Mosaic BRCA1 promoter methylation contribution in hereditary breast/ovarian cancer pedigrees.

PURPOSE: Mosaic BRCA1 promoter methylation (BRCA1meth) increases the risk of early-onset breast cancer, triple-negative breast cancer and ovarian cancer. As mosaic BRCA1meth are believed to occur de novo, their role in family breast/ovarian cancer has not been assessed. PATIENTS: Blood-derived DNA from 20 unrelated affected cases from families with aggregation of breast/ovarian cancer, but with no germline pathogenic variants in BRCA1/2, PALB2 or RAD51C/D, were screened by methylation-sensitive high-resolution melting. CpG analysis was performed by pyrosequencing on blood and buccal swab. Two probands carried a pathogenic variant in a moderate-penetrance gene (ATM and BARD1), and 8 of 18 others (44%) carried BRCA1meth (vs none of the 20 age-matched controls). Involvement of BRCA1 in tumourigenesis in methylated probands was demonstrated in most tested cases by detection of a loss of heterozygosity and a homologous recombination deficiency signature. Among the eight methylated probands, two had relatives with breast cancer with detectable BRCA1meth in blood, including one with high methylation levels in two non-tumour tissues. CONCLUSIONS: The high prevalence of mosaic BRCA1meth in patients with breast/ovarian cancer with affected relatives, as well as this first description of a family aggregation of mosaic BRCA1meth, shows how this de novo event can contribute to hereditary breast/ovarian cancer pedigrees.

Male breast cancer: No evidence for mosaic BRCA1 promoter methylation involvement.

Breast cancers (BC) are rare in men and are often caused by constitutional predisposing factors. In women, mosaic BRCA1 promoter methylations (MBPM) are frequent events, detected in 4-8% of healthy subjects. This constitutional epimutation increases risk of early-onset and triple-negative BC. However, the role of MBPM in male BC predisposition has never been assessed. We screened 40 blood samples from men affected by BC, and performed extensive tumour analysis on MBPM-positive patients. We detected two patients carrying MBPM. Surprisingly, tumour analysis revealed that neither of these two male BCs were caused by the constitutional BRCA1 epimutations carried by the patients.

The BRCA2 R2645G variant increases DNA binding and induces hyper-recombination.

BRCA2 tumor suppressor protein ensures genome integrity by mediating DNA repair via homologous recombination (HR). This function is executed in part by its canonical DNA binding domain located at the C-terminus (BRCA2CTD), the only folded domain of the protein. Most germline pathogenic missense variants are located in this highly conserved region which binds to single-stranded DNA (ssDNA) and to the acidic protein DSS1. These interactions are essential for the HR function of BRCA2. Here, we report that the variant R2645G, identified in breast cancer and located at the DSS1 interface, unexpectedly increases the ssDNA binding activity of BRCA2CTDin vitro. Human cells expressing this variant display a hyper-recombination phenotype, chromosomal instability in the form of chromatid gaps when exposed to DNA damage, and increased PARP inhibitor sensitivity. In mouse embryonic stem cells (mES), this variant alters viability and confers sensitivity to cisplatin and Mitomycin C. These results suggest that BRCA2 interaction with ssDNA needs to be tightly regulated to limit HR and prevent chromosomal instability and we propose that this control mechanism involves DSS1. Given that several missense variants located within this region have been identified in breast cancer patients, these findings might have clinical implications for carriers.

Germline HPF1 retrogene insertion in RB1 gene involved in cancer predisposition.

About half of the human genome is composed of repeated sequences derived from mobile elements, mainly retrotransposons, generally without pathogenic effect. Familial forms of retinoblastoma are caused by germline pathogenic variants in RB1 gene. Here, we describe a family with retinoblastoma affecting a father and his son. No pathogenic variant was identified after DNA analysis of RB1 gene coding sequence and exon-intron junctions. However, RB1 mRNA analysis showed a chimeric transcript with insertion of 114 nucleotides from HPF1 gene inside RB1 gene. This chimeric transcript led to an insertion of 38 amino acids in functional domain of retinoblastoma protein. Subsequent DNA analysis in RB1 intron 17 revealed the presence of a full-length HPF1 retrogene insertion in opposite orientation. Functional assay shows that this insertion has a deleterious impact on retinoblastoma protein function. This is the first report of a full-length retrogene insertion involved in human Mendelian disease leading to a chimeric transcript and a non-functional chimeric protein. Some retrogene insertions may be missed by standard diagnostic genetic testing, so contribution of retrogene insertions to human disease may be underestimated. The increasing use of whole genome sequencing in diagnostic settings will help to get a more comprehensive view of retrogenes.

