Multilocus inherited neoplasia alleles syndrome: a retrospective review from a Canadian single institution

Abstract

Genetic testing in hereditary cancer is evolving from single-gene-focused approaches in affected individuals to multi-gene panel testing for affected individuals and unaffected relatives. The widespread use of multi-gene panel testing has led to the identification of individuals with two or more pathogenic or likely pathogenic variants in hereditary cancer susceptibility genes (CSGs), termed Multilocus Inherited Neoplasia Allele Syndrome (MINAS) carriers. It remains unclear whether MINAS carriers are at increased risk of multiple, atypical, or more severe cancer phenotypes, and currently, there is no consensus on how best to identify and manage cancer risk. In this retrospective study, we identified 54 MINAS carriers at Princess Margaret Cancer Center in Toronto, Canada. Demographic, clinical, and genetic data were extracted from medical records. Group comparisons were performed using Fisher’s exact or chi-square tests for categorical variables and independent t tests for age at cancer diagnosis. Statistical significance was set at p ≤ 0.05. The majority of affected MINAS carriers had a cancer consistent with the expression of at least one pathogenic variant. Approximately 28% of MINAS carriers were diagnosed with one atypical cancer. The most frequent gene pair combinations included hereditary breast cancer genes, with some carriers exhibiting earlier age of breast cancer onset than single CSG variants reported in the literature. Our study indicates that the cancer spectrum associated with CSGs is expanding and suggests that more intensive cancer surveillance for subgroups of MINAS carriers with hereditary breast cancer CSGs may be warranted.

Introduction

Hereditary cancer syndromes account for approximately 5–10% of all cancers and describe a spectrum of cancers that are caused by inherited pathogenic or likely pathogenic variants in cancer susceptibility genes (CSGs) [1]. Hereditary cancer syndromes include BRCA1 and BRCA2 associated breast, ovarian, and prostate cancers, and lesser-known cancer syndromes, including CDH1-associated hereditary diffuse gastric cancer. Traditionally, hereditary cancers have been identified by recognizing clusters of related individuals with a high burden of cancer, followed by targeted testing of individual genes. With the advent of next generation sequencing, there has been a paradigm shift toward multi-gene panel testing, which allows for simultaneous detection of multiple CSGs. This in turn has enabled detection of pathogenic variants not predicted by phenotype or family history. Further, multi-gene panel testing has led to the detection of carriers of two or more pathogenic or likely pathogenic variants in CSGs, termed Multilocus Inherited Neoplasia Allele Syndrome, or MINAS [2]. The estimated prevalence of MINAS carriers is 0.2–2.4% [3–5]. Although early MINAS studies included both autosomal dominant (AD) and recessive (AR) gene pair combinations [3], more recent studies have analyzed AD-AD and AD-AR gene pair combinations separately, citing higher penetrance and higher frequency of malignancies in AD-AD gene pair combinations [4, 5].

Hereditary cancer syndromes have been associated with an elevated lifetime risk for specific cancers and more severe cancer phenotypes [6]. Carriers of single pathogenic variants in CSGs undergo lifelong specialized cancer surveillance. These genetic test results provide information on cancer risks and implications for other family members, such as the potential need for cascade testing [7]. The results of surveillance are actionable, guiding decision making around preventive or risk-reducing measures and the use of certain therapeutic agents in cancer treatments. Compared to single CSG variants, it has long been hypothesized that MINAS carriers are at elevated risk of multiple, atypical, and more severe cancers because of synergistic interactions CSGs may have in related tumourigenic pathways or in chromosomal proximity [2]. Synergistic interactions in MINAS cases refer to CSGs interacting in a way that may produce a more severe phenotype than would be expected from each CSG acting alone. In order to evaluate for synergistic effects of CSGs, studies have compared the clinical spectrum and characteristics of cancers in MINAS cases to single gene hereditary cancers. Most MINAS studies have focused on hereditary breast cancer genes, including BRCA1 and BRCA2. Some studies have found a younger age of onset and multiple malignancies in hereditary breast cancer MINAS cases compared to single gene hereditary breast cancers [3, 4, 8–13]. The largest study to date by Rebbeck et al. comparing BRCA1 and BRCA2 MINAS carriers found that the mean age of breast cancer onset was not significantly different from single BRCA variants [14]. Hereditary breast cancers differ in hormone receptor status based on the specific germline mutation, which significantly impacts both prognosis and treatment, as hormone receptor status is an independent predictor of survival [15]. BRCA1-associated cancers are predominantly triple-negative, while BRCA2, ATM, and CHEK2-associated cancers are predominantly ER-positive. Li-Fraumeni-associated cancers show increased HER2-positivity but variable hormone receptor status [16].

