, and was the only mutation associated with shorter PFS on multivariate analysis. The presence of a TP53 mutation was associated with worse OS. Evaluation of ctDNA is feasible in a general breast cancer population and has prognostic impact; further correlation of these findings with tumor samples is ongoing. Citation Format: Li H, Wang J, Rugo HS, Zhang Y, Yang L, Liu X, Shao B, Xu Y, Yang L, Zhang R, Ran R, Chang L. Biomarker analysis by next generation circulating tumor DNA (ctDNA) sequencing in patients with advanced breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P2-02-06.