Combined Tumor-Based BRCA1/2 and TP53 Mutation Testing in Ovarian Cancer.

Somatic/germline BRCA1/2 mutations (m)/(likely) pathogenic variants (PV) (s/gBRCAm) remain the best predictive biomarker for PARP inhibitor efficacy. As >95% of high-grade serous ovarian cancers (HGSOC) have a somatic TP53m, combined tumor-based BRCA1/2 (tBRCA) and TP53 mutation testing (tBRCA/TP53m) may improve the quality of results in somatic BRCAm identification and interpretation of the 'second hit' event, i.e., loss of heterozygosity (LOH). A total of 237 patients with HGSOC underwent tBRCA/TP53m testing. The ratio of allelic fractions (AFs) for tBRCA/TP53m was calculated to estimate the proportion of cells carrying BRCAm and to infer LOH. Among the 142/237 gBRCA results, 16.2% demonstrated a pathogenic/deleterious variant (DEL) gBRCA1/2m. Among the 195 contributive tumor samples, 43 DEL of tBRCAm (22.1%) were identified (23 gBRCAm and 20 sBRCAm) with LOH identified in 37/41 conclusive samples. The median AF of TP53m was 0.52 (0.01-0.93), confirming huge variability in tumor cellularity. Initially, three samples were considered as wild type with <10% cellularity. However, additional testing detected a very low AF (<0.05) in both BRCA1/2m and TP53m, thus reidentifying them as sBRCA1/2m. Combined tBRCA/TP53m testing is rapid, sensitive, and identifies somatic and germline BRCA1/2m. AF TP53m is essential for interpreting sBRCA1/2m in low-cellularity samples and provides indirect evidence for LOH as the 'second hit' of BRCA1/2-related tumorigenesis.

Adaptive nanopore sequencing to determine pathogenicity of BRCA1 exonic duplication.

BRCA1 and BRCA2 are tumour suppressor genes that have been characterised as predisposition genes for the development of hereditary breast and ovarian cancers among other malignancies. The molecular diagnosis of this predisposition syndrome is based on the detection of inactivating variants of any type in those genes. But in the case of structural variants, functional consequences can be difficult to assess using standard molecular methods, as the precise resolution of their sequence is often impossible with short-read next generation sequencing techniques. It has been recently demonstrated that Oxford Nanopore long-read sequencing technology can accurately and rapidly provide genetic diagnoses of Mendelian diseases, including those linked to pathogenic structural variants. Here, we report the accurate resolution of a germline duplication event of exons 18-20 of BRCA1 using Nanopore sequencing with adaptive sampling target enrichment. This allowed us to classify this variant as pathogenic within a short timeframe of 10 days. This study provides a proof-of-concept that nanopore adaptive sampling is a highly efficient technique for the investigation of structural variants of tumour suppressor genes in a clinical context.

WWOX binds MERIT40 and modulates its function in homologous recombination, implications in breast cancer.

The tumor suppressor gene WWOX is localized in an unstable chromosomal region and its expression is decreased or absent in several types of cancer. A low expression of WWOX is associated with a poor prognosis in breast cancer (BC). It has recently been shown that WWOX contributes to genome stability through its role in the DNA damage response (DDR). In breast cancer cells, WWOX inhibits homologous recombination (HR), and thus promotes the repair of DNA double-stranded breaks (DSBs) by non-homologous end joining (NHEJ). The fine-tuning modulation of HR activity is crucial. Its under or overstimulation inducing genome alterations that can induce cancer. MERIT40 is a positive regulator of the DDR. This protein is indispensable for the function of the multi-protein complex BRCA1-A, which suppresses excessive HR activity. MERIT40 also recruits Tankyrase, a positive regulator of HR, to the DSBs to stimulate DNA repair. Here, we identified MERIT40 as a new molecular partner of WWOX. We demonstrated that WWOX inhibited excessive HR activity induced by overexpression of MERIT40. We showed that WWOX impaired the MERIT40-Tankyrase interaction preventing the role of the complex on DSBs. Furthermore, we found that MERIT40 is overexpressed in BC and that this overexpression is associated to a poor prognosis. These results strongly suggest that WWOX, through its interaction with MERIT40, prevents the deleterious impact of excessive HR on BC development by inhibiting MERIT40-Tankyrase association. This inhibitory effect of WWOX would oppose MERIT40-dependent BC development.

