ABSTRACT
Germline pathogenic/likely pathogenic (P/LP) variants inform cancer treatment and prevention, yet real‐world detection patterns and downstream clinical management remain poorly characterized in East Asian hereditary cancer cohorts. This retrospective cohort study included 5829 patients with cancer who underwent germline testing at the National Cancer Center, Republic of Korea, between April 2006 and April 2023 using BRCA1/2‐only testing or multigene panel testing. We examined P/LP variant prevalence, detection patterns under the revised national reimbursement criteria, gene group‐stratified post‐test clinical management, and variant reclassification outcomes. Among 5829 patients (96.6% female; median age, 51 years), 987 (16.9%) carried P/LP variants. Detection rates differed across reimbursement criteria, with the highest rates in patients with ovarian cancer (24.2%) and those meeting multiple criteria (up to 39.3%). Among P/LP carriers, 773 (78.3%) received at least one post‐test clinical action; however, action rates differed substantially across gene groups, ranging from 84.3% among BRCA1/2 carriers to 46.0% among mismatch repair gene carriers. Of 1384 initial variants of uncertain significance, 25 (1.8%) were reclassified as P/LP, identifying 46 additional carriers. Overall variant frequencies were strongly shaped by the breast and ovarian cancer–predominant cohort composition, whereas documented post‐test clinical actions varied across gene groups. Variant reclassification identified additional P/LP carriers, supporting the need for structured recontact and follow‐up pathways.
Keywords: genetic counseling, genetic testing, genetic variation, hereditary breast and ovarian cancer syndrome
In 5829 Korean patients undergoing hereditary cancer germline testing, 987 (16.9%) carried pathogenic/likely pathogenic variants. Post‐test management varied substantially across gene groups, while variant reclassification identified 46 additional carriers, underscoring the need for gene‐specific management and structured recontact.
Abbreviations
- B/LB
benign/likely benign
- HRR
homologous recombination repair
- MGPT
multigene panel testing
- MMR
mismatch repair
- NGS
next‐generation sequencing
- P/LP
pathogenic/likely pathogenic
- PARP
poly(ADP‐ribose) polymerase
- RRM
risk‐reducing mastectomy
- RRSO
risk‐reducing salpingo‐oophorectomy
- TNBC
triple‐negative breast cancer
- VUS
variants of uncertain significance
1. Introduction
Germline testing for pathogenic/likely pathogenic (P/LP) variants in cancer susceptibility genes informs a range of clinical decisions, including risk‐stratified surveillance, risk‐reducing surgery, poly(ADP‐ribose) polymerase (PARP) inhibitor therapy, and cascade testing of at‐risk family members [1]. Although the recommended management for BRCA1/2 carriers is well established and widely implemented, evidence‐based management for carriers of non‐BRCA genes is still evolving, and the extent to which variant findings are translated into clinical practice has not been systematically characterized [2, 3].
With the adoption of next‐generation sequencing (NGS)‐based multigene panel testing (MGPT), variant detection has expanded beyond BRCA1/2 to include homologous recombination repair (HRR), mismatch repair (MMR), and other cancer susceptibility genes [4]. As the proportion of P/LP variants in non‐BRCA genes increases, real‐world data on the clinical management that follows are needed. Moreover, as evidence accumulates, variants of uncertain significance (VUS) are periodically reclassified as P/LP, expanding the population of carriers who may benefit from clinical management [5].
To date, most available evidence on post‐test clinical management has been derived from Western clinical settings [6]. Differences in testing indications, reimbursement systems, variant‐interpretation resources, and care pathways may limit the direct generalizability of these findings to East Asian practice, thus necessitating independent evaluation.
Korea provides a unique setting, with a single‐payer national insurance system and clearly defined reimbursement changes. National health insurance reimbursement for BRCA1/2 testing was first announced in 2012 [7] and revised in July 2020 [8], generating longitudinal data spanning the transition from single‐gene testing to MGPT within a consistent framework.
This study aimed to characterize the P/LP variant landscape across cancer types and gene groups, evaluate detection patterns according to revised national reimbursement criteria, describe post‐test clinical management among variant carriers, and assess clinical follow‐up after variant reclassification in Korean patients undergoing hereditary cancer testing—providing evidence for cancer type–aware interpretation of variant frequencies and structured post‐test care pathways in East Asian practice.
2. Materials and Methods
2.1. Study Population
A total of 6739 individuals underwent germline testing at the National Cancer Center (Goyang, Republic of Korea) between April 2006 and April 2023. Patients without a cancer diagnosis (n = 756), those who underwent single‐gene testing other than BRCA1/2 (n = 118), and those tested through familial cascade testing (n = 36) were excluded, resulting in 5829 patients being included in the analysis (Figure 1A).
FIGURE 1.
Study flow diagram and P/LP variant frequency by cancer type and gene. (A) Flow diagram of patient selection and variant classification by testing modality. (B) Bubble plot of P/LP variant frequencies by cancer type and gene. Bubble size reflects variant frequency (%). Bubble shape indicates penetrance level (circle: high; square: moderate; diamond: low/uncertain) and color indicates functional gene classification. aIncluding 26 patients who received additional TP53‐targeted testing. CRC, colorectal cancer; HRR, homologous recombination repair; MMR, mismatch repair; P/LP, pathogenic/likely pathogenic.
For breast and ovarian cancer, patients were referred for testing in accordance with the national insurance reimbursement criteria for BRCA1/2 testing (Table S1). During this period, approximately 137,000 patients were diagnosed with cancer at the center, of whom approximately 19,000 had breast and approximately 7000 had ovarian, fallopian tube, or primary peritoneal cancer; the tested cohort therefore represented approximately 4% of all patients with cancer but a higher fraction of those with breast (approximately 17%) and ovarian (approximately one‐third) cancer. For other types of cancer, testing was performed when hereditary cancer was clinically suspected based on early‐onset disease, multiple primary tumors, or a clinically suggestive family history of cancer. In 2016, NGS‐based MGPT was introduced at our institution, and eligible patients were given the option to undergo either BRCA1/2‐only testing or MGPT. This study was approved by the Institutional Review Board (IRB) of the National Cancer Center of Korea (IRB No. NCC2023‐0136), and the requirement for written informed consent was waived owing to the retrospective nature of the study.
2.2. Genetic Testing and Variant Classification
Genetic testing was classified into two groups: (i) BRCA1/2‐only testing and (ii) NGS‐based MGPT covering 9–73 cancer susceptibility genes, with panel size varying by testing period (Table S2). Sequencing platforms, variant calling pipelines, and variant classification procedures are summarized here; detailed information can be found in Document S1. All variants were classified according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology guidelines [9]. Variants initially classified as VUS were re‐evaluated in July 2025 based on updated evidence and combined with institutional reclassification records accumulated during routine clinical practice. For patient‐level analysis, patients were categorized based on the highest‐significance variant identified: patients with at least one P/LP variant were classified as P/LP carriers; those without P/LP variants but with at least one VUS were classified as the VUS group; and the remainder were classified as the benign/likely benign (B/LB) group [9].
2.3. Clinical Data Collection
Clinical data were extracted from electronic medical records, including primary cancer diagnosis, age at diagnosis, histological subtype, immunohistochemical characteristics, bilateral disease, multiple primary malignancies, and family history of cancer among first‐, second‐, and third‐degree relatives. For P/LP variant carriers, post‐test clinical management was ascertained from the date of germline testing to the last available follow‐up and included PARP inhibitor use, risk‐reducing salpingo‐oophorectomy (RRSO), risk‐reducing mastectomy (RRM), department‐specific surveillance visits (breast, gynecology, gastroenterology, and other departments), and cascade genetic testing of family members. At our institution, germline test results are typically disclosed to patients approximately 2–3 weeks after testing; the germline testing date was therefore used as a consistently available proxy for the date of result disclosure. For reclassified variants, post‐reclassification management was assessed from the date of reclassification documentation in the medical record. PARP inhibitor use was defined as any documented administration after the germline testing date, regardless of whether the prescription was explicitly germline‐guided or indication‐based. Cascade testing was captured when genetic testing of relatives was documented in institutional records or reported by the proband and recorded in the medical record.
