Genetic testing for hereditary cancer syndromes in Tunisian patients: Impact on health system

Highlights

• •A phenotypic overlap between cancer syndromes was observed which may cause clinical misdiagnosis. This underscores the crucial role of genetic testing and the necessity of a multidisciplinary approach to enhance the effectiveness of the cancer health system.
• •Within Tunisian patients with cancer, genetic analysis allowed the identification of multiple disease-causing variants accounting for 44.68% of patients, with different inheritance patterns (autosomal dominant or recessive transmission).
• •Two pathogenic TP53 mutations previously unreported in the Tunisian population were identified. Phenotype-genotype correlation suggests that the (c.638G>A) mutation carrier may exhibit an atypical/mild LFS phenotype.
• •A novel homozygous mutation (c.3254dupT) in the BLM gene was identified in a patient with multiple primary cancers, along with other phenotypes suggestive of a bloom syndrome diagnosis.
• •Results provided in this study may contribute to the establishment of a national database on known and novel hereditary cancer syndromes in Tunisia and the broader African context.

Abstract

Introduction

Cancer management in Africa faces diverse challenges due to limited resources, health system challenges, and other matters. Identifying hereditary cancer syndromic cases is crucial to improve clinical management and preventive care in these settings. This study aims to explore the clinicopathological features and genetic factors associated with hereditary cancer in Tunisia, a North African country with a rising cancer burden

Materials and methods

Clinicopathological features and personal/family history of cancer were explored in 521 patients. Genetic analysis using Sanger and next-generation sequencing was performed for a set of patients

Results

Hereditary breast and ovarian cancer syndrome was the most frequent cluster in which 36 BRCA mutations were identified. We described a subgroup of patients with likely ‘’breast cancer-only syndrome’’ among this cluster. Two cases of Li-Fraumeni syndrome with distinct TP53 mutations namely c.638G>A and c.733G>A have been identified. Genetic investigation also allowed the identification of a new BLM homozygous mutation (c.3254dupT) in one patient with multiple primary cancers. Phenotype-genotype correlation suggests the diagnosis of Bloom syndrome. A recurrent MUTYH mutation (c.1143_1144dup) was identified in three patients with different phenotypes

Conclusion

Our study calls for comprehensive genetic education and the implementation of genetic screening in Tunisia and other African countries health systems, to reduce the burden of hereditary diseases and improve cancer outcomes in resource-stratified settings.

Graphical abstract

Introduction

Cancer remains a leading cause of mortality worldwide. In Africa, it has been suggested that by 2030 there will be a 70% increase in new cancer cases [1] Hereditary factors of cancer account for approximately 5–10% of all diagnosed cases [2]. This evidence derives originally from observations of cancer clustering in families with genetically determined syndromes. The typical clinical features of hereditary cancer include cancer diagnosis in several family members, early onset of the disease, and multiple primary cancers [3]. Several genes involved in known inherited cancer susceptibility syndromes have been identified. BRCA1/2 are the most prevalent, and well-known genes associated with Hereditary Breast and Ovarian Cancer Syndrome1(HBOC). BRCA genes are mutated in approximately 25% of patients with breast and ovarian cancers [4]. This syndrome is also associated with an increased risk of developing other malignancies such as melanoma, prostate, and pancreatic cancers [5]. Beyond BRCA-related disorder, other syndromes conferring a high risk of cancer exist. They include Li-Fraumeni Syndrome (LFS), Lynch Syndrome2 (LS), Hereditary Diffuse Gastric Cancer3(HDGC), Peutz-Jeghers Syndrome4 (PJS), and Cowden Syndrome5 (CS). Those diseases are transmitted in an autosomal dominant mode [6] and predispose to a diverse spectrum of phenotypes (Supplementary Table 1). An additional 2–3% of cancer cases are due to mutations in genes associated with more rare genetic diseases such as Bloom syndrome which is caused by homozygous or compound heterozygous mutations in the BLM gene [7]. Moreover, three Fanconi's anemia genes (FANCD1, FANCN, and FANCJ) have been proven to be cancer-susceptibility genes when found in a heterozygous state [8]. The FANCF gene mutations have also been found in breast cancer, acute leukemia, and patients with ovarian cancer [9]. So far, half of patients with familial cancer are negative for all known cancer-predisposing genes, suggesting the existence of ultra-rare mutations on novel candidate genes [9]. Analysing co-occurrence patterns of complex diseases using accurate bioinformatics tools, especially in under-investigated populations, may help to identify new cancer-predisposing syndromes and to assess clinical associations between Mendelian diseases and cancer [10]. This has the potential to enhance patient care, improve healthcare systems, and provide opportunities for developing targeted therapies. In Tunisia, most studies have focused on investigating HBOC syndrome by analyzing BRCA genes in patients with breast cancer, [11] which makes a substantial part of hereditary cancer missed. In the present study, we aim to explore the clinicopathological and genetic features associated with different hereditary cancer syndromes in Tunisian patients with cancer.

