The concept of a broader phenotypic spectrum for HBOC syndrome was first proposed in 1994 by Ford et al., who identified patients with P/LP BRCA1 variants at an increased risk for cancers beyond BC and OC, including colorectal and prostate cancers [10]. Large-scale studies have since supported this notion, showing that individuals with P/LP BRCA1 variants have an elevated risk of various cancers, particularly GC [6, 11, 12]. Variability in cancer presentation in HBOC syndrome has been linked to the specific gene involved (BRCA1 or BRCA2), the location and type of the variant within the gene [13, 14], ancestry, and the geographic distribution of carriers due to environmental factors or genetic ancestral backgrounds [14, 15].
Here, we present the phenotypic oncological spectrum of a large Colombian family carrying the BRCA1: c.3331_3334del, p.Gln1111Asnfs*5 (rs80357903) pathogenic variant in exon 10, which is predicted to undergo NMD (Nonsense Mediated Decay) mechanism. This variant corresponds to a founder mutation in the Andean region, originating from Iberia and introduced to Colombia during the American colonization [16–20]. It has been associated with the typical spectrum of cancers linked to P/LP BRCA1 variants, such as BC and OC. Notably, this variant has also been identified in families with GC and colorectal cancers [18, 19, 21, 22]. Torres et al. [18] reported five families segregating this variant, with three showing members with GC and two with colorectal cancer. Sahasrabudhe et al [21]. reported two cases with this variant and diffuse GC.
The association between BRCA genes and GC risk has been extensively studied (Table 2). The first population study to identify this association was conducted by the Breast Cancer Linkage Consortium in 1999, confirming a significant risk for BRCA2 carriers (RR 2.59, CI 1.46–4.61) [23]. Subsequent studies have supported these findings. Differences in cancer presentation by age have also been reported: at 50 years (0.15%, CI 0.04–0.39) and 70 years (1.88%, CI 1.05–3.12) [24]. Momozawa et al. (2022) highlighted significant odds ratios for both BRCA1 (OR 5.2, CI 2.6–10.5) and BRCA2 (OR 4.7, CI 1.8–30), noting different peak ages of cancer presentation for BRCA1 (30–39 years) and BRCA2 (60–69 years) [6]. The age of GC onset in P/LP BRCA variants carriers typically aligns with known patterns: women with BRCA1 variants have a higher risk of developing BC rapidly until about age 30–40, while BRCA2 diagnosis increases until about age 40–50 [25].
Table 2
Studies Reporting the Association between BRCA1/2 Mutations and Increased Risk of Gastric and Salivary Gland Cancers
Cancer Type | Year | Gene | Estimated Incidence | Estimated Risk | Risk Association | Clinical information | Ref |
Gastric | 1994 | BRCA1 | 1 observed case vs 0.76 expected cases | RR 1.11 (p > 0.05) | Not significant | - | Ford et al. [10] |
1999 | BRCA2 | - | RR 2.59 (CI 1.46–4.61) | Confirmed | - | Breast Cancer Linkage Consortium [23] |
2001 | BRCA1/2 | (BRCA1) 4.9%; (BRCA2) 1.8% | (BRCA1) RR 6.2 (CI 2.0–19); (BRCA2) RR 2.3 (CI 0.30–18) | Confirmed | - | Risch et al. [36] |
2004 | BRCA1/2 | 1.88% (CI 1.05–3.12%) vs 0.92% | - | Confirmed | Study for both genes, risk calculated by age: 50 years (0.15%, CI 0.04–0.39); 70 years (1.88%, CI 1.05–3.12). | Bermejo et al. [24] |
2012 | BRCA1/2 | Family GC incidence 24.7% vs 20.5% | RR 0.947 (CI 0.288–1.091) | Not significant | - | Noh et al. [37] |
2019 | BRCA1/2 | Family GC incidence 13.8% vs 7.4% | OR 1.666 (CI 1.183–2.345) | Confirmed | - | Kim et al. [11] |
2020 | BRCA1/2 | Family GC incidence 11.9% vs 5.5% | p > 0.001 | Confirmed | Squamous cell tumor | Sun et al. [12] |
2021 | BRCA1/2 | - | BRCA1: RR 1.70 (CI 0.93–3.09); BRCA2: RR 2.15 (CI 1.98–2.33) | Confirmed | Only BRCA2 | Lee et al. [38] |
2022 | BRCA1/2 | - | BRCA1: OR 5.2 (CI 2.6–10.5); BRCA2: OR 4.7 (CI 1.8–30) | Confirmed | For BRCA1, the most frequent age of gastric cancer presentation is 30–39 years, and for BRCA2 it is 60–69 years. | Momozawa et al. [6] |
Salivary gland | 2014 | BRCA1/2 | 3 out of 5754 (0.052%); incidence 0.003% | p < 0.001 | Confirmed | Non-Clinical Characterization | Shen et al. [8] |
2020 | BRCA1 | Case report | - | Suggested | Diagnosis at 55 years. Genetic testing identified: BRCA1: p.Ser157fs*1. | De Barros et al. [9] |
2021 | BRCA1 | Case Series (5 families). | - | Suggested | Only two cases (Diagnosis at 55 and 65 years) with genetic testing identified: BRCA1: p.Ser267Lysfs19 and BRCA2: p.Lys3161. | Ripamonti et al. [7] |
(-) No additional information reported. |
Interestingly, reported cases of GC onset for BRCA1: p.Gln1111Asnfs*5 carriers are usually above 50 years (range 50–70) [18]. In our family, GC was diagnosed in members ranging from 40 to 82 years, with a mean age of 60 years (Table 1). This suggests that this variant may be associated with GC at an advanced age, possibly related to a strong interaction between BRCA1 carriers and environmental factors.