Identification of a large intra-exonic deletion in BRCA2 exon 18 in a pancreatic ductal adenocarcinoma.

By 2030, pancreatic cancer will become the second leading cause of cancer-related deaths in the United States and in Europe. The management of patients with advanced pancreatic cancer relies on chemotherapy and poly (ADP-ribose) polymerase inhibitors for patients who carry BRCA1/2 inactivating alterations. Some variants, such as large insertion/deletions (Indels), inactivating BRCA1/2 and therefore of clinical relevance can be hard to detect by next-generation sequencing techniques. Here we report a 47-year-old patient presenting with pancreatic cancer whose tumour harbours a large somatic intra-exonic deletion of BRCA2 of 141 bp. This BRCA2 deletion, located in the C-terminal domain, can be considered as pathogenic and consequently affect tumorigenesis because it is involved in the interaction between the DSS1 protein and DNA. Thanks to the optimized bioinformatics algorithm, this intermediate size deletion in BRCA2 was identified, enabling personalized patient management via the inclusion of the patients in a clinical trial.

Association and performance of polygenic risk scores for breast cancer among French women presenting or not a familial predisposition to the disease.

BACKGROUND: Three partially overlapping breast cancer polygenic risk scores (PRS) comprising 77, 179 and 313 SNPs have been proposed for European-ancestry women by the Breast Cancer Association Consortium (BCAC) for improving risk prediction in the general population. However, the effect of these SNPs may vary from one country to another and within a country because of other factors. OBJECTIVE: To assess their associated risk and predictive performance in French women from (1) the CECILE population-based case-control study, (2) BRCA1 or BRCA2 (BRCA1/2) pathogenic variant (PV) carriers from the GEMO study, and (3) familial breast cancer cases with no BRCA1/2 PV and unrelated controls from the GENESIS study. RESULTS: All three PRS were associated with breast cancer in all studies, with odds ratios per standard deviation varying from 1.7 to 2.0 in CECILE and GENESIS, and hazard ratios varying from 1.1 to 1.4 in GEMO. The predictive performance of PRS313 in CECILE was similar to that reported in BCAC but lower than that in GENESIS (area under the receiver operating characteristic curve (AUC) = 0.67 and 0.75, respectively). PRS were less performant in BRCA2 and BRCA1 PV carriers (AUC = 0.58 and 0.54 respectively). CONCLUSION: Our results are in line with previous validation studies in the general population and in BRCA1/2 PV carriers. Additionally, we showed that PRS may be of clinical utility for women with a strong family history of breast cancer and no BRCA1/2 PV, and for those carrying a predicted PV in a moderate-risk gene like ATM, CHEK2 or PALB2.

A likelihood ratio approach for utilizing case-control data in the clinical classification of rare sequence variants: application to BRCA1 and BRCA2.

A large number of variants identified through clinical genetic testing in disease susceptibility genes, are of uncertain significance (VUS). Following the recommendations of the American College of Medical Genetics and Genomics (ACMG) and Association for Molecular Pathology (AMP), the frequency in case-control datasets (PS4 criterion), can inform their interpretation. We present a novel case-control likelihood ratio-based method that incorporates gene-specific age-related penetrance. We demonstrate the utility of this method in the analysis of simulated and real datasets. In the analyses of simulated data, the likelihood ratio method was more powerful compared to other methods. Likelihood ratios were calculated for a case-control dataset of BRCA1 and BRCA2 variants from the Breast Cancer Association Consortium (BCAC), and compared with logistic regression results. A larger number of variants reached evidence in favor of pathogenicity, and a substantial number of variants had evidence against pathogenicity - findings that would not have been reached using other case-control analysis methods. Our novel method provides greater power to classify rare variants compared to classical case-control methods. As an initiative from the ENIGMA Analytical Working Group, we provide user-friendly scripts and pre-formatted excel calculators for implementation of the method for rare variants in BRCA1, BRCA2 and other high-risk genes with known penetrance.

SPiP: Splicing Prediction Pipeline, a machine learning tool for massive detection of exonic and intronic variant effects on mRNA splicing.

