A Founder Mutation of the MSH2 Gene and Hereditary Nonpolyposis Colorectal Cancer in the United States

ORIGINAL CONTRIBUTION

Henry T. Lynch, MD Context Hereditary nonpolyposis colorectal cancer (HNPCC), also known as Lynch Stephanie M. Coronel, MPH syndrome, is caused by mutations in the mismatch repair genes and confers an ex- Ross Okimoto, BS traordinarily high risk of colorectal, endometrial, and other cancers. However, while carriers of these mutations should be identified, counseled, and offered clinical sur- Heather Hampel, MS, CGC veillance, at present the mutations are not tested for in mutation analyses. Kevin Sweet, MS, CGC Objective To describe the prevalence of a large genomic deletion encompassing ex- Jane F. Lynch, BSN ons1to6oftheMSH2 gene that is widespread in the US population as a result of a founder effect. Ali Barrows, BS Design, Setting, and Patients Ongoing genealogical and historical study con- Juul Wijnen, PhD ducted to assess the origin and spread of an MSH2 mutation previously identified in 9 Heleen van der Klift, MS apparently unrelated families with putative HNPCC and living in widely different geographic locations in the United States. Patrick Franken, HLO Main Outcome Measures Classification of family members as carriers or noncar- Anja Wagner, PhD, MD riers of the MSH2 mutation; spread of the mutation across the continental United Riccardo Fodde, PhD States. Albert de la Chapelle, MD, PhD Results To date, 566 family members of the 9 probands have been identified to be at risk and counseled; 137 of these have been tested, and 61 carry the founder mu- HE ANNUAL WORLDWIDE INCI- tation. Three families have been genealogically shown to descend from a German im- dence of colorectal cancer is es- migrant family that arrived and first settled in Pennsylvania in the early 1700s. Move- timated at 944717 cases,1 of ments of branches of the family from Pennsylvania through North Carolina, Alabama, Kentucky, Missouri, Iowa, Nebraska, Utah, Texas, and California have been docu- which about 10% (94472 cases) mented, and carriers of the mutation have already been diagnosed in 14 states. In areT estimated to be hereditary.2,3 Heredi- contrast, the deletion was not found among 407 European and Australian families with tary nonpolyposis colorectal cancer HNPCC. (HNPCC), also known as Lynch syn- Conclusion The postulated high frequency and continent-wide geographic distri- drome, is the most common form of he- bution of a cancer-predisposing founder mutation of the MSH2 gene in a large, out- reditary colorectal cancer, accounting for bred (as opposed to genetically isolated) population, and the ease with which the mu- somewhere between 2% (18900 cases) tation can be detected, suggest that the routine testing of individuals at risk for HNPCC and 7% (66130 cases) of the annual in the United States should include an assay for this mutation until more is learned worldwide incidence of colorectal about its occurrence. cancer.4 JAMA. 2004;291:718-724 www.jama.com Most tumors from patients with HNPCC harbor a characteristic abnor- mality of an instability at short tandem- Author Affiliations: Department of Preventive Department of Clinical Genetics, Erasmus University Medicine, Creighton University School of Medicine, Medical Centre, Rotterdam, the Netherlands (Dr repeat sequences, ie, microsatellites, in Omaha, Neb (Dr H. T. Lynch; Mss Coronel, J. F. Wagner). Mr Franken and Dr Fodde are currently the genome.5-7 Cancer susceptibility in Lynch, and Barrows; and Mr Okimoto); Human Can- affiliated with the Department of Pathology, Jose- cer Genetics Program, Comprehensive Cancer Cen- phine Nefkens Institute, Erasmus University Medical patients with HNPCC is believed to be ter, The Ohio State University, Columbus (Ms Ham- Centre. caused by malfunction of postreplica- pel, Mr Sweet, and Dr de la Chapelle); Department Corresponding Author: Henry T. Lynch, MD, De- of Human Genetics, Sylvius Laboratory, Leiden Uni- partment of Preventive Medicine, Creighton Univer- tive mismatch repair, as evidenced by in- versity, Leiden, the Netherlands (Drs Wijnen and sity School of Medicine, 2500 California Plaza, Omaha, herited mutations in 1 of 5 different mis- Fodde, Ms van der Klift, and Mr Franken); and NE 68178 ([email protected]).

