Breast cancer (BC) is the most commonly diagnosed cancer in women, with an estimated 2.3 million new cases (11.7%) in Global Cancer Statistics in 2020 (1). A proportion of 5% of BC cases are due to germline pathogenic (P) or likely pathogenic (LP) variants in cancer predisposition genes, and significant progress has been made in the identification of inherited genetic factors underlying hereditary cancers (2). BRCA1 and BRCA2 are the most commonly mutated BC susceptibility genes that convey a high-risk of breast and ovarian cancer and represent 25%–28% of hereditary BC (3). In addition, other germline pathogenic variants in genes that confer moderate to high-risk in carriers may also explain other mechanisms of hereditary BC (4).

In Brazil, there is a high prevalence of Li-Fraumeni syndrome due to a founder mutation in TP53 gene, c.1010; p.(Arg337His), that is present in 0.3% of all South and Southeastern Brazilian population (5–7). Age at the time of tumor diagnosis and family history are the best clinical predictors for the identification of carriers (8). Studies have demonstrated that the inclusion of morphologic and immunohistopathologic BC characteristics in models of risk estimation also improves the accuracy in predicting the BRCA1/BRCA2 status (9).

The non-BRCA1– and BRCA2-associated hereditary BCs are a heterogeneous group of tumors, with different phenotypes yet to be established. Pathogenic variants in TP53, PTEN, CDH1, STK11, and PALB2 have also been shown to confer high-risk for BC. ATM, CHEK2, and BRIP1, among others, are known to confer moderate risk (10). The clinical impact on BC susceptibility in NBN, PSM2, and other genes has been described in certain populations, but its impact is still unclear (11–13). The majority of high- and moderate-penetrance genes have an autosomal dominant pattern, in which single heterozygous pathogenic variants lead to cancer risk.

Because the completion of the Human Genome Project, advances in the genomic technology became largely available and the cost of next-generation sequencing (NGS) tests have decreased significantly. In the current context of molecular testing for hereditary cancer predisposition genes, frequently testing only one gene is not enough to detect what is the underlying germline pathogenic variant in a high-risk patient. Therefore, NGS multilane panels became the choice as a less expensive technique, especially if more genes need to be evaluated.

The use of larger NGS panels and the increased use of exome/genome sequencing presented a new challenge to geneticists; the detection of individuals who harbor two or more pathogenic variants in cancer predisposing genes or MINAS (multilocus inherited neoplasia allele syndrome) (14). There are few reports in the literature on the co-occurrence and the cancer phenotypes associated with double heterozygous (DH) pathogenetic variants in cancer predisposing genes and if they combinations appear to have synergistic effects in cancer risk. The co-occurrence of more than one germline pathogenic variant in BC genes is a very rare condition, and its effects on cancer risk in carriers are still unknown.

As a result of the expansion toward larger gene panels and more individuals being tested, the identification of DH events is becoming more frequent. Currently, there are only a few reports in medical literature addressing the detection of DH (14, 18–22). It is still unknown whether the presence of a DH in high- and moderate-penetrance genes would have an additive or synergic effect on lifetime risk of tumor development. In this study, we conducted a large cancer predisposition gene panel in a cohost of early onset or familial BC. A total of 19.6% (226 of 1,156) of women with BC carried at least one pathogenic variant in a moderate- or high-penetrance BC-risk genes. DH variants in BC-moderate to high-risk genes were observed in a subset of 14 patients, representing 1.2% (14 of 1,156) of the BC cohort. Most of the DH carriers carried at least one pathogenic variant in BRCA1 or BRCA2 genes (10 of 14, 71.4%). A systematic review recently reported 385 individuals carrying pathogenic variants in two or more of 94 cancer susceptibility genes (14). Although more cancer-risk genes were considered and their cohort was not limited to BC, 74.5% (287 of 385) were also carriers of variants in BRCA1 or BRCA2 genes.

