The landscape of BRCA1 and BRCA2 large rearrangements in an international cohort of over 20 000 ovarian tumors identified using next-generation sequencing

BACKGROUND

Approximately half of ovarian tumors have defects within the homologous recombination repair pathway. Tumors carrying pathogenic variants (PVs) in BRCA1/BRCA2 are more likely to respond to poly-ADP ribose polymerase (PARP) inhibitor treatment. Large rearrangements (LRs) are a challenging class of variants to identify and characterize in tumor specimens and may therefore be underreported. This study describes the prevalence of pathogenic BRCA1/BRCA2 LRs in ovarian tumors and discusses the importance of their identification using a comprehensive testing strategy.

METHODS

Sequencing and LR analyses of BRCA1/BRCA2 were conducted in 20 692 ovarian tumors received between March 18, 2016 and February 14, 2023 for MyChoice CDx testing. MyChoice CDx uses NGS dosage analysis to detect LRs in BRCA1/BRCA2 genes using dense tiling throughout the coding regions and limited flanking regions.

RESULTS

Of the 2217 PVs detected, 6.3% (N = 140) were LRs. Overall, 0.67% of tumors analyzed carried a pathogenic LR. The majority of detected LRs were deletions (89.3%), followed by complex LRs (5.7%), duplications (4.3%), and retroelement insertions (0.7%). Notably, 25% of detected LRs encompassed a single or partial single exon. This study identified 84 unique LRs, 2 samples each carried 2 unique LRs in the same gene. We identified 17 LRs that occurred in multiple samples, some of which were specific to certain ancestries. Several cases presented here illustrate the intricacies involved in characterizing LRs, particularly when multiple events occur within the same gene.

CONCLUSIONS

Over 6% of PVs detected in the ovarian tumors analyzed were LRs. It is imperative for laboratories to utilize testing methodologies that will accurately detect LRs at a single exon resolution to optimize the identification of patients who may benefit from PARP inhibitor treatment.

Homologous Recombination Deficiency as an Ovarian Cancer Biomarker in a Real-World Cohort: Validation of Decentralized Genomic Profiling

The diagnostic evaluation of homologous recombination deficiency (HRD) is central to define targeted therapy strategies for patients with ovarian carcinoma. We evaluated HRD in 514 ovarian carcinoma samples by next-generation sequencing of DNA libraries, including BRCA1/BRCA2 and 26,523 single-nucleotide polymorphisms using the standardized Myriad HRD assay, with the predefined cut point of ≥42 for a positive genomic instability score (GIS). All samples were measured in the central Myriad laboratory and in an academic molecular pathology laboratory. A positive GIS was detected in 196 (38.1%) of tumors, whereas 318 (61.9%) were GIS negative. Combining GIS and BRCA mutations, a total of 200 (38.9%) of the 514 tumors were HRD positive. A positive GIS was significantly associated with high-grade serous histology (P < 0.000001), grade 3 tumors (P = 0.001), and patient age <60 years (P = 0.0003). The concordance between both laboratories for the GIS status was 96.9% (P < 0.000001), with a sensitivity of 94.6% and a specificity of 98.4%. Concordance for HRD status was 97.1% (499 of 514 tumors). The percentage of HRD-positive tumors in our real-life cohort was similar to the proportion observed in the recently published PAOLA-1 trial, with high concordance between central and local laboratories. Our results support introduction of the standardized HRD assay in academic molecular pathology laboratories, thus broadening access to personalized oncology strategies for patients with ovarian cancer worldwide.

Abstract P5-10-02: Development and validation of a polygenic score to predict breast cancer risk in unaffected Hispanic women negative for mutations on a multi-gene hereditary cancer panel

Background: Breast cancer (BC) is the most commonly diagnosed cancer and the leading cause of cancer-related death among Hispanic women in the United States. For women of European ancestry, genome-wide association studies (GWAS) have identified common variants, primarily single-nucleotide polymorphisms (SNPs), that individually confer modest risk but together explain a significant proportion of genetic BC predisposition. For Hispanic women, the genetic contribution of SNPs to BC risk is not well understood. In these studies, we aim to develop and validate a polygenic score to improve risk assessment for Hispanic women who test negative for mutations in known BC susceptibility genes.

Methods: Genotypes and clinical histories were collected from consecutive development and validation cohorts of patients referred for hereditary cancer testing. Study subjects include women who report strictly Hispanic or Latin American ancestry, and who test negative for mutations in 11 genes associated with breast cancer (BRCA1, BRCA2, TP53, PTEN, STK11, CDH1, PALB2, CHEK2, ATM, NBN, BARD1).