Modeling splicing is essential for tackling the challenge of variant interpretation as each nucleotide variation can be pathogenic by affecting pre-mRNA splicing via disruption/creation of splicing motifs such as 5'/3' splice sites, branch sites, or splicing regulatory elements. Unfortunately, most in silico tools focus on a specific type of splicing motif, which is why we developed the Splicing Prediction Pipeline (SPiP) to perform, in one single bioinformatic analysis based on a machine learning approach, a comprehensive assessment of the variant effect on different splicing motifs. We gathered a curated set of 4616 variants scattered all along the sequence of 227 genes, with their corresponding splicing studies. The Bayesian analysis provided us with the number of control variants, that is, variants without impact on splicing, to mimic the deluge of variants from high-throughput sequencing data. Results show that SPiP can deal with the diversity of splicing alterations, with 83.13% sensitivity and 99% specificity to detect spliceogenic variants. Overall performance as measured by area under the receiving operator curve was 0.986, better than SpliceAI and SQUIRLS (0.965 and 0.766) for the same data set. SPiP lends itself to a unique suite for comprehensive prediction of spliceogenicity in the genomic medicine era. SPiP is available at: https://sourceforge.net/projects/splicing-prediction-pipeline/.

Value of the loss of heterozygosity to BRCA1 variant classification.

At least 10% of the BRCA1/2 tests identify variants of uncertain significance (VUS) while the distinction between pathogenic variants (PV) and benign variants (BV) remains particularly challenging. As a typical tumor suppressor gene, the inactivation of the second wild-type (WT) BRCA1 allele is expected to trigger cancer initiation. Loss of heterozygosity (LOH) of the WT allele is the most frequent mechanism for the BRCA1 biallelic inactivation. To evaluate if LOH can be an effective predictor of BRCA1 variant pathogenicity, we carried out LOH analysis on DNA extracted from 90 breast and seven ovary tumors diagnosed in 27 benign and 55 pathogenic variant carriers. Further analyses were conducted in tumors with PVs yet without loss of the WT allele: BRCA1 promoter hypermethylation, next-generation sequencing (NGS) of BRCA1/2, and BRCAness score. Ninety-seven tumor samples were analyzed from 26 different BRCA1 variants. A relatively stable pattern of LOH (65.4%) of WT allele for PV tumors was observed, while the allelic balance (63%) or loss of variant allele (15%) was generally seen for carriers of BV. LOH data is a useful complementary argument for BRCA1 variant classification.

Cancer Risks Associated With BRCA1 and BRCA2 Pathogenic Variants.

PURPOSE: To provide precise age-specific risk estimates of cancers other than female breast and ovarian cancers associated with pathogenic variants (PVs) in BRCA1 and BRCA2 for effective cancer risk management. METHODS: We used data from 3,184 BRCA1 and 2,157 BRCA2 families in the Consortium of Investigators of Modifiers of BRCA1/2 to estimate age-specific relative (RR) and absolute risks for 22 first primary cancer types adjusting for family ascertainment. RESULTS: BRCA1 PVs were associated with risks of male breast (RR = 4.30; 95% CI, 1.09 to 16.96), pancreatic (RR = 2.36; 95% CI, 1.51 to 3.68), and stomach (RR = 2.17; 95% CI, 1.25 to 3.77) cancers. Associations with colorectal and gallbladder cancers were also suggested. BRCA2 PVs were associated with risks of male breast (RR = 44.0; 95% CI, 21.3 to 90.9), stomach (RR = 3.69; 95% CI, 2.40 to 5.67), pancreatic (RR = 3.34; 95% CI, 2.21 to 5.06), and prostate (RR = 2.22; 95% CI, 1.63 to 3.03) cancers. The stomach cancer RR was higher for females than males (6.89 v 2.76; P = .04). The absolute risks to age 80 years ranged from 0.4% for male breast cancer to approximately 2.5% for pancreatic cancer for BRCA1 carriers and from approximately 2.5% for pancreatic cancer to 27% for prostate cancer for BRCA2 carriers. CONCLUSION: In addition to female breast and ovarian cancers, BRCA1 and BRCA2 PVs are associated with increased risks of male breast, pancreatic, stomach, and prostate (only BRCA2 PVs) cancers, but not with the risks of other previously suggested cancers. The estimated age-specific risks will refine cancer risk management in men and women with BRCA1/2 PVs.

Novel germline MET pathogenic variants in French patients with papillary renal cell carcinomas type I.