2.4. P/LP Variant Frequency Analysis
P/LP variant frequencies were calculated for each gene using the number of patients whose testing panel included that gene as the denominator, and frequencies were then stratified by cancer type and gene group. Each gene was further classified by penetrance level—high, moderate, or low/uncertain—as defined by Srinivasan et al. [10]. Genes with P/LP variants were categorized into four functional groups: BRCA1/2, non‐BRCA HRR, MMR, and Others (Table S2); this classification was applied throughout all subsequent analyses. Detection rates were also evaluated among patients with breast or ovarian cancer according to each criterion of the BRCA1/2 testing reimbursement guidelines revised in July 2020. These revised criteria were applied retrospectively to the entire 2006–2023 cohort to enable uniform comparison across the testing period. Gene‐specific frequencies were also compared with large‐scale MGPT studies from other populations.
2.5. Clinical Management Patterns Among P/LP Variant Carriers
Post‐test clinical management among P/LP variant carriers was assessed across five domains: PARP inhibitor therapy, RRSO, RRM, cancer surveillance (department visits or imaging), and cascade genetic testing. For PARP inhibitor therapy, RRSO, and RRM, rates were calculated within prespecified analysis populations informed by National Comprehensive Cancer Network guideline‐supported gene‐ and cancer‐specific management considerations (Table S3) [11]. PARP inhibitor use was assessed as documented treatment exposure rather than as a direct measure of germline‐guided therapeutic eligibility. Selected non‐BRCA HRR carriers were analyzed separately as an exploratory group. RRSO was analyzed among female carriers of BRCA1, BRCA2, BRIP1, RAD51C, RAD51D, or PALB2 variants without prior ovarian cancer. Lynch syndrome genes were excluded from the denominator‐based RRSO analysis, but documented RRSO events in these carriers were retained in the overall management analysis. Surveillance and cascade testing rates were calculated among all P/LP variant carriers and compared across four gene groups. Surveillance visits were defined as gene‐ or syndrome‐specific risk‐stratified follow‐up performed in accordance with documented hereditary cancer management pathways and were considered distinct from routine post‐treatment oncologic follow‐up. For gene group–stratified analyses, each carrier was assigned to a single, mutually exclusive gene group, taken as the highest‐priority group among their P/LP variants in the order BRCA1/2 > non‐BRCA HRR > MMR > Others.
2.6. Statistical Analysis
Descriptive statistics were used to summarize the clinical and demographic characteristics of the patients. Categorical variables were compared using the chi‐square or Fisher's exact test (the latter when any expected cell count was < 5), and continuous variables using the Wilcoxon rank‐sum test. Statistical significance was defined at two‐tailed p < 0.05. All statistical analyses and data visualizations were performed using R (v.4.3.1; R Core Team 2023, Vienna, Austria). For the time‐to‐event analysis of post‐test clinical actions, the cumulative probability of receiving any documented clinical action was estimated as one minus the Kaplan–Meier survival function, with the date of germline testing as the time origin and death or last follow‐up treated as censoring events; gene groups were compared using the log‐rank test. Sensitivity analyses of the overall P/LP carrier rate were performed in the HBOC‐spectrum subset (patients with breast, ovarian, pancreatic, or prostate cancer) and after excluding patients with ovarian cancer.
3. Results
3.1. Clinical Characteristics
The analysis included 5829 patients with cancer. Of these, 2262 (38.8%) underwent BRCA1/2‐only testing (including 26 patients who received additional TP53‐targeted testing) and 3567 (61.2%) underwent NGS‐based MGPT (Figure 1A). Most patients were female (n = 5628; 96.6%), with a median age of 51 years at the time of genetic testing. The most common cancer diagnoses at the time of testing were breast (n = 3272; 56.1%) and ovarian (n = 2206; 37.8%) cancer, followed by thyroid (n = 153), pancreatic (n = 148), endometrial (n = 145), and colorectal (n = 110) cancer, among others. At the time of testing, 474 (8.1%) patients had multiple primary cancers, increasing to 543 patients (9.3%) during the follow‐up period, with a higher rate among P/LP variant carriers than among non‐carriers (13.2% vs. 8.5%; p < 0.001). Detailed clinical characteristics are presented in Table 1.
TABLE 1.
Clinical characteristics of the study population.
| Characteristic | No. (%) | p a | ||
|---|---|---|---|---|
| Total (n = 5829) | Patients with P/LP variants (n = 987) | Patients with VUS or B/LB variants (n = 4842) | ||
| Sex | ||||
| Female | 5628 (96.6) | 959 (97.2) | 4669 (96.4) | 0.248 |
| Male | 201 (3.4) | 28 (2.8) | 173 (3.6) | |
| Age at first genetic testing | ||||
| Mean ± SD (year) | 51.4 ± 11.8 | 52.8 ± 11.0 | 51.2 ± 12.0 | < 0.001 |
| ≤ 40 | 1167 (20.0) | 145 (14.7) | 1022 (21.1) | |
| 41–50 | 1656 (28.4) | 282 (28.6) | 1374 (28.4) | |
| 51–60 | 1673 (28.7) | 318 (32.2) | 1355 (28.0) | |
| > 60 | 1333 (22.9) | 242 (24.5) | 1091 (22.5) | |
| Age at first cancer diagnosis | ||||
| Mean ± SD (year) | 50.0 ± 11.9 | 51.0 ± 11.3 | 49.8 ± 12.0 | 0.006 |
| ≤ 40 | 1408 (24.2) | 203 (20.6) | 1205 (24.9) | |
| 41–50 | 1711 (29.4) | 282 (28.6) | 1429 (29.5) | |
| 51–60 | 1567 (26.9) | 300 (30.4) | 1267 (26.2) | |
| > 60 | 1143 (19.6) | 202 (20.5) | 941 (19.4) | |
| Personal history of cancer at genetic testing b | ||||
| Breast cancer | 3272 (56.1) | 412 (41.7) | 2860 (59.1) | < 0.001 |
| Bilateral breast cancer | 487 (8.4) | 65 (6.6) | 422 (8.7) | 0.028 |
| Ovarian cancer c | 2206 (37.8) | 553 (56.0) | 1653 (34.1) | < 0.001 |
| Thyroid cancer | 153 (2.6) | 21 (2.1) | 132 (2.7) | 0.284 |
| Endometrial cancer | 145 (2.5) | 33 (3.3) | 112 (2.3) | 0.058 |
| Pancreatic cancer | 148 (2.5) | 15 (1.5) | 133 (2.7) | 0.026 |
| Colorectal cancer | 110 (1.9) | 26 (2.6) | 84 (1.7) | 0.058 |
| Prostate cancer | 73 (1.3) | 7 (0.7) | 66 (1.4) | 0.092 |
| Gastric cancer | 45 (0.8) | 10 (1.0) | 35 (0.7) | 0.342 |
| Lung cancer | 40 (0.7) | 3 (0.3) | 37 (0.8) | 0.110 |
| Cervical cancer | 39 (0.7) | 5 (0.5) | 34 (0.7) | 0.492 |
| Biliary tract cancer | 21 (0.4) | 1 (0.1) | 20 (0.4) | 0.237 |
| Kidney cancer | 18 (0.3) | 1 (0.1) | 17 (0.4) | 0.341 |
| Sarcoma | 16 (0.3) | 3 (0.3) | 13 (0.3) | 0.744 |
| Others | 46 (0.8) | 7 (0.7) | 39 (0.8) | 0.755 |
| Personal history of cancer after genetic testing b | ||||
| Breast cancer | 3286 (56.4) | 418 (42.4) | 2868 (59.2) | < 0.001 |
| Bilateral breast cancer | 526 (9.0) | 74 (7.5) | 452 (9.3) | 0.066 |
| Ovarian cancer c | 2216 (38.0) | 560 (56.7) | 1656 (34.2) | < 0.001 |
| Thyroid cancer | 166 (2.8) | 23 (2.3) | 143 (3.0) | 0.283 |
| Endometrial cancer | 149 (2.6) | 35 (3.5) | 114 (2.4) | 0.031 |
| Pancreatic cancer | 151 (2.6) | 16 (1.6) | 135 (2.8) | 0.035 |
| Colorectal cancer | 115 (2.0) | 27 (2.7) | 88 (1.8) | 0.059 |
| Prostate cancer | 75 (1.3) | 8 (0.8) | 67 (1.4) | 0.145 |
| Gastric cancer | 50 (0.9) | 11 (1.1) | 39 (0.8) | 0.337 |
| Lung cancer | 54 (0.9) | 6 (0.6) | 48 (1.0) | 0.252 |
| Cervical cancer | 40 (0.7) | 5 (0.5) | 35 (0.7) | 0.453 |
| Biliary tract cancer | 22 (0.4) | 2 (0.2) | 20 (0.4) | 0.566 |
| Kidney cancer | 21 (0.4) | 2 (0.2) | 19 (0.4) | 0.560 |
| Sarcoma | 17 (0.3) | 3 (0.3) | 14 (0.3) | > 0.999 |
| Others | 51 (0.9) | 12 (1.2) | 39 (0.8) | 0.207 |
| Multiple primary cancer at genetic testing | ||||
| Yes | 474 (8.1) | 103 (10.4) | 371 (7.7) | 0.004 |
| No | 5355 (91.9) | 884 (89.6) | 4471 (92.3) | |
| Multiple primary cancer after genetic testing | ||||
| Yes | 543 (9.3) | 130 (13.2) | 413 (8.5) | < 0.001 |
| No | 5286 (90.7) | 857 (86.8) | 4429 (91.5) | |
| Family history of cancer within TDR | ||||
| Any cancer | 4016 (68.9) | 721 (73.0) | 3295 (68.1) | 0.002 |
| Breast cancer | 1792 (30.7) | 328 (33.2) | 1464 (30.2) | 0.063 |
| Ovarian cancer | 320 (5.5) | 142 (14.4) | 178 (3.7) | < 0.001 |
| Pancreatic cancer | 372 (6.4) | 61 (6.2) | 311 (6.4) | 0.776 |
| Prostate cancer | 195 (3.3) | 30 (3.0) | 165 (3.4) | 0.558 |
Note: Statistically significant p‐values are indicated in bold (i.e., p < 0.05).