In this single-center retrospective cohort study, we identified carriers with AD-AD (i.e., MINAS) and AD-AR gene pair combinations who were diagnosed at Princess Margaret Cancer Center from January 1, 2017, to September 30, 2024. We extracted demographic, clinical, and genetic data from medical records and made comparisons using Fisher’s exact test, chi-square tests, or independent t tests, with statistical significance set at p ≤ 0.05. The objectives of this study were to characterize the clinical spectrum of cancers in an adult cohort of MINAS carriers and compare our findings to the hereditary cancer and MINAS literature. The overall aim of this study was to work toward building comprehensive and evidence-based guidelines for screening, prognostication, and clinical management of cancers in MINAS carriers.

Materials and methods

Study cohort and ethics committee approval

We conducted a retrospective cohort study of individuals referred for genetic counseling and testing between January 1, 2017, and September 30, 2024, to identify and characterize AD-AD and AD-AR carriers, defined as those harboring germline pathogenic variants in two or more different CSGs. Individuals were selected for genetic testing based on Ontario Health Hereditary Cancer testing guidelines based on strong personal or family history of cancer [17]. A total of 11,344 patients underwent genetic testing during the study period at the Bhalwani Familial Cancer Clinic at the Princess Margaret Cancer Center, University Health Network (UHN), Toronto, Canada. There were 2588 patients identified through the Progeny database as testing positive for a germline genetic variant. The type of testing (single gene, vs multi-gene panel and the gene composition) is not detailed in our database. A total of 65 eligible cases from 60 families were identified as MINAS in the Progeny database of the Bhalwani Familial Cancer Clinic. Of the 19 unaffected MINAS individuals included in our study, 10 were relatives of affected probands. Ethics approval for this retrospective chart-review study was granted by the UHN Research Ethics Board (ID: 24-5884). The study was classified as no-risk and exempt from written informed consent.

Data collection

We collected detailed information across four domains by reviewing data from the Progeny database and the Electronic Medical Records (EMR): (i) Demographic data, including sex and self-reported ancestry from maternal and paternal lineages; (ii) Clinical data, including affected/unaffected status, cancer type(s), number of primary cancers, age at each diagnosis, and breast cancer receptor status when applicable; (iii) Genetic data, including HUGO gene symbols, HGVS nomenclature for each pathogenic variant, pathogenicity interpretation, inheritance pattern, and cancer risk category per gene (high risk (HR), moderate risk (MR), or low increased risk (LR)) based on the National Comprehensive Cancer Network Guidelines and estimates of relative risk (high risk defined as a greater than 4 fold increase, moderate risk as a 2–4 fold increase and low risk defined as a less than twofold increase in disease risk compared to the general population [18–20] (Supplementary Table 1). All APC carriers in this study have the p.I1301K allele and are classified as LR; iv) Family history among first- and second-degree relatives. The clinical team assessed, for each individual, whether none (0), one (1), or both (2) variants were phenotypically expressed in the proband or family members, based on their established association with diagnosed cancer types, reported family history, and known inheritance patterns (autosomal dominant (AD) or autosomal recessive (AR)) (Supplementary Table 2). The presence of an atypical phenotype was determined if there was one or more cancers phenotypically expressed that were not associated with pathogenic variant expression. A dominant inheritance mode was assigned to genes with both AD and AR implications if the heterozygous state was linked to cancer susceptibility (e.g., ATM). To safeguard confidentiality, all records were de-identified using unique Study IDs. Detailed clinical and genetic information for each MINAS carrier can be found in Supplementary Table 3, 4.

Genetic counseling and test results

All patients were referred to the Princess Margaret Genetics Clinic or had their charts reviewed by certified genetic counselors, the medical genetics team, or an oncologist to determine the most appropriate genetic test(s) based on each patient’s personal and family history. Referrals were received from a variety of sources, including oncologists for patients with active cancer with features consistent with genetic testing guidelines and primary care where healthy people with a strong family history of cancer. Decisions were guided by institutional protocols and provincial guidelines [17, 21]. Genetic testing was performed at Clinical Laboratory Improvement Amendments (CLIA)-certified laboratories. Following result disclosure, all individuals received post-test counseling and were referred to appropriate multidisciplinary teams for personalized risk management and clinical follow-up.