Material and methods

Patients and ethics statement

A total of 521 patients with histologically confirmed cancers (breast, ovarian, endometrial, prostate, and pancreatic cancers) that have occurred in the context of familial predisposition were included in this study. Probands were recruited from different medical or surgical oncology departments in Tunisia, including Abderrahman Mami Hospital of Ariana, Military Hospital of Tunis, Salah Azaiez Institute of Cancer, and Farhat Hached University Hospital of Sousse. All patients gave written informed consent. The study was approved by the biomedical ethics committee of Pasteur Institute (2020/24/I/LR16IPT). Data on personal and family history of diseases and clinicopathological characteristics of cancer were collected from questionnaires and medical records.

Patients clustering

Patients were clustered in different hereditary cancer syndromes based on the National Comprehensive Cancer Network (NCCN) clinical practice guidelines (version 3.2019 and 2.2021). The detailed criteria are illustrated in supplementary Table 1.

Statistical analysis and data mining

The clinicopathological features of patients with breast and ovarian cancer in each characterized syndrome were analyzed using R software. Text mining for the unclassified patients was performed using the ngram R package version 4.0.3.

DNA isolation

Genomic DNA was isolated from peripheral blood using the QIAamp blood DNA mini QIAcube Kit from Qiagen according to the manufacturer's instructions.

Genetic investigation

To explore the genetic basis of cancer in patients clustered in the different syndromes, more stringent criteria were used and allowed the selection of 94 patients that meet at least one of the following criteria: 1/Multiple primary cancers. 2/Age at diagnosis ≤ 36 years. 3/Breast cancer before 46 years, and at least one close blood relative with cancer associated with LFS spectrum (supplementary Table 1) before 56 years, or with multiple primary tumors. 4/Three or more close-blood relatives with cancer of all ages. 5/Two close blood relatives with cancers, including one cancer case < 40 years, or 1 male breast cancer, or 1 case of ovarian, prostate, or pancreatic cancer 6/Triple-negative breast cancer ≤ 40 years. 7/Patient with epithelial ovarian cancer or a first-degree relative with epithelial ovarian cancer. 8/Lobular breast cancer with close blood relative diagnosed by gastric cancer, 9/Patient with endometrial cancer. A study flowchart is illustrated in Fig. 1.

Fig. 1.

Study design. HBOC S: Hereditary Breast and Ovarian Cancer Syndrome, LS: Lynch syndrome, LFS: Li-Fraumeni syndrome, NCCN: National Comprehensive Cancer Network.

Sanger and next-generation sequencing

Before performing Next Generation Sequencing (NGS), Sanger sequencing was used to screen known recurrent mutations associated with hereditary predisposition to cancer in the Tunisian population. Sanger sequencing was also performed for targeted screening of knowing mutation within those families.

NGS was then performed on 79 patients (56 breast cancer cases and 23 with other cancer types). TGS (targeted gene sequencing) was conducted in the majority of cases, given that most patients met HBOC criteria. WES (whole exome sequencing) was used for a set of patients with heterogeneous family history of diseases for a more comprehensive genetic investigation.