Hereditary papillary renal cell carcinoma (HPRC) is a rare inherited renal cancer syndrome characterized by bilateral and multifocal papillary type 1 renal tumors (PRCC1). Activating germline pathogenic variants of the MET gene were identified in HPRC families. We reviewed the medical and molecular records of a large French series of 158 patients screened for MET oncogenic variants. MET pathogenic and likely pathogenic variants rate was 12.4% with 40.6% among patients with familial PRCC1 and 5% among patients with sporadic PRCC1. The phenotype in cases with MET pathogenic and likely pathogenic variants was characteristic: PRCC1 tumors were mainly bilateral (84.3%) and multifocal (87.5%). Histologically, six out of seven patients with MET pathogenic variant harbored biphasic squamoid alveolar PRCC. Genetic screening identified one novel pathogenic variant MET c.3389T>C, p.(Leu1130Ser) and three novel likely pathogenic variants: MET c.3257A>T, p.(His1086Leu); MET c.3305T>C, p.(Ile1102Thr) and MET c.3373T>G, p.(Cys1125Gly). Functional assay confirmed their oncogenic effect as they induced an abnormal focus formation. The genotype-phenotype correlation between MET pathogenic variants and PRCC1 presentation should encourage to widen the screening, especially toward nonfamilial PRCC1. This precise phenotype also constitutes a strong argument for the classification of novel missense variants within the tyrosine kinase domain when functional assays are not accessible.

Risks of breast and ovarian cancer for women harboring pathogenic missense variants in BRCA1 and BRCA2 compared with those harboring protein truncating variants.

PURPOSE: Germline genetic testing for BRCA1 and BRCA2 variants has been a part of clinical practice for >2 decades. However, no studies have compared the cancer risks associated with missense pathogenic variants (PVs) with those associated with protein truncating (PTC) variants. METHODS: We collected 582 informative pedigrees segregating 1 of 28 missense PVs in BRCA1 and 153 pedigrees segregating 1 of 12 missense PVs in BRCA2. We analyzed 324 pedigrees with PTC variants in BRCA1 and 214 pedigrees with PTC variants in BRCA2. Cancer risks were estimated using modified segregation analysis. RESULTS: Estimated breast cancer risks were markedly lower for women aged >50 years carrying BRCA1 missense PVs than for the women carrying BRCA1 PTC variants (hazard ratio [HR] = 3.9 [2.4-6.2] for PVs vs 12.8 [5.7-28.7] for PTC variants; P = .01), particularly for missense PVs in the BRCA1 C-terminal domain (HR = 2.8 [1.4-5.6]; P = .005). In case of BRCA2, for women aged >50 years, the HR was 3.9 (2.0-7.2) for those heterozygous for missense PVs compared with 7.0 (3.3-14.7) for those harboring PTC variants. BRCA1 p.[Cys64Arg] and BRCA2 p.[Trp2626Cys] were associated with particularly low risks of breast cancer compared with other PVs. CONCLUSION: These results have important implications for the counseling of at-risk women who harbor missense PVs in the BRCA1/2 genes.

Classification of 101 BRCA1 and BRCA2 variants of uncertain significance by cosegregation study: A powerful approach.

Up to 80% of BRCA1 and BRCA2 genetic variants remain of uncertain clinical significance (VUSs). Only variants classified as pathogenic or likely pathogenic can guide breast and ovarian cancer prevention measures and treatment by PARP inhibitors. We report the first results of the ongoing French national COVAR (cosegregation variant) study, the aim of which is to classify BRCA1/2 VUSs. The classification method was a multifactorial model combining different associations between VUSs and cancer, including cosegregation data. At this time, among the 653 variants selected, 101 (15%) distinct variants shared by 1,624 families were classified as pathogenic/likely pathogenic or benign/likely benign by the COVAR study. Sixty-six of the 101 (65%) variants classified by COVAR would have remained VUSs without cosegregation data. Of note, among the 34 variants classified as pathogenic by COVAR, 16 remained VUSs or likely pathogenic when following the ACMG/AMP variant classification guidelines. Although the initiation and organization of cosegregation analyses require a considerable effort, the growing number of available genetic tests results in an increasing number of families sharing a particular variant, and thereby increases the power of such analyses. Here we demonstrate that variant cosegregation analyses are a powerful tool for the classification of variants in the BRCA1/2 breast-ovarian cancer predisposition genes.

Intrinsic Disorder and Phosphorylation in BRCA2 Facilitate Tight Regulation of Multiple Conserved Binding Events.