Individuals with two or more germline variants in different genes, classified as pathogenic, likely pathogenic, or risk alleles (for example, APC c.3920 T > A (p.Ile1307Lys), which is risk factor for colorectal cancer in individuals of Ashkenazi Jewish ancestry and carries specific screening recommendations), were included in this study. Variant classification followed the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines [22]. All variant nomenclature was standardized according to Human Genome Variation Society (HGVS) conventions using the VariantValidator tool (https://variantvalidator.org).

Statistical analysis

Descriptive statistics were used to summarize demographic, clinical, and genetic characteristics of the study cohort. Categorical variables were described using frequencies and percentages. Age at first cancer diagnosis was the only continuous variable analyzed, and it was described using measures of central tendency or categorized into age ranges. Normality and homogeneity of variance for age were assessed using the Shapiro-Wilk and Levene’s tests, respectively. Implicated genes were assigned a cancer risk level (high risk (HR), moderate risk (MR), or low increased risk (LR)) and an inheritance pattern (AD or AR) (Supplementary Table 1). Based on the inheritance pattern of the implicated genes per individual (Supplementary Table 2), two groups were formed: AD-AD combinations and AD-AR combinations. Group comparisons for categorical variables, which often included small cell counts ( < 5), were performed using Fisher’s exact test (Fischer-Freeman-Halton extension for rxc tables); otherwise, chi-square tests were applied. Differences in age at diagnosis between groups were assessed using independent t tests. Besides gene combination category, cases were also stratified by clinical status (affected vs. unaffected), and the number of phenotypically expressed variants (0, 1, or 2) in the individual and family based on their cancer history. Statistical significance was set at p ≤ 0.05. Cancer co-occurrence, as well as gene co-occurrence patterns, were visualized using circular chord diagrams (circlize package). All analyses were conducted in R (version 4.4.2).

Results

Clinical characteristics of MINAS carriers and relatives

We identified 65 individuals based on the search criteria (Supplementary Table 3, 4). Of the 65 cases, a total of 54 individuals from 49 families were categorized as MINAS (AD-AD gene combinations), while 11 individuals were categorized as AD-AR. Most MINAS carriers were female (57%; 31/54). Self-reported ancestry across both lineages was primarily White (63%), followed by Ashkenazi Jewish (17%) and South Asian (7.5%) (Table 1).

Table 1.

Comparison of AD-AD and AD-AR gene combinations across demographic, clinical, and genetic variables.

Variable	Overall N = 65a	AD-AD N = 54a	AD-AR N = 11a	p valueb
Sex				0.181
Female	40 (62%)	31 (57%)	9 (82%)
Male	25 (38%)	23 (43%)	2 (18%)
Age of first cancer				0.522
N Non-missing	46	38	8
Mean (SD)	47 (18)	48 (19)	44 (10)
Median (Q1, Q3)	49 (33, 59)	52 (31, 62)	40 (36, 54)
Min, Max	4, 83	4, 83	31, 59
Self-reported ancestry (both lineages)				0.802
White	76 (62%)	67 (63%)	9 (56%)
Ashkenazi Jewish	22 (18%)	18 (17%)	4 (25%)
South Asian	10 (8.2%)	8 (7.5%)	2 (13%)
Southeast Asian	4 (3.3%)	4 (3.8%)	0 (0%)
Latino	3 (2.5%)	2 (1.9%)	1 (6.3%)
Middle Eastern	3 (2.5%)	3 (2.8%)	0 (0%)
Black or African descent	2 (1.6%)	2 (1.9%)	0 (0%)
East Asian	2 (1.6%)	2 (1.9%)	0 (0%)
Number of cancers				0.762
0	19 (29%)	16 (30%)	3 (27%)
1	32 (49%)	25 (46%)	7 (64%)
2	9 (14%)	8 (15%)	1 (9.1%)
3	5 (7.7%)	5 (9.3%)	0 (0%)
Cancer type				0.402
Breast	22 (34%)	18 (32%)	4 (44%)
Prostate	7 (11%)	7 (13%)	0 (0%)
Hematologic Disorders	6 (9.2%)	5 (8.9%)	1 (11%)
Melanoma	5 (7.7%)	5 (8.9%)	0 (0%)
Ovarian	5 (7.7%)	4 (7.1%)	1 (11%)
Renal	5 (7.7%)	5 (8.9%)	0 (0%)
Endometrial	2 (3.1%)	2 (3.6%)	0 (0%)
Neuroendocrine	2 (3.1%)	1 (1.8%)	1 (11%)
Sarcoma	2 (3.1%)	2 (3.6%)	0 (0%)
Thyroid (Papillary)	2 (3.1%)	2 (3.6%)	0 (0%)
Adenoid Cystic	1 (1.5%)	1 (1.8%)	0 (0%)
Pancreatic	1 (1.5%)	0 (0%)	1 (11%)
PGL/PCC	1 (1.5%)	1 (1.8%)	0 (0%)
Testicular	1 (1.5%)	0 (0%)	1 (11%)
Small Bowel	1 (1.5%)	1 (1.8%)	0 (0%)
Thyroid (Medullary)	1 (1.5%)	1 (1.8%)	0 (0%)
Bladder	1 (1.5%)	1 (1.8%)	0 (0%)
Breast cancer receptors				0.119
ER/PR positive	9 (50%)	8 (57%)	1 (25%)
Triple negative	4 (22%)	4 (29%)	0 (0%)
HER2-enriched	3 (17%)	1 (7.1%)	2 (50%)
Triple positive	1 (5.6%)	1 (7.1%)	0 (0%)
ND	1 (5.6%)	0 (0%)	1 (25%)
Personal history indicative of CSGs expression				0.089
0	29 (45%)	22 (41%)	7 (64%)
1	20 (31%)	16 (30%)	4 (36%)
2	16 (25%)	16 (30%)	0 (0%)
Family history indicative of CSG expression				0.000
0	17 (26%)	10 (19%)	7 (64%)
1	18 (28%)	14 (26%)	4 (36%)
2	30 (46%)	30 (56%)	0 (0%)

