Identification of BRCA1/2 mutation female carriers using circulating microRNA profiles

Abstract

Identifying germline BRCA1/2 mutation carriers is vital for reducing their risk of breast and ovarian cancer. To derive a serum miRNA-based diagnostic test we used samples from 653 healthy women from six international cohorts, including 350 (53.6%) with BRCA1/2 mutations and 303 (46.4%) BRCA1/2 wild-type. All individuals were cancer-free before and at least 12 months after sampling. RNA-sequencing followed by differential expression analysis identified 19 miRNAs significantly associated with BRCA mutations, 10 of which were ultimately used for classification: hsa-miR-20b-5p, hsa-miR-19b-3p, hsa-let-7b-5p, hsa-miR-320b, hsa-miR-139-3p, hsa-miR-30d-5p, hsa-miR-17-5p, hsa-miR-182-5p, hsa-miR-421, hsa-miR-375-3p. The final logistic regression model achieved area under the receiver operating characteristic curve 0.89 (95% CI: 0.87-0.93), 93.88% sensitivity and 80.72% specificity in an independent validation cohort. Mutated gene, menopausal status or having preemptive oophorectomy did not affect classification performance. Circulating microRNAs may be used to identify BRCA1/2 mutations in patients of high risk of cancer, offering an opportunity to reduce screening costs.

Fig. 1. miRNA expression data from healthy subjects with known germline BRCA1/2 mutation status.

a Overview of datasets. BWH Brigham and Women’s Hospital, CCGP Center for Cancer Genetics and Prevention at DFCI, DGO Department of Gynaecological Oncology, Tata Medical Center, Kolkata, India, IHCC International Hereditary Cancer Center of the Pomeranian Medical University, Poland, DFCI DFCI/BWH biobank, UPenn University of Pennsylvania. b Scheme of statistical analysis design.

Fig. 2. Development of the BRCA-mt signature.

UMAP representation of BRCA-mt and BRCA-wt samples from all evaluated cohorts without (a) and after (b) batch adjustment (N = 653). Volcano plots showing differentially expressed miRNAs between BRCA-mutated and wild-type samples without (c) and after (d) batch adjustment (N = 521); red markings represent miRNAs with P < 0.01 and FC > 1.5 or <0.66; purple markings denote ones that were significant in both comparisons. Limma package was used for between-group miRNA expression comparison, presented unadjusted P values, and FCs were calculated by limma algorithm. e Heatmap of 19 miRNAs with convergent BRCA-mt; BRCA-wt profiles regardless of data preprocessing. Clustering primarily by BRCA status with no clear pattern of BRCA1 or BRCA2 or interference by prior oophorectomy (N = 521). Euclidean distance and complete linkage were used to determine cluster structure.

Fig. 3. Performance and BRCA1/2 mutation probability estimated by the final logistic regression model.

a Training (sample size N = 391) ROC curve of the final logistic regression model; the area under the curve equaled 0.89 (95% CI: 0.87–0.93); b predicted BRCA1/2 mutation probability in training, testing, and validation sets according to reference mutational status; c predicted BRCA1/2 mutation probability in the context of menopausal status; d predicted BRCA1/2 mutation probability in the context of having ovaries at the time of testing. c, d show that both menopausal status and lack of ovaries did not influence predicted BRCA1 or BRCA2 mutation probability. In boxplots median is marked as central line, boxes indicate the first and third quartile and whiskers present 1.5× IQR. b–d present all N = 653 samples, statistics were derived using all samples in respective subgroups.