Paediatric intestinal cancer and polyposis due to bi-allelic PMS2 mutations: case series, review and follow-up guidelines.

BACKGROUND: Bi-allelic germline mutations of one of the DNA mismatch repair genes, so far predominantly found in PMS2, cause constitutional MMR-deficiency syndrome. This rare disorder is characterised by paediatric intestinal cancer and other malignancies. We report the clinical, immunohistochemical and genetic characterisation of four families with bi-allelic germline PMS2 mutations. We present an overview of the published gastrointestinal manifestations of CMMR-D syndrome and propose recommendations for gastro-intestinal screening. METHODS AND RESULTS: The first proband developed a cerebral angiosarcoma at age 2 and two colorectal adenomas at age 7. Genetic testing identified a complete PMS2 gene deletion and a frameshift c.736_741delinsTGTGTGTGAAG (p.Pro246CysfsX3) mutation. In the second family, both the proband and her brother had multiple intestinal adenomas, initially wrongly diagnosed as familial adenomatous polyposis. A splice site c.2174+1G>A, and a missense c.137G>T (p.Ser46Ile) mutation in PMS2 were identified. The third patient was diagnosed with multiple colorectal adenomas at age 11; he developed a high-grade dysplastic colorectal adenocarcinoma at age 21. Two intragenic PMS2 deletions were found. The fourth proband developed a cerebral anaplastic ganglioma at age 9 and a high-grade colerectal dysplastic adenoma at age 10 and carries a homozygous c.2174+1G>A mutation. Tumours of all patients showed microsatellite instability and/or loss of PMS2 expression. CONCLUSIONS: Our findings show the association between bi-allelic germline PMS2 mutations and severe childhood-onset gastrointestinal manifestations, and support the notion that patients with early-onset gastrointestinal adenomas and cancer should be investigated for CMMR-D syndrome. We recommend yearly follow-up with colonoscopy from age 6 and simultaneous video-capsule small bowel enteroscopy from age 8.

Recurrence and variability of germline EPCAM deletions in Lynch syndrome.

Recently, we identified 3' end deletions in the EPCAM gene as a novel cause of Lynch syndrome. These truncating EPCAM deletions cause allele-specific epigenetic silencing of the neighboring DNA mismatch repair gene MSH2 in tissues expressing EPCAM. Here we screened a cohort of unexplained Lynch-like families for the presence of EPCAM deletions. We identified 27 novel independent MSH2-deficient families from multiple geographical origins with varying deletions all encompassing the 3' end of EPCAM, but leaving the MSH2 gene intact. Within The Netherlands and Germany, EPCAM deletions appeared to represent at least 2.8% and 1.1% of the confirmed Lynch syndrome families, respectively. MSH2 promoter methylation was observed in epithelial tissues of all deletion carriers tested, thus confirming silencing of MSH2 as the causative defect. In a total of 45 families, 19 different deletions were found, all including the last two exons and the transcription termination signal of EPCAM. All deletions appeared to originate from Alu-repeat mediated recombination events. In 17 cases regions of microhomology around the breakpoints were found, suggesting nonallelic homologous recombination as the most likely mechanism. We conclude that 3' end EPCAM deletions are a recurrent cause of Lynch syndrome, which should be implemented in routine Lynch syndrome diagnostics.

Risk of colorectal and endometrial cancers in EPCAM deletion-positive Lynch syndrome: a cohort study.