AD autosomal dominant, AR autosomal recessive.

^an (%).

^bFisher’s exact test (Fisher–Freeman–Halton extension for r×c tables); Two Sample t test.

Seventy percent of MINAS carriers were affected with at least one cancer at the time of data collection. Forty-six percent (25/54) had one cancer, 15% (8/54) had two cancers and 9.3% (5/54) had three or more cancers. The most common cancer types were breast (32%, 18/54) and prostate (13%, 7/54), followed by hematologic disorders, melanoma and renal cancers at 8.9% each, and ovarian cancers with 7.1%. In individuals with two or more primary cancers, nearly 20 distinct cancer pairings were observed. Age of first cancer diagnosis ranged widely among MINAS carriers, with nearly 40% (15/38) of affected individuals diagnosed before age 41 (Table 1).

We compared MINAS carriers to those with AD–AR gene combinations across demographic, clinical, and genetic variables (Table 1). No significant differences were observed between groups in sex distribution (p = 0.181), age at first cancer diagnosis (mean 48 vs. 44 years; p = 0.522), self-reported ancestry (p = 0.802), number of cancers (p = 0.807), or cancer type (p = 0.402). Although not statistically significant, three or more primary cancers were only observed in MINAS carriers, suggesting potentially higher cancer burden in AD-AD compared to AD-AR gene pair combinations.

We also compared breast cancer receptor expression in MINAS and AD-AR carriers. Breast cancer receptor subtypes did not differ significantly between groups (p = 0.119) (Table 1). In MINAS carriers, most breast cancers were ER/ PR-positive/ HER-2 negative (57%). HER2-enriched tumors were more frequent in AD-AR than MINAS carriers (50% vs. 7.1%, respectively). Triple negative breast cancers were the second most frequent subtype among the MINAS group (29%). There were zero occurrences of triple negative breast cancers in the AD-AR group.

Genes and gene combinations in MINAS carriers and correlation with clinical characteristics

In MINAS carriers, we identified pathogenic variants in 26 CSGs (Supplementary Table 4). In MINAS carriers with cancer (herein referred to as “affected”) (n = 38), the most frequently identified genes were CHEK2 (n = 19), BRCA1 (n = 12) and BRCA2 (n = 12), followed by ATM (n = 6), PALB2 (n = 3) and RAD51C (n = 3). In MINAS carriers unaffected by cancer (herein referred to as “unaffected”) (n = 16), the most frequently identified genes were CHEK2 (n = 8), BRCA2 (n = 5), APC (n = 3, all of which corresponds to the risk allele c.3920 T > A (p.Ile1307Lys)) and SDHB (n = 3, all of which are healthy relatives from the affected proband). In AD-AR carriers, the most frequently identified genes included MUTYH (n = 5), CHEK2 (n = 3) and BRCA1 (n = 3), with varying gene combinations observed across affected and unaffected patients.

