Cancer Predisposition Syndromes and Medulloblastoma in the Molecular Era

Medulloblastoma is the most common malignant brain tumor in children. In addition to sporadic cases, medulloblastoma may occur in association with cancer predisposition syndromes. This review aims to provide a complete description of inherited cancer syndromes associated with medulloblastoma. We examine their epidemiological, clinical, genetic, and diagnostic features and therapeutic approaches, including their correlation with medulloblastoma. Furthermore, according to the most recent molecular advances, we describe the association between the various molecular subgroups of medulloblastoma and each cancer predisposition syndrome. Knowledge of the aforementioned conditions can guide pediatric oncologists in performing adequate cancer surveillance. This will allow clinicians to promptly diagnose and treat medulloblastoma in syndromic children, forming a team with all specialists necessary for the correct management of the other various manifestations/symptoms related to the inherited cancer syndromes.


INTRODUCTION
Medulloblastoma (MB) is the most frequent malignant tumor of the central nervous system (CNS) in childhood, representing 15-20% of all CNS neoplasms (1). It mainly affects the pediatric age with a 10-fold higher frequency than in adults (2). Children are diagnosed generally between 2 and 8 years old (median of 6 years old), with 50% of cases occurring in children under 5 years old and with a male/female ratio of 2:1 (3).

METHODS
The authors conducted a literature search describing the issue of CNS tumors and cancer predisposition syndromes. Research studies were selected based on research topics ("cancer predisposition syndrome," "brain tumor genetics," "brain tumor cancer predisposition syndrome," "medulloblastoma predisposition syndromes," "medulloblastoma in childhood") found in PubMed considering the last 10 years until April 2020. These studies were classified according to their relevance. In the selected studies the data were carefully evaluated, and they are described in detail and discussed in the following sections. The association between the different cancer predisposition syndromes described below and the related molecular subgroups of MB is summarized in Figure 1. The main cancer predisposition syndromes associated to pediatric MB and their related molecular, pathological, clinical, and prognostic features are summarized in Table 1.

Medulloblastoma Molecular Subgroups
Main features of MB subgroups are: • Wingless (WNT) accounts for about 10% of diagnoses and is found mainly in girls with a peak between 10 and 12 years of age. The most common histological variant is classic. Approximately 85-90% of MB WNT harbor somatic mutations in exon 3 of Catenin beta 1 (CTNNB1), which causes stabilization and nuclear accumulation of b-catenin leading to uncontrolled activation of WNT signaling (23,30). Patients with MB WNT without CTNNB1 mutations can harbor a mutant APC tumor suppressor gene, which is involved in the ubiquitination and consequently degradation of b-catenin (22). MB WNT have a low tendency to metastasize and patients under 16 years of age have an excellent prognosis. Therefore, some ongoing clinical trials, PNET5 and SJMB12, are currently investigating de-escalation of therapy (19). • Sonic hedgehog (SHH) accounts for about 30% of all MB diagnoses and has a bimodal distribution, with peaks in children <3 years of age and in young adults >16 years of age (21). This subgroup affects both sexes almost equally with a slight predominance in males among infants (31). The histological variant is frequently desmoplastic/nodular. MBs-SHH harbor germline or somatic mutations in genes involved in SHH signaling pathway, leading to its constitutive activation, such as deletions or loss-of-function alterations in Patched 1 (PTCH1) (43% of patients) or Suppressor of fused (SUFU) (10%), activating mutations in Smoothened (SMO) (9%), amplification of GLI1/GLI2 (9%) or MYCN (7%) (23,32). More recently, four SHH subtypes have been identified (SHH a , SHH b , SHH g , SHH d ) with distinct biological and clinical features (33). Older children with MB SHH can harbor germline or somatic Tumor Protein 53 (TP53) mutations, associated with a poor prognosis (25,32). • Group 3 accounts for about 25-28% of all MB diagnoses and is exclusively found in childhood, with a male sex predominance. It is associated with metastatic disease at diagnosis and with large cell/anaplastic histological variant. About 17% of Group 3 MBs harbor MYC amplification. Among MB subgroups, Group 3 is characterized by the poorest prognosis, especially in the presence of metastatic disease, isochromosome 17q, and MYC amplification (19). • Group 4 is the most common MB molecular subgroup, accounting for about 35% of diagnoses. It is mostly found in males and more frequently associated to classic histological variant. It is characterized by an overall intermediate prognosis; however, a subset of patients with either chromosome 11 loss or 17 gain have an excellent prognosis (19).

Clinical Phenotype
GS is characterized by the onset of multiple jaw keratocysts, most frequent in the second decade of life, and/or basal cell carcinomas (BCCs), generally starting from the third decade. Sixty percent of all patients have a recognizable phenotype. More than 100 features have been associated with GS, and the most representative are listed in Table 2 (39,40,43).