BACKGROUND: Lynch syndrome is caused by germline mutations in MSH2, MLH1, MSH6, and PMS2 mismatch-repair genes and leads to a high risk of colorectal and endometrial cancer. We previously showed that constitutional 3' end deletions of EPCAM can cause Lynch syndrome through epigenetic silencing of MSH2 in EPCAM-expressing tissues, resulting in tissue-specific MSH2 deficiency. We aim to establish the risk of cancer associated with such EPCAM deletions. METHODS: We obtained clinical data for 194 carriers of a 3' end EPCAM deletion from 41 families known to us at the Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands and compared cancer risk with data from a previously described cohort of 473 carriers from 91 families with mutations in MLH1, MSH2, MSH6, or a combined EPCAM-MSH2 deletion. FINDINGS: 93 of the 194 EPCAM deletion carriers were diagnosed with colorectal cancer; three of the 92 women with EPCAM deletions were diagnosed with endometrial cancer. Carriers of an EPCAM deletion had a 75% (95% CI 65-85) cumulative risk of colorectal cancer before the age of 70 years (mean age at diagnosis 43 years [SD 12]), which did not differ significantly from that of carriers of combined EPCAM-MSH2 deletion (69% [95% CI 47-91], p=0·8609) or mutations in MSH2 (77% [64-90], p=0·5892) or MLH1 (79% [68-90], p=0·5492), but was higher than noted for carriers of MSH6 mutation (50% [38-62], p<0·0001). By contrast, women with EPCAM deletions had a 12% [0-27] cumulative risk of endometrial cancer, which was lower than was that noted for carriers of a combined EPCAM-MSH2 deletion (55% [20-90], p<0·0001) or of a mutation in MSH2 (51% [33-69], p=0·0006) or MSH6 (34% [20-48], p=0·0309), but did not differ significantly from that noted for MLH1 (33% [15-51], p=0·1193) mutation carriers. This risk seems to be restricted to deletions that extend close to the MSH2 gene promoter. Of 194 carriers of an EPCAM deletion, three had duodenal cancer and four had pancreatic cancer. INTERPRETATION: EPCAM deletion carriers have a high risk of colorectal cancer; only those with deletions extending close to the MSH2 promoter have an increased risk of endometrial cancer. These results underscore the effect of mosaic MSH2 deficiency, leading to variable cancer risks, and could form the basis of an optimised protocol for the recognition and targeted prevention of cancer in EPCAM deletion carriers.

Mononucleotide precedes dinucleotide repeat instability during colorectal tumour development in Lynch syndrome patients.

A progressive accumulation of genetic alterations underlies the adenoma-carcinoma sequence of colorectal cancer. This accumulation of mutations is driven by genetic instability, of which there are different types. Microsatellite instability (MSI) is the predominant type present in the tumours of Lynch syndrome patients and in a subset of sporadic tumours. It is generally accepted that MSI can be found in the early stages of tumour progression, such as adenomas; however, the frequencies reported vary widely among studies. Moreover, data on the qualitative differences between adenomas and carcinomas, or between tumours of hereditary and sporadic origin, are scarce. We compared MSI in samples of colorectal adenoma and colorectal carcinoma in order to identify possible differences along the adenoma-carcinoma sequence. We compared germline mismatch repair (MMR) gene mutation carriers and non-carriers, to address possible differences of instability patterns between Lynch syndrome patients and patients with sporadic tumours. We found a comparable relative frequency of mono- and dinucleotide instability in sporadic colorectal adenomas and carcinomas, dinucleotide instability being observed most frequently in these sporadic tumours. In MMR gene truncating mutation carriers, the profile was different: colorectal adenomas showed predominantly mononucleotide instability and in colorectal carcinomas, also more mononucleotide than dinucleotide instability was detected. We conclude that MSI profiles differ between sporadic and Lynch syndrome tumours, and that mononucleotide marker instability precedes dinucleotide marker instability during colorectal tumour development in Lynch syndrome patients. As mononucleotide MSI proves to be highly sensitive for detecting mutation carriers, we propose the use of mononucleotide markers for the identification of possible Lynch syndrome patients.

Germline hypermethylation of MLH1 and EPCAM deletions are a frequent cause of Lynch syndrome.

It was shown that Lynch syndrome can be caused by germline hypermethylation of the MLH1 and MSH2 promoters. Furthermore, it has been demonstrated very recently that germline deletions of the 3' region of EPCAM cause transcriptional read-through which results in silencing of MSH2 by hypermethylation. We wanted to determine the prevalence of germline MLH1 promoter hypermethylation and of germline and somatic MSH2 promoter hypermethylation in a large group of Lynch syndrome-suspected patients. From a group of 331 Lynch Syndrome-suspected patients we selected cases, who had no germline MLH1, MSH2, or MSH6 mutation and whose tumors showed loss of MLH1 or MSH2, or, if staining was unavailable, had a tumor with microsatellite instability. Methylation assays were performed to test these patients for germline MLH1 and/or MSH2 promoter hypermethylation. Two patients with germline MLH1 promoter hypermethylation and no patients with germline MSH2 promoter hypermethylation were identified. In the subgroup screened for germline MSH2 promoter hypermethylation, we identified 3 patients with somatic MSH2 promoter hypermethylation in their tumors, which was caused by a germline EPCAM deletion. In the group of 331 Lynch Syndrome-suspected patients, the frequencies of germline MLH1 promoter hypermethylation and somatic MSH2 promoter hypermethylation caused by germline EPCAM deletions are 0.6 and 0.9%, respectively. These mutations, therefore, seem to be rather infrequent. However, the contribution of germline MLH1 hypermethylation and EPCAM deletions to the genetically proven Lynch syndrome cases in this cohort is very high. Previously 27 pathogenic mutations were identified; the newly identified mutations now represent 16% of all mutations.