Abbreviations: B/LB, benign/likely benign; P/LP, pathogenic/likely pathogenic; TDR, third‐degree relative; VUS, variant of uncertain significance.
Wilcoxon rank‐sum test for continuous variables; Pearson chi‐square test for categorical variables; Fisher's exact test applied when any expected cell count was < 5.
Cancer types are not mutually exclusive; patients with multiple primary cancers are counted in each applicable category.
Ovarian cancer includes primary peritoneal, ovarian, and fallopian tube cancer.
3.2. Landscape of Germline P/LP Variants
Of the 5829 patients tested, 987 (16.9%) carried P/LP variants, 1402 (24.1%) carried VUS without P/LP variants, and 3440 (59.0%) carried only B/LB variants (Figure 1A). These counts include 46 patients newly identified as P/LP variant carriers through variant reclassification, as described in the next section. The proportion of VUS carriers was substantially higher in the MGPT group than in the BRCA1/2‐only group (35.8% vs. 5.5%). Within the MGPT group, the VUS rate generally increased with panel size, from 7.4% for the 9‐gene panel to 30.4%–32.6% for intermediate panels and 47.7% for the largest (73‐gene) panel (Table S4). The most frequently identified P/LP variant genes were BRCA1 (454 of 5829; 7.8%), BRCA2 (317 of 5829; 5.4%), RAD51D (28 of 2829; 1.0%), ATM (27 of 3513; 0.8%), and PALB2 (22 of 3513; 0.6%), with gene‐specific frequencies calculated using the number of patients tested for each gene as the denominator (Table S5). Specific locations of the identified variants are presented in lollipop plots (Figure S1). In sensitivity analyses, the P/LP carrier rate was 16.8% (940 of 5589) in the HBOC‐spectrum subset and 12.0% (434 of 3623) after excluding patients with ovarian cancer.
In breast cancer, BRCA2 (166 of 3272; 5.1%) and BRCA1 (154 of 3272; 4.7%) were the most common, followed by PALB2 (19 of 1673; 1.1%), MUTYH (12 of 1673; 0.7%), and CHEK2 (11 of 1673; 0.7%). In ovarian cancer, BRCA1 (330 of 2206; 15.0%) and BRCA2 (152 of 2206; 6.9%) predominated, followed by RAD51D (18 of 1306; 1.4%), BRIP1 (10 of 1460; 0.7%), and RAD51C (8 of 1460; 0.5%). ATM (7 of 140; 5.0%) and BRCA2 (4 of 73; 5.5%) were the most frequently detected genes in pancreatic and prostate cancer, respectively. P/LP variant frequencies by cancer type and gene, stratified by penetrance level, are presented in Figures 1B and S2.
3.3. Variant Reclassification Outcomes
To assess the contribution of variant reclassification to the P/LP carrier counts described above, variants from the initial laboratory reports were re‐evaluated. Among 941 patients with an initial P/LP classification, 370 unique P/LP variants were identified. Of the 1384 initial VUS identified, 237 (17.1%) were reclassified as B/LB and 25 (1.8%) as P/LP, with reclassification occurring predominantly in BRCA1/2 (Figure 2A,B). The 25 variants reclassified as P/LP spanned several cancer predisposition genes, including BRCA1 (n = 11; 44.0%) and BRCA2 (n = 5; 20.0%) (Table 2). At the patient level, upward reclassification of a VUS affected 49 patients: 46 (93.9%) were newly identified as P/LP variant carriers following the upward reclassification of a VUS, whereas the remaining 3 (6.1%) already carried a P/LP variant before the reclassification of an additional VUS (Figure 2C,D and Table S6). Representative pedigrees of two patients who underwent cascade testing after variant reclassification are presented in Figure S3.
FIGURE 2.
Reclassification outcomes of VUS. Sankey diagrams showing reclassification outcomes of initial VUS by gene category and testing modality. (A) Initial BRCA1/2 VUS (n = 312): 132 (42.3%) were reclassified, including 16 to P/LP and 116 to B/LB. (B) Initial non‐BRCA VUS (n = 1072): 130 (12.1%) were reclassified, including 9 to P/LP and 121 to B/LB. (C) BRCA1/2‐tested patients in the VUS group (n = 526): 402 (76.4%) had at least one variant reclassified, including 25 to P/LP. (D) NGS‐tested patients in the VUS group (n = 1531): 265 (17.3%) had at least one variant reclassified, including 21 to P/LP. Panels (A)–(B) represent variant‐level analysis; panels (C)–(D) represent patient‐level analysis. Panels C and D show the 46 patients newly classified as P/LP carriers and exclude three patients who already carried another P/LP variant. B/LB, benign/likely benign; P/LP, pathogenic/likely pathogenic; VUS, variant of uncertain significance.
TABLE 2.