Genetic Basis
Heterozygous germline mutations leading to the aberrant activation of SHH signaling are involved in GS, most frequently PTCH1, followed by SUFU. PTCH1 and SUFU mutations work at different levels by disabling SHH pathway signaling, which is normally active during brain development, thus promoting proliferation and inhibiting apoptosis (24,(44)(45)(46)(47).

Correlation With Medulloblastoma
In 1963 Herzberg and Wiskemann first described the association between GS and MB that has been also confirmed by various published studies (48).
In the first large population based study of GS, Evans et al. investigated the incidence of GS in 173 consecutive cases of MB in the North-West of England between 1954 and 1989; they observed a 5% incidence of GS in MB patients with less than 5 years of age, conversely, the incidence of MB in the GS population considered in this study was 3.6% (49). The mean age at MB diagnosis was 2 years in GS patients, earlier than that described in the general population with sporadic MB (38). The desmoplastic/nodular and the extensive nodularity subtypes of MB are the most frequently described (50,51). The risk of MB in subjects with germline mutations of PTCH1 reported in a large series of 115 individuals with related GS-PTCH1 was <2%, while individuals with GS and SUFU germline mutations presented an approximately 20 times higher risk (33%) (24).

Diagnosis
Many individuals with GS are only recognized in adulthood. However, there are clinical signs that could appear early and guide the diagnosis, such as the presence of odontogenic keratocysts in children <20 years of age, basal cell carcinomas in persons <20 years of age, palmar or plantar pits, lamellar calcification of the falx cerebri, and MB with desmoplastic histology in combination with other major or minor criteria (52). Current diagnostic criteria for GS are summarized in Table 3. Diagnosis can be made if 2 major or 1 major and 2 minor criteria are fulfilled (36).

Cancer Surveillance
Surveillance protocols for individuals affected by GS have been proposed by several authors. As suggested in the consensus statement from the first international colloquium on GS, all individuals with GS should perform annually an assessment with a geneticist. A dermatological evaluation is also recommended annually until the first basal cell carcinoma is found, and then every 6 months. Baseline digital Panorex of jaw should be performed starting from the age of 3 years (or as soon as tolerated) and repeated annually before the detection of a first jaw cyst, and then every 6 months (until no jaw cyst for 2 years or until the age of 21).
A baseline echocardiographic evaluation is recommended to exclude cardiac fibromas; in females a pelvic ultrasound for fibromas is also recommended, starting from puberty.
A baseline spine film should be performed at age 1 or at time of diagnosis, and if a skeletal anomaly is found, it must be

Skin anomalies
Pits in the palm of the hand Other anomalies * Ectopic calcifications, frequently in the falx cerebri in more than 90% of patients by age 20 years repeated every 6 months, or sooner if necessary. A routine developmental screening, including an assessment of vision, hearing, and speech, is recommended annually. Annual brain MRI with contrast has been recommended until the age of 8 (52).
However, Smith and colleagues recently described the risk stratification of MB development between PTCH1 and SUFU mutation carriers, recommending the performance of brain MRI only for patients carrying SUFU mutation (24).
Expert consensus recommendations for tumor surveillance of gene carrier and family members were proposed in 2016 based on a literature review and discussion in the AACR Childhood Cancer Predisposition Workshop held in Boston, Massachusetts, in October 2016 (see Table 4) (53).

Therapeutic Approaches
Vismodegib and Sonidegib are selective antagonists of the SHH pathway that act by binding to the transmembrane activator SMO, inhibiting the activation of the downstream SHH pathway.
Vismodegib is the first SHH pathway inhibitor approved by U.S. Food and Drug Administration (FDA) in 2012 and by European Medicines Agency in 2013 for the treatment of advanced or metastatic basal cell carcinomas (54,55).
Sonidegib is approved by the FDA in adult patients for the treatment of locally advanced recurrent basal-cell carcinomas after radiation or surgery or for patients that cannot undergo surgery or radiotherapy (56).
A systemic review and meta-analysis about phase I and phase II Sonidegib and Vismodegib clinical trials highlighted that they are both well tolerated and with anti-tumor activity in MB SHH . The efficacy of Sonidegib was better than Vismodegib in pediatric MB SHH ; however, this has been observed in 3 pediatric patients and further studies are needed for a reliable result (57).
Since SHH signaling has a crucial role during development, along with reports of younger patients treated with SMO inhibitors that show various growth plate complications, their use is not recommended in skeletally immature patients (58).