The relationship between BRCA1 and environmental factors, particularly Helicobacter pylori infection, has been extensively studied. Usui et al. suggest that H. pylori infection can significantly increase GC risk, especially in individuals carrying P/LP variants in the Homologous Recombination Repair (HRR) pathway, which includes BRCA1 [26]. Families carrying P/LP variants in HRR genes often share the same H. pylori strains, potentially amplifying the risk [26]. This study found a significantly higher prevalence of BRCA1/2 and ATM variants in Japanese GC patients with H. pylori infection compared to those without infection (OR 20.25 vs. 5.76), suggesting an additive effect where H. pylori-induced genomic instability may further elevate GC risk in BRCA1 variants carriers. Identifying such high-risk populations could be crucial for early detection and prevention strategies.
Salivary gland tumors are considered rare entities, accounting for 3–10% of all head and neck tumors [27]. They are mostly sporadic and not typically related to hereditary cancer syndromes [28]. However, similarities between salivary glands and mammary glands have suggested possible similarities in their carcinogenesis. Both salivary and mammary glands are exocrine tissues, sharing a similar embryological origin as tubuloacinar exocrine glands, and their tumors often exhibit overlapping morphological and immunohistochemical features [29, 30]. Salivary duct carcinomas share similar structures and histologic features with certain forms of BC, including overexpression of ERBB2 and the presence of estrogen or progesterone receptors. They also exhibit pathologic features, such as sclerosing polycystic adenosis, which resembles fibrocystic changes in the breast [28]. Additionally, ERBB2-positive salivary duct carcinomas have been reported to respond well to trastuzumab, a treatment also used for HER2-positive BC [31]. These observations have supported potential genetic links, particularly with BRCA genes [7, 9, 32].
Though evidence is limited (Table 2), the association between BRCA genes and salivary gland cancer began to be explored by Shen et al. (2014) [8], who identified an incidence of 0.052% of salivary gland cancers among 5754 carriers of BRCA1/2 variants (p < 0.001). Since then, only a few studies have reported associations between pathogenic/likely pathogenic variants in BRCA genes and salivary gland cancers. De Barros et al. [9] reported a case of a Brazilian female diagnosed with metachronous tumors, including bilateral BC (first at 33 years), primary papillary thyroid carcinoma (53 years), and primary mucoepidermoid carcinoma in the right parotid (55 years), carrying a pathogenic BRCA1 variant: c.470_471delCT, p.Ser157fs*1, found in exon 7 and predicted to lead to NMD. Although segregation of the variant was not proven, the authors described four relatives with untested BRCA1, including one case of non-smoking-related lung cancer, two cases of smoking-related laryngeal cancer, and a paternal aunt with GC diagnosed at 50 years old.
Ripamonti et al. [7] described a case series of five families from the Istituto Nazionale Tumori of Milan carrying BRCA variants and salivary gland cancers. Although only two cases were well characterized and lacked detailed family descriptions, the first case involved a 56-year-old man with a parotid myoepithelial carcinoma with lung and bone metastases. Genetic testing revealed a BRCA1 variant: c.798_799delTT, p.Ser267Lysfs19, in exon 10, predicted to undergo NMD. The second case involved a 66-year-old male diagnosed with submandibular papillary cystadenocarcinoma, with genetic testing showing a nonsense variant BRCA2: c.9481A > T, p.Lys3161, also predicted to undergo NMD.
Our case is the first well-characterized large family with a diverse oncologic spectrum, including SGCs, segregating a BRCA1 variant. All reported cases, including ours, identified BRCA1 frameshift variants in exons not typically considered hotspots, particularly exon 10, usually a “cold spot” [33, 34]. These BRCAs variants generate null alleles and undergo NMD. HBOC presentation varies by variant type and location; BC is more associated with NMD variants, while OC is linked to non-NMD variants [14, 15]. This pattern matches our family, which has no OC cases, and the case reported by De Barros et al. [9]. Like GC, SGCs in BRCA1 carriers are often diagnosed after 50 years, indicating a specific presentation in these individuals. However, further research is needed to understand the clinical, pathological, and molecular patterns described in this family and the particularities of the association of SGC and GC in BRCA-related HBOC. Genotype-phenotype family characterization is a valid approach to understanding the expanded spectrum of hereditary cancer syndromes like HBOC, enhancing personalized follow-up and treatment.
More importantly, considering the introduction of PARP inhibitors (PARPi), which improve progression-free and distant disease-free survival in individuals with germline and somatic P/LP variants in HRR pathway genes, particularly BRCA1/2 [35], recognizing the expanded spectrum of BRCA1 oncological phenotypes could broaden the benefits of PARPi treatment to patients with tumors beyond the classic HBOC spectrum who harbor BRCA1 mutations. Thus, we suggest that clinicians consider BRCA1 mutation screening in patients with SGTs and GC, especially those with a family history suggestive of HBOC.