TGS including the following genes (BRCA1, BRCA2, PALB2, CDH1, PTEN, TP53, RAD51D, MLH1, MSH2, MSH6, PMS2, EPCAM, ATM, BRIP, CHEK2, STK11, MRE11A, NBN, RAD50, BARD1, BLM, XRCC, MUTYH) and whole exome sequencing (WES) were performed on 60 and 19 patients respectively. Whole Exome Sequencing was performed on Illumina HiSeq2000 platform using TruSeq v3 chemistry with paired-end (2 × 100 bp). Samples were prepared according to Agilent SureSelect Protocol Version 1.2, and enrichment was carried out according to Agilent SureSelect protocols.

Targeted gene sequencing was carried out using the GeneReader NGS System. Enrichment steps were carried out using an automated protocol on the GeneRead QIAcube (QIAGEN, Hilden, Germany) using the GeneRead Clonal Amp Q Kit (QIAGEN, Hilden, Germany), according to the manufacturer's protocol. All coding regions in addition to flanking regions in adjacent introns (20 bp), were amplified. For variant classification, the ACMG standards were used.

Results

From a total of 480 breast cancer cases, 85.4% (410 cases) met the clinical criteria for at least one of the investigated syndromes. HBOC syndrome was the most frequent group representing 84.5% of the breast cancer cohort, followed by LFS (15.6%) and LS (10.6%). Several patients met inclusion criteria for more than one syndrome (resulting in the ad being more than 100%). Almost all patients included in the LFS group have fulfilled HBOC-related criteria (98.6%). Overlap between HBOC, LFS, and LS clusters occurred in 9.7% of all breast cancer cases (Fig. 2). All patients with ovarian, pancreatic, and prostate cancer were clustered in the HBOC syndrome, and 10.52% of patients with ovarian cancer also met LS criteria (Fig. 2).

Fig. 2.

Clustering and overlap between the investigated syndromes among patients with A) breast cancer; B) ovarian cancer. The figure was generated using VENNY web server (https://bioinfogp.cnb.csic.es/tools/venny/index.html). BC: breast cancer, HBOC: hereditary breast and ovarian cancer, OC: ovarian cancer, LFS: Li Fraumeni syndrome, LS: lynch syndrome.

No patients with PJS, CS, or HDGC syndrome have been identified based on available data. In addition to the rarity of those syndromes, it might be due to the lack of exhaustive clinical data.

1/ Clinicopathological features of clusters

Considering all patients with breast cancer, luminal B subtype, invasive ductal carcinoma, a high-grade SBR (III), and the presence of lymph node invasion were the most frequent features observed in all syndromic groups. Detailed characteristics are described in Table 1.

Table 1.

The distribution of clinicopathological parameters of patients with breast cancer in each hereditary cancer syndrome.

Features Syndrome	Mean diagnosis age	Molecular state				SBR grade			Mean ki 67% index	Mean Tumor size(mm)	Histological type				Lymph node invasion		Metastasis
		LumA%	LumB%	TN%	HER2+%	I%	II%	III%			IDC%	ILC%	MLDC%	Others	N−%	N+%	M−%	M+%
HBOC	41 years	19.13	62.44	28.57	12.32	10.33	5.16	44.94	38%	36.76	91.28	3.13	2.43	3.13	45.73	54.27	25.95	74.05
LS	37 years	5.88	47.18	20.58	14.70	10.00	43.33	64.66	43%	23.77	84.37	3.12	6.25	6.25	47	53	66.66	33.34
LFS	29 years	12.28	55.76	24.56	47.40	9.25	35.18	55.55	36.97%	22.88	94.82	1.72	0	3.44	48.64	51.36	66.66	33.34

HBOC: hereditary breast and ovarian cancer, IDC: invasive ductal carcinoma, ILC: invasive lobular carcinoma, LS: Lynch syndrome, LFS: Li-Fraumeni syndrome, LumA: luminalA, LumB: luminalB, MLDC: mixed lobular and ductal carcinoma, TN: Triple negative.

Regarding HBOC cluster, bilateral breast cancer was detected in 8.33%, and male breast cancer (MBC) in nearly 2% (n = 5) of patients. The mean age at diagnosis for MBC patients was 59 years, with all being diagnosed with positive lymph node status and Luminal subtype. Two patients developed Ewing sarcoma in their childhood at 13 and 14 years. Breast cancer was diagnosed at 28 and 35 years respectively and showed triple-negative subtype and node-positive tumors for both cases. Among HBOC-group, breast cancer appeared as a specific site in patients and their family members in 4.24% of cases. No specific features were observed, meaning that they shared the same clinicopathological features as the global HBOC group. For patients with ovarian cancer, the mean age at diagnosis was 52 years and 80% have serous ovarian carcinoma.

