Sequential and co-occurring DNA damage response genetic mutations impact survival in stage III colorectal cancer patients receiving adjuvant oxaliplatin-based chemotherapy

• Published in: BMC cancer Vol. 21; no. 1; pp. 217 - 11
• Main Authors: Lin, Peng-Chan, Yeh, Yu-Min, Chan, Ren-Hao, Lin, Bo-Wen, Chen, Po-Chuan, Pan, Chien-Chang, Shen, Meng-Ru
• Format: Journal Article
• Language: English
• Published: London BioMed Central 02.03.2021; BioMed Central Ltd; Springer Nature B.V; BMC
• Subjects: Adenomatous polyposis coli; Adenomatous Polyposis Coli Protein - genetics; Aged; Biomedical and Life Sciences; Biomedicine; BRCA2 protein; BRCA2 Protein - genetics; Cancer; Cancer Research; Chemotherapy; Clinical outcomes; Colorectal cancer; Colorectal carcinoma; Colorectal Neoplasms - drug therapy; Colorectal Neoplasms - genetics; Colorectal Neoplasms - mortality; Colorectal Neoplasms - pathology; Decision trees; Deoxyribonucleic acid; Development and progression; DNA; DNA Damage; DNA damage response gene; DNA repair; Drug resistance; Drug therapy; Evolution; Feature selection; Female; Gene mutations; Genetic aspects; Health aspects; Health Promotion and Disease Prevention; Humans; Learning algorithms; Machine learning; Male; Medical prognosis; Medicine/Public Health; Middle Aged; Mismatch repair; Mutation; Neoplasm Staging; Oncology; Oxaliplatin; Oxaliplatin - administration & dosage; p53 Protein; Patients; Phylogenetics; Phylogeny; Proto-Oncogene Proteins p21(ras) - genetics; Research Article; Risk groups; Sequential somatic mutations; Surgical Oncology; Survival; Tumor evolution; Tumor heterogeneity; Taiwan; Tumor heterogeneity; Tumor evolution; Sequential somatic mutations; DNA damage response gene; Colorectal cancer
• ISSN: 1471-2407; 1471-2407
• DOI: 10.1186/s12885-021-07926-1

Background Certain sequences of genomic mutations can lead to cancer formation and affect treatment outcomes and drug resistance. We constructed a cancer evolutionary tree using bulk-targeted deep sequencing to explore the impact of sequential and co-occurring somatic mutations on patients with stage III colorectal cancer (CRC). Methods A total of 108 stage III CRC patients from National Cheng Kung University Hospital (NCKUH) were recruited for this study between Jan. 2014 and Jan. 2019. Clinical information and tumor-targeted deep sequencing data were collected. Phylogenetic trees were reconstructed for evolutionary trajectories. We used a machine learning model for survival analysis. Results Six sequential somatic mutations stratified patients into seven subgroups based on survival. Patients carrying sequential germline followed by DNA damage response-related ATM or BRCA2 somatic mutations or non- TP53, APC somatic mutations had a better outcome than those without such mutations. The 4-year recurrence-free survival (RFS) probability was 88% in the low-risk group (G1) and 46% in the high-risk group (G2) (log-rank p -value 2e-05). The predictive efficacy by the area under the curve (AUC) was 0.73, 0.7, 0.797, and 0.88 at 2, 4, 6, and 8 years, respectively. The mutation status of mismatch repair (MMR) genes was not associated with RFS. Different genomic features were found between the groups. The orders of APC , KRAS and APC , BRCA2 sequential somatic mutations were associated with clinical outcomes. The occurrence of somatic mutations in BRCA2 , such as TP53 somatic mutations, affected recurrence-free survival. Conclusions According to the evolution model, DNA damage response (DDR)-related ATM or BRCA2 somatic mutations are promising biomarkers for assessing the response of stage III CRC patients to oxaliplatin-based chemotherapy. The sequential order and co-occurring DDR somatic mutations are associated with recurrence-free survival.