Pathogenic variants in BRCA1, BRCA2, and TP53 were the most commonly found in this cohort, corresponding to 69.4% of all pathogenic variants (PV). PVs in BRCA1 were detected in 20.8% (47 of 226) of all carriers, BRCA2 in 18.1% (41 of 226), and TP53 in 30.5% (69 of 226). A population-based study conducted in 32,247 women with BC screened for a panel of cancer predisposing genes found the majority of variants to occur in BRCA1 and BRCA2 genes. Nevertheless, limited numbers of women with pathogenic variants in TP53 did not allow associations with BC in this large international cohort (PMID: 33471974). The high frequency of patients with BC carrying TP53 variants in our cohort is probably due to the Brazilian founder variant c.1010G>A (p.Arg337His) present in 0.2%–0.3% of the South and Southeastern Brazilian population (6, 7). In fact, c.1010G>A (p.Arg337His) comprised 79.7% (55 of 69) of all TP53 variants detected in our cohort. This study was conducted is the main reference center for Li-Fraumeni syndrome in the country, and this may have caused a recruitment bias.

In this cohort, carriers of BRCA1 and TP53 variants had BC at an early age of onset, whereas ATM variant carriers were more likely to be diagnosed later in life, in agreement with prior reports (23–25). Patients with BC who carry a germline pathogenic variant in BRCA1 gene mutation are more likely to have triple-negative BC (TNBC) (PMID: 19298662). Previous studies have suggested that BC in TP53 pathogenic variant carriers is highly likely to be HER2 positive (PMID: 20805372).

The age of BC diagnosis in the DH carriers was similar to those with a high- or moderate-penetrance pathogenic variant in one gene. The BC onset in DH was comparable to reports of hereditary BC related to BRCA1 and TP53 germline variant carriers (26). Bilateral BC was found mostly in carriers of at least one BRCA1 and/or BRCA2 pathogenic variant (28%). In previous reports in the literature, the occurrence of bilateral BC in BRCA1 and BRCA2 carriers may be as high as 53% (27). Although TP53 germline mutation carriers are at a higher risk of developing bilateral BC (28), carriers of DH with at least one TP53 variant in this study did not present bilateral BC. Although there was a higher prevalence of bilateral BC in DH compared with carriers of a single pathogenic variant (35.7% vs. 18.9%, respectively), there was no putative additive effect of DH (p = 0.161).

The advances in DNA sequencing technologies and the decreased costs of genetic assays lead to expressive increase in the number of genes incorporated in genetic testing for BC risk prediction (20). Most gene panels include high-penetrance genes, such as BRCA1, BRCA2, CDH1, PALB2, PTEN, STK11, and TP53, and moderate-penetrance genes, such as ATM, BRIP1, CHEK2, FANCD2, RAD51C, NBN, and PMS2, which also associates with lifetime BC risk (29). Although the clinical impact of NBN and PSM2 on BC susceptibility has remained uncertain, their introduction in clinical genetic testing for suspected hereditary BC has been recently supported (13, 30). NBN was shown to increase about three-fold the risk in BC (odds ratio, 3.1; 95% CI, 1.4–6.6) and indicates that the 657del5 deletion [c.657_661del (p.Lys219fs)] and perhaps the R215W substitution contribute to inherited BC susceptibility (11). We identified two cases harboring NBN pathogenic variants: the aforementioned c.657_661del (p.Lys219fs) and c.1142del (p.Pro381fs). Pathogenic variants in PMS2 and BC risk remain a controversial association. In a large cohort of women with Lynch syndrome identified through panel testing, there was no evidence for increased risk of BC compared with the general US population (12). However, when evaluating by gene, the age-standardized BC risks for PMS2 (standard incidence ratio = 2.92; 95% CI, 2.17–3.92) were associated with a statistically significant risk for BC (30). In our cohort, we described one patient who carried a BRCA2 (c.2516dupA) and PMS2 (c.2182_2184delinsG) and who developed a BC at age 49.

In a cohort of 8,162 breast/ovarian cancer families from the German Consortium for Hereditary Breast and Ovarian Cancer, eight female DH patients for BRCA1 and BRCA2 mutations were described and analyzed for their phenotypic features (22). The authors did not find association of DHs with younger age at BC diagnosis compared with single heterozygous index patients. Nevertheless, it is suggested DH patients had a more severe disease than their female relatives carrying a single BRCA1/2 mutation. Likewise, a study including 5,391 Slavic women carriers of BRCA1, CHEK2, NBN, ATM, and BLM variants identified 17 patients with DH BC. No differences in age of onset and risk of bilateral BC were found between single variant and DH carriers (18).