Variants reclassified from uncertain significance to pathogenic or likely pathogenic.
| No. | Gene | Nucleotide change | Protein change | Position a | Ref | Alt | Carriers, N (%) b | Type of evidence c | Reclassification | ClinVar d |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | ATM | c.4109+5G>A | p.? | 108158447 | G | A | 1 (0.03) | PM2, PP3, PP5_Strong | LP | CIP [P(3):US(1)] |
| 2 | BARD1 | c.2143C>T | p.Gln715Ter | 215593591 | G | A | 1 (0.03) | PVS1, PM2 | LP | P/LP |
| 3 | BRCA1 | c.4485‐2A>G | p.? | 41226540 | T | C | 1 (0.02) | PVS1, PM2, PP5 | P | P/LP |
| 4 | BRCA1 | c.5014_5016del | p.His1673del | 41219679 | TGTG | T | 6 (0.10) | PM2, PP5_Strong | LP | P/LP |
| 5 | BRCA1 | c.5099C>T | p.Thr1700Ile | 41215944 | G | A | 3 (0.05) | PM2, PP3, PP5_Strong | LP | CIP [LP(1):US(1)] |
| 6 | BRCA1 | c.5152+6T>C | p.? | 41215885 | A | G | 1 (0.02) | PM2, PP3, PP5_Strong | LP | LP |
| 7 | BRCA1 | c.5165C>A | p.Ser1722Tyr | 41215378 | G | T | 1 (0.02) | PM2, PP3, PP5_Strong | LP | CIP [LP(4):US(2)] |
| 8 | BRCA1 | c.5215G>C | p.Asp1739His | 41209131 | C | G | 1 (0.02) | PM2, PP3, PP5_Strong | LP | LP |
| 9 | BRCA1 | c.5332G>A | p.Asp1778Asn | 41203080 | C | T | 1 (0.02) | PM2, PP3, PP5_Strong | LP | P/LP |
| 10 | BRCA1 | c.5339T>C | p.Leu1780Pro | 41201205 | A | G | 37 (0.63) | PVS1, PM2, PP5 | P | P/LP |
| 11 | BRCA1 | c.5363G>T | p.Gly1788Val | 41201181 | C | A | 1 (0.02) | PVS1, PM2, PP5 | P | P |
| 12 | BRCA1 | c.5516T>C | p.Leu1839Ser | 41197771 | A | G | 1 (0.02) | PM2, PP3, PP5_Strong | LP | P |
| 13 | BRCA1 | c.81‐9C>G | p.? | 41267805 | G | C | 1 (0.02) | PVS1, PM2, PP5 | P | P |
| 14 | BRCA2 | c.5434G>T | p.Glu1812Ter | 32913926 | G | T | 1 (0.02) | PVS1, PM2, PP5 | P | P |
| 15 | BRCA2 | c.8165C>T | p.Thr2722Ile | 32937504 | C | T | 1 (0.02) | PM2, PP3, PP5_Strong | LP | CIP [LP(5):US(1)] |
| 16 | BRCA2 | c.8350C>T | p.Arg2784Trp | 32944557 | C | T | 2 (0.03) | PM2, PP3, PP5_Strong | LP | P/LP |
| 17 | BRCA2 | c.8488‐1G>A | p.? | 32945092 | G | A | 4 (0.07) | PVS1, PM2, PP5 | P | P |
| 18 | BRCA2 | c.9154C>T | p.Arg3052Trp | 32954180 | C | T | 5 (0.09) | PVS1, PM2, PP5 | P | P |
| 19 | CHEK2 | c.846+4_846+7del | p.? | 29105986 | ATACT | A | 1 (0.03) | PM2, PP3, PP5_Strong | LP | P/LP |
| 20 | MSH6 | c.3431T>G | p.Met1144Arg | 48030817 | T | G | 1 (0.03) | PM2, PP3, PP5_Strong | LP | LP |
| 21 | MUTYH | c.643G>A | p.Val215Met | 45798293 | C | T | 1 (0.03) | PM2, PP3, PP5_Strong | LP | P/LP |
| 22 | NF1 | c.4934T>C | p.Leu1645Pro | 29652936 | T | C | 1 (0.04) | PM2, PP3, PP5_Strong | LP | CIP [P(1):US(1)] |
| 23 | PALB2 | c.3350+5G>A | p.? | 23619180 | C | T | 6 (0.17) | PVS1, PM2, PP5 | P | P/LP |
| 24 | RAD51C | c.934C>T | p.Arg312Trp | 56801430 | C | T | 4 (0.11) | PM2, PP3, PP5_Strong | LP | CIP [P(1):LP(6):US(3)] |
| 25 | RAD51D | c.904‐2A>T | p.? | 33428057 | T | A | 10 (0.35) | PVS1, PM2 | LP | P/LP |
Abbreviations: CIP, conflicting interpretations of pathogenicity; P/LP, pathogenic/likely pathogenic; US, uncertain significance; VUS, variant of uncertain significance.
Genomic position in GRCh37/hg19 coordinates.
Carrier counts represent all cohort patients carrying each variant after final reclassification; Table S6 includes only patients whose original VUS report was upgraded, so counts do not sum to the 49 reclassification‐affected patients.
Classified by the 2015 American College of Medical Genetics and Genomics and the Association for Molecular Pathology guidelines.
Accessed July 30, 2025.
3.4. Comparison of Germline P/LP Variant Frequencies With International Cohorts
In this study, the prevalence of germline P/LP variants in cancer susceptibility genes was analyzed among Korean patients with cancer and compared with that in large‐scale international cohorts, overall and stratified by cancer type (Table S7) [12, 13, 14, 15]. In four studies breast cancer accounted for 52%–69% of participants, broadly similar to the 56.1% in the present cohort, whereas ovarian cancer accounted for 8%–25%, substantially lower than the 37.8% in this cohort. BRCA1 and BRCA2 P/LP variants were detected in 7.8% and 5.4% of patients in the pan‐cancer comparison. However, cancer type–stratified analysis showed that these frequencies were broadly comparable with those reported in international cohorts for the same cancer type—among patients with breast cancer (BRCA1 4.7%, BRCA2 5.1%) [16, 17, 18, 19, 20, 21, 22] and ovarian cancer (BRCA1 15.0%, BRCA2 6.9%) [23, 24], comparable to Western ovarian‐cancer cohorts—suggesting that the elevated pan‐cancer BRCA1/2 frequencies reflect the enrichment of this referral‐based cohort for breast and ovarian cancer rather than population‐level ethnic differences. The breast‐cancer BRCA1 frequency was within the reported range, whereas the breast‐cancer BRCA2 frequency (5.1%) was modestly higher than in the selected comparator cohorts.
In contrast, CHEK2 and ATM P/LP variants were less frequent (0.5% and 0.8%, respectively) than those observed in Western cohorts, as reported by LaDuca et al. and Bychkovsky et al. [13, 14]. RAD51D P/LP variants were identified in 1.0% of patients; this pan‐cohort frequency was influenced by the cohort's ovarian‐cancer predominance, and in cancer type–stratified comparison the ovarian‐cancer RAD51D frequency fell within the range reported for international ovarian‐cancer cohorts (Table S7) [23, 24].
3.5. Frequency of P/LP Variants by BRCA1/2 Testing Reimbursement Criteria
Among patients with breast or ovarian cancer, P/LP variant detection rates were evaluated according to each criterion of the reimbursement guidelines revised in July 2020, combining both BRCA1/2‐only and NGS‐tested patients (Figure 3). Patients meeting the ovarian cancer criterion had the highest single‐criterion detection rate (24.2%), followed by those meeting the triple‐negative breast cancer (TNBC; 14.2%), family history (breast, ovarian, pancreatic, or prostate cancer; 10.2%), bilateral breast cancer (8.1%), and young‐onset breast cancer (6.6%) criteria. Patients meeting multiple criteria showed further elevated detection rates, reaching up to 39.3% for a specific combination of multiple criteria. Across all criteria, BRCA1/2 variants accounted for most of the detected P/LP variants, whereas non‐BRCA HRR, MMR, and Others comprised a variable but consistent proportion. The previous version of the reimbursement criteria is provided in Figure S4.
FIGURE 3.
P/LP variant detection rates among patients with breast or ovarian cancer according to each criterion of the national insurance reimbursement guidelines for BRCA1/2 testing after the July 2020 revision. The lower panel is an UpSet plot in which each column is a unique criteria combination (connected dots); bar height indicates the number of patients, with P/LP carriers in pink and non‐carriers in gray. Only combinations with ≥ 20 patients are displayed. Percentages above bars indicate the P/LP detection rate for each combination. The upper panel shows the distribution of P/LP variants by gene group (BRCA1/2, non‐BRCA HRR, MMR, and Others) for each combination. The horizontal bar chart on the left indicates the set size (total number of patients meeting each individual criterion). BC, breast cancer; FH, family history; HRR, homologous recombination repair; MMR, mismatch repair; OC, ovarian cancer; P/LP, pathogenic/likely pathogenic; PC, pancreatic cancer; PRC, prostate cancer; TNBC, triple‐negative breast cancer.