LI-FRAUMENI SYNDROME
Li-Fraumeni Syndrome (LFS) (OMIM #151623) is one of the most aggressive cancer predisposition syndromes, first described in 1969 by Frederick Li and Joseph Fraumeni Jr (59). LFS is a rare autosomal dominantly inherited disorder caused by germline mutation of TP53, the "guardian of the genome" (60)(61)(62). Loss of p53 function in affected individuals is responsible for an increased risk of developing various solid and hematologic cancers (63). LFS has an estimated prevalence of 1 in 5,000 to 1 in 20,000 (64,65). However, according to Andrade et al., prevalence estimates of the LFS could be higher (1 in 3,555-5,476), reflecting the complexity linked to a wide phenotype and a variable penetrance (66).

Genetic Basis
TP53 gene is located at chromosome 17p13.1 and is composed by 14 coding exons: 10 encode TP53 protein, one a non-coding exon, and three alternative exons (67). TP53 acts as a tumor suppressor gene: in unstressed cells TP53 is unstable and, after exposure to genotoxic stressors, it accumulates and induces the expression of various target genes involved in the regulation of critical cellular processes (growth suppression, apoptosis, DNA repair). Various mechanisms have been proposed to explain how the mutated TP53 protein contributes to tumor formation, including loss of TP53 tumor suppressor function and consequently the dysregulation of its target genes, the "dominant negative" effect in which the mutated TP53 protein inhibits wild-type TP53 protein and the "gain-of-function effect" in which the altered TP53 protein acquires new oncogenic properties.

Clinical Phenotype
Both children and adults affected by LFS have an increased risk of developing multiple primary tumors (68). The most frequent six "core" cancers, their relative prevalence estimates, and other less frequent types of tumor reported in LFS are summarized in Table 5 (60,69,70). Considering all ages, the most frequent tumor reported in LFS families is breast cancer, with a median age at onset of 33 years in females (65,(70)(71)(72)(73). Soft tissue sarcomas and osteosarcoma are the most common tumors in children and adolescents with LFS (65,70,74). The most common type of CNS tumors is glioblastoma/astrocytoma (65,71). Choroid plexus carcinomas (CPC) are more tightly associated with LFS since 45-100% of children with CPC show a germline TP53 mutation (65,(75)(76)(77)(78).

Correlation With Medulloblastoma
Although MB has been described in families with LFS, its prevalence in TP53 carriers is not well known (79). About 5-10% of MBs present TP53 mutations; however, most of these are somatic and only 1% of MBs have been associated with germline TP53 mutations (22,23,(80)(81)(82).
The correlation between TP53 mutation (both somatic and germline) and MB molecular subgroup has been investigated. In 2013, Zhukowa et al. analyzed a cohort of 397 individuals affected by MB (age 1.1 to 45 years) and reported a TP53 mutation almost exclusively in WNT and SHH subgroups while it was virtually

PTCH1 mutation carriers
Basal cell carcinoma screening annually by age 10, with increased frequency after first basal cell carcinoma observed Baseline echocardiogram in infancy, dental exams with jaw X-ray every 12 to 18 months beginning at age 8, and an ovarian ultrasound by age 18 Low risk of medulloblastoma: no radiographic screening unless concerning neurologic exam, head circumference change, or other unusual signs or symptoms If medulloblastoma: radiation-sparing treatment given risk of radiation-induced skin cancers SUFU mutation carriers Same as PTCH1 mutation carriers, with the exception of no jaw X-rays, as keratocysts have not been described Additional medulloblastoma screening: consider every 4 month brain MRI through age 3 and then every 6 month brain MRI until the age of 5 a . Radiation-sparing treatments are again recommended if a brain tumor should occur a Data to support optimal frequency and timing of imaging are not currently available. absent in subgroups 3 and 4. They described a high difference in age distribution between MB SHH /TP53 mutated, which are almost exclusively between ages 5 and 18 years, and MB SHH /TP53 wildtype, that showed a bimodal distribution with peaks before 9 and after 18 years of age. Another interesting fact was that all individuals with TP53 germline mutation, therefore affected by LFS, had MB SHH , and no germline mutations were observed in MB WNT /TP53 mutated. For individuals with TP53 mutant tumors, a dramatic association between biologic subgroups and survival was observed. Patients with MB SHH /TP53 mutated showed a lower 5-year OS than those MB SHH without TP53 alteration (41% +/-9% vs 81% +/-5% respectively); on the contrary, individuals with MB WNT /TP53 mutated showed an almost similar 5-year OS than those MB WNT without TP53 alteration (90% +/-9% vs 97% +/-3% respectively), demonstrating that TP53 mutation status is much more crucial in the SHH subgroup. Within the limitation of the small cohort, no significant difference was observed between LFS children with MB SHH and MB SHH with somatic mutations of TP53 (25).