In MINAS carriers, a total of 58 gene pair combinations were identified (Supplementary Table 4, Fig. 1). Nearly all carriers (52/54) had two pathogenic variants, leading to one possible gene pair combination, while two carriers had three pathogenic variants, leading to three gene pair combinations per carrier. Of these 58 gene pair combinations, the most frequently identified gene pairs were BRCA1 + BRCA2 (n = 5, 100% affected), followed by BRCA2 + CHEK2 (n = 5, 80% affected), CHEK2 + SDHB (n = 4, 25% affected), APC + CHEK2 (n = 3, 33% affected). Of the 11 gene pair combinations in AD-AR carriers, nine were unique gene pair combinations. The most frequent gene pair combinations were BRCA1 + FANCC (n = 2, 50% affected) and CHEK2 + MUTYH (n = 2, 100% affected). A small portion of both MINAS (13/54) and AD-AR (1/11) gene pair combinations were associated with two or more cancers (Fig. 2, Supplementary Table 3 & 4).

Fig. 1. Chord diagrams showing combinations of CSGs in AD-AD carriers (n = 54).

A Gene pairs found in affected individuals B. The Gene pairs found in unaffected individuals. The width of each chord reflects the frequency of co-occurrence of gene pairs. Gene colors represent their associated cancer risk category (high, moderate, or low).

Fig. 2. Chord diagram depicting the cancer gene pair combinations in AD-AD carriers affected by two and three cancers (n = 13).

Multiple combinations may apply to the same individual. Breast(1) refers (primary breast cancer), breast(2) (secondary primary breast cancer), PGL/ PCC (pheochromocytoma/ paraganglioma).

We next analyzed the correlation of the gene and gene combination with clinical presentation. In this analysis, we defined a pathogenic variant as being phenotypically expressed if a CSG was known to have an association with the cancer or a key trait diagnosed in the individual (i.e. café au lait macules in individual in NF1 variant). Conversely, a pathogenic variant was not phenotypically expressed if the diagnosed cancer had no known associations with the CSG.

The majority of MINAS carriers (60%; 32/54) had a cancer (31/32) or a key trait (i.e. café au lait macules in individual in NF1 variant, 1/32) consistent with expression of at least one of the CSGs (Table 1). Of the MINAS carriers with no CSG expression, the majority (68%, 15/22) were unaffected at time of data collection. Notably, 39% (15/38) of MINAS carriers with cancer were diagnosed with at least one atypical cancer with no known strong association to the CSGs (Supplementary Table 5). Of the 15 MINAS carriers with atypical cancers, the only gene pair present more than once was CHEK2 + SEC23B, which was present in two members of the same family who were diagnosed with essential thrombocythemia.

Comparing MINAS to AD-AR carriers, significantly fewer AD-AR individuals had a cancer consistent with expression of at least one of the CSGs, while the majority (64%) had no CSG expression (p = 0.089) (Table 1). Notably, a significantly greater proportion of AD–AR carriers compared to MINAS carriers lacked a family history indicative of pathogenic variant expression (64% vs. 19%, p < 0.001), highlighting family history as a key differentiating feature of MINAS carriers (Table 1).

We then classified genes as high (HR), moderate (MR), or low increased risk (LR) (see methods for details) and analyzed gene-pair combinations with respect to pathogenic variant expression (Table 2, Fig. 2). The distribution of gene risk combinations differed across CSGs phenotypically expressed (p = 0.028). Combinations involving HR genes, either paired with other HR or MR genes, were the most frequent in individuals with two CSGs phenotypically expressed (45% and 50%, respectively). In individuals with one CSG phenotypically expressed, the most common combinations included MR genes paired with either other MR or HR genes (31% and 50%, respectively). Interestingly, in cases with no pathogenic variant expression, the most common gene risk combination was HR + MR (55%). The HR-MR group mostly includes unaffected individuals, including healthy relatives of probands identified through cascade testing as MINAS carriers, but who have not developed cancer. In line with this, the median age at last data capture for unaffected carriers was relatively young (48 years, range 31–73) compared to affected individuals (58 years, range 30–84) (p = 0.037) (Table 3). These findings highlight the importance of ongoing surveillance, as many unaffected carriers may still be within the age range of risk for developing related cancers.