Based on a development cohort ascertained through June 2017, we evaluated an 86-SNP Residual Risk Score (RRS) that was previously developed and validated for women of European ancestry. In the same cohort we are developing a Hispanic Residual Risk Score (HRRS) optimized for women of Hispanic ancestry. BC associations of individual SNPs are being established through meta-analysis of the development cohort and published Hispanic studies.

Multivariate logistic regression models were used to evaluate the 86-SNP RRS, and were the primary statistical tool for evaluation of individual SNPs and candidate polygenic scores. All models included personal/family cancer history and age as independent variables. P-values are based on likelihood ratio test statistics, and reported as two-sided. The development and validation studies are being conducted according to a protocol approved by the Quorum Institutional Review Board.

Results: The development cohort included 5,454 Hispanic women, 24% of whom reported a personal history of BC. The 86-SNP RRS was significantly associated with a personal history of BC after accounting for personal and family cancer history (p&lt;10-19) with odds ratio per unit standard deviation 1.39 (95% CI = 1.30-1.50). To date, more than 5,000 Hispanic women have been ascertained for inclusion in the validation cohort. Results comparing discriminatory accuracy of the RRS and the HRRS will be presented.

Conclusions: The implementation of a clinically validated polygenic score may improve risk assessment and medical management of Hispanic women who test negative for monogenic BC mutations. The HRRS will be validated in an independent study population according to a pre-specified statistical analysis plan.

Citation Format: Hughes ER, Wagner S, Pruss D, Gallagher SK, Swedlund B, Bulka K, Hoff R, Jammulapati S, Morris B, Perry T, Lanchbury JS, Gutin A. Development and validation of a polygenic score to predict breast cancer risk in unaffected Hispanic women negative for mutations on a multi-gene hereditary cancer panel [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-10-02.

A candidate prostate cancer susceptibility gene at chromosome 17p

It is difficult to identify genes that predispose to prostate cancer due to late age at diagnosis, presence of phenocopies within high-risk pedigrees and genetic complexity. A genome-wide scan of large, high-risk pedigrees from Utah has provided evidence for linkage to a locus on chromosome 17p. We carried out positional cloning and mutation screening within the refined interval, identifying a gene, ELAC2, harboring mutations (including a frameshift and a nonconservative missense change) that segregate with prostate cancer in two pedigrees. In addition, two common missense variants in the gene are associated with the occurrence of prostate cancer. ELAC2 is a member of an uncharacterized gene family predicted to encode a metal-dependent hydrolase domain that is conserved among eukaryotes, archaebacteria and eubacteria. The gene product bears amino acid sequence similarity to two better understood protein families, namely the PSO2 (SNM1) DNA interstrand crosslink repair proteins and the 73-kD subunit of mRNA 3' end cleavage and polyadenylation specificity factor (CPSF73).

BRG1, a component of the SWI-SNF complex, is mutated in multiple human tumor cell lines

Human BRG1 is a component of the evolutionarily conserved SWI-SNF chromatin remodeling complex. BRG1 has been implicated in growth control through its interaction with the tumor suppressor pRb and may consequently serve as a negative regulator of proliferation. Postulating that BRG1 may itself be a tumor suppressor gene, we screened a panel of tumor cell lines to determine whether the gene is targeted for mutation. We report that the COOH-terminal region of BRG1 is homozygously deleted in two carcinoma cell lines, prostate TSU-Pr1 and lung A-427. In addition, biallelic inactivations of BRG1 were observed in four other cell lines derived from carcinomas of the breast, lung, pancreas, and prostate; their mutations in BRG1 included three frameshift lesions and one nonsense lesion. Point mutations were also discovered in a number of other cell lines, however in most cases any effect of these mutations on BRG1 function remains to be established. A variety of different mutations within BRG1, in several cell lines, suggest that BRG1 may be targeted for disruption in human tumors. Significantly, reintroduction of BRG1 into cells lacking BRG1 expression was sufficient to reverse their transformed phenotype inducing growth arrest and a flattened morphology. These data strongly support the model that BRG1 may function as a tumor suppressor and strengthen the hypothesis that the regulation of gene expression through chromatin remodeling is critical for cancer progression. It will be important to confirm these observations in primary tumors.