Diagnosis
The original definition of LFS requires one individual with a sarcoma diagnosed under the age of 45 that has at least one firstdegree relative (parent, sibling, or child) with a cancer of any kind diagnosed under the age of 45 and a third family member who is either a firstor second-degree relative in the same parental lineage (grandparent, aunt, uncle, niece, nephew, or grandchild) with any cancer diagnosed under the age of 45, or a sarcoma at any age (83,84). The finding of TP53 mutations that did not fully respect classical criteria for LFS diagnosis led to the formulation of revised Chompret criteria. Individuals who meet classic and/or revised Chompret diagnostic criteria (Appendix A) should undergo TP53 genetic testing (65,68,71,85).

Cancer Surveillance
Cancer screening in LFS individuals is challenging due to the wide range of associated tumors. Villani et al. in a prospective observational follow-up study of a comprehensive clinical surveillance protocol identified 89 carriers of TP53 pathogenic variants in 39 unrelated families and divided them in two groups: carriers who accepted surveillance (45%) and carriers who did not accept (55%); 21% of patients crossed over from the nonsurveillance to the surveillance group for a total of 66% patients undergoing surveillance for a median of 32 months (86). Over an 11-year period, they identified 40 asymptomatic tumors in 32% of individuals who underwent surveillance and 60 symptomatic neoplasms in 88% patients who initially declined surveillance. The authors highlighted a significant survival advantage in individuals who underwent surveillance reporting 5-year OS of 88.8% in patients with the surveillance group and 59.6% in patients in the non-surveillance group. The Villani et al. 2016 version of the surveillance protocol for children with germline TP53 pathogenic variants is summarized in Table 6 (86). According to Ballinger et al. baseline whole-body magnetic resonance imaging can be used to identify early tumors in a highly cancer-prone population such as LFS patients, although further studies are needed (87).

Therapeutic Approaches
Currently, there is no targetable therapy against tumors of LFS patients available. Generally, it is recommended to avoid use of DNA-damaging agents such as ionizing radiation in order to reduce the risk of secondary tumors with the exception of high grade CNS tumors. Notably, CNS tumor patients with LFS tend to show an overall worse outcome when compared to patients with the same CNS tumors but without TP53 alteration (78,88,89). Even though no guidelines exist, LFS patients should be subjected to physical examination annually with particular attention to neurologic functions. Radiologic approaches without ionizing radiation such as whole-body MRI are currently under investigation (81,86).

TURCOT SYNDROME
Turcot syndrome (TS) is defined by the association of colorectal cancer (CRC) and primary brain tumors and is one of the clinical manifestations of the mismatch repair cancer syndrome (OMIM # 276300). The first clinical report of the association of primary brain tumor and colorectal polyposis dates back to 1949 by Crail et al. Complete physical examination every 3-4 months, including anthropometric measurements plotted on a growth curve (with particular attention to rapid acceleration in weight or height), signs of virilization (pubic hair, axillary moisture, adult body odor, androgenic hair loss, clitoromegaly, or penile growth) and full neurological assessment Prompt assessment with primary care physician for any medical concerns *Serial specimens obtained at the same time of day and processed in the same laboratory. (90). Ten years later Jacques Turcot described two siblings both affected by adenomatous colorectal polyposis and a malignant tumor of CNS, suggesting a common origin for this association (91). Two types of TS are known in literature. Type 1 (TS1) is characterized by the association between hereditary non-polyposis colorectal cancer (HNPCC), also called Lynch syndrome (LS), genetically related to the mutation of the mismatch repair (MMR) genes and CNS tumor (most frequently glioma). Type 2 (TS2) is characterized by the association of brain tumor and colorectal cancer due to familial adenomatous polyposis (FAP), caused by the mutation of the adenomatous polyposis coli (APC) gene, a suppressor gene in the long arm of chromosome 5 (92). Up to 10% of all CRC are inherited and among them a small number, commonly HNPCC or FAP, would be TS (93). Brain tumors in TS are mainly glioblastomas, associated with MMR genes mutations (TS1), and MB, associated with APC gene mutations (TS2).

Genetic Basis
There is a strong association between TS1 and LS. Lynch syndrome is caused by heterozygous germline mutations, inherited in an autosomal-dominant manner, in any of the MMR genes (MLH1; MSH2, MSH6, PMS1; PMS2), which are involved in DNA repair pathway. Unlike LS, TS1 is caused by homozygous mutations in the aforementioned genes (94,95).

Clinical Phenotype
TS1 can clinically manifest with both gastrointestinal (diarrhea, constipation, and/or a positive fecal occult blood test) and neurological symptoms depending on which tumor arises first (95). Lynch syndrome is characterized by an average age of onset that is earlier than in sporadic cases (45 vs 63 years) and by CRC that develops most frequently proximal to splenic flexure and can often be synchronous and metachronous (94). Regarding the development of extracolonic cancers the most frequent are represented by carcinoma of the endometrium, ovary, stomach, small bowel, pancreas, hepatobiliary tract, brain, upper uroepithelial tract, sebaceous adenomas and carcinomas, and multiple keratoacanthomas (94). TS1 patients may have skin signs such as cafe-au-lait spots, resembling type 1 neurofibromatosis, which instead are not reported in TS2 patients (95).