Do microsatellite instability profiles really differ between colorectal and endometrial tumors?

Microsatellite instability (MSI) occurs in more than 90% of the tumors of Lynch syndrome patients, and in 15-25% of sporadic colorectal (CRC) and endometrial carcinomas (EC). Previous studies comparing EC and CRC using BAT markers showed that the frequency of unstable markers is lower in EC, and that the size of the mutations is smaller in EC. In the present study, we analyzed the type (insertions/deletions), size, and frequency of mutations occurring at three BAT and three dinucleotide markers in CRC and EC, to elucidate whether it is possible to establish different MSI profiles in carcinomas of different tissue origin. We show that mononucleotide markers nearly always become shorter whereas dinucleotide markers can become shorter or longer, in both CRC and EC. We therefore conclude that the type of mutation is a marker-dependent feature rather than tissue-dependent. However, we observed that the size of the deletions/insertions differs between CRC and EC, with EC having shorter alterations. The frequency of mono- and dinucleotide instability found in both tissues is comparable, with mononucleotide and dinucleotide markers being affected at similar rates. We conclude that it is not possible to define clearly different MSI profiles that could distinguish MSI-high in CRC and EC. We propose that the differences observed might indicate different durations of tumor development and/or differences in tissue turnover between colorectal and endometrial epithelium, rather than reflecting truly different MSI profiles. We therefore suggest that the same MSI tests can be used for both tumor types.

PMS2 involvement in patients suspected of Lynch syndrome.

It is well-established that germline mutations in the mismatch repair genes MLH1, MSH2, and MSH6 cause Lynch syndrome. However, mutations in these three genes do not account for all Lynch syndrome (suspected) families. Recently, it was shown that germline mutations in another mismatch repair gene, PMS2, play a far more important role in Lynch syndrome than initially thought. To explore this further, we determined the prevalence of pathogenic germline PMS2 mutations in a series of Lynch syndrome-suspected patients. Ninety-seven patients who had early-onset microsatellite instable colorectal or endometrial cancer, or multiple Lynch syndrome-associated tumors and/or were from an Amsterdam Criteria II-positive family were selected for this study. These patients carried no pathogenic germline mutation in MLH1, MSH2, or MSH6. When available, tumors were investigated for immunohistochemical staining (IHC) for PMS2. PMS2 was screened in all patients by exon-by-exon sequencing. We identified four patients with a pathogenic PMS2 mutation (4%) among the 97 patients we selected. IHC of PMS2 was informative in one of the mutation carriers, and in this case, the tumor showed loss of PMS2 expression. In conclusion, our study confirms the finding of previous studies that PMS2 is more frequently involved in Lynch syndrome than originally expected.

Biochemical characterization of MLH3 missense mutations does not reveal an apparent role of MLH3 in Lynch syndrome.

So far 18 MLH3 germline mutations/variants have been identified in familial colorectal cancer cases. Sixteen of these variants are amino acid substitutions of which the pathogenic nature is still unclear. These substitutions are known as unclassified variants or UVs. To clarify a possible role for eight of these MLH3 UVs identified in suspected Lynch syndrome patients, we performed several biochemical tests. We determined the protein expression and stability, protein localization and interaction of the mutant MLH3 proteins with wildtype MLH1. All eight MLH3 UVs gave protein expression levels comparable with wildtype MLH3. Furthermore, the UV-containing proteins, in contrast to previous studies, were all localized normally in the nucleus and they interacted normally with wildtype MLH1. Our different biochemical assays yielded no evidence that the eight MLH3 UVs tested are the cause of hereditary colorectal cancer, including Lynch syndrome.

A database to support the interpretation of human mismatch repair gene variants.

Germline mutations in the mismatch repair (MMR) genes MLH1, MSH2, MSH6, or PMS2 can cause Lynch syndrome. This syndrome, also known as hereditary nonpolyposis colorectal cancer (HNPCC), is an autosomal dominantly-inherited disorder predominantly characterized by colorectal and endometrial cancer. Truncating MMR gene mutations generally offer a clear handle for genetic counseling and allow for presymptomatic testing. In contrast, the clinical implications of most missense mutations and small in-frame deletions detected in patients suspected of having Lynch syndrome are unclear. We have constructed an online database, the Mismatch Repair Gene Unclassified Variants Database (www.mmruv.info), for information on the results of functional assays and other findings that may help in classifying these MMR gene variants. Ideally, such mutations should be clinically classified by a broad expert panel rather than by the individual database curators. In addition, the different MMR gene mutation databases could be interlinked or combined to increase user-friendliness and avoid unnecessary overlap between them. Both activities are presently being organized by the International Society for Gastrointestinal Hereditary Tumours (InSiGHT; www.insight-group.org).

