Comprehensive clinical characterization of patients with BRCA1: c.5017_5019del germline variant

It is widely known that cancer is a disease in which abnormal cells proliferate uncontrolled due to changes in DNA. The causative genes for specific cancers have been identified, and the survival rate has risen due to advances in cancer screening, treatment, and prevention. BRCA1 and BRCA2 discovered in the 1990s normally belong to the DNA repair genes, which assume a regulatory function in the cell cycle [1]. These genes intrinsic to all human beings are mutated in some individuals and are representative of the causative genes of hereditary breast and ovarian cancer syndrome (HBOC) [2].

The germline pathogenic variant (PV)/likely pathogenic variant (LPV) in BRCA1/2 has become one of the most important keys in treatment, surveillance, whether patients choose risk-reducing surgery, and the counseling of the HBOC family members [3, 4]. Because poly ADP ribose polymerase (PARP)-inhibitors and prophylactic treatments, such as risk-reducing agents, risk-reducing bilateral mastectomy, and risk-reducing salpingo-oophorectomy, are widely known to be effective in BRCA1/2-positive patients, the BRCA1/2 test is considered essential for high-risk breast cancer patients and ovarian cancer patients [5, 6, 7].

From this point of view, the need for genetic tests, especially for BRCA1/2 genes, has increased. In addition, with advances in genetic testing methods such as next generation sequencing (NGS), more tests can be performed. Along with the increased number of tests, new deleterious genetic mutations have been discovered and the numbers of variants of uncertain significance (VUSs) have also increased [8]. Although previous studies showed that the frequency of VUS in BRCA1/2 differed depending upon the study population [9], generally a frequency of 5%–10% VUSs has been reported [10, 11]. Some VUSs were reclassified as LPVs or likely benign variants based on accumulated evidence [12, 13]. The BRCA1:c.5017_5019del variant was classified as a VUS but recently, was reclassified as an LPV [14]. Here, we present the clinicopathological characteristics of patients with BRCA1:c.5017_5019del and basic information for the interpretation of this variant.

BRCA1/2 gene tests have been performed more frequently with the development of testing methods and the importance of treating HBOC. Accordingly, more novel variants have been discovered and the detection of VUSs is also increasing. Since the BRCA1/2 test is essential for not only the diagnosis but also the treatment of HBOC, the misinterpretation of VUSs can cause significant damage to the carriers, requiring continuous efforts to reclassify VUSs [20, 21]. When a VUS is reclassified as a PV/LPV, genetic counseling or appropriate management including prophylactic strategies and treatment should be discussed with the patients and their relatives. Olaparib, a PARP inhibitor, has been approved for the treatment of patients with BRCA1/2 mutated advanced ovarian cancer [22]. Recently, it was proven to be effective as adjuvant therapy after the completion of local treatment and chemotherapy for patients with HER2-negative early breast cancer and BRCA1/2 PV or LPVs [23]. Here, we report the characteristics of patients with a specific BRCA1 gene variant (c.5017_5019del) from multiple centers in Korea.

The frequencies of VUSs in BRCA1/2 have been reported to be 5%–10% on average, but Myriad Genetics Inc. (Salt Lake City, UT, USA) reported that this decreased to 2.1% using accumulated data [24]. Therefore, studies for the reclassification of VUSs in BRCA1/2 have been conducted in Korea. In early studies by Park et al. [25] and Korean Hereditary Breast Cancer (KOHBRA) study [26], which included breast cancer patients, the BRCA1:c.5017_5019del variant was still reported as a VUS. However, So et al. [12] reclassified this variant into an LPV, which included both breast and ovarian cancer patients, and this result could be understood as consistent with a study by Zuntini et al. [14]. That study reported 14 families including 20 cases of breast cancer and 23 cases of ovarian cancer in Italy. Here, we present 17 patients from 15 families identified as having the BRCA1:c.5017_5019del variant and a total of 19 cases of breast cancer and 14 cases of ovarian cancer in these families. This was the largest number of cases for a specific variant depending on the multicenter study of breast cancer patients in Korea. Ten breast cancer patients (58.8%) with this variant showed similar clinical characteristics to those with BRCA1 mutation-related breast cancer, including 8 patients (80.0%) with a rich family history of breast or ovarian cancer, 7 patients (70.0%) with a cancer diagnosis prior to 50 years of age, and 5 patients (50.0%) categorized into the TNBC subgroup. Seven patients (41.2%) were diagnosed with ovarian cancer, and the ratio of breast-to-ovarian cancer was 1.3:1. Also, ovarian cancer patients with this variant showed strong family histories of breast and/or ovarian cancer.