Regarding LS group, a family history of colorectal cancer appeared in almost all patients with breast cancer. One patient was carrying bilateral breast cancer and endometrial carcinoma.

Among all cases with breast cancer, none fulfilled HDGC syndrome-associated criteria, although 10 patients had lobular breast cancer (LBC). Family history of breast and gastric cancers was documented in 50% and 30% respectively, and 20.3% presented a bifocal lobular breast cancer.

2/ Patients with likely new syndromes

Within the Breast cancer cohort,14.6% of patients with a cancer family history did not match any of the known syndromes. The mean diagnosis age was 58 years. Analysis of cancer types among their close relatives showed 3 clusters: Cluster 1: Breast, lung, esophagus, and colon. Cluster 2: Breast, lung, pharynx, liver, and brain. Cluster 3: Breast, lung, stomach, uterus, thyroid, and bones. In addition to cancer spectrum, questionnaires and medical records examination showed the co-occurrence of a heterogeneous spectrum of other diseases among patients' close relatives (Fig. 3). Nephropathy, neuropathy (Alzheimer, Parkinson/ Epilepsy), asthma, anemia, and deafness were the most frequent comorbidities that occurred in the first cancer cluster. Neuropathy was also observed in cluster 2 along with ocular and dermatological diseases. Thyroid diseases and pneumopathy appeared exclusively with cancer types belonging to cluster 3. Anemia occurred among all three clusters. Metabolic pathologies such as cardiovascular diseases and diabetes occurred respectively in 64.86% and 50% of all the investigated families. Epidemiological features of the unclassified patient showed a high rate of consanguinity and/or endogamy (40.90%). Infertility was observed in 38 % of cases. Around 46% of patients had used oral contraceptives and the majority (88.88%) had a stressful life status. Moreover, 46.29% of patients had a high Body mass index (>30) and 36.6% had a high-fat diet.

Fig. 3.

Clustering and disease occurrence among patients with breast cancer having likely new syndromes. Cluster 1 (Yellow): Breast, lung, esophagus, and colon cancers. Cluster 2 (Pink): Breast, lung, pharynx, liver, and brain cancers. Cluster 3 (Blue): Breast, lung, stomach, uterus, thyroid, and bones cancers.

3/ Genetic investigation and genotype-phenotype correlation

Genetic analysis allowed the identification of multiple disease-causing variants accounting for 44.68% of patients, with different inheritance patterns (autosomal dominant or recessive transmission) (Fig. 4). A total of 36 heterozygous BRCA mutations have been identified [12]. Those mutations and the associated phenotypes are illustrated in supplementary Table 2. Two heterozygous TP53 mutations, 1 BLM homozygous mutation, 3 MUTYH mutations (1 homozygous and 2 heterozygous), and one FANCF homozygous variation were also detected (Table 2). Additional variations in CHEK2, ATM, BRIP, and PALB2 genes were also identified (unpublished data).

Fig. 4.

Main genetic results and their associated hereditary syndromes.

Table 2.

Correlation between the identified mutations, phenotype spectrum and cancer family history.

Gene	Associated Syndrome	ACMG clasification	Mutation, protein change	State	Phenotype	Diagnosis age (years)	Family history of cancer	Consanguinity//Endogamy
TP53	Li-Fraumeni	Pathogenic	c.733G>A (p.Gly245Ser)	Heterozygous	Breast cancer	22	1Breast	No
TP53	Li-Fraumeni	Pathogenic	c.638G>A (p.Arg213Gln)	Heterozygous	Ovarian cancer	77	1Gastric	No
FANCF	Fanconi Anemia	Benign	c.*1338dup	Homozygous	Pancreatic Cancer	63	6Breast, 1Pharynx, 1Brain, 1Colorectum, 1Leukemia, and (2Fanconia anemia)	No
MUTYH	MUTYH-associated polyposis	Pathogenic	c.1143_1144dup (p.Glu382fs)	Homozygous	Endometrium cancer	50	3Colon, 2Breast	Endogamy
MUTYH	MUTYH-associated polyposis	Pathogenic	c.1143_1144dup (p.Glu382fs)	Heterozygous	Breast cancer	31	2Ovarian, 2 esophagus, 1Colon,	Endogamy
MUTYH	MUTYH-associated polyposis	Pathogenic	c.1143_1144dup (p.Glu382fs)	Heterozygous	Pancreatic cancer	68	1Ovarian, 1Meyloma, 1Brain, 1 esophagus	No
BLM	Bloom Syndrome	Likely Pathogenic	c.3254dupT (Arg1086Lysfs*7)	Homozygous	Endometrium/Breast/colon cancers	45, 47, 50	No	Consanguinity