Correlation With Medulloblastoma
MB cases within TS1 are less frequently described than those reported in the setting of TS2, while gliomas are the most frequently reported brain tumors in TS1 (96)(97)(98)(99).
In 2007, Scott et al. described a 13-year-old girl with two colonic carcinomas and MB diagnosed at the age of 7 years caused by constitutional biallelic mutations in the mismatch repair gene MSH6, the first case of MB reported in literature that was caused by the aforementioned biallelic alteration (100). Another report by Lindsay et al. described a 12-yearold with colonic adenocarcinoma and classic MB due to biallelic deletion in PMS2 gene (101). To our knowledge, a correlation between TS1 and various subgroups of MB has not yet been highlighted.

Diagnosis
Some aspects should be considered in TS1 diagnosis: individuals with TS1 are offspring of consanguineous in 20% of cases, with no family history of brain tumors or colon; in TS1 polyps are larger and less numerous than in TS2; in TS1 skin lesions are cafe-au-lait spots while in TS2 they resemble epidermal cysts (95).
According to the American College of Gastroenterology all newly diagnosed CRCs should be studied for MMR deficiency with immunohistochemical testing for the MLH1, MSH2, MSH6, PMS2 proteins and/or with testing for microsatellite instability. Individuals with a history of a tumor that is suspected to be determined by MMR deficiency, a known family mutation associated with LS, or a risk ≥5% of LS obtained with risk prediction models should undergo genetic testing: discovering LS may sometimes be the first step toward diagnosing TS1 (102).

Cancer Surveillance
Cancer surveillance guidelines for patients at risk of or affected by LS have been published while, to our knowledge, no specific guidelines regarding the brain tumor surveillance in patients with TS1 have been established (102).

Therapeutic Approaches
Immunotherapeutic agents such as checkpoint inhibitors have been used in children with biallelic MMR deficiency glioblastoma multiforme, with encouraging results in some studies (26,103). Checkpoint inhibitors seems to be effective in patients whose tumors harbor a high mutation load, resulting in the expression of neoantigens that act as a target for immunotherapy. Checkpoint inhibitors, through different mechanisms, activate T cells that recognize cancer cells as foreign by destroying them. Nivolumab is an anti-programmed death-1 (PD-1) directed checkpoint inhibitor, approved for the treatment of non-smallcell lung cancer and melanoma, and is being tested in various adult and pediatric tumors (103). Ipilimumab is an anticytotoxic T lymphocyte-associated protein 4 (CTLA-4) approved for the treatment of advanced melanoma and renal cell carcinoma and is also under clinical investigation in multiple adult and pediatric cancers (26). To our knowledge, there are no studies that have demonstrated the effectiveness of checkpoint inhibitors in children with MB, and therefore in those associated with MMR deficiency. Nivolumab and Ipilimumab are currently under investigation in a phase II trial of pediatric patients with high-grade CNS malignancies, including medulloblastoma (NCT03130959) (104).

Turcot Syndrome Type 2
Genetic Basis APC mutation is generally inherited with an autosomal dominant manner for the development of FAP, while TS2 seems to require a biallelic loss of the APC gene (92,105). Indeed, in patients with a germ-line alteration of APC, inactivation of the second copy of the gene seems to be crucial for brain tumor development.

Clinical Phenotype
Clinical findings are those typically associated with colorectal cancer and brain tumors, which can occur at different times. Patients with TS2 tend to develop a number of polyps, around thousands, and they frequently manifest gastrointestinal symptoms (similar to those mentioned for TS1). Either before or after the polyps are found, various neurological symptoms and signs can arise, depending on the location of the tumor: headache, vomiting, visual and/or hearing problems, and sensorimotor deficits. In TS2 patients brain tumors can occur without polyposis, and this could be explained by the hypothesis that affected individuals die before adenomatous polyps have time to develop. Skin lesions can also occur in patients with TS2 and are most commonly epidermal cysts.

Correlation With Medulloblastoma
About 40% of patients with TS develop MB (95). According to Hamilton et al., the relative risk of MB in patients with FAP was 92 times higher than in the general population (92). Surun et al. in their multicentric retrospective review of 12 patients, treated between 1988 and 2018 for MB with an identified or highly suspected APC germline pathogenic variant, described some recurrent features such as a constant classic histopathology, a frequent lateral location, and a predominant nonmetastatic status. They highlighted a strong correlation between APCmutated MB and WNT subgroup, demonstrating their excellent outcome, as indeed have wild-type-MB WNT (106). An international multicenter study by Waszak et al., which included 1022 patients with MB, highlighted a close association between APC germline mutations and WNT subgroup; in this study germline APC mutations were found in five (71%) of seven CTNNB1-wild type MB WNT cases, representing 7.6% of all MB WNT , which together with the counterpart constituted by somatic mutations of CTNNB1 (89.4%), account for 97% of all MB WNT (22).

