Genetic Characterization of Hereditary Cancer Syndromes Based on Targeted Next-Generation Sequencing

A hereditary cancer syndrome is a genetic predisposition to cancer which is caused by a disease-causing mutation in the genes. The majority of these syndromes are inherited in an autosomal dominant manner [ACOG, 2019]. Hereditary cancers are responsible for 5–10% of all the cancer cases. To date, more than 200 hereditary cancer syndromes have been noted in the literature [Macháčková et al., 2016; Tsaousis et al., 2019]. Lynch syndrome, familial adenomatous polyposis, MUTYH-associated polyposis, hamartomatous polyposis, and hereditary breast-ovarian cancer are the most common hereditary cancer syndromes [Samadder et al., 2019]. MLH1, MSH2, APC, BRCA1, BRCA2, and TP53 are well-defined high-risk cancer genes [Susswein et al., 2016].

Detecting the hereditary defect is important for both the patient and relatives at risk in cancer families. Identifying the mutation-causing cancer susceptibility could be a guide in both treatment and secondary cancer prevention. In addition, cancer risk estimation in relatives at risk enables appropriate screening, surveillance, and interventions in cancer prevention management [Samadder et al., 2019; Tsaousis et al., 2019].

BRCA1 and BRCA2 are most common causative genes for hereditary breast and ovarian cancer. The proteins encoded by these genes have an important role in DNA repair and transcriptional regulation [Yoshida and Miki, 2004]. Most of pathogenic/likely pathogenic variants in BRCA1/BRCA2 lead to immature stop codons, truncating proteins, and reduce expression of proteins [Thorat and Balasubramanian, 2020; Yoshida, 2021]. Also, ATM and CHEK2 genes play a role in the DNA repairing system. The CHEK2 gene encodes an ezyme which is a cell cycle checkpoint kinase [Filippini and Vega, 2013].

Errors during DNA replication and recombination are corrected by the mismatch repair system. MLH1, MSH2, MSH6, and PMS2 are the major proteins in this system. These proteins act as heterodimers; MSH2 interacts with MSH6 or MSH3 while MLH1 interacts with PMS2 or MLH3. Defects in DNA mismatch repair system lead to Lynch syndrome [Baretti and Le, 2018]. Additionally, germline deletion in EPCAM (epithelial cell adhesion molecule) has been associated with Lynch syndrome [Kanth et al., 2017; Wells and Wise, 2017].

The protein encoded by APC gene acts as tumor supressor and has an important role in migration, transcription, and cell cycle control. Mutation in APC gene causes classic familial adenomatous polyposis [Nieuwenhuis and Vasen, 2007]. The MUTYH protein is involved in base excision repair system, corrects errors caused by guanine oxidation, and acts as a cell protective factor. Biallelic germline mutation in the MUTYH gene causes MUTYH-associated polyposis which is characterized by adenomatous polyps in the upper and lower gastrointestinal tract [Half et al., 2009]. Also, heterozygous individuals have a slightly increased risk of gastric, liver, endometrial, breast, and especially colorectal cancer [Curia et al., 2020]. The PTEN enzyme plays a role in the oncogenic PI3K/AKT/mTOR signal pathway as it metabolizes PIP3 [Álvarez-Garcia et al., 2019]. The SMAD4 and BMPR1A genes are involved in the BMP/TGF-beta signalling pathway. These 2 genes and STK11 are tumor suppressor genes [Brosens et al., 2011; Daniell et al., 2018; Zhao et al., 2018]. PTEN, STK11, SMAD4, BMPR1A, and GREM1 mutations are associated with hamartomatous polyposis syndromes [Jelsig et al., 2014].

The protein encoded by the RET gene acts as a tyrosine kinase receptor, in which gain of function predisposes to neoplasia. Germline mutations of this gene have been reported in different cancer types such as multiple endocrine neoplasia, rhabdomyosarcoma, prostate and breast cancer [Sponziello et al., 2018; Larouche et al., 2019].

