Search

ObjectiveThe status of homologous recombination repair (HRR) gene mutations and their impact on the survival of patients with Chinese epithelial ovarian cancer (EOC) are still unclear. In this study, we retrospectively analyzed the mutations of HRR genes in tumor tissues and evaluated their values for predicting the survival of Chinese EOC patients.MethodsA total of 273 primary EOC patients from five different hospitals between 2015 and 2016 were recruited. All patients received staging surgeries or debulking surgeries combined with systemic platinum-based chemotherapy. DNA was extracted from formalin-fixed, paraffin-embedded sections and analyzed for mutations using a 21-gene panel (including 13 well-known HRR genes) by next-generation sequencing.ResultsHigh-grade serous carcinoma (HGSOC) accounted for 76.2% of the cohort. A total of 34.1% (93/273) cases had 99 deleterious mutations in 9 HRR genes, namely, BRCA1 (56/273, 20.5%), BRCA2 (20/273, 7.3%), ATM (5/273, 1.8%), RAD51C (5/273, 1.8%), RAD51D (5/273, 1.8%), BRIP1 (2/273, 1.8%), CHEK2 (2/273, 0.7%), FANCI (2/273, 0.7%), and RAD54L (1/273, 0.4%). There is a strong mutual exclusion between HRR genes. The mutation landscape revealed several unappreciated deleterious variants in BRCA1/2 and other HRR genes reported previously. Estimated according to the mutation allele frequency, about 4.8% of the patients had potential somatic HRR gene mutations, which might be underestimated. Moreover, HRR mutations mainly exist in HGSOC (83/208, 39.9%), clear cell (2/30, 6.7%), and endometroid subtypes (8/20, 40%), but not seen in other rare subtypes. BRCA1 mutations tend to be present in younger patients with family history or multiple primary foci. Patients with BRCA1/2 mutations tend to have a longer progression-free survival and overall survival, while other HRR mutation carriers tend to have a shorter progression-free survival, but no significant difference in overall survival.ConclusionThis study revealed the distribution of HRR gene mutations in Chinese EOC tissues. BRCA1/2 account for the majority of HRR gene mutations and predict long prognosis in HGSOC. Non-BRCA HRR mutations also account for a very important proportion and might be associated with poor prognosis in HGSOC. It is suggested that HRR gene mutations need to be detected in EOC tissues and germline status be further clarified in clinical algorithm for potential targeted therapy, genetic screening, and prognosis prediction.

Background:Expert consensus on BRCA1/2 genetic testing and clinical application in Chinese breast cancer patients recommends that BRCA1/2 testing should be performed in those with clinical risk factors, such as an early onset, triple-negative breast cancer (TNBC) or family history of cancer. With the increasing application of multigene panels, testing for genes beyond BRCA1/2 has become more prevalent. However, the non-BRCA mutation status of Chinese high-risk breast cancer patients has not been fully explored.Methods: A total of 230 high-risk breast cancer patients from Fudan University Shanghai Cancer Center who had undergone peripheral blood germline 72 genes next-generation sequencing (NGS) from June 2018 to June 2020 were enrolled for retrospective analysis. The 72 genes include common hereditary breast cancer genes, such as homologous recombination repair (HRR) genes and other DNA damage repair genes. High-risk factors included: 1) TNBC; 2) male breast cancer; 3) primary bilateral breast cancer; 4) diagnosed with breast cancer at age less than or equal to 40 years; or 5) at least one first- and/or second-degree relative with BRCA-related cancer (breast or ovarian or prostate or pancreatic cancer).Results: The germline pathogenic or likely pathogenic mutation rate was 29.6% (68/230) in high-risk breast cancer patients. Among them, 44 (19.1%, 44/230) were identified as harboring BRCA1/2 mutation, and 28 (12.2%, 28/230) patients carried non-BRCA germline variants. Variants were detected in 16 non-BRCA genes, including PALB2 (5, 2.2%), ATM (4, 1.7%), RAD51D (3, 1.3%), TP53 (3, 1.3%), CHEK2 (2, 0.9%), FANCA (2, 0.9%) and ATR, BARD1, BRIP1, ERCC3, HOXB13, MLH1, MRE11, PMS2, RAD51C, RAD54L (1, 0.4%). Besides, 22 (9.6%, 22/230) patients were non-BRCA HRR gene mutation (including ATM, ATR, BARD1, BRIP1, CHEK2, FANCA, MRE11, PALB2, RAD51C RAD51D and RAD54L) carriers. Among high-risk factors, family history showed a correlation with both BRCA (p = 0.005) and non-BRCA HRR gene mutation status (p = 0.036). In addition, TNBC showed a correlation with BRCA1 gene mutation status (p = 0.038). However, other high-risk factors have not shown significantly related to BRCA1/2, non-BRCA genes and non-BRCA HRR gene mutations (p > 0.05). In addition, 312 unique variants of uncertain significance (VUS) were identified among 175 (76.1%, 175/230) patients and 65 different genes.Conclusions: Non-BRCA gene mutations are frequently identified in breast cancer patients with high risk factors. Family history showed a correlation with both BRCA (p = 0.005) and non-BRCA HRR gene mutation status (p = 0.036), so we strongly suggest that breast cancer patients with a BRCA-related family history receive comprehensive gene mutation testing in China, especially HRR genes, which are not only related to high risk of breast cancer, but also potentially related to poly ADP ribose polymerase inhibitor (PARPi) targeted therapy. The exact relationship of rare gene mutations to breast cancer predisposition and the pathogenicity of VUS need to be further investigated.

Objective: To investigate whether machine learning analysis of multiparametric MR radiomics can help classify immunohistochemical (IHC) subtypes of breast cancer.Study design: One hundred and thirty-four consecutive patients with pathologically-proven invasive ductal carcinoma were retrospectively analyzed. A total of 2,498 features were extracted from the DCE and DWI images, together with the new calculated images, including DCE images changing over six time points (DCEsequential) and DWI images changing over three b-values (DWIsequential). We proposed a novel two-stage feature selection method combining traditional statistics and machine learning-based methods. The accuracies of the 4-IHC classification and triple negative (TN) vs. non-TN cancers were assessed.Results: For the 4-IHC classification task, the best accuracy of 72.4% was achieved based on linear discriminant analysis (LDA) or subspace discrimination of assembled learning in conjunction with 20 selected features, and only small dependent emphasis of Kendall-tau-b for sequential features, based on the DWIsequential with the LDA model, yielding an accuracy of 53.7%. The linear support vector machine (SVM) and medium k-nearest neighbor using eight features yielded the highest accuracy of 91.0% for comparing TN to non-TN cancers, and the maximum variance for DWIsequential alone, together with a linear SVM model, achieved an accuracy of 83.6%.Conclusions: Whole-tumor radiomics on MR multiparametric images, DCE images changing over time points, and DWI images changing over different b-values provide a non-invasive analytical approach for breast cancer subtype classification and TN cancer identification.