3.6. Clinical Management Patterns Among P/LP Variant Carriers
Among the 987 P/LP variant carriers, 773 (78.3%) received at least one form of post‐test clinical management (Figure 4A). PARP inhibitor therapy was administered to 39.4% (301 of 764) of BRCA1/2 carriers and 12.7% (8 of 63) of selected non‐BRCA carriers analyzed as an exploratory group. RRSO was performed in 67.4% (186 of 276) of eligible BRCA1/2 carriers and 46.2% (12 of 26) of eligible non‐BRCA carriers. RRM rates were low in both groups (43 of 696 [6.2%] for BRCA1/2; 1 of 26 [3.8%] for non‐BRCA; p = 1.00) (Figure 4B).
FIGURE 4.
Post‐test clinical management patterns among P/LP variant carriers. (A) Proportion of P/LP variant carriers (n = 987) who received any type of clinical management. (B) Clinical management rates by gene group and management domain. For PARP inhibitor therapy, RRSO, and RRM, rates were calculated within prespecified analysis populations; for surveillance and cascade testing, rates were calculated among all P/LP variant carriers in each gene group. (C) Cascade genetic testing yield among P/LP carriers and their family members. Gene groups were mutually exclusive: carriers harboring P/LP variants in both BRCA1/2 and non‐BRCA HRR genes were assigned to the BRCA1/2 group. In panel B, “Non‐BRCA” denotes the intervention‐specific eligible genes for PARP inhibitor therapy, RRSO, or RRM, whereas “non‐BRCA HRR” denotes the predefined HRR gene group used for the surveillance and cascade‐testing comparisons. HRR, homologous recombination repair; MMR, mismatch repair; P/LP, pathogenic/likely pathogenic; PARP, poly(ADP‐ribose) polymerase; RRM, risk‐reducing mastectomy; RRSO, risk‐reducing salpingo‐oophorectomy.
Cancer surveillance rates varied across the four gene groups: 66.1% (508 of 769) for BRCA1/2, 44.5% (53 of 119) for non‐BRCA HRR, 32.0% (16 of 50) for MMR, and 46.9% (23 of 49) for Others. Cascade genetic testing was initiated by 305 of 987 (30.9%) probands, with rates of 33.4% (257 of 769) for BRCA1/2, 17.6% (21 of 119) for non‐BRCA HRR, 24.0% (12 of 50) for MMR, and 30.6% (15 of 49) for Others, respectively. Of these 305 probands, 291 were identified from institutional testing records and 14 from patient‐reported external testing documented in the medical record. Among the 603 family members tested (mean: 2.0 per proband), 291 (48.3%) carried a P/LP variant (Figure 4C). Most were children (50.2%) and siblings (35.3%), whereas parents accounted for only 5.5% (Figure S5). In a cancer‐type‐stratified analysis, among BRCA1/2 carriers, PARP inhibitor uptake was highest in ovarian cancer (58.2%) and lower in breast cancer (11.6%); among non‐BRCA HRR carriers with ovarian cancer, the corresponding rate was 19.1% (Table S8). The higher rate of clinical action among BRCA1/2 carriers relative to other gene groups persisted in a time‐to‐event (cumulative‐probability) analysis that accounted for differences in follow‐up duration (Figure S6).
4. Discussion
This retrospective study of 5829 Korean patients with cancer tested over 17 years identifies three clinically relevant findings: a germline variant landscape strongly shaped by the cohort's breast and ovarian cancer predominance, detection yield across national reimbursement eligibility criteria, and gene group‐dependent differences in documented post‐test clinical management. Overall, 987 (16.9%) patients harbored P/LP variants, with BRCA1 (7.8%) and BRCA2 (5.4%) being the most common. These frequencies exceeded those measured in comparable MGPT cohorts from Western and Latin American populations [12, 13, 14, 15, 25], most likely due to referral enrichment, as 37.8% of this cohort had ovarian cancer (P/LP frequency 25.1%). In contrast, CHEK2 and ATM frequencies were lower than those in European‐descent cohorts [13, 14], and that of RAD51D was likewise elevated in the overall cohort; however, in cancer type–stratified comparison the ovarian‐cancer RAD51D frequency fell within the range reported for international ovarian‐cancer cohorts (Table S7), indicating that this pattern reflected the cohort's ovarian‐cancer predominance rather than population‐specific enrichment. Although RAD51D was not enriched after accounting for cancer type, its ovarian cancer–specific frequency was consistent with its established role in hereditary ovarian cancer susceptibility [26, 27, 28]. These findings underscore that variant frequencies in referral‐based, cancer type–enriched testing cohorts must be interpreted in light of cancer‐type composition when informing panel design.
Detection rates under the July 2020 revised criteria ranged from 6.6% (young‐onset breast cancer only) to 24.2% (ovarian cancer) and up to 39.3% for specific multi‐criteria combinations, suggesting enrichment for patients with a higher probability of carrying P/LP variants. The inclusion of TNBC—supported by domestic data reporting a 13.1% BRCA1/2 mutation rate in unselected Korean patients with TNBC [29]—demonstrates how domestic epidemiological evidence can drive evidence‐based policy revision.
The substantially higher VUS rate observed in the MGPT group (35.8% vs. 5.5% in the BRCA1/2‐only group) reflects broader gene coverage and underscores the need for adequate pre‐ and post‐test counseling as testing expands beyond BRCA1/2 [30, 31]. However, whether findings are acted upon depends on the clarity of the result and the gene involved. The most clinically consequential finding is the gene group‐dependent heterogeneity in post‐test management as the implicated gene extends beyond BRCA1/2, which may reflect gene‐specific management indications and incomplete capture of care outside the institution, beyond implementation gaps alone. Any clinical action was documented in 84.3% of BRCA1/2 carriers, compared with 58.0% for non‐BRCA HRR carriers, 67.3% for carriers of other genes, and 46.0% for MMR carriers (p < 0.001). This pattern was consistent across specific interventions, with RRSO performed in 67.4% of eligible BRCA1/2 carriers compared with 46.2% of eligible non‐BRCA carriers. This difference most likely reflects the more established evidence base for BRCA1/2‐directed surgery and is consistent with prior studies reporting lower rates of risk‐reducing surgery among non‐BRCA gene variant carriers [2, 3].
PARP inhibitor use was also lower in non‐BRCA carriers (12.7% vs. 39.4%), consistent with more limited and gene‐specific therapeutic actionability outside BRCA1/2. However, as niraparib is approved in Korea as first‐line maintenance therapy for advanced ovarian cancer regardless of BRCA status, a proportion of documented PARP inhibitor use may reflect indication‐based rather than germline‐guided prescribing, which may have introduced confounding in this comparison. Consistent with this, the cancer‐type‐stratified uptake reported above (Table S8) was concentrated in ovarian cancer, indicating that a considerable proportion of documented PARP inhibitor use reflects tumor‐type‐based indications rather than strictly germline‐guided prescribing. RRM rates were low in both BRCA1/2 and non‐BRCA groups (BRCA1/2: 6.2%; non‐BRCA: 3.8%), because most carriers did not have breast cancer (predominantly ovarian cancer), a clinical setting in which risk‐reducing mastectomy is not indicated; the procedure was almost entirely confined to carriers already diagnosed with breast cancer, reflecting the predominance of patients presenting with ovarian or other active malignancies, for whom the timing and expected benefit of the procedure require individualized consideration.
Beyond surgery and targeted therapy, cancer surveillance rates were lowest among MMR carriers (32.0%), possibly reflecting implementation barriers, variation in gene‐specific pathways, and incomplete capture of surveillance outside the institution, rather than the absence of established Lynch syndrome recommendations. Cascade genetic testing was initiated by only 30.9% of probands—with the lowest rate observed for non‐BRCA HRR carriers (17.6%)—despite a P/LP yield of 48.3% among the 603 tested family members (mean: 2.0 per proband). These findings, derived from institutional records supplemented by patient‐reported data documented in medical records, fall within the wide range of cascade testing uptake rates (approximately 20%–70%) reported in previous studies, which vary substantially according to counseling delivery model, healthcare system, and patient cohort [32, 33, 34].