A novel MSH2 germline mutation in a Druze HNPCC family.

Germline mutations in DNA mismatch repair (DNA-MMR) genes, mainly MLH1, MSH2, and MSH6, underlie Hereditary non-polyposis colorectal cancer (HNPCC) and are mostly family-specific, with few reported founder mutations in MSH2 (Ashkenazim) MLH1 (Finnish). No mutations in colon cancer susceptibility genes have ever been reported in Druze individuals, a Moslem related faith encompassing approximately 1,000,000 individuals worldwide. A novel MSH2 mutation is described in a Druze HNPCC family: a multigenerational family with 10 members in 4 generations affected with colorectal cancer (mean age of diagnosis 46.5 years), two with gastric cancer and one--endometrial cancer. Mutational analysis of the MSH2 gene using denaturing gradient gel electrophoresis (DGGE) and direct sequencing revealed the c.702delA mutation in codon 234 of exon 4 of the MSH2 gene leading to a premature early stop in codon 245, p.Thr234fsX245. Analysis of mutation-carrying or presumed carriers individuals' offspring, revealed 11/42 asymptomatic mutation carriers, age range 17-50 years. The mutation was not present in two additional Druze HNPCC families and 20 Druze sporadic colon cancer patients. This is the first mutation ever reported in a colon cancer susceptibility gene in a Druze family and it appears not to be a founder mutation in Druze individuals with HNPCC.

Functional analysis helps to clarify the clinical importance of unclassified variants in DNA mismatch repair genes.

Hereditary nonpolyposis colorectal cancer (HNPCC) or Lynch syndrome is caused by DNA variations in the DNA mismatch repair (MMR) genes MSH2, MLH1, MSH6, and PMS2. Many of the mutations identified result in premature termination of translation and thus in loss-of-function of the encoded mutated protein. These DNA variations are thought to be pathogenic mutations. However, some patients carry other DNA mutations, referred to as unclassified variants (UVs), which do not lead to such a premature termination of translation; it is not known whether these contribute to the disease phenotype or merely represent rare polymorphisms. This is a major problem which has direct clinical consequences. Several criteria can be used to classify these UVs, such as: whether they segregate with the disease within pedigrees, are absent in control individuals, show a change of amino acid polarity or size, provoke an amino acid change in a domain that is evolutionary conserved and/or shared between proteins belonging to the same protein family, or show altered function in an in vitro assay. In this review we discuss the various functional assays reported for the HNPCC-associated MMR proteins and the outcomes of these tests on UVs identified in patients diagnosed with or suspected of having HNPCC. We conclude that a large proportion of MMR UVs are likely to be pathogenic, suggesting that missense variants of MMR proteins do indeed play a role in HNPCC.

MUTYH and the mismatch repair system: partners in crime?

Biallelic germline mutations of MUTYH-a gene encoding a base excision repair protein-are associated with an increased susceptibility of colorectal cancer. Whether monoallelic MUTYH mutations also increase cancer risk is not yet clear, although there is some evidence suggesting a slight increase of risk. As the MUTYH protein interacts with the mismatch repair (MMR) system, we hypothesised that the combination of a monoallelic MUTYH mutation with an MMR gene mutation increases cancer risk. We therefore investigated the prevalence of monoallelic MUTYH mutations in carriers of a germline MMR mutation: 40 carriers of a truncating mutation (group I) and 36 of a missense mutation (group II). These patients had been diagnosed with either colorectal or endometrial cancer. We compared their MUTYH mutation frequencies with those observed in a group of 134 Dutch colorectal and endometrial cancer patients without an MMR gene mutation (0.7%) and those reported for Caucasian controls (1.5%). In group I one monoallelic MUTYH mutation was found (2.5%). In group II five monoallelic germline MUTYH mutations were found (14%), four of them in MSH6 missense mutation carriers (20%). Of all patients with an MMR gene mutation, only those with a missense mutation showed a significantly higher frequency of (monoallelic) MUTYH mutations than the Dutch cancer patients without MMR gene mutations (P = 0.002) and the published controls (P = 0.001). These results warrant further study to test the hypothesis of mutations in MMR genes (in particular MSH6) and MUTYH acting together to increase cancer risk.

Pigmentary mosaicism following the lines of Blaschko in a girl with a double aneuploidy mosaicism: (47,XX,+7/45,X).

