Abstract
3 min readAbstract TNBC account for ∼15% of breast cancers and are most difficult to treat. However, TNBC patient outcome is very heterogeneous. In an effort to characterize the biological characteristics of TNBC, we examined 190 routinely diagnosed tumor tissues from early high-risk breast cancer patients treated with adjuvant chemotherapy (anthracyclines and/or taxanes). Highly multiplexed PCR primer pools were designed for 43 genes previously implicated in TNBC with Ion Ampliseq optimized for FFPE samples. Upon library construction and clonal amplification, amplicons were massively sequenced on Ion Proton PI chips and base-called (Torrent Suite 3.6). Variants were called and annotated (Ion Reporter 1.6), and accepted for analysis upon stringent read quality filtering at p<0.001. Informative results were obtained in 183 cases (96.3%). Deleterious and hot-spot mutations (Ingenuity and Oncomine databases) were observed in 39/43 genes, and were more frequent in TP53 (77%), CDH1 (29%), PIK3CA (16.4%), ARID1B (10.4%). TP53 was affected as a single gene in 55/183 (30%) of the cases. Most TNBC (103/183, 56.3%) were mutant in more than one genes; double deleterious/damaging TP53 mutations were observed in 59 (32.2%) tumors at relatively high incidence; and, 29 (15.8%) tumors had extremely variable mutation profiles with >4 mutations per sample, most of them at low incidence but still indicative of dynamic clonal expansion. CDH1 mutations seldom occurred alone (2% of all tumors) and were probably non-founders, while mutations in some genes, e.g., AKT1 (9/12) and ARID1B (13/19) preferentially occurred in CDH1mutant tumors (p<0.001). In comparison to the 12 patients with tumors free of mutations in any of the genes tested (no events up to 120 mo after treatment start), patients with tumors bearing any number of mutations (n = 171) had significantly shorter disease-free survival (DFS, median: 56 mo, log-rank p = 0.031). Mutations in single genes were not significantly associated with disease outcome. The 29 patients with tumors bearing >4 mutations had longer median DFS (91 mo) as compared to those with 1 mutation (n = 68, median DFS 46.5 mo) or with 4 mutations (n = 21, median DFS 39 mo) (HR: 0.4; 95%CI: 0.2-0.8; Wald's p = 0.010). In comparison to patients with TP53&CDH1 non-mutant tumors (n = 26, median DFS 62 mo), patients with TP53mutant/CDH1non-mutant tumors relapsed significantly earlier (n = 103, median DFS 48 mo, HR 3.9, 95%CI 1.2-12.9, Wald's p = 0.023) and tended to have shorter overall survival. CDH1mutant/TP53non-mutant and tumors mutated in both genes did not show such associations (interaction Wald's p = 0.044). No interactive effects on patient outcome were observed between mutation markers and treatment with taxanes, adjuvant radiotherapy, or with standard clinicopathologic parameters. In conclusion, TNBC may be assigned as a TP53 disease. However, the present data underline the need for a broad assessment of tumor mutational profiles, including non-founder mutations, since they may interfere with patient outcome upon standard treatments. The application of targeted parallel sequencing on routinely processed FFPE tissue samples seems feasible and may help in assessing clinically relevant genomic variant profiles of these highly intra- and inter-heterogeneous tumors. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-04-10.
Discussion(0)
No comments yet. Be the first to comment.