MMAC1/PTEN mutations in primary tumor specimens and tumor cell lines

A candidate tumor suppressor gene, MMAC1/PTEN, located in human chromosome band 10q23, was recently identified based on sequence alterations observed in several glioma, breast, prostate, and kidney tumor specimens or cell lines. To further investigate the mutational profile of this gene in human cancers, we examined a large set of human tumor specimens and cancer cell lines of many types for 10q23 allelic losses and MMAC1 sequence alterations. Loss of heterozygosity (LOH) at the MMAC1 locus was observed in approximately one-half of the samples examined, consistent with the high frequency of 10q allelic loss reported for many cancers. Of 124 tumor specimens exhibiting LOH that have been screened for MMAC1 alterations to date, we have detected variants in 13 (approximately 10%) of these primary tumors; the highest frequency of variants was found in glioblastoma specimens (approximately 23%). Novel alterations identified in this gene include a missense variant in a melanoma sample and a splicing variant and a nonsense mutation in pediatric glioblastomas. Of 76 tumor cell lines prescreened for probable LOH, microsequence alterations of MMAC1 were detected in 12 (approximately 16%) of the lines, including those derived from astrocytoma, leukemia, and melanoma tumors, as well as bladder, breast, lung, prostate, submaxillary gland, and testis carcinomas. In addition, in this set of tumor cell lines, we detected 11 (approximately 14%) homozygous deletions that eliminated coding portions of MMAC1, a class of abnormality not detected by our methods in primary tumors. These data support the occurrence of inactivating MMAC1 alterations in multiple human cancer types. In addition, we report the discovery of a putative pseudogene of MMAC1 localized on chromosome 9.

Human mitogen-activated protein kinase kinase 4 as a candidate tumor suppressor

Mitogen-activated protein kinases function in signal transduction pathways that are involved in controlling key cellular processes in many organisms. A mammalian member of this kinase family, MKK4/JNKK1/SEK1, has been reported to link upstream MEKK1 to downstream stress-activated protein kinase/JNK1 and p38 mitogen-activated protein kinase. This mitogen-activated protein kinase pathway has been implicated in the signal transduction of cytokine- and stress-induced apoptosis in a variety of cell types. Here, we report that two human tumor cell lines, derived from pancreatic carcinoma and lung carcinoma, harbor homozygous deletions that eliminate coding portions of the MKK4 locus at 17p, located approximately 10 cM centromeric of p53. In addition, in a set of 88 human cancer cell lines prescreened for loss of heterozygosity, we detected two nonsense and three missense sequence variants of MKK4 in cancer cell lines derived from human pancreatic, breast, colon, and testis cells. In vitro biochemical assays revealed that, when stimulated by MEKK1, four of the five altered MKK4 proteins lacked the ability to phosphorylate stress-activated protein kinase. Thus, the incidence of coding mutations of MKK4 in the set of cell lines is 6 of 213 (approximately 3%). These findings suggest that MKK4 may function as a suppressor of tumorigenesis or metastasis in certain types of cells.

Identification of a candidate tumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers

Deletions involving regions of chromosome 10 occur in the vast majority (> 90%) of human glioblastoma multiformes. A region at chromosome 10q23-24 was implicated to contain a tumour suppressor gene and the identification of homozygous deletions in four glioma cell lines further refined the location. We have identified a gene, designated MMAC1, that spans these deletions and encodes a widely expressed 5.5-kb mRNA. The predicted MMAC1 protein contains sequence motifs with significant homology to the catalytic domain of protein phosphatases and to the cytoskeletal proteins, tensin and auxilin. MMAC1 coding-region mutations were observed in a number of glioma, prostate, kidney and breast carcinoma cell lines or tumour specimens. Our results identify a strong candidate tumour suppressor gene at chromosome 10q23.3, whose loss of function appears to be associated with the oncogenesis of multiple human cancers.

Low incidence of BRCA2 mutations in breast carcinoma and other cancers

Inherited mutant alleles of familial tumour suppressor genes predispose individuals to particular types of cancer. In addition to an involvement in inherited susceptibility to cancer, these tumour suppressor genes are targets for somatic mutations in sporadic cancers of the same type found in the familial forms. An exception is BRCA1, which contributes to a significant fraction of familial breast and ovarian cancer, but undergoes mutation at very low rates in sporadic breast and ovarian cancers. This finding suggests that other genes may be the principal targets for somatic mutation in breast carcinoma. A second, recently identified familial breast cancer gene, BRCA2 (refs 5-8), accounts for a proportion of breast cancer roughly equal to BRCA1. Like BRCA1, BRCA2 behaves as a dominantly inherited tumour suppressor gene. Individuals who inherit one mutant allele are at increased risk for breast cancer, and the tumours they develop lose the wild-type allele by heterozygous deletion. The BRCA2 coding sequence is huge, composed of 26 exons that span 10,443 bp. Here we investigate the rate of BRCA2 mutation in sporadic breast cancers and in a set of cell lines that represent twelve other tumour types. Surprisingly, mutations in BRCA2 are infrequent in cancers including breast carcinoma. However, a probable germline mutation in a pancreatic tumour cell line suggests a role for BRCA2 in susceptibility to pancreatic cancer.