Mutation of the TP53 gene, encoding the tumour suppressor p53 protein, causes Li-Fraumeni syndrome . This syndrome is characterized by early-onset cancer in multiple organ systems [McBride et al., 2014]. 8–12% of all advanced prostate cancer patients have a germline mutation in tumor suppressor genes. Current studies suggest that BRCA2, ATM, CHEK2, BRCA1, PALB2, and mismatch repair genes are responsible for susceptibility to prostate cancer. Additionally, a germline mutation in BRCA1 and BRCA2 increases the risk of pancreatic cancer as well as metastatic prostate cancer [Zhen et al., 2018].

As the clinical usage of NGS becomes widespread, the number of hereditary cancer-related mutations are increasing. The aim of this study is the analysis and reporting of significant variants associated with cancer susceptibility.

A single-center study was conducted in the hospital of Ataturk University. The study group was composed of individuals who were referred to our clinic for routine genetic tests between April 2018 and December 2020. Individuals were selected based on the age at onset of cancer (before age 60) and on the family history. History of cancer was defined as having at least 2 cancer patients (different types) in the family and at least one patient being a second-degree relative. Results of 280 affected and 25 unaffected individuals were analyzed. Unaffected individuals had no cancer diagnosis, but they had a family history of cancers which were different cancer types onset before age 60. These unaffected individuals were included in the study to highlight the importance of the NGS method in identifying not only cancer patients but also individuals with a familial history of cancer. Information about age at diagnosis, age at testing, demographics, clinical history, family history of cancer and pedigrees were collected (Table 1).

Genomic DNA was isolated from peripheral blood samples of the patients using the QIAcube, automated purification system (Qiagen, Hilden, Germany). Genomic DNA samples were validated by fluorescence-based quantitation assay. Only high quality samples were included in the study. Extracted DNA was fragmented, genomic targets were molecularly barcoded and enriched, and libraries were constructed according to the manufacturers’ protocol. Amplicon-based enrichment was used. Prepared libraries were validated by QIAxcel DNA analyzing system (Qiagen, Hilden, Germany). Hereditary cancer panel, including 33 genes (online suppl. Table 1; see www.karger.com/doi/10.1159/000518927), was performed on MiSeq Illumina NGS system from prepared suitable libraries. Sequencing files were fed into a cloud-based data analysis pipeline, which filters, maps, aligns reads, and counts unique molecular barcodes associated with targeted genomic regions. Variants were called by a barcode-aware algorithm [Ravi et al., 2018].

Variant calling and analysis was performed on Qiagen Clinical Insight Interpret software. This analyzing system detects single nucleotide variants, small indels, and copy number variations. According to the gnomAD database, variants with population frequency value above 1% were considered as single nucleotide polymorphisms. All of the reportable variants were classified according to the 2015 American College of Medical Genetics and Genomics (ACMG) guidelines [Richards et al., 2015]. Disease-specific information for variants was retrieved from ClinVar, OMIM, and the literature [Landrum et al., 2018]. Deletion/duplication analysis of BRCA1 and BRCA2 genes was performed using samples of 27 patients with familial breast/ovarian cancer who had negative results in hereditary cancer panel.

Three hundred five individuals were included in the study. Of these patients, 280 had a diagnosis of various cancer types. Twenty-five unaffected individuals had no cancer diagnosis, but they had a family history of cancer. Of all individuals, 164 were female and 141 were male. The median testing age of all 305 individuals was 49, with a range of 18–84 years (Table 1). Information about the diagnosis of the patients is summarized in Figure 1. The majority of the patients were diagnosed with breast, colon, and prostate cancer. Twenty-two percent of all variants detected were classified as pathogenic/likely pathogenic, and about one-third of the individuals had negative results (Fig. 2). Most of the unaffected individuals included in the study had a family history of breast cancer (Table 1).