We report on a 6-year-old girl with linear streaks of apparent hypopigmentation and hyperpigmentation following the Blaschko lines, growth retardation, bupthalmos of the left eye, and mild mental retardation. She had a 45,X karyotype in lymphocytes. In cultured fibroblasts a double aneuploidy mosaicism was detected, consisting of a cell line with trisomy for chromosome 7 and a cell line with monosomy for the X-chromosome and no cell line with a normal karyotype. Cutis tricolor or three levels of pigmentation in different skin areas suggested presence of a third, probably normal cell line. Double aneuploidy mosaicism of a cell line with monosomy X and a cell line with trisomy of an autosome is a rare finding. The combination of monosomy X with trisomy of chromosomes 8, 10, 13, 18, and 21 has been reported, but not the combination with trisomy 7. In the 45,X cell line, microsatellite analysis showed loss of the maternal X-chromosome, and presence of a maternal and paternal chromosome 7. The 47,XX,+7 cell line showed a paternal and a maternal X-chromosome, and a paternal and two identical maternal chromosomes 7. Mechanisms that might explain this double aneuploidy mosaicism are discussed.

Colorectal cancer and the CHEK2 1100delC mutation.

The CHEK2 1100delC mutation was recently identified as a low-penetrance breast cancer susceptibility allele. The mutation occurred more frequently in families with clustering of breast and colorectal cancers (CRCs) than in families with clustering of breast cancer only. Hence, the 1100delC mutation could also be a low-penetrance CRC susceptibility allele. To test this hypothesis, we examined the mutation in 629 unselected CRC cases, 230 controls, and 105 selected CRCs diagnosed in patients before age 50. The mutation was observed in 1.6% of unselected patients and in 0.3% of controls (Not significant (NS)). After stratifying unselected patients according to defined genetic risk (on the basis of age at diagnosis and family history of colorectal and endometrial cancer), the highest frequency was observed in high-risk patients (12.5%), followed by moderate-risk patients (3.3%), and was lowest in low-risk patients (1.0%, P(trend) 0.014). In selected patients, 1.6% carried the mutation (NS). Subgroup analyses for tumor localization, gender, and age at diagnosis did not reveal an association with the 1100delC genotype. In addition, a pooled analysis, combining data of one published study in unselected CRC cases and our study, also did not reveal an association. In conclusion, the frequency of the 1100delC genotype was neither significantly increased in unselected CRC patients nor in selected CRC patients diagnosed before age 50. However, after stratifying unselected CRC patients according to defined genetic risk, a significant trend of increasing frequency was observed. Together, the results are consistent with a low-penetrance effect (OR 1.5-2.0) of the CHEK2 1100delC on CRC risk. Large case-control studies are required to clarify the exact role of the CHEK2 1100delC mutation in CRC.

BRAF-V600E is not involved in the colorectal tumorigenesis of HNPCC in patients with functional MLH1 and MSH2 genes.

Recently, it was shown that the oncogenic activation of BRAF, a member of the RAS/RAF family of kinases, by the V600E mutation is characteristic for sporadic colon tumors with microsatellite instability. Further, it was shown to associate with the silencing of the mismatch repair (MMR) gene MLH1 by hypermethylation. Moreover, BRAF mutations proved to be absent in tumors from hereditary nonpolyposis colorectal cancer syndrome (HNPCC) families with germline mutations in the MMR genes MLH1 and MSH2. These data suggest that the oncogenic activation of BRAF is involved only in sporadic colorectal tumorigenesis. In order to further support this hypothesis, we have extended the analysis of the BRAF gene to a different subset of HNPCC families without germline mutations in MLH1 and MSH2. BRAF-V600E mutations were analysed by automatic sequencing in 38 tumors from HNPCC families with germline mutations in the MSH6 gene and also in HNPCC (suspected) families that do not have mutations in the MMR genes MLH1, MSH2 and MSH6. All patients belong to different families. No mutations were detected in 14 tumors from HNPCC patients with germline mutations in MSH6. Further, no mutations of BRAF were found in tumors from 23 MMR-negative families, from which 13 fulfilled the Amsterdam criteria (HNPCC) and 10 were suspected for HNPCC as they were positive for the Bethesda criteria. Overall, our data reinforce the concept that BRAF is not involved in the colorectal tumorigenesis of HNPCC. The detection of a positive BRAF-V600E mutation in a colorectal cancer suggests a sporadic origin of the disease and the absence of germline alterations of MLH1, MSH2 and also of MSH6. These findings have a potential impact in the genetic testing for HNPCC diagnostics and suggest a potential use of BRAF as exclusion criteria for HNPCC or as a molecular marker of sporadic cancer.