Table 2.

Gene risk combination counts and CSG expression among MINAS carriers.

		CSG expression in the individual
Variable	Overall N = 58a	0 N = 22a	1 N = 16a	2 N = 20a	p valueb
Clinical status					0.000
Affected	42 (72%)	7 (32%)	15 (94%)	20 (100%)
Unaffected	16 (28%)	15 (68%)	1 (6.3%)	0 (0%)
Gene risk combination					0.028
HR + HR	12 (21%)	1 (4.5%)	1 (6.3%)	10 (50%)
HR + MR	28 (48%)	12 (55%)	8 (50%)	8 (40%)
HR + LR	2 (3.4%)	1 (4.5%)	1 (6.3%)	0 (0%)
MR + MR	11 (19%)	5 (23%)	5 (31%)	1 (5.0%)
LR + MR	5 (8.6%)	3 (14%)	1 (6.3%)	1 (5.0%)
Family history indicative of CSG expression					0.033
0	12 (21%)	1 (4.5%)	5 (31%)	6 (30%)
1	14 (24%)	6 (27%)	6 (38%)	2 (10%)
2	32 (55%)	15 (68%)	5 (31%)	12 (60%)

N number of gene pairs among MINAS cases, HR high-risk, MR moderate-risk, LR low-risk.

^an (%); Includes all gene pair combinations identified in MINAS cases (n = 54).

^bFisher’s exact test (Fisher–Freeman–Halton extension for r×c tables).

Table 3.

Age at last data-entry (years) by clinical status of MINAS carriers.

Variable	Overall N = 54a	Affected N = 38a	Unaffected N = 16a	p valueb
Age at last data-entry (years)				0.037
N Non-missing	54	38	16
Mean (SD)	55 (16)	58 (16)	48 (13)
Median (Q1, Q3)	54 (41, 69)	58 (43, 72)	48 (37, 54)
Min, Max	30, 84	30, 84	31, 73

^an (%).

^bTwo sample t test.

Correlation with clinical characteristics in hereditary breast cancer gene pair combinations

The most frequent gene pair combinations in our MINAS dataset were BRCA1 + BRCA2 (n = 5) and BRCA2 + CHEK2 (n = 5). Notably, all BRCA1 + BRCA2 MINAS carriers were affected by breast or prostate cancer. In BRCA1 + BRCA2 carriers with breast cancer, 75% (3/4) had receptor expression profiles consistent with BRCA1 breast cancers (triple negative), while 25% (1/4) had an expression profile consistent with BRCA2 breast cancers (ER/PR +). All four carriers with BRCA1 + BRCA2 gene pair combinations who developed cancer were diagnosed before age 39, with a median age of onset of 33.5 years, which is earlier than reported for single BRCA1 or BRCA2 variant carriers in the literature (40 and 42 years, respectively) [23]. MINAS carriers with BRCA2 + CHEK2 had variable phenotypes. Sixty percent (3/5) were affected by one primary cancer (sarcoma, melanoma or breast cancer, n = 1), while the remaining remained unaffected. Age of onset of breast cancer was 61 years in the BRCA2 + CHEK2 carrier, which is not significantly earlier than breast cancer onset in the general population.

Thirty six percent (5/14) of MINAS carriers with breast cancer had two variants in other hereditary breast cancer genes. These gene pair combinations included BRCA1 + TP53, BRCA2 + PALB2, BRCA2 + RAD51D, PALB2 + CHEK2 and ATM + CHEK2 (one of each, respectively). Pathogenic variants in TP53 are associated with Li-Fraumeni syndrome, a hereditary cancer predisposition syndrome associated with multiple cancers, including breast cancer onset at a median age of 34 [24]. The BRCA1 + TP53 carrier in our cohort developed bilateral invasive ductal carcinoma breast cancer at age 33. Carriers with BRCA2 + PALB2, BRCA2 + RAD51D, ATM + CHEK2, PALB2 + CHEK2 had breast cancer (or bilateral breast cancer, in the case of the ATM + CHEK2 and PALB2 + CHEK2 carriers) diagnosed at age 28, 28, 42 and 51, respectively. Notably, median age of breast cancer diagnosis in single variants of PALB2, RAD51C, ATM and CHEK2 is reported as 52, 43, 47 and 47, respectively, in the literature [25–27]. Although our data set is small, our results suggest that most carriers affected by breast cancer and with two or more variants in hereditary breast cancer genes have an earlier onset of breast cancer.