We identified three carriers of ATM who also had either a BRCA1 or BRCA2 variant. A BRCA1/ATM carrier (case 5) developed BC at a young age (33 years old) and BRCA2/ATM carriers (cases 9 and 10, FY0073) developed BC at 32 and 45 years, respectively. Case 10 developed bilateral BC (a second BC was diagnosed at 52 years). There are no reports in the literature indicating an additive effect of both germline variants in female patients. An in vitro study was performed to evaluate the combinations of wild type and hemizygous genotypes for BRCA1 and ATM and its interference in cell transformation and apoptosis induced by radiation (31). The results indicated that double heterozygosity for ATM and BRCA1 leads to an additive effect on tumor development than single heterozygosity. In a murine model of induced BRCA1 and ATM DH loss, a synergistic effect in mammary gland tumorigenesis and tumor aggressiveness was observed. The results suggested that these DH combinations lead to a newly characterized developmental defect during glandular maturation (32). As only a few individuals in this study were found to carry DH in BRCA1/ATM or BRCA2/ATM, the same synergistic effect could not be confirmed.

In our cohort, three individuals who carried BRCA1 and PALB2 pathogenic variants (cases 1, 2, and 3) developed BC at young age (32, 40, and 32, respectively). Case 3 developed a second BC at 37 years. Multiple combinations of BRCA1/2, PALB2, and TP53 variants were also tested in a murine mammary gland model using parallel conditional knockout (33). Curiously, the combined loss of BRCA1 and PALB2 was demonstrated to increase reactive oxygen species and apoptosis, which could be responsible for a delayed tumorigenesis in this condition. Carriers of multiple combinations of BRCA1, PALB2, and TP53 did not present any differences than what is expected of carriers of only one pathogenic variant in any of the three genes. Although in vitro and in vivo models are useful to predict clinical outcomes, cases and series reports in humans have shown inconclusive results (34–36).

Curiously, we detected two patients with BC in our registry who carried three pathogenic variants in high-penetrance BC-risk genes. The presence of three pathogenic germline variants in the same individual is an extremely rare event, and there are only a few case reports in the literature (19, 21). In this cohort, two sisters (identified as patients 1 and 2) tested positive for concomitant BRCA1 (c.5266dupC), PALB2 (c.3771C>T), and TP53 (c.1010G>A) variants. They both had BC at early ages (≤40 years). Further molecular studies on family members were not performed, although multiple cancers of the Li-Fraumeni and BRCA1/BRCA2-spectrum were referred in the maternal family history.

However, we believe that DH carriers may benefit from more intensive surveillance programs/follow-up care and family members should perform cascade testing. Risk-reducing procedures must be well integrated and individualized in these populations.

In conclusion, the multi-gene hereditary cancer panel testing comes with the challenge of counseling and managing patients who may carry variants in multiple germline cancer susceptibility genes. Functional studies should be conducted to improve our comprehension of how germline variant combinations may interact and impact in the tumor development, reflecting on the BC onset and increased risk for multiple tumors. Larger studies are needed to estimate whether the presence of DH has a synergistic interaction or additive effect on cancer risk in BC. It is important to investigate these rare DH genotypes to provide better personalized clinical management for carriers and determine the benefit of risk-reducing procedures in these populations.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

This study was reviewed and approved by Research Ethics Committee (ID1747) and the National Research Ethics Committee (number 39512620.9.0000.5461). The patients/participants provided their written informed consent to participate in this study.

Concept and design: TM, MB-F, and MA. Acquisition, analysis, or interpretation of data: TM, MB-F, JP, and MA. Drafting of the manuscript: TM, MB-F, and MA. Critical revision of the manuscript for important intellectual content: TM, MB-F, JP, and MA. Supervision: MA. All authors have read, edited and approved the final manuscript, and agree to be accountable for the content of the work.

This work was supported by Stand Up to Cancer (SU2C - subaward number 0266-3190-4609).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.