Among the 46 patients newly identified as P/LP carriers following reclassification, 34 (73.9%) received at least one clinical action, comparable with the 78.5% observed among initially identified carriers (p = 0.46). The only significant difference was PARP inhibitor uptake among eligible patients (18.4% vs. 40.5%; p = 0.006), whereas the RRSO, RRM, surveillance, and cascade testing rates did not differ significantly between groups. This exploratory difference may reflect differences in cancer type, treatment era, follow‐up duration, and opportunity to receive therapy and should not be interpreted causally. These comparisons should be interpreted with caution given the limited sample size of newly identified carriers (n = 46) and the resulting limited statistical power.
This study has several limitations. First, it was based on a single tertiary cancer center strongly enriched for breast and ovarian cancer through clinical referral and reimbursement‐driven testing; the resulting variant frequencies reflect a high‐risk referral population rather than population‐level prevalence, and generalization should account for the HBOC‐restricted and ovarian‐excluded sensitivity analyses reported above. Second, gene‐specific denominators varied across testing eras and panel compositions, introducing potential ascertainment bias in frequency comparisons and temporal heterogeneity in clinical management. Third, PARP inhibitor use could not be distinguished between germline‐guided and indication‐based prescribing; because niraparib is approved for advanced ovarian cancer irrespective of BRCA status in Korea, aggregate uptake should be interpreted descriptively. Fourth, follow‐up duration varied across testing eras, with patients tested in 2022–2023 having a shorter observation window that may have underestimated clinical‐action rates, particularly after reclassification; a time‐to‐event analysis yielded patterns broadly consistent with the overall gene group–dependent heterogeneity. Fifth, several panel genes have uncertain or context‐dependent associations with hereditary cancer (e.g., SPINK1, RAD51B), and carriers of a single MUTYH variant do not meet criteria for MUTYH‐associated polyposis; excluding SPINK1 and RAD51B changed the overall carrier rate only marginally (16.9% to 16.8%). Finally, cascade testing was partly ascertained from patient‐reported family communication documented in medical records; when restricted to institutional records, the initiation rate (29.5%, 291 of 987) was consistent with the overall estimate.
In conclusion, in this breast and ovarian cancer–predominant Korean testing cohort, overall germline variant frequencies were strongly influenced by cancer‐type composition, whereas documented post‐test clinical actions showed gene group‐dependent heterogeneity. Variant reclassification identified additional P/LP carriers and underscored the need for structured recontact and follow‐up. These findings support cancer type–aware interpretation of variant frequencies and more systematic pathways for translating germline test results into clinical care (Figure 5).
FIGURE 5.
Graphical abstract summarizing the germline variant landscape and post‐test clinical management. Among 5829 Korean patients tested for suspected hereditary cancer, 987 (16.9%) carried pathogenic/likely pathogenic variants. Post‐test management varied substantially across gene groups, and variant reclassification identified 46 additional carriers, underscoring the need for gene‐specific management and structured recontact. Created in BioRender. Kong, S. (2026) https://BioRender.com/cmnd3bc.
Author Contributions
Min‐Chae Kang: conceptualization, data curation, formal analysis, visualization, writing – original draft, writing – review and editing. Ha‐Eun Lee: conceptualization, data curation, visualization, formal analysis, writing – original draft, writing – review and editing. Sang‐Yoon Park: investigation, writing – review and editing. Seeyoun Lee: investigation, writing – review and editing. So‐Youn Jung: investigation, writing – review and editing. Myong Cheol Lim: investigation, writing – review and editing. Han‐Sung Kang: investigation, writing – review and editing. Chang Ahn Seol: conceptualization, methodology, validation, resources. Jun‐Ha Jang: conceptualization. Sang Myung Woo: investigation, writing – review and editing. Jae Young Joung: investigation, writing – review and editing. Wonyoung Choi: investigation, writing – review and editing. Sung Chan Park: investigation, writing – review and editing. Sun‐Young Kong: conceptualization, funding acquisition, investigation, methodology, project administration, supervision, validation, writing – review and editing.
Funding
This work was supported by a grant from the National Cancer Center, funded by the Ministry of Health & Welfare (Republic of Korea) (Grant No. NCC2410821), and by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS‐2026‐25483431). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Ethics Statement
Approval of the research protocol by an Institutional Review Board: This study was approved by the Institutional Review Board (IRB) of the National Cancer Center of Korea (IRB No. NCC2023‐0136). Informed Consent: Informed consent was waived owing to the retrospective nature of this study.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Doc S1: Supplementary Methods.
Figure S1: Lollipop plots of the germline pathogenic/likely pathogenic variants in the top 10 frequently detected genes. Lollipop plots are presented for the top 10 genes with the highest number of pathogenic variants generated by cBioportal (http://www.cbioportal.org/): (A) BRCA1, (B) BRCA2, (C) RAD51D, (D) ATM, (E) PALB2, (F) MUTYH, (G) CHEK2, (H) MSH2, (I) BRIP1, and (J) MLH1. The x‐axis reflects the number of amino acid residues, and the y‐axis shows the total number of patients with each pathogenic/likely pathogenic (P/LP) variant.
Figure S2: Heatmap of pathogenic/likely pathogenic variant frequencies by cancer type and gene. Frequencies (%) were calculated using the number of patients whose testing panel included each gene as the denominator (panel‐specific gene denominators are provided in Table S5). Genes are ordered by penetrance (high, moderate, low/uncertain) and colored by functional classification. Color intensity within cells reflects variant frequency. Empty cells indicate no P/LP variant detected or gene not included in the testing panel. GPV, germline P/LP variant; P/LP, pathogenic/likely pathogenic; PTs, patients.
Figure S3: Pedigrees of two patients with reclassified pathogenic/likely pathogenic variants. (A) The family pedigree of patient PT05 was constructed during the follow‐up period. (B) The family pedigree of PT28 was obtained at the time of genetic testing.
Figure S4: Distribution of patients and pathogenic/likely pathogenic variant frequency before the revision of reimbursement criteria for BRCA1/2 testing. BC, breast cancer; FH, family history; GPV, germline P/LP variants; HRR, homologous recombination repair; MMR, mismatch repair; OC, ovarian cancer; P/LP, pathogenic/likely pathogenic; PTs, patients.
Figure S5: Cascade genetic testing distribution among family members by relationship type. FDR, first‐degree relatives; SDR, second‐degree relatives; TDR, third‐degree relatives.
Figure S6: Cumulative probability of any post‐test clinical action over time, by gene group. Curves show the cumulative probability of receiving any documented clinical action, estimated as one minus the Kaplan–Meier survival function, with the date of germline testing as the time origin and deaths treated as censoring events. Estimated cumulative probabilities at 1, 3, and 5 years were 75.0%, 83.9%, and 87.0% for BRCA1/2; 50.6%, 57.0%, and 64.3% for non‐BRCA HRR; 45.9%, 45.9%, and 45.9% for MMR; and 63.4%, 68.6%, and 68.6% for Others (overall, 70.0%, 78.5%, and 82.2%). The difference across gene groups was significant by the log‐rank test (p < 0.001). HRR, homologous recombination repair; MMR, mismatch repair.
Table S1: Insurance reimbursement criteria for BRCA1/2 testing before and after July 2020.
Table S2: Germline tests performed during the study period and their gene composition.
Table S3: Definitions of analysis populations and clinical action criteria for five actionability domains.
Table S4: Variant of uncertain significance (VUS) rate by testing panel. Proportion of patients with a VUS‐only result (a VUS but no P/LP variant), by testing panel; the multigene panel testing (MGPT) tiers are further broken down by panel size.
Table S5: Gene‐specific testing denominators and pathogenic/likely pathogenic (P/LP) variant frequencies. For each gene, the number of P/LP variant carriers (patients with P/LP variants in more than one gene are counted once for each gene), the panel‐coverage‐adjusted number of patients tested, and the resulting frequency.
Table S6: Clinical characteristics of patients with variants reclassified from variants of uncertain significance to pathogenic or likely pathogenic variants (n = 49).
