Abstract Background: Ovarian carcinoma is comprised of distinct histological subtypes with different etiology, molecular, genomic and clinical attributes. Patterns of genomic diversity and different treatment responses differentiate each ovarian cancer histotype. The relative patterns of both mutational, copy number and structural variation have not been studied with relation to each disease phenotype. We hypothesized that global genomic architectures will stratify ovarian cancer patients and reveal different treatment response groups. Methods: Whole genome sequencing was performed on 133 ovarian tumors, including 123 carcinomas (59 high-grade serous (HGSC), 35 clear cell (CCOC), 29 endometrioid (ENOC)) and 10 granulosa cell tumours (GCT). Profiles of copy number aberrations, loss of heterozygosity (LOH), mutations (SNVs and INDELs) and structural variations were assessed. Mutational characteristics including mutation signatures derived from tri-nucleotide substitution patterns together with genomic structural characteristics, such as the relative proportion of rearrangement types, reflective of specific DNA repair processes were calculated for each patient. Results: Integrative clustering of the 133 patients according to their mutation and structural signatures resulted in seven distinct subgroups of patients. LOH and the homologous recombination deficiency mutation signature mainly distinguished HGSC cases from non-serous histotypes. HGSC cases were further clustered into two main subgroups. One subgroup (n = 23, 39%) showed a high prevalence of foldback inversions with homology size >5bp, while the other group (n = 25, 42%) was enriched in tandem duplications and deletions, and associated with microhomology (<3bp). Survival analysis revealed that the foldback inversion group associated with poor overall and progression-free survival (logrank p-value = 0.016 and 0.015). CCOC cases were characterized by tandem duplications (Median = 39%, p-value <0.001). The mutation signatures further identified two main subgroups of CCOC; one (n = 10, 29%) showing prevalence of kataegis events typically associated with an APOBEC mutational signature, and the other (n = 17, 49%) characterized by an age-related signature. Enrichment of a mis-match repair defect signature identified a microsatellite instable subgroup of ENOC (n = 8). A signature related to breast cancers uniquely identified GCT cases. Conclusion: Our results suggest that mutational and chromosomal structural variant signatures (rearrangement and copy number profiles) constitute new and defining features of ovarian carcinoma that relate to different DNA repair mechanisms. Our results provide insight into divergent etiologies within histotypes and suggest a novel structure on which to base treatment. Citation Format: Yikan Wang, Ali Bashashati, Michael S. Anglesio, Dawn Cochrane, Diljot Grewal, Hugo Horlings, Anthony Karnezis, Anne-Marie Mes-Masson, Aikou Okamoto, Satoshi Yanagida, Nozomu Yanaihara, Misato Saito, Blake Gilks, Jessica McAlpine, Samuel Aparicio, David Huntsman, Sohrab Shah. Genomic consequences of aberrant DNA repair stratify ovarian cancer histotypes. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-324.
更多