Diagnosis
A key point in the diagnosis of TS2 patients is represented by family history. Individuals who have one or both parents with CRC diagnosed at an early age should be monitored for precancerous colorectal polyps. According to the American College of Gastroenterology an individual with a history of ≥ 10 colorectal adenomatous polyps, or suggestive extracolonic manifestations, without a family history of an underlying pathogenic mutation, should be referred for genetic testing. In addition, the referral for genetic testing is also indicated for relatives of an individual with a known pathogenic mutation in order to establish the presence or absence of that specific mutation and to understand whether the relatives should be considered at-risk subjects (102).

Cancer Surveillance
The identification of family history of FAP and/or APC gene mutations may allow the clinician to perform surveillance in order to promptly identify the possible appearance of a brain tumor. An early diagnosis can allow an earlier treatment. However, it seems there is no advantage in terms of cost-effectiveness since not all individuals who present a CRC at an early age then develop a brain tumor and inversely (27,95). Cancer surveillance guidelines for patients with FAP have been published, while, to our knowledge, no specific guidelines regarding brain tumor surveillance in patients with TS2 have been established (95,102,107).

Therapeutic Approaches
There is currently no targeted therapy available against tumors arising in the setting of TS2.

RECENTLY IDENTIFIED GENETIC SYNDROMES ASSOCIATED WITH MEDULLOBLASTOMA PREDISPOSITION
Pathogenic germline mutations in BRCA2, PALB2, GPR161, and ELP genes have been recently associated with an increased risk of developing MB.

Germline BRCA2 and PALB2 Mutations
The international multicenter study by Waszak et al. identified germline BRCA2 mutations in 11 (1%) of 1022 patients with MB, 10 children and one adult, with a median age at diagnosis of 5.7 years (22). They observed compound heterozygosity at BRCA2 in 4 (36%) of 11 patients, of which all developed MB SHH and showed a worse Progression-Free Survival (PFS) and OS (25% at 5 years, respectively) compared to patients with heterozygous germline BRCA2 mutations, which instead showed a 100% OS and PFS, without secondary neoplasms. Germline mutations in BRCA2, compared with 53105 controls, were associated with increased risk of MB SHH and MB Group3/4 (22).
BRCA2 biallelic mutations are known to be responsible for Fanconi Anemia (FA). The association of FA with MB has been described in literature (108). FA is a syndrome characterized by a chromosomal instability associated with congenital anomalies, bone marrow failure, and an increased risk of developing acute myeloid leukemia, myelodysplastic syndrome, and a number of solid tumors. It is a genetically and phenotypically heterogeneous disorder, inherited with an autosomal recessive pattern (rarely Xlinked). We reported a novel BRCA2 mutation (c.2944_2944delA.) in a 35-month-old female with FA and diagnosis of two distinct MBs that had been previously treated for a nephroblastoma at the age of 15 months. Genetic testing on the patient's DNA extracted from both peripheral blood and MB cells revealed the presence of compound heterozygosis for BRCA2 frameshift mutations. Molecular analysis showed a MB SHH for both the firstand the second-diagnosed MB. However differences in localization, more aggressive histology, and distinct gene expression pattern led to hypothesize a second distinct tumor rather than a distant relapse from the first one (109). The identification of SHH subgroup in FA patients may play a crucial role for their treatment with the use of targeted therapies, especially in these individuals extremely sensitive to conventional treatments.
In 2016 we described a case report of a 7-year-old girl with a classic histotype MB WNT and whose family history was negative for cancer (28). After six years of complete remission from MB the patient developed a secondary glioblastoma. Genetic testing for cancer predisposition syndromes was performed despite a negative family history for neoplasms, and we identified a maternal inherited heterozygous germline BRCA2 mutation, an unusual finding, since cases described in literature were non-WNT subgroups and, to our knowledge, this was the first case of BRCA2-mutated MB WNT reported so far.
Waszak et al. also reported pathogenic heterozygous germline PALB2 mutations in five (<1%) of 1022 patients with MB, of which there were 3 with MB SHH , 1 with MB Group3 , and 1 with MB Group4 . Five-year OS and PFS for patients with germline PALB2 mutations was 75% (22).
Interestingly, a correlation was described between germline BRCA2 and PALB2 mutations and homologous recombination repair deficiency (HRD)-like mutation spectrum, specifically for pediatric MB SHH (89% of cases), revealing HRD as potential biomarker for cancer predisposition in this subgroup (22). Furthermore, the association between germline BRCA2 and PALB2 with HRD-like mutation spectrum can be exploited to evaluate the susceptibility to combination therapies with PARP inhibitors.