Table S7: Germline P/LP variant frequencies compared with international studies, overall and stratified by cancer type. Values are carriers, n (%), verified against the original article or supplement. Frequencies use each study's own denominator (many use gene‐specific, panel‐coverage‐adjusted denominators; our cohort likewise uses panel‐adjusted denominators, with BRCA1 and BRCA2 calculated over the full (sub)cohort and other genes over NGS‐panel‐tested patients, counting each gene once per patient). Cancer‐type composition of the pan‐cancer cohorts (percentage of the cohort affected by each cancer type; patients may have more than one cancer): This study, breast 56.1% and ovarian 37.8%; Susswein 2016, breast 52.4% and ovarian 8.5% (a further 25.2% were unaffected individuals tested because of family history); LaDuca 2020, breast 54.4% and ovarian 7.6%; Bychkovsky 2022 (a cohort of patients with multiple primary cancers), breast 68.1%, ovarian 25.1%, and endometrial 21.2%; Cheo 2025, breast 53.2% and ovarian 11.4%.
Table S8: PARP inhibitor uptake stratified by cancer type and gene group. PARP inhibitor uptake by cancer type and gene group among P/LP variant carriers with each qualifying cancer. BRCA1/2 carriers with an approved cancer indication and non‐BRCA HRR carriers with the same cancers are shown. Patients with more than one qualifying cancer are counted in each applicable row.
Data Availability Statement
All data generated or analyzed during this study are available from the corresponding author upon reasonable request within the range of IRB approval. In addition, individual‐level germline variant data are contributed to international hereditary cancer research consortia, including the Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA) and the Evidence‐based Network for the Interpretation of Germline Mutant Alleles (ENIGMA), in accordance with institutional and applicable consortium data‐governance requirements.
References
- 1. Imyanitov E. N., Kuligina E. S., Sokolenko A. P., et al., “Hereditary Cancer Syndromes,” World Journal of Clinical Oncology 14, no. 2 (2023): 40–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Lee S. S., Karpel H. C., Oh C., Smith J., and Pothuri B., “Beyond BRCA: Patterns of Risk‐Reducing Surgery for Non‐BRCA, Homologous Recombination Repair Pathway Gene Variant Carriers,” Gynecologic Oncology 170 (2023): 234–240. [DOI] [PubMed] [Google Scholar]
- 3. Vysotskaia V., Kaseniit K. E., Bucheit L., Ready K., Price K., and Johansen Taber K., “Clinical Utility of Hereditary Cancer Panel Testing: Impact of PALB2, ATM, CHEK2, NBN, BRIP1, RAD51C, and RAD51D Results on Patient Management and Adherence to Provider Recommendations,” Cancer 126, no. 3 (2020): 549–558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hall M. J., Obeid E. I., Schwartz S. C., Mantia‐Smaldone G., Forman A. D., and Daly M. B., “Genetic Testing for Hereditary Cancer Predisposition: BRCA1/2, Lynch Syndrome, and Beyond,” Gynecologic Oncology 140, no. 3 (2016): 565–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kobayashi Y., Chen E., Facio F. M., et al., “Clinical Variant Reclassification in Hereditary Disease Genetic Testing,” JAMA Network Open 7, no. 11 (2024): e2444526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Oak N., Cherniack A. D., Mashl R. J., et al., “Ancestry‐Specific Predisposing Germline Variants in Cancer,” Genome Medicine 12, no. 1 (2020): 51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Health Insurance Review & Assessment Service , Scope of Accreditation on BRCA (BRCA1, BRCA2) Gene Mutation Test (Health Insurance Review & Assessment Service, 2012). [Google Scholar]
- 8. Ministry of Health and Welfare , Notification No. 2020‐135 of the Ministry of Health and Welfare (Ministry of Health and Welfare, Korea, 2020). [Google Scholar]
- 9. Richards S., Aziz N., Bale S., et al., “Standards and Guidelines for the Interpretation of Sequence Variants: A Joint Consensus Recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology,” Genetics in Medicine 17, no. 5 (2015): 405–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Srinivasan P., Bandlamudi C., Jonsson P., et al., “The Context‐Specific Role of Germline Pathogenicity in Tumorigenesis,” Nature Genetics 53, no. 11 (2021): 1577–1585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. National Comprehensive Cancer Network , “Genetic/Familial High‐Risk Assessment: Breast, Ovarian, Pancreatic, and Prostate Version 3.2026,” https://www.nccn.org/professionals/physician_gls/pdf/genetics_bop.pdf. [DOI] [PubMed]
- 12. Susswein L. R., Marshall M. L., Nusbaum R., et al., “Pathogenic and Likely Pathogenic Variant Prevalence Among the First 10,000 Patients Referred for Next‐Generation Cancer Panel Testing,” Genetics in Medicine 18, no. 8 (2016): 823–832. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. LaDuca H., Polley E. C., Yussuf A., et al., “A Clinical Guide to Hereditary Cancer Panel Testing: Evaluation of Gene‐Specific Cancer Associations and Sensitivity of Genetic Testing Criteria in a Cohort of 165,000 High‐Risk Patients,” Genetics in Medicine 22, no. 2 (2020): 407–415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Bychkovsky B. L., Lo M. T., Yussuf A., et al., “Prevalence and Spectrum of Pathogenic Variants Among Patients With Multiple Primary Cancers Evaluated by Clinical Characteristics,” Cancer 128, no. 6 (2022): 1275–1283. [DOI] [PubMed] [Google Scholar]
- 15. Cheo S. W., Zhao J. J., Ong P. Y., et al., “Pathogenic Germline Variants in Cancer Predisposition Genes in Patients With Multiple Primary Cancers in an Asian Population and the Role of Extended Panel Genetic Testing,” ESMO Open 10, no. 3 (2025): 104495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Couch F. J., Shimelis H., Hu C., et al., “Associations Between Cancer Predisposition Testing Panel Genes and Breast Cancer,” JAMA Oncology 3, no. 9 (2017): 1190–1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Sun J., Meng H., Yao L., et al., “Germline Mutations in Cancer Susceptibility Genes in a Large Series of Unselected Breast Cancer Patients,” Clinical Cancer Research 23, no. 20 (2017): 6113–6119. [DOI] [PubMed] [Google Scholar]
- 18. Palmer J. R., Polley E. C., Hu C., et al., “Contribution of Germline Predisposition Gene Mutations to Breast Cancer Risk in African American Women,” Journal of the National Cancer Institute 112, no. 12 (2020): 1213–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Yadav S., Hu C., Nathanson K. L., et al., “Germline Pathogenic Variants in Cancer Predisposition Genes Among Women With Invasive Lobular Carcinoma of the Breast,” Journal of Clinical Oncology 39, no. 35 (2021): 3918–3926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Fasching P. A., Yadav S., Hu C., et al., “Mutations in BRCA1/2 and Other Panel Genes in Patients With Metastatic Breast Cancer—Association With Patient and Disease Characteristics and Effect on Prognosis,” Journal of Clinical Oncology 39, no. 15 (2021): 1619–1630. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Huang H., Couch R. E., Karam R., et al., “Pathogenic Variants in Cancer Susceptibility Genes Predispose to Ductal Carcinoma In Situ of the Breast,” Clinical Cancer Research 31, no. 1 (2025): 130–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kamdee K., Roothumnong E., Thongnoppakhun W., et al., “Comprehensive Germline and Somatic Profiling of High‐Risk Thai Breast Cancer via Next‐Generation Sequencing,” Scientific Reports 15, no. 1 (2025): 11427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Lhotova K., Stolarova L., Zemankova P., et al., “Multigene Panel Germline Testing of 1333 Czech Patients With Ovarian Cancer,” Cancers (Basel) 12, no. 4 (2020): 956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Morgan R. D., Burghel G. J., Flaum N., et al., “Extended Panel Testing in Ovarian Cancer Reveals BRIP1 as the Third Most Important Predisposition Gene,” Genetics in Medicine 26, no. 10 (2024): 101230. [DOI] [PubMed] [Google Scholar]