Discussion

In this single-center retrospective cohort study, we identified 54 MINAS carriers. To our knowledge, this is the largest cohort of MINAS carriers at a single center identified to date. In line with previous studies, a significant proportion of MINAS carriers (77%) had at least one hereditary breast cancer gene, most commonly BRCA1 (n = 14) or BRCA2 (n = 17). The high frequency of hereditary breast CSGs in our cohort is reflective of several factors, including the high proportion of patients with personal or family history of breast cancer referred to Princess Margaret Cancer Center and offered genetic testing (i.e. ascertainment bias), and more generally, the high penetrance, well-established screening protocols and genetic panels for hereditary breast cancers compared to other hereditary cancers, such as colorectal and pancreatic cancer.

Interestingly, compared to MINAS carriers, AD-AR carriers are less likely to have a family history of cancer but are otherwise similar in terms of demographic features, cancer type and age of first cancer diagnosis. Our results are similar to work done by Yuen et al., who found that AD-AD carriers have a higher burden of cancer than AD-AR carriers, and suggest that cancer risk management should be guided by the AD gene [4]. Notably, our AD-AR cohort is small (n = 11) and clinically heterogeneous, with nine unique gene pair combinations. Further work is required to clarify the role of AR CSGs in cancer risk management.

Approximately 28% (15/ 54) of MINAS carriers affected by cancer in our study presented with at least one atypical cancer with no known strong association to the CSGs. In previous studies of MINAS carriers, atypical tumors have been found in 14.5- 15.8% of cohorts [3, 4]. Given how rare these CSG combinations are present, it is not possible to say whether the resulting cancer is coincidental or a reflection of novel CSG synergistic interaction. Unfortunately, rare gene pair combinations are an inherent challenge with MINAS cases. Open access MINAS databases, such as the Global Variome LOVD created by Whitworth et al., are essential in continuing to document rare gene pair combinations and investigate atypical tumor presentations [2].

There is great interest in determining whether MINAS carriers exhibit more severe disease due to synergistic interactions of CSGs. The most well studied gene pair combinations include BRCA1 and BRCA2. In our study, the median age of breast cancer diagnosis in BRCA1 and BRCA2 carriers was 33.5 years, which is nearly a decade younger than single BRCA1 and BRCA2 variants cited in the literature [23]. The largest study of BRCA1 + BRCA2 carriers and cancer risk was done by the Consortium of Investigators of Modifiers of BRCA1 and BRCA2 (CIMBA) in 2012 [23]. In their study of 93 BRCA1 + BRCA2 carriers, Rebbeck et al. found that BRCA1 + BRCA2 carriers were more likely to be diagnosed with breast cancer than either BRCA1 or BRCA2 single variants; however, the mean age of breast cancer diagnosis was not statistically different than carriers of a single BRCA1 variant (40.4 vs. 41.9 years) [23]. Further loss of heterozygosity studies supported an additive rather than synergistic effect in BRCA1 + BRCA2 carriers [23]. A more recent study by Yuen et al. analyzed 89 BRCA1 + BRCA2 carriers and found that BRCA1 + BRCA2 carriers had a higher percentage of multiple malignancies and an earlier onset of cancer than single CSG variants (but did not specifically compare BRCA1 + BRCA2 to single BRCA1 or BRCA2 variant carriers) [4]. In addition to these studies, there are multiple case reports of small cohorts of BRCA1 + BRCA2 carriers that have yielded conflicting results [10, 11, 13]. There are methodological limitations that may limit comparison between MINAS studies. Primarily, there is significant clinical heterogeneity. For example, studies such as Rebbeck et al. have a large Ashkenazi Jewish population and subsequently higher rates of particular founder variants (i.e. c.68_69delAG; p.Glu23Valfs17*) in their BRCA1 + BRCA2 cohort compared to studies with predominantly Asian cohorts [4, 11, 13, 23]. Additionally, guidelines for genetic testing vary between countries, testing centers and even individual providers, leading to variability in which gene panels are selected for genetic testing.