GPR161 Mutations
Germline G protein-coupled receptor 161 (GPR161) mutations have recently been described by Begemann et al. as variants predisposing to pediatric MB (110). GPR161 is located on chromosome 1q24.2 and is involved in various aspects of embryonic development, including granule cell proliferation (111,112). Proliferation of granule cells in cerebellum is regulated by SHH ligand and becomes abnormal when SHH-signaling pathway is constitutively activated. GPR161 acts as a SHH-pathway suppressor and its loss of function causes MB development (113). The frequency of germline GPR161 mutations in the general population is about 6 in 10,000 individuals (110). GPR161 biallelic inactivation, most frequently by copy-neutral loss of heterozygosity of chromosome 1q in individuals with heterozygous germline mutation, in the absence of other driver somatic events, has been associated with early TP53-wild-type-MB SHH development (110). According to Begemann et al., overall prevalence of germline GPR161 mutations among pediatric (age<18 years) and infant (age<4 years) patients with MB SHH was 3.4% and 5.5%, respectively (110). Copy-neutral loss of heterozygosity of chromosome 1q was never reported in GPR161 wild-type MB SHH ; therefore, it can be considered a molecular feature (110).

Germline ELP1 Mutations
Germline loss of function (LOF) variants in ELP1 have recently been identified in strong association with MB in pediatric age (114). ELP1 is a molecule that is part of the Elongator Complex, involved in epitranscriptomic tRNA modifications, whose main function is to modify wobble base uridines in the anticodon loop of tRNAs in order to ensure a correct translational elongation (29,(115)(116)(117). The loss of even a single subunit causes the dysregulation of the Elongator Complex with consequent proteome instability. The cerebellum is described as the site of greatest ELP1 expression during brain development (118,119).
According to Waszak et al., three consecutive mutational events are probably required for the development of ELP1-associated MB SHH: a heterozygous germline ELP1LOF variant; somatic biallelic inactivation of ELP1 with monoallelic inactivation of PTCH1 via loss of chromosome arm 9q and biallelic inactivation of the residual PTCH1 allele via a somatic mutation or focal deletion (114). Interestingly, Waszak et al. found a strong association between germline LOF variants in ELP1 and MB SHH subgroup, especially with SHHa subtype (114). Patients with ELP1-associated MB SHH showed a median age at diagnosis of 6.3 years, older than patients with MB SHH and germline SUFU or PTCH1 LOF variants and younger than those with MB SHH and germline TP53 mutations. These patients most frequently presented a desmoplastic nodular histotype and showed a favorable clinical outcome with 92% 5-year OS (114).
Rubinstein-Taybi syndrome (RSTS) is an extremely rare genetic disease, with an incidence of 1 in 100,000 to 125,000 live births, characterized by intellectual disability, unusual behavior, postnatal growth retardation, and multiple congenital anomalies, most frequently of the face and distal limbs (120,121). RSTS is caused by a heterozygous mutation in cyclic-AMP regulated enhancer binding protein (CREBBP) gene, a transcriptional coactivator gene on chromosome 16p13.3, in about 60% of affected individuals (122), a submicroscopic deletion on chromosome 16p13.3 in about 10% of individuals (RSTS1, OMIM #180849) (123), alteration of E1A binding protein p300 (EP300) on chromosome 22q13.2 in about 5-10% of individuals (RSTS2, OMIM #613684) (124,125). CREBBP gene and EP300 genes act as transcriptional co-activators and are involved in DNA repair, cellular growth, differentiation, apoptosis, and tumor suppression (126). According to Boot et al. that reviewed the literature from 1963 to 2017, a total of 132 tumors have been reported in 115 individuals with RSTS and MB was the second most frequent CNS neoplasm with 6 reported cases, after meningioma (121). However, an increased risk for malignant tumors in RSTS could not be confirmed given the small numbers of affected individuals reported in literature, and additional studies are warranted.