“Breast cancer only” syndrome

Among BRCA+ mutation carriers with breast cancer, there are 5 patients (4 women, 1 male) with personal and/or family history of breast cancer only. All patients showed a luminal B subtype and the mean age at diagnosis was 35.2 years. Four of these patients carried the same BRCA2 mutation (c.1310_1313delAAGA), while one patient harbored BRCA1_ c.211dupA mutation. Both variants were reported as recurrent mutations in the Tunisian population and were identified in other patients with breast cancer but not in cases of ovarian cancer. Both mutations are located in breast cancer cluster regions (BCCR) among BRCA2 and BRCA1 genes.

TP53 mutations

Two pathogenic TP53 mutations reported for the first time in the Tunisian population were detected.

The first mutation, c.733G>A is located in exon 6 and results in the replacement of glycine with serine at codon 245 of the TP53 protein (p.Gly245Ser). It was identified in one patient with early-onset breast cancer at the age of 22 years (fulfilling both HBOC and LFS criteria). Based on the genetic result, HBOC syndrome was excluded and the diagnosis of LFS was confirmed. Anatomopathological examination revealed a PR+, ER+, HER2+ tumor, invasive ductal carcinoma, SBR grade II, with Ki-67 of 30%. The tumor size was 65 mm, and multiple sites of metastasis were detected. Family history showed a grandmother who died from breast cancer at 60 years old.

The second TP53 mutation; c.638G>A (p.Arg213Gln) is located in exon 5 of the TP53 gene which corresponds to the DNA binding domain of P53 protein. The arginine at codon 213 was replaced by a glutamine an amino acid with different properties, which induced gene loss of transactivation capacity, resulting in an inability to up-regulate cell cycle arrest and apoptosis. It was identified in a patient diagnosed with epithelial ovarian cancer at 77 years, who had a cousin diagnosed with gastric cancer. According to NCCN guidelines, this patient fulfills HBOC syndrome suggestive criteria and not those of the LFS (supplementary Table 1). c.638G>A had an allele frequency of 0.00001, and was reported as associated with LFS, excluding a CHIP case. Thus, this patient has been considered as an atypical LFS case.

MUTYH mutation

Genetic analysis showed the presence of homozygous frameshift duplication in the MUTYH gene: c.1143_1144dup (p.Glu382fs), classified as Pathogenic/Likely pathogenic in the ClinVar database. It was identified in one patient with metastatic endometrial cancer at 50 years. Our findings confirmed the diagnosis of MUTYH-associated polyposis syndrome (MAP), although her clinical data were suggestive of LS. Indeed, this patient's family history showed 3 cases of colorectal cancer and 2 members with breast cancer diagnosed at a young age. the same mutation was identified in a heterozygous state in two patients. The first proband was a 31-year-old breast cancer woman with invasive ductal carcinoma, a Ki 67 of 60%, a luminal B subtype, with HER2 amplification, and a T2N0M0 stade, who carries a family history of ovarian, esophagus, and colorectal cancers.

The second carrier was a man diagnosed with pancreatic adenocarcinoma at 68 years. Family history showed the presence of ovarian, Myeloma, esophagus, and brain cancers. All those patients derived from the same southern region of Tunisia, characterized by a high endogamy rate.