After the 2015 AMP-ACMG guidelines were published, genetic interpretation became more systematic. Even when deleterious variants are highly suspected, all variants should be interpreted based on the guidelines, and familial analysis and functional studies like messenger RNA studies should be performed if needed. ClinGen Sequence Variant Interpretation Workgroup refined the AMP-ACMG guidelines and pathogenic criteria (PVS1) were suggested based on the available evidence for each variant type [27]. According to the PVS1 strength level [27], these criteria were divided into PVS1, PVS1_Strong, PVS1_Moderate, and PVS1_Supporting. An in-frame deletion, especially a single amino acid deletion should be applied carefully not to be an overestimation. Here, the BRCA1:c.5017_5019del variant was an amino acid deletion variant and we could only add the evidence for protein length (PM4) added to PM2 evidence for population frequency. However, detection in patients with the same phenotypes should be accessed for other evidences. PS4 can be used for typical case-control studies when the relative risk or odds ratio is >5.0, and the confidence interval (CI) does not include 1.0 and has been calculated to assess whether a variant is likely to be associated with a particular phenotype. Table 4 shows evidence codes according to the 2015 ACMP and AMP guidelines [16]. Also, PP4, which represents patient phenotypes or family history for highly specific phenotypes, can be considered as other evidence. Another consideration in genetic interpretation is the interpretation of in-frame deletions, especially, single amino acid deletion variants. There were 4 kinds of single amino acid deletion variants in BRCA1/2 genes, besides BRCA1:c.5017_5019del, which were found in Korean patients during this period. Those were 1 variant in the BRCA1 gene, c.3327_3329del (p.Lys1110del) and 3 variants in the BRCA2 gene, c.644_646del (p.Glu215del), c.7981_7983del (p.Asp2661del), and c.8437_8439del (p.Gly2813del). All those 4 variants were reported as VUS based on the 2015 AMP-ACMG guidelines. We applied the evidence for allele frequencies and protein length change to these deletions. We plan to continuously compare and analyze the clinical data to clarify the clinical significance of each variant in the future. In addition, we will continue to consider the genetic domain and functional studies of the variants.

Table 4
2015 ACMP/AMP evidence codes for classifying sequence variants organized data type and strength

• Click for larger image

Currently, germline testing for the BRCA1/2 genes is performed in only indicated patients such as those with a family history, certain age, etc. The optimal management for patients and carriers who have germline variants depends upon accurate risk estimation and surveillance. There are well-known overall risks for breast and ovarian cancers until the age of 80 years. Moreover, the annual incidence and cumulative risk of breast, ovarian and contralateral breast cancer has also been reported [28].

One limitation of the study was the lack of family studies of BRCA1/2 tests and few familial history updates due to the limitations of the retrospective study. In the same context, there was a lack of information on ovarian cancer patients, since the study was conducted in multicenter breast cancer departments. Since several studies that previously suggested the reclassification of this variant into a PV/LPV included gynecological cancer patients [12, 14, 29] and this study had a large proportion of ovarian cancer patients, a multicenter study including gynecological cancer patients should be conducted. An interesting point was the location of the variant in what is known as the breast cancer cluster region (BCCR), but it was detected more often in ovarian cancer patients. A few points should be considered to understand the study results. First, there was a very small number of patients from which to draw statistical judgments on breast and ovarian cancer patients. Moreover, regions such as the BCCR and ovarian cancer cluster region were defined based on the ratio of breast-to-ovarian cancer incidence and there were exceptions even within the same cluster. The broad region that includes this variant (c.5017_5019del) was defined as the BCCR, but the specific region, which was defined as bin (c.4946_5123), showed a hazard ratio of 1.07 (breast-to-ovarian) (95% CI, 0.83–1.39) [30]. Another thing to consider is that this variant was previously reported as a VUS and excluded to define these regions. Therefore, we expect that additional studies will be conducted and the clinical information of the patients with this variant will be accumulated.

Another limitation is that we could not perform functional studies to make up the evidence based on the 2015 AMP-ACMG guideline for the corresponding variant. There are 8 categories including population data, computational/predictive data, functional data, segregation data, de novo data, allelic data, other databases, and other data. Public data such as population data and computational/predictive data could be used in interpreting variants; however, functional data is not easy to be used. Moreover, in this study, we described only a variant and we did not analyze other VUSs. In the future, as the BRCA1/2 analysis continues, the authors expect and make efforts so that the interpretation of the variants could be in-depth, including a functional study, with the patients’ clinical conditions.

In conclusion, we present clinical evidence for the reclassification of BRCA1:c.5017_5019del as an LPV according to ACMG standards and guidelines. Because the BRCA1:c.5017_5019del variant has been classified as a VUS, probands have not been given the option of prophylactic treatment and family testing has not been recommended. However, when reclassified as an LPV, prophylactic treatment and medical surveillance may be provided to the probands, family testing recommended, and appropriate genetic counseling provided to the family.