Prevalence and predictors of germline BRCA1 and BRCA2 mutations among young patients with breast cancer in Jordan

BRCA1 and BRCA2 mutations are not uncommon in breast cancer patients. Western studies show that such mutations are more prevalent among younger patients. This study evaluates the prevalence of germline mutations in BRCA1 and BRCA2 among breast cancer patients diagnosed at age 40 or younger in Jordan. Blood samples of patients with breast cancer diagnosed at age 40 years or younger were obtained for DNA extraction and BRCA sequencing. Mutations were classified as benign/likely benign (non-carrier), pathogenic/likely pathogenic variant (carrier) and variant of uncertain significance (VUS). Genetic testing and counseling were completed on 616 eligible patients. Among the whole group, 75 (12.2%) had pathogenic or likely pathogenic variants; two of the BRCA2 mutations were novel. In multivariate analysis, triple-negative disease (Odd Ratio [OR]: 5.37; 95% CI 2.88–10.02, P < 0.0001), breast cancer in ≥ 2 family members (OR: 4.44; 95% CI 2.52–7.84, P < 0.0001), and a personal history ≥ 2 primary breast cancers (OR: 3.43; 95% CI 1.62–7.24, P = 0.001) were associated with higher mutation rates. In conclusion, among young Jordanian patients with breast cancer, mutation rates are significantly higher in patients with triple-negative disease, personal history of breast cancer and those with two or more close relatives with breast cancer.

www.nature.com/scientificreports/ and BRCA2 are currently detected. Such mutations include ATM, CDH1, CHEK2, PALB2, PTEN, STK11, and TP53 [10][11][12] . Studies had shown that both BRCA1 and BRCA2 mutations are associated with a high penetrance rate. The cumulative risk estimates for developing breast cancer by age 80 are 70-90% for carriers of BRCA1 pathogenic variants and 60-70% for BRCA2 carriers. The cumulative risk for developing ovarian cancer is a little lower; 40-50% for BRCA1 carriers and around 20% for BRCA2 carriers 13,14 . Additionally, the risk of contralateral breast cancer, 20 years after the initial diagnosis, is 40% and 26% for BRCA1 and BRCA2 mutation carriers, respectively 15 .
Because of this high penetrance rate and its associated significant consequences, identifying such mutations should be actively sought in high-risk patients identified by international guidelines 13 . Risk-reduction interventions, like bilateral mastectomies and salpingo-oophorectomies, are highly recommended for patients with BRCA1 or BRCA2 pathogenic variant carriers, especially so among younger patients.
In addition to its value in preventing breast and ovarian cancers, identification of mutation carriers may have therapeutic importance in patients with breast cancer, too. Recent data had suggested that patients with advanced-stage breast cancer associated with BRCA1 or BRCA2 mutations may benefit from PARP (poly ADP ribose polymerase) inhibitors like olaparib and talazoparib; both are currently approved for such situation [16][17][18] .
Data related to hereditary breast cancer among Arabs, particularly Jordanians, is scarce. Reported pathogenic variant carrier rates vary [19][20][21][22] . It is unknown if inherited germline mutations account for earlier age at breast cancer diagnosis in our region. We recently reported our experience on 517 high risk patients treated and followed at our institution; a total of 72 (13.9%) patients had pathogenic or likely pathogenic BRCA1 or BRCA2 mutations, while 53 (10.3%) others had a variant of uncertain significance (VUS) 23 .
The diagnosis of breast cancer in young women and its possible genetic implications have potentially serious consequences for patients and their family members, too. Physicians and genetic counselors can help navigate such complex medical and psychosocial issues. In this paper, we aim to study the prevalence and pattern of germline BRCA1 and BRCA2 mutations among a group of young Jordanian patients with breast cancer thought to be at higher risk for such mutations.

Methods
Jordanian breast cancer patients aged 40 years or younger at the time of diagnosis were invited for BRCA1 and BRCA2 testing as part of our clinical practice guidelines. Family history or personal history of breast, or other cancers, were not mandated for eligibility. All patients had their diagnosis, treatment, and follow-up at our center.
Eligible patients were identified at their first encounter by a medical oncologist or following the weekly breast multidisciplinary team meetings. Eligible patients who consented to be tested were then referred to a specialized genetic counseling clinic where all potential psychosocial and clinical consequences of positive test results were discussed.
As recommended by international guidelines 15 , BRCA1 and BRCA2 variants were classified as benign/likely benign (non-carrier), pathogenic/likely pathogenic (carrier) and VUS. Clinical details and pathological characteristics of the tumors were reviewed. Additionally, a detailed 3-generation family history was also obtained. Estrogen (ER) or progesterone receptors (PR) were positive if tumor cell nuclei staining is ≥ 1%. Human epidermal growth factor receptor-2 (HER-2) was tested using a standardized immune histochemical staining (IHC), and tumor cells were considered negative with scores of 0 or + 1, and positive for those with + 3 scores. Fluorescence in situ hybridization (FISH) was performed for equivocal samples with + 2 scores. Triple-negative tumors are those which tested negative for ER, PR, and HER-2.
Blood samples were obtained for DNA extraction, full-gene sequencing, and deletion/duplication analysis for BRCA1 and BRCA2 using next-generation sequencing technology (NGS) and/or Multiplex Ligation-dependent Probe Amplification (MLPA) analysis were performed at three reference labs: Myriad Genetics laboratory (Salt Lake City, UT), Leeds Cancer Center (Leeds, United Kingdom) and invitae (San Francisco, CA).
Our study was carried out in accordance with the code of ethics of the World Medical Association (Declaration of Helsinki) and was approved by the Institutional Review Board (IRB) at King Hussein Cancer Center. All patients signed informed consent.