Fourty-nine pathogenic/likely pathogenic variants were detected in 75 individuals (Table 2). Thirty percent of these mutations were in the MUTYH gene. CHEK2, BRCA2, BRCA1, APC, MSH2, and TP53 were the genes with the most pathogenic mutations, respectively (Fig. 2). Nine novel mutations in BRCA1, BRCA2, GALNT12, ATM, MLH1, MSH2, APC, and KIT genes, 5 frameshift, 2 splicing, 1 nonsense, and 1 missense, were assessed as pathogenic/likely pathogenic in 9 patients (Table 2).

One hundred twelve variants reported in previous studies were assessed as variants of uncertain clinical significance in 106 individuals (Fig. 2). Thirty-one novel variants, not previously reported, were assessed as variants of uncertain clinical significance in 34 individuals.

One hundred twenty-three individuals did not have any pathogenic, likely pathogenic, or variants of uncertain significance.

In our study, 81 patients were diagnosed with breast cancer, of these patients 15 had a pathogenic/likely pathogenic variant. Three different BRCA1 variants and 5 different BRCA2 variants were identified. Five patients had a mutation in CHEK2 and 4 of the CHEK2 mutations were c.1556C>T (p.T519M). Also, a patient with breast cancer, stomach cancer, and soft tissue sarcoma had a TP53 mutation.

MLH1 and MSH2 are responsible for Lynch syndrome in the majority of the patients. In this study, there were 55 patients with colorectal cancer, harboring synchronous occurrence of colon and other neoplasms in 2 patients. Twenty-three of 55 patients had a pathogenic/likely pathogenic variant. The majority of them had a mutation in the MUTYH gene, particularly c.800C>T (p.P267L) mutation. Other identified genes in this group were MSH2, MLH1, CHEK2, and APC. In the patients with colorectal cancer, pathogenic/likely pathogenic novel variants were detected in the MSH2, BRCA1, and APC genes including 3 frameshift variants and a nonsense variant.

Twenty patients had a diagnosis of polyposis. Eleven variants in MUTYH, 4 variants in APC, 1 variant in BRCA2, and 1 variant in the ATM gene were detected in these patients. Moreover, 3 patients had a negative result. Almost all polyposis patients with pathogenic/likely pathogenic variants had a family history.

Fourty-five patients who were diagnosed with metastatic prostate cancer and a patient with both prostate and breast cancer were included in the study. Six of these patients had a pathogenic/likely pathogenic mutation in the CHEK2, BRCA2, ATM, and RET genes.

Three of the unaffected 25 individuals had a pathogenic/likely pathogenic mutation. One of them had a novel splicing MLH1 c.307–1G>T variant, others had a mutation in ATM and BRCA1, respectively.

The BRCA1/BRCA2 deletion-duplication analysis performed on patients diagnosed with breast or ovarian cancer, revealed only 2 patients with a deletion in these genes.

The aim of this study was to evaluate genetic results related to most commonly encountered hereditary cancer predisposition syndromes, including hereditary breast cancer syndromes, familial adematous polyposis, hereditary non-polyposis colorectal cancer, Li-Fraumeni syndrome, and hereditary diffuse gastric cancer [Garber and Offit, 2005].

A prevalence of 4% germline mutations, in non-BRCA1/2 breast cancer susceptibility genes, has been reported in patients with breast cancer in previous studies. Tung et al. [2016] reported that the majority of non-BRCA1/2 germline mutations were in the CHEK2 gene, in which most of them were c.1100delC variant [Tung et al., 2016; Apostolou and Papasotiriou, 2017]. In our study, we found a higher rate of CHEK2 mutations in breast cancer patients compared to previous studies. Interestingly, 4 of 5 CHEK2 mutations were rare, p.(T519M) variant, also known as p.T476M (NM_007194.4). This mutation has been classified as likely pathogenic in the literature several times [Roeb et al., 2012]. These results show that the p.(T519M) variant in CHEK2 is at least as important as c.1100delC variant for breast cancer. Although different variants, especially c.1100delC, have been reported in malignant melanoma, to the best of our knowledge, the c.538C>T (p.R180*) variant, which we detected, has been reported for the first time together with this disease [Weischer et al., 2012; Abdel-Rahman et al., 2020]. In addition, the p.R180* variant has been previously reported as pathogenic in different cancers [Momozawa et al., 2018].