GENETIC TESTING OF CANCER PREDISPOSITION SYNDROMES
With the advent of next generation sequencing (NGS) and implementation of genetic testing for adult cancer predisposition syndromes into routine clinical practice, cancer genetics research has extended the use of molecular testing for tumor and germline analysis in pediatric cancer patients. Molecular diagnosis of cancer predisposition syndromes can influence cancer screening initiation or frequency, to either prevent or detect cancer at an earlier and more treatable stage, and directly impact treatment decisions. However, even if medulloblastoma can be associated with rare hereditary cancer predisposition syndromes, screening guidelines for genetic counseling and testing of pediatric patients are not available (23). For genetic testing of cancer predisposition syndromes, different approaches are being used, and, currently, most molecular diagnostics laboratories that offer NGS are performing targeted gene panel testing or clinical whole exome sequencing (WES), more rarely whole genome sequencing (WGS). A multi-gene panel usually includes high and moderate penetrance genes and, sometimes, some low or of yet unknown risk genes, offering the advantage of identifying germline pathogenic variants in genes that would normally not be tested based on the patient's diagnosis. However, it is possible that variants in genes not included in the panels contribute to the cancer risk and WES or WGS can be used to explore other genetic basis of familial syndromes in a more extensive way, permitting to identify new high-and moderate-risk genes of cancer predisposition. Genome-wide approaches generate huge amounts of genetic data and it remains a challenge to interpret the identified variants. Such data interpretation needs close collaboration among molecular geneticists, bioinformaticians, and clinicians. However, as sequencing costs are decreasing and computer and technological resources are expanding, genomewide analysis in clinical practice will become more common.

CONCLUSIONS
MB is the most frequent malignant CNS tumor in children, and additionally to the sporadic form, MB can occur in association with a cancer predisposition syndrome. Knowledge of the clinical findings, etiopathogenic basis, and diagnostic criteria of each syndrome described in this review allow the pediatrician to make a correct diagnosis, start cancer surveillance, and suspect precociously a MB on its onset, providing a prompt treatment. Conversely, when MB is diagnosed, the correct identification/ detection of a cancer predisposition syndrome can allow the clinician to make a more appropriate and complete management of treatment involving several medical specialists in a multidisciplinary team. The molecular studies conducted in the last years have evidenced an association between the various cancer predisposition syndromes and the different MB subgroups. Knowing these relationships can help further clarify the difference not only from a biological point of view but also in prognostic terms. Notably, the extremely poor outcome of MB SHH in children expressing germline TP53 mutations has already been reported. Based on the findings described by Waszak et al., pediatric MB SHH development could be explained by a high genetic predisposition (about 40%); therefore, the effort to carry out genetic testing and surveillance program for affected patients and families in this subgroup becomes even more crucial.
According to Waszak et al. we suggest that patients with MB SHH should be tested for germline TP53 (when older than 3 years), SUFU and PTCH1 mutations (when younger than 3 years), and if negative, also for germline mutations in BRCA2 and PALB2. Furthermore, we suggest that patients with MB SHH should be tested for germline ELP1 mutations, especially those presenting outside of infancy, and for germline GPR161 mutations, particularly those presenting in infancy. We suggest, also, genetic counselling for germline APC mutations in children with MB WNT .
Considering that only 5-6% of MB are associated with cancer predisposition syndromes, our current knowledge is probably still limited. Given the importance that the recognition of a cancer predisposition syndrome can have in the management of a child with MB, we suggest to extend genetic testing also in patients with family history for cancer and/or finding of a dysmorphic phenotype. Knowledge of the associations between molecular subgroups and cancer predisposition syndromes can also be useful in clarifying the differences in terms of therapeutic vulnerability, guiding the development of new targeted therapies. Finally, the comprehension of these biological and molecular differences can help to further improve cancer surveillance measures, with the aim of guaranteeing the best quality of care for the patients.

AUTHOR CONTRIBUTIONS
RC and GB equally contributed to this manuscript. EM, AP, ZMB, FN, GC, EP, EA, MR, ML, and AC contributed to the finishing of the work. FN provided the figure. AM, ACar, EF, LB, and FL revised it critically for important intellectual content. All authors finally approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors contributed to the article and approved the submitted version.
APPENDIX A. LI-FRAUMENI SYNDROME CLASSIC DIAGNOSTIC CRITERIA AND REVISED CHOMPRET CRITERIA

Classic diagnostic criteria
A proband with Sarcoma diagnosed under the age of 45 years AND A first degree relative with any cancer under 45 years AND Another first or second degree relative with either cancer under 45 years or a sarcoma at any age Chompret diagnostic criteria (revised) A proband with an LFS spectrum tumor (soft tissue sarcoma, osteosarcoma, brain tumors, pre-menopausal breast cancer, adrenal cortical carcinoma, leukemia, lung bronchoalveolar cancer) before 46 years AND one of the following criteria: At least one firstor second-degree relative with an LFS tumor (except breast cancer, if the proband has breast cancer) before 56 years or with multiple primary tumors OR A proband with multiple primary tumors (except multiple breast tumors), two of which belong to the LFS tumor spectrum and the first of which occurred before 46 years OR A proband with adrenal cortical carcinoma or choroid plexus carcinoma or embryonal anaplastic subtype rhabdomyosarcoma independent of the family history OR Breast cancer before the age of 31 years