BLM mutation

A new BLM mutation was identified: NM_000057.4 (c.3254dupT_Arg1086Lysfs*7) in a homozygous state in one patient. It is located in exon 17 of BLM gene, causing premature truncation of the BLM protein at amino acid position 1092. It, therefore, affects BLM protein linear sequence and structure. This mutation has never been reported in previously published studies neither in Tunisia nor in other populations. It was identified in one patient diagnosed with three primary cancers who fulfilled both LS and LFS criteria. The first diagnosed cancer was an endometrial carcinoma at 45 years. Two years later, a diagnosis of breast tumor was made followed by colon cancer after three years. This patient had a small size (short stature: 149 cm, 52 kg). She was infertile, had early menarche (9 years) and premature menopause (32 years). Thus, phenotype-genotype correlation was in favor of Bloom syndrome diagnosis. Consanguineous marriage was noted in her family. No family history of cancer was reported among the patient's relatives.

FANCF variation

One FANCF gene variation: NM_022725:c.*1338dup, was identified in one woman diagnosed with pancreatic cancer at 63 years. It was described for the first time in Tunisian population. This mutation is localized on the 3 UTR region and classified as a variant of unknown significance in the ClinVar database. Cancer was diagnosed in many family members (maternal side): breast cancer in 1 aunt, 1 niece, and 4 cousins; pharynx cancer in 1 uncle and 1 aunt; brain cancer in 1 aunt; colorectal cancer in 1 aunt; leukemia in 1 aunt. Fanconi Anemia was diagnosed in her 2 nephews (Children of consanguineous marriage) who died at 18 and 36 years respectively.

Familial adenomatous polyposis and inconclusive genetic results

No pathogenic/likely pathogenic mutations were identified in MMR and CDH1 genes. Among patients with negative genetic results, there was one patient diagnosed with breast and ovarian cancer at 58 years. Genetic testing on her brother diagnosed with Familial adenomatous polyposis (FAP) revealed the presence of a new heterozygous germline mutation in the APC gene: 3771_3775del (exon15), causing a stop coding at the nucleotides 3822_3824. Family history showed many other cases of heterogeneous carcinoma among her first, second, and third-degree relatives: Breast cancer at a young age (3 cases), lymphomas (2 cases), lung carcinoma (1 case) liver cancer (2 cases), and colorectal cancer (3 cases). The APC mutation was absent in this patient, excluding PAF diagnosis.

Discussion

In the current report, we have investigated the clinicopathological features and the genetic basis of hereditary cancer syndromes in the Tunisian population. Disease-causing variants within different genes and different inheritance modes were identified in 44.68% of patients, reflecting the genetic heterogeneity of hereditary cancer forms in Tunisia. The complex patterns of cancer predisposition are related to the key roles of affected genes in DNA damage repair [13]. Complete loss of BRCA1 or TP53 protein function is lethal during embryonic development, and individuals with heterozygous mutations that retain some DNA repair capacity, survive with a high risk of cancer development [14,15]. However, homozygous mutations of other genes like BLM and FANCA are associated with rare autosomal recessive syndromes known as ‘’Chromosomal Breakage Syndromes’’, namely Bloom syndrome and Fanconi Anaemia [16]. Genes involved in those diseases interact together for the maintenance of genomic integrity. Thus, individuals with homozygous mutations survive with a wide clinical spectrum and multiple impairments, with cancer being the most frequent one [[17], [18]]. In this study, we identified a novel homozygous frameshift mutation, BLM:c.3254dupT_(Arg1086Lysfs*7) that seems to cause BLM protein truncation/absence, which leads to chromosome instability, and the expression of Bloom Syndrome-clinical features. Moreover, we have identified a homozygous mutation in MUTYH gene known as a recurrent mutation in patients diagnosed with MAP [[19], [20]]. The same mutation was identified in a heterozygous state in two patients with distinct phenotypes, confirming that homozygous mutations in genes associated with rare genetic diseases predispose to cancer when found in a heterozygous state. Indeed, a subset of Fanconi Anaemia-genes namely FANCN (PALB2) and FANCJ (BRIP1), are associated with hereditary breast and/or ovarian cancer predisposition. Accordingly, cancer risk has been evaluated, and clinical management is also well-established [21]. Nevertheless, the role of heterozygous mutations in other Fanconi Anemia genes regarding cancer predisposition is a matter of discussion [22]. Paradoxically, in this study, we identified a homozygous variation in the FANCF gene in one patient with pancreatic cancer. Beyond the ACMG classification, this mutation might be associated with cancer and Fanconi Anemia development. However, due to the lack of biological samples of the patient's relatives, its role in the different phenotypic expressions could not be properly assessed.