The novel KIT c.2383G>T (p.A795S) variant was detected in a patient with breast cancer. The CADD score of this variant was 27.4, and it was predicted as a disease-causing mutation by in silico tools. The pathogenicity of the variant, in the metal ion-binding region of the protein encoded by KIT gene, was 92% according to the UniProt database. We reported this variant as likely pathogenic, according to the ACMG criteria (PM1 and PM2: located in a mutational hot spot and/or critical, well-established functional domain, and variant not found in gnomAD exomes, respectively). Although, somatic mutations of the KIT gene have been reported in breast cancer, to our knowledge, a germline mutation of KIT has never been reported yet [Janostiak et al., 2018]. Unfortunately, we do not have enough data on the clinical effect of the KIT variant.

A novel frameshift mutation of the GALNT12 gene, p.H356fs*29, was identified in a patient with polyposis and assessed as likely pathogenic. Interestingly, this patient also had a homozygous MUTYH mutation. Although the relationship between GALNT12 and polyposis has been reported in the literature, we can not comment about the effect on the patient’s phenotype [Lorca et al., 2017].

A heterozygous mutation in the ATM gene leads to an increased risk of various cancers, especially breast and lymphoma [Choi et al., 2016]. Moreover, previous studies have reported that mutations in this gene are closely related to lung cancer [Liu et al., 2014]. A novel splicing variant in ATM (c.1802+1G>T) was assessed as likely pathogenic in a patient who had lung cancer, renal cell carcinoma, and non-Hodgkin lymphoma. The coexistence of lung cancer and lymphoma in our patient suggests that this mutation is responsible for the patient’s phenotype and that the ATM gene causes the coexixtence of multiple cancers such as the TP53 gene.

The protein encoded by the BLM gene plays an important role in regulation of homologous recombination and genomic integrity. A heterozygous mutation, which alters BLM function, is closely related to an increased susceptibility to cancer [Broberg et al., 2009]. Although heterozygous BLM mutations have been reported in breast cancer and other cancer types in several studies, they have rarely been reported in ovarian cancer [Schubert et al., 2019; Kaneyasu et al., 2020]. We identified a W881* variant in BLM in a patient with ovarian cancer. The variant was nonsense and has been previously reported as pathogenic.

CDH1 encodes E-cadherin glycoprotein which plays a key role in a cell-cell adhesion and invasion suppression. Loss-of-function mutations in the CDH1 gene cause hereditary diffuse gastric cancer and lobular breast cancer. Also, it has rarely been reported that loss-of-function mutations lead to susceptibility of colorectal cancer and other cancer types [Figueiredo et al., 2019]. We identified a missense mutation in the CDH1 gene in a 22-year-old male patient with colon cancer.

Remarkably, the male patient who was diagnosed with both prostate and breast cancer had no variant in cancer-related genes and BRCA1/BRCA2 deletion-duplication analysis was also normal.

In our study, we obtained interesting and novel variants which could be related to hereditary cancer syndromes. Although there are different studies on hereditary cancer syndromes in the literature, still more studies are needed. Here, we confirmed that NGS is an indispensable method for risk assessment in cancer families, since multiple genes can be studied simultaneously in one panel, and from a single sample. The results of this study will provide useful data for future studies on the familial transmission of cancer.

We would like to thank the patients for their cooperation.

The protocols used in this study were in compliance with the Declaration of Helsinki. This study was approved by the Ethics Committee of the faculty of medicine at Ataturk University Data were obtained retrospectively for clinical purposes.

The authors declare no conflicts of interest.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

P.E. was involved in the collection of clinical data, clinical examination, and wrote the manuscript. C.Y.K. and A.T. conducted clinical evaluations and managed the patients. G.O. prepared the samples for sequencing.

The data that support the findings of this study are available from the corresponding author upon reasonable request.