The complete BRCA2 gene and mutations in chromosome 13q-linked kindreds

Breast carcinoma is the most common malignancy among women in developed countries. Because family history remains the strongest single predictor of breast cancer risk, attention has focused on the role of highly penetrant, dominantly inherited genes in cancer-prone kindreds (1). BRCA1 was localized to chromosome 17 through analysis of a set of high-risk kindreds (2), and then identified four years later by a positional cloning strategy (3). BRCA2 was mapped to chromosomal 13q at about the same time (4). Just fifteen months later, Wooster et al. (5) reported a partial BRCA2 sequence and six mutations predicted to cause truncation of the BRCA2 protein. While these findings provide strong evidence that the identified gene corresponds to BRCA2, only two thirds of the coding sequence and 8 out of 27 exons were isolated and screened; consequently, several questions remained unanswered regarding the nature of BRCA2 and the frequency of mutations in 13q-linked families. We have now determined the complete coding sequence and exonic structure of BRCA2 (GenBank accession #U43746), and examined its pattern of expression. Here, we provide sequences for a set of PCR primers sufficient to screen the entire coding sequence of BRCA2 using genomic DNA. We also report a mutational analysis of BRCA2 in families selected on the basis of linkage analysis and/or the presence of one or more cases of male breast cancer. Together with the specific mutations described previously, our data provide preliminary insight into the BRCA2 mutation profile.

Sorbitol as the Primary Carbon Source for the Growth of Embryogenic Callus of Maize

The effects of various carbon sources on initiation and maintenance of embryogenic callus of maize (Zea mays L.) and on the regeneration of plants from embryogenic callus were studied. Growth of embryogenic callus tissue on media containing sucrose was typified by the subsequent growth of both embryogenic (regenerable) and nonembryogenic (nonregenerable) callus. Growth of embryogenic callus on sorbitol was unique among the carbon sources tested in that sorbitol supported the subsequent growth of only embryogenic callus. Further experiments demonstrated that embryogenic callus grown on sorbitol had a greater regenerative capacity (more plants produced per gram fresh weight of callus) than callus grown on sucrose. Sorbitol dehydrogenase was detected in embryogenic callus of maize at a specific activity roughly equivalent to that found in zygotic embryos of developing seeds. Nonembryogenic callus did not contain significant levels of sorbitol dehydrogenase activity.

Somatic embryogenesis and plant regeneration in two-year old cultures of Zea diploperennis

Immature embryos and immature leaf tissues were used to establish embryogenic cultures of Zea diploperennis. Callus was induced on media containing MS salts and vitamins, sucrose (2% for leaves, 6% for embryos), 5% coconut milk and 1-6 mg/l 2, 4-D. Embryogenic callus was maintained by subculturing on media containing MS salts and vitamins, 2% sucrose, 500 mg/l casein hydrolysate and 1 mg/l 2,4-D. Regeneration occurred when the 2,4-D level was reduced to 0.25 mg/l. Kinetin added at 0.25 mg/l further stimulated regeneration. Root tip squashes on 10 plants regenerated after 2 years in culture indicated a normal 2n=20 chromosome number.

Cytogenetic characterization of embryogenic callus and regenerated plants of Pennisetum americanum (L.) K. Schum

Embryogenic calli were derived from cultured segments of immature inflorescences of Pennisetum americanum (pearl millet). The original explants as well as the embryogenic calli and the plants regenerated via somatic embryogenesis were examined cytogenetically. Embryogenic calli were predominantly diploid (2n=14) after one month and six months in culture (92% and 76%, respectively). Tetraploid and aneuploid cells were observed in the original explant (2.5% and 1.2%) as well as in one (4.0% and 4.0%) and six-month-old calli (10.0% and 14.0%). Plants were regenerated from calli that had been in continuous culture for two, four and six months. Of the 101 regenerants, 100 were diploid and 1 was tetraploid. The tetraploid was an albino as were three of the diploid regenerants. Examination of 30 of the regenerants in meiotic diakinesis, anaphase I, anaphase II and quartet stages revealed no cytogenetic differences between control and regenerated plants. Gel electrophoresis for total protein content and alcohol dehydrogenase and malate dehydrogenase activity also did not reveal any differences between the controls and regenerants. The results of this study show that a slight shift toward aneuploidy and polyploidy may occur in embryogenic cultures, but there also is a strong selection in favor of plant regeneration from cytogenetically normal cells.