Beyond BRCA1 + BRCA2 gene pair combinations, there is limited information on cancer risk management in other gene pair combinations. In their recent study, Yuen et al. commented that MINAS carriers with hereditary breast cancer genes beyond BRCA1 + BRCA2 are also more likely to be associated with multiple malignancies and earlier onset of cancer [4]. Consistent with this, we found several hereditary breast gene pair combinations with an earlier onset of breast cancer than either of the single variants. Because of the relatively small sample size, we are unable to directly infer causality or interaction effects from this dataset. Based on our results and others, we believe a closer examination of gene pair combinations with two or more hereditary breast CSGs is warranted in order to determine the benefit of enhanced surveillance and prophylactic surgeries at an earlier age. Importantly, there are no large-scale prospective cohort studies or case series that provide robust, quantitative estimates of cancer incidence, age of onset or outcomes.

There are several limitations to our study. While a total of 11,344 patients underwent germline testing, the true prevalence of MINAS cases was not captured for a number of reasons. Our study period spans January 2017 to September 2024. The clinical uptake of multi-gene panel testing became more pervasive after April 2021 with the development of provincial genetic testing eligibility criteria that became the standard of care [17]. Prior to April 2021, multi-gene panel testing was conducted for a limited number of indications. Due to these limitations and the gap of routine retesting [28, 29], the true prevalence of MINAS is not possible to calculate. Additionally, classification of variants was ascertained from the official report at the time the report was received. Variants of uncertain significance were not systematically reclassified over time and were only reclassified if the lab actively notified the team (due to an internal lab quality improvement) or if the patient was re-referred for a variant of uncertain significance classification. Our study may therefore reflect an underestimation of MINAS cases and is a limitation of our cohort analysis.

Our study has limitations that are present in MINAS studies. Due to the relatively rare nature of MINAS, our cohort size is relatively small and gene pair combinations relatively infrequent. We are not able to determine definitively if one or more of the MINAS CSGs are contributing to tumor occurrence. As discussed by McGuigan et al., detailed tumor studies, specifically in the form of loss of heterozygosity or other tumor profiling strategies including immunohistochemistry of CSG gene products and microsatellite instability, are needed clarify the role of MINAS CSGs in tumourigenesis [3]. Further, individuals with more severe presentations (i.e. multiple cancers) are more likely to be referred for genetic testing and undergo multi-gene panel testing, therefore reflecting selection bias. Inclusion of unaffected individuals (usually identified through cascade testing of affected relatives) may confound the analysis, as they are likely to be captured at a younger age and have not yet developed cancer.

Conclusions

Here, we provide information on the largest single center cohort of MINAS carriers. In our study, a significant proportion of MINAS carriers are associated with atypical cancer presentations that cannot be explained by the single CSG alone, which may point to novel synergistic interactions of CSGs and expand the spectrum of cancers associated with CSGs. We find that MINAS carriers with two or more hereditary breast CSGs developed earlier onset of breast cancer than single variants reported in the literature; however, further studies are necessary to determine if earlier surveillance and intervention is warranted. Importantly, research is needed to interrogate novel cancer associations and disease severity in MINAS carriers with hereditary breast CSGs, in the form of larger, prospective cohort studies, case series and tumor profiling studies.

Author contributions

KO analyzed data and wrote the manuscript. MS performed statistical analysis and helped to write the manuscript. MH, KA, RM, LP contributed to study design and data extraction. RHK conceived the study. All authors read and approved the final version of the manuscript before submission.

Funding

This work was supported in part by the Bhalwani Family Charitable Foundation, Goldie R. Feldman, Karen Green and George Fischer Genomics and Genetics Fund, Lindy Green Family Foundation, FDC Foundation, Shar Foundation, The Devine/Sucharda Charitable Foundation, Leslie E. Born, Hal Jackman Foundation, Nicol Family Foundation, James and Christine Nicol, Janice Fukakusa and Greg Belbeck, Jack and Buschie Kamin Foundation, Marcus Tzaferis, Paul Bronfman Family Foundation, The Honey and Leonard Wolfe Family Charitable Foundation, Arman Alie and Margarette Nory, The Princess Margaret Cancer Foundation.

Data availability

Data generated or analyzed during this study can be found within the published article and its supplementary file. The genetic data that support the findings of this study have been submitted to be made publicly available in ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/). The ClinVar accession numbers corresponding to the data reported in this article are SCV007602285–SCV007602372.

Competing interests

The authors declare no competing interests.

Ethical approval

Ethics approval for this retrospective chart-review study was granted by the UHN Research Ethics Board (ID: 24-5884). The study was classified as no-risk and exempt from written informed consent.

Supplementary information

The online version contains supplementary material available at 10.1038/s41431-026-02142-6.