Moreover, the risk of autosomal recessive genetic disorders is higher in endogamous populations [23]. In Tunisia, despite the educational, demographic, and behavioral changes that have taken place during the last decades, consanguinity and endogamy rates are still high and may reach 98% in some geographic isolates [12]. This may explain the high rates of autosomal recessive diseases observed in this country [24,25] and may influence the prevalence and frequency of novel and founder mutations associated with autosomal dominant syndromes. In the current study, consanguinity and endogamy have been observed in the two patients carrying the BLM and MUTYH homozygous mutations. Furthermore, all MUTYH:c.1143_1144dup carriers derived from the same geographic region, suggesting a possible founder effect of this mutation.

So far, 29 heterozygous BRCA mutations have been reported in the Tunisian population, including 11 novel variations identified recently by our research group [12]. Based on our results, HBOC is probably the most frequent cause of hereditary breast and ovarian cancers in Tunisia. Among the HBOC cluster, we have identified a set of patients with likely ‘’breast cancer only syndrome’’ which may affect both genders. Those patients seem to carry a specific BRCA2 mutation. However, this result needs to be explored in a larger pedigree to assume that there is no ovarian cancer risk. Two distinct TP53 pathogenic mutations associated with LFS were also identified. Unlike HBOC syndrome, LFS is an under-explored disease in Tunisia, with only 2 reports being published so far [24,25]. Thus, this study, along with other findings, will contribute to the establishment of a national database on hereditary cancer syndromes in Tunisia, and encompassing the broader North African and African region. The (c.638G>A) was previously described in both classic and non-classic LFS families presenting cancers that are outside the narrow spectrum of LFS, and with late-onset malignancies, suggesting a variable expression and penetrance of this mutation [26]. The carrier of this mutation had ovarian cancer diagnosed at 77 years. The variant allele frequency of this variant was under 0.02 excluding CHIP case and suggesting an atypical/mild LFS phenotype [26,27]. The second TP53 mutation (c.733G>A) is also described for the first time in the Tunisian population. It is segregated in multiple families with LFS or LFS-like syndrome [28,29]. Clinicopathological characters of this mutation carrier showed an emerging phenotype of a classic LFS-breast cancer [30].

Furthermore, we clustered patients in different cancer syndromes based on standard phenotypic criteria. Interestingly, Patients with breast cancer that did not match any of the known syndromes-criteria showed common epidemiological features. Analysis of medical family history revealed the association between cancer and other disorders. Thus, we hypothesize i) The emergence of new syndromes with their own clinical and genetic pattern. Ii) The existence of phenotypic causality, in which one disease is a direct cause of another (e.g. neuropathy). Iii) Those patients share the same environmental factors /lifestyle that influence their epigenetic mechanisms leading to a common health status (occurrence of both cancer and non-cancer diseases). Iv) The presence of known cancer hereditary syndromes such as LFS and LS, with atypical phenotypes specific to the Tunisian population. This calls for extensive clinicopathological, genetic, and epigenetic investigations to bring new insights into the landscape of complex and rare diseases in under-investigated populations, such as North Africa. Indeed, analysis of the co-occurrence of complex pathologies and Mendelian diseases has the potential to elucidate disease etiology and identify new comorbidities [31]. Moreover, African populations may carry a distinct genetic background that influences disease expression and treatment.