- 25. Ossa Gomez C. A., Achatz M. I., Hurtado M., et al., “Germline Pathogenic Variant Prevalence Among Latin American and US Hispanic Individuals Undergoing Testing for Hereditary Breast and Ovarian Cancer: A Cross‐Sectional Study,” JCO Global Oncology 8 (2022): e2200104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Yang X., Song H., Leslie G., et al., “Ovarian and Breast Cancer Risks Associated With Pathogenic Variants in RAD51C and RAD51D,” Journal of the National Cancer Institute 112, no. 12 (2020): 1242–1250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Miwa M., Kitagawa M., Asami Y., et al., “Prevalence and Outcomes of Germline Pathogenic Variants of Homologous Recombination Repair Genes in Ovarian Cancer,” Cancer Science 115, no. 12 (2024): 3952–3962. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Yao H., Li N., and Yuan H., “Clinical Characteristics and Survival Analysis of Chinese Ovarian Cancer Patients With RAD51D Germline Mutations,” BMC Cancer 22, no. 1 (2022): 1337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Ryu J. M., Choi H. J., Kim I., et al., “Prevalence and Oncologic Outcomes of BRCA 1/2 Mutations in Unselected Triple‐Negative Breast Cancer Patients in Korea,” Breast Cancer Research and Treatment 173, no. 2 (2019): 385–395. [DOI] [PubMed] [Google Scholar]
- 30. Robson M. E., Bradbury A. R., Arun B., et al., “American Society of Clinical Oncology Policy Statement Update: Genetic and Genomic Testing for Cancer Susceptibility,” Journal of Clinical Oncology 33, no. 31 (2015): 3660–3667. [DOI] [PubMed] [Google Scholar]
- 31. Tung N. M., Boughey J. C., Pierce L. J., et al., “Management of Hereditary Breast Cancer: American Society of Clinical Oncology, American Society for Radiation Oncology, and Society of Surgical Oncology Guideline,” Journal of Clinical Oncology 38, no. 18 (2020): 2080–2106. [DOI] [PubMed] [Google Scholar]
- 32. Schmidlen T. J., Bristow S. L., Hatchell K. E., Esplin E. D., Nussbaum R. L., and Haverfield E. V., “The Impact of Proband Indication for Genetic Testing on the Uptake of Cascade Testing Among Relatives,” Frontiers in Genetics 13 (2022): 867226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Samadder N. J., Riegert‐Johnson D., Boardman L., et al., “Comparison of Universal Genetic Testing vs Guideline‐Directed Targeted Testing for Patients With Hereditary Cancer Syndrome,” JAMA Oncology 7, no. 2 (2021): 230–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Evans D. G., Green K., Burghel G. J., et al., “Cascade Screening in HBOC and Lynch Syndrome: Guidelines and Procedures in a UK Centre,” Familial Cancer 23, no. 2 (2024): 187–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Doc S1: Supplementary Methods.
Figure S1: Lollipop plots of the germline pathogenic/likely pathogenic variants in the top 10 frequently detected genes. Lollipop plots are presented for the top 10 genes with the highest number of pathogenic variants generated by cBioportal (http://www.cbioportal.org/): (A) BRCA1, (B) BRCA2, (C) RAD51D, (D) ATM, (E) PALB2, (F) MUTYH, (G) CHEK2, (H) MSH2, (I) BRIP1, and (J) MLH1. The x‐axis reflects the number of amino acid residues, and the y‐axis shows the total number of patients with each pathogenic/likely pathogenic (P/LP) variant.
Figure S2: Heatmap of pathogenic/likely pathogenic variant frequencies by cancer type and gene. Frequencies (%) were calculated using the number of patients whose testing panel included each gene as the denominator (panel‐specific gene denominators are provided in Table S5). Genes are ordered by penetrance (high, moderate, low/uncertain) and colored by functional classification. Color intensity within cells reflects variant frequency. Empty cells indicate no P/LP variant detected or gene not included in the testing panel. GPV, germline P/LP variant; P/LP, pathogenic/likely pathogenic; PTs, patients.
Figure S3: Pedigrees of two patients with reclassified pathogenic/likely pathogenic variants. (A) The family pedigree of patient PT05 was constructed during the follow‐up period. (B) The family pedigree of PT28 was obtained at the time of genetic testing.
Figure S4: Distribution of patients and pathogenic/likely pathogenic variant frequency before the revision of reimbursement criteria for BRCA1/2 testing. BC, breast cancer; FH, family history; GPV, germline P/LP variants; HRR, homologous recombination repair; MMR, mismatch repair; OC, ovarian cancer; P/LP, pathogenic/likely pathogenic; PTs, patients.
Figure S5: Cascade genetic testing distribution among family members by relationship type. FDR, first‐degree relatives; SDR, second‐degree relatives; TDR, third‐degree relatives.
Figure S6: Cumulative probability of any post‐test clinical action over time, by gene group. Curves show the cumulative probability of receiving any documented clinical action, estimated as one minus the Kaplan–Meier survival function, with the date of germline testing as the time origin and deaths treated as censoring events. Estimated cumulative probabilities at 1, 3, and 5 years were 75.0%, 83.9%, and 87.0% for BRCA1/2; 50.6%, 57.0%, and 64.3% for non‐BRCA HRR; 45.9%, 45.9%, and 45.9% for MMR; and 63.4%, 68.6%, and 68.6% for Others (overall, 70.0%, 78.5%, and 82.2%). The difference across gene groups was significant by the log‐rank test (p < 0.001). HRR, homologous recombination repair; MMR, mismatch repair.
Table S1: Insurance reimbursement criteria for BRCA1/2 testing before and after July 2020.
Table S2: Germline tests performed during the study period and their gene composition.
Table S3: Definitions of analysis populations and clinical action criteria for five actionability domains.
Table S4: Variant of uncertain significance (VUS) rate by testing panel. Proportion of patients with a VUS‐only result (a VUS but no P/LP variant), by testing panel; the multigene panel testing (MGPT) tiers are further broken down by panel size.
Table S5: Gene‐specific testing denominators and pathogenic/likely pathogenic (P/LP) variant frequencies. For each gene, the number of P/LP variant carriers (patients with P/LP variants in more than one gene are counted once for each gene), the panel‐coverage‐adjusted number of patients tested, and the resulting frequency.
Table S6: Clinical characteristics of patients with variants reclassified from variants of uncertain significance to pathogenic or likely pathogenic variants (n = 49).
Table S7: Germline P/LP variant frequencies compared with international studies, overall and stratified by cancer type. Values are carriers, n (%), verified against the original article or supplement. Frequencies use each study's own denominator (many use gene‐specific, panel‐coverage‐adjusted denominators; our cohort likewise uses panel‐adjusted denominators, with BRCA1 and BRCA2 calculated over the full (sub)cohort and other genes over NGS‐panel‐tested patients, counting each gene once per patient). Cancer‐type composition of the pan‐cancer cohorts (percentage of the cohort affected by each cancer type; patients may have more than one cancer): This study, breast 56.1% and ovarian 37.8%; Susswein 2016, breast 52.4% and ovarian 8.5% (a further 25.2% were unaffected individuals tested because of family history); LaDuca 2020, breast 54.4% and ovarian 7.6%; Bychkovsky 2022 (a cohort of patients with multiple primary cancers), breast 68.1%, ovarian 25.1%, and endometrial 21.2%; Cheo 2025, breast 53.2% and ovarian 11.4%.
Table S8: PARP inhibitor uptake stratified by cancer type and gene group. PARP inhibitor uptake by cancer type and gene group among P/LP variant carriers with each qualifying cancer. BRCA1/2 carriers with an approved cancer indication and non‐BRCA HRR carriers with the same cancers are shown. Patients with more than one qualifying cancer are counted in each applicable row.
Data Availability Statement
All data generated or analyzed during this study are available from the corresponding author upon reasonable request within the range of IRB approval. In addition, individual‐level germline variant data are contributed to international hereditary cancer research consortia, including the Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA) and the Evidence‐based Network for the Interpretation of Germline Mutant Alleles (ENIGMA), in accordance with institutional and applicable consortium data‐governance requirements.