Statistical analysis.
Patients' clinical and pathologic characteristics were collected, tabulated, and described by ranges, medians, or percentages. Relatives diagnosed with breast cancer and tested after the family's index case were not enrolled and were excluded from the analysis. Chi-square tests were used to compare the proportion of BRCA1 and BRCA2 pathogenic/likely pathogenic variant carriers according to age (≤ 30 versus > 30), triple-negative status, and family history. Multivariate analysis using a logistic regression model was performed. Odds ratios and their related 95% confidence intervals (CI) were calculated. P-value ≤ 0.05 was considered significant. Analyses were conducted using Minitab Statistical Software version 18 Table S1). Patients with at least two breast cancer primaries (n = 48) had a significantly high mutation rate (n = 8, 29.2%). Table 2 presents mutation rates according to different categories.
Family history. The majority of the patients enrolled (n = 499, 81.0%) had a positive family history of breast cancer in first-, second-or third-degree family members. Women with two or more close relatives diagnosed at any age with breast cancer (Group-A, n = 97) had the highest mutation rate (n = 29, 30.0%). In contrast, women with one or more family members diagnosed with breast cancer before the age of 50 years (Group-B, n = 153) had a mutation rate of 17.6%, P = 0.011. The mutation rate was lower (15.1%, P = 0.001) among women with one or more family members diagnosed at any age (Group-C, n = 305), Fig. 1.
Age at diagnosis. We studied the contribution of age to mutation rate in two ways. First, we compared mutation rates across the median age of our cohort; BRCA1 or BRCA2 pathogenic variants were reported in 14.7% of 341 patients aged ≤ 35 years, compared to 9.1% in 275 patients older than 35 years, P = 0.017. Second, we compared mutation rates across two age groups: < 30 years and those aged 31-40 years; mutation rates were 17.4% and 10.9% (P = 0.05), respectively, Fig. 2. Triple-negative disease. Patients with triple-negative disease (n = 69) had significantly higher rates (n = 23, 33.3%) compared to 9.5% among non-triple negative patients, P < 0.001. Most of the pathogenic variants     (Tables 3 and 4). To our knowledge, two mutations in BRCA2 (c.6193C > T in exon 11 and c.1013del in exon 10) have not been reported previously in any database. Additionally, five unrelated females in our cohort were found to harbor two concomitant mutations in BRCA2 exon11 (c.2254_2257del) and (c.5351dup), simultaneously (Table 4). These two mutations appeared separately in a very limited number of studies [24][25][26] . Except for mutations c.1233dup and c.9257-1G>A/IVS24-1G>A, for which two family members were tested for each, all other variants have been detected in different families. Nineteen (25.3%) of the mutations detected in our patients were either (c.2254_2257del) or (exon 5-11 duplication); both in BRCA2 gene and were detected in 11 and 8 different patients, respectively.

Discussion
Our study confirms that younger patients are at a higher risk of harboring pathogenic or likely pathogenic mutations and such risk is higher for patients younger than 30 years at the time of breast cancer diagnosis. However, differences in mutation rates between patients above or below 40 years is less obvious. In one of our previous studies, the mutation rate among 333 younger patients (≤ 40 years) was 13.2% compared to 15.2% among 184 older ones, P = 0.53 23 . Our findings of two novel mutations that have been detected in our database as well as a higher frequency of certain mutations like (c.2254_2257del) and (exon 5-11 duplication) will probably have an important consequence for the genetic testing of BRCA genes in Jordan where consanguineous marriage is relatively common. In one study, researchers reviewed published and unpublished data to identify population-specific founder BRCA pathogenic sequence variants (PSVs) in Middle East, North Africa, and Southern Europe; 232 PSVs in BRCA1 and 239 in BRCA2 were identified 27 .
It is also worth highlighting that our study identifies three risk factors, the presence of any of which in younger patients increases the pathogenic variant carrier rate to almost one in three tested patients. These include, patients with triple-negative disease, women with at least two breast primaries, and those with a family history of breast cancer in two or more close relatives diagnosed at any age. Such findings might help simplify our efforts to educate both patients and health care providers about the importance of genetic testing and counseling for such patients.
Our VUS rate (9.3%) is higher than what had been reported among Caucasian patients 28 . This rate will probably go even higher with the wider implementation of multi-gene testing. Several studies had shown higher VUS rates among African-Americans, Hispanics and patients of Ashkenazi-Jewish descent [29][30][31] . www.nature.com/scientificreports/ We have built a good experience in dealing with patients before and after testing. Ensuring confidentiality was never a problem in our current daily practice. Very few patients refused genetic testing and counseling because of their fear of stigmatization and labeling. However, prophylactic bilateral mastectomies and oophorectomies with reconstructive surgery can be a challenge. Studies addressing the psychosocial consequences of pathogenic variants especially among younger patients in our region, are highly needed.
Though our study represents a single-center, we believe it reflects the whole country as our institution treats most of the country's breast cancer cases. However, our study is not without limitations; issues related to psychosocial aspects related to pathogenic variant carrier state, risk-reduction surgeries, fertility-related issues, and outcome of family members at-risk of index cases need to be followed and addressed.

Conclusions
BRCA1 and BRCA2 mutation rates among patients 40 years or younger are relatively high but not necessarily higher than older patients. However, personal and family risk factors can identify subgroups of younger patients with much higher mutation rates.

Data availability
Data will not be available online as it might contain sensitive information. Data will be available through the corresponding author on reasonable requests.