Despite being caused by mutations in different genes, diseases with the same phenotypic expression such as ‘cancer syndromes’ are clinically similar [32]. This is based on the assumption that genes mutation indirectly cause ‘cancer’ by directly affecting biological pathways such as DNA repair, cell cycle control, and apoptosis [32,33]. As an example, MAP and LS showed several phenotypic and physio-pathological overlaps [34]. Both the base-excision-repair and the mismatch-repair pathways are involved in the removal of oxidative DNA damage, in different ways or in a synergistic manner [35]. In this study, we identified MUTYH homozygous mutation in an ‘LS-like’ patient, diagnosed with endometrial cancer and with a family history of colorectal and breast cancers. The same for the BRCA1: c.2433delC carrier diagnosed with breast and endometrium cancers. We also described two cases of breast cancer with a personal history of Ewing sarcoma. Clinical features of cancer showed a BRCA1-deficient-like phenotype. Nevertheless, genetic testing was negative suggesting that the development of breast cancer as a second tumor may be a radiation-induced event, without the presence of a hereditary component. Breast cancer has previously been reported as a secondary tumor in Ewing sarcoma patients, in whom the risk is likely attributed to radiation fields in the chest area [36]. Moreover, it has been shown that the fusion protein EWS/FLI1 formed (known as the molecular cause of Ewing sarcoma) acts as a pathogenic transcription factor, which causes R‐loops and blocks BRCA1 repairing pathways [36].

Finally, the assessment of hereditary susceptibility to cancer holds great significance from the understanding of cancer causation to the entire health system, shaping prevention, and therapy options for individuals and families [37]. For patients suffering from cancer with DNA repair disorders, conventional chemo or radiotherapy would lead to profound genomic instability and consequently, enhanced cancer development. Thus, for Bloom syndrome and LFS patients, chemo- or radiotherapy should be avoided if there are other possible options [[38], [39], [40]]. Whole body RMI was recommended for BLM and LFS patients for the early detection of other cancers. Moreover, based on our genetic results, colonoscopy with polypectomy has been recommended every one to two years for MAP patient and her family members, to reduce the risk of colorectal cancer [41]. While genetic screening has clear benefits to increased cancer screening and prevention options, there are potential ethical issues and psychological harm, especially for unaffected family members [42]. Moreover, sociocultural taboos surrounding ‘cancer’ prevent people from talking about their family health history, particularly in underdeveloped countries [43]. Psychological management and social well-being of families with hereditary diseases become interesting challenges for health system stakeholders. A multidisciplinary specialist comprising genetic counselors, clinicians, researchers, social workers, psychologists, and support groups, plays a crucial role in raising awareness among patients and families [44]. Despite enduring socio-cultural and financial challenges, including the costs of genetic screening, our 'resource-stratified' strategy, aiming to integrate genomics into cancer control and clinical management, seems to be efficient and has the potential to influence the traditional health system policies in Tunisia.

Conclusion

This study highlights several important topics from both clinical and research sides, revealing significant issues within the health system. It may contribute to the establishment of a national database on hereditary cancer syndromes in Tunisia and the broader African context. We showed that there is a phenotypic overlap between cancer syndromes which may cause clinical misdiagnosis, highlighting the role of multipanel/WES genetic testing and the necessity of more collaborations between clinicians and geneticists to properly diagnose, personalize treatment, and orient screening and surveillance policies. Our study also calls for comprehensive genetic education and premarital genetic counseling programs, to reduce the burden of hereditary diseases and cancer development in inbred communities.

Ethics statement

This study was approved by the biomedical ethics committee of Pasteur Institute (2020/24/I/LR16IPT).

-Written informed consent was obtained from all the patients for publication. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

CRediT authorship contribution statement

Nouha Jandoubi: Methodology, Investigation, Data curation. Maroua Boujemaa: Methodology. Najah Mighri: Methodology. Nesrine Mejri: Investigation. Sonia Ben Nasr: Investigation. Hanen Bouaziz: Investigation. Yosra Berrazega: Investigation. Haifa Rachdi: Investigation. Nouha Daoud: Investigation. Aref Zribi: Investigation. Jihene Ayari: Investigation. Houda El Benna: Investigation. Soumaya Labidi: Investigation. Abderazzek Haddaoui: Investigation. Ridha Mrad: Investigation. Slim Ben Ahmed: Investigation. Hamouda Boussen: Investigation, Funding acquisition. Sonia Abdelhak: Supervision, Project administration, Funding acquisition. Samir Boubaker: Writing – review & editing, Supervision, Funding acquisition. Yosr Hamdi: Writing – review & editing, Supervision, Project administration, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

• The data used to support the findings of this study are included within the article or the supplementary information file. More details are available on request from the corresponding author.