The maintenance of genome integrity in the cell is an essential process for the accurate transmission of the genetic material. BRCA2 participates in this process at several levels, including DNA repair by homologous recombination, protection of stalled replication forks, and cell division. These activities are regulated and coordinated via cell-cycle dependent modifications. Pathogenic variants in BRCA2 cause genome instability and are associated with breast and/or ovarian cancers. BRCA2 is a very large protein of 3418 amino acids. Most well-characterized variants causing a strong predisposition to cancer are mutated in the C-terminal 700 residues DNA binding domain of BRCA2. The rest of the BRCA2 protein is predicted to be disordered. Interactions involving intrinsically disordered regions (IDRs) remain difficult to identify both using bioinformatics tools and performing experimental assays. However, the lack of well-structured binding sites provides unique functional opportunities for BRCA2 to bind to a large set of partners in a tightly regulated manner. We here summarize the predictive and experimental arguments that support the presence of disorder in BRCA2. We describe how BRCA2 IDRs mediate self-assembly and binding to partners during DNA double-strand break repair, mitosis, and meiosis. We highlight how phosphorylation by DNA repair and cell-cycle kinases regulate these interactions. We finally discuss the impact of cancer-associated variants on the function of BRCA2 IDRs and more generally on genome stability and cancer risk.

5' Region Large Genomic Rearrangements in the BRCA1 Gene in French Families: Identification of a Tandem Triplication and Nine Distinct Deletions with Five Recurrent Breakpoints.

BACKGROUND: Large genomic rearrangements (LGR) in BRCA1 consisting of deletions/duplications of one or several exons have been found throughout the gene with a large proportion occurring in the 5' region from the promoter to exon 2. The aim of this study was to better characterize those LGR in French high-risk breast/ovarian cancer families. METHODS: DNA from 20 families with one apparent duplication and nine deletions was analyzed with a dedicated comparative genomic hybridization (CGH) array, high-resolution BRCA1 Genomic Morse Codes analysis and Sanger sequencing. RESULTS: The apparent duplication was in fact a tandem triplication of exons 1 and 2 and part of intron 2 of BRCA1, fully characterized here for the first time. We calculated a causality score with the multifactorial model from data obtained from six families, classifying this variant as benign. Among the nine deletions detected in this region, eight have never been identified. The breakpoints fell in six recurrent regions and could confirm some specific conformation of the chromatin. CONCLUSIONS: Taken together, our results firmly establish that the BRCA1 5' region is a frequent site of different LGRs and highlight the importance of the segmental duplication and Alu sequences, particularly the very high homologous region, in the mechanism of a recombination event. This also confirmed that those events are not systematically deleterious.

A new hybrid record linkage process to make epidemiological databases interoperable: application to the GEMO and GENEPSO studies involving BRCA1 and BRCA2 mutation carriers.

BACKGROUND: Linking independent sources of data describing the same individuals enable innovative epidemiological and health studies but require a robust record linkage approach. We describe a hybrid record linkage process to link databases from two independent ongoing French national studies, GEMO (Genetic Modifiers of BRCA1 and BRCA2), which focuses on the identification of genetic factors modifying cancer risk of BRCA1 and BRCA2 mutation carriers, and GENEPSO (prospective cohort of BRCAx mutation carriers), which focuses on environmental and lifestyle risk factors. METHODS: To identify as many as possible of the individuals participating in the two studies but not registered by a shared identifier, we combined probabilistic record linkage (PRL) and supervised machine learning (ML). This approach (named "PRL + ML") combined together the candidate matches identified by both approaches. We built the ML model using the gold standard on a first version of the two databases as a training dataset. This gold standard was obtained from PRL-derived matches verified by an exhaustive manual review. Results The Random Forest (RF) algorithm showed a highest recall (0.985) among six widely used ML algorithms: RF, Bagged trees, AdaBoost, Support Vector Machine, Neural Network. Therefore, RF was selected to build the ML model since our goal was to identify the maximum number of true matches. Our combined linkage PRL + ML showed a higher recall (range 0.988-0.992) than either PRL (range 0.916-0.991) or ML (0.981) alone. It identified 1995 individuals participating in both GEMO (6375 participants) and GENEPSO (4925 participants). CONCLUSIONS: Our hybrid linkage process represents an efficient tool for linking GEMO and GENEPSO. It may be generalizable to other epidemiological studies involving other databases and registries.