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

match repair genes: MSH2, MLH1, of 1:10 at the founding increases or de- the screening of an additional cohort of MSH6, MLH3, and PMS2. More than 400 creases, or perhaps remains un- 11 families with HNPCC from Ohio State different pathogenic mutations have changed. The more isolated the popu- University’s resource, where this dele- been registered in the international da- lation, the more pronounced will be the tion was identified in 2 additional cases.26 tabase of mutations in HNPCC kin- effects of genetic drift. For these rea- Because of the extremely high risk for co- dreds (available at http://www.nfdht sons most founder mutations of this lorectal (80%-85%), endometrial (56%), .nl). Detection of mutations is usually type have been described in popula- ovarian (12%), and a number of other performed by sequencing. In the con- tions that have remained isolated while cancers in patients with HNPCC, it is de- text of the present report it is notewor- growing rapidly. Prime examples are the sirable to diagnose carriers of these mu- thy that by sequencing and other com- Finns (“founded” [ie, the main bottle- tations in order to identify high-risk in- monly used genetic testing methods, 1 neck occurred] approximately 2000 dividuals needing targeted clinical class of mutations, ie, large structural re- years ago13), Icelanders (approxi- surveillance. Our purpose is to describe arrangements such as large deletions, is mately 1100 years ago14), Ashkenazi the existence of a mutation that is wide- difficult to detect. These mutations can Jews (600-800 years ago15), and French spread in the US population as a result be readily detected by Southern hybrid- Canadians and Amish (250-400 years of a founder effect and its implications ization,8 multiplex ligation-dependent ago16). Hereditary breast cancer in the for the early detection and prevention of probe amplification,9 and after conver- Ashkenazi Jews, with cancer-predis- cancers associated with HNPCC. sion to haploidy.10,11 posing founder mutations in BRCA1 Among patients who meet the diag- (185delAG and 5382insC) and BRCA2 METHODS nostic criteria for HNPCC, about 40% to (6174delT), is a recent well-published This study was approved by the insti- 60% test positive for mutation, demon- example of a founder effect.17,18 tutional review boards of Creighton strated by a germline mutation in 1 of the At least 5 examples of founder muta- University and the Ohio State Univer- DNA mismatch repair genes; more than tions have been described in cases with sity. Family members provided in- 90% of these mutations are in MLH1 or HNPCC. In Finns, a 3.5-kilobase (kb) de- formed consent for separate institu- MSH2, while approximately 10% are letion of MLH1 comprising exon 16 ac- tional review board–approved colon MSH6 mutations.12 Up to 15% of all dis- counts for as many as half of all cases of cancer genetics studies at each partici- ease-causing mutations in patients with HNPCC, and a splice-site mutation af- pating institution. Genealogical, medi- HNPCC are believed to be large dele- fecting exon 6 of MLH1 accounts for 15% cal, and pathological reports were col- tions, especially in MSH2.8 to 20% of all cases.19 A splice-site muta- lected for members of the 9 extended A founder mutation arises in a single tion affecting exon 5 of MSH2 was first families over a period of more than 30 individual whose offspring each have detected in a large kindred in Newfound- years. Each family displayed inherit- a 50% chance of inheriting the muta- land20 and later turned out to be wide- ance patterns of cancer consonant with tion. The fate of the mutation in the spread in that population through a HNPCC, in concert with the presence subsequent generations will depend on founder effect.21 Interestingly, this mu- of the mutation. 2 main factors, namely, selection and tation has been observed in many other chance. If the mutation leads to selec- populations as well, and actually arises Genealogical Studies tive advantage, it may increase in fre- de novo with appreciable frequency22,23; Available key family members from quency. If it leads to selective disad- thus, this mutation is a recurrent one each of these mutation-verified HNPCC vantage, eg, reduced reproduction, it worldwide but its spread in Newfound- families were personally interviewed may disappear. land is by the founder mechanism. A when possible and/or completed de- In the absence of selection, chance nonsense mutation leading to a stop tailed family history questionnaires. In events known as genetic drift can codon in exon 19 of MLH1 is wide- order to decipher the relationships greatly influence its prevalence in the spread in the Valais region of Switzer- among the 9 families, genealogical re- population. Typically the incidence of land.24 A fifth founder mutation, cords from 3 databases were inten- founder mutations can be increased at 1906G→CinMSH2, is a major contribu- sively searched: the Church of Jesus population bottlenecks. If, for in- tor to HNPCC in the Ashkenazi Jewish Christ of Latter-Day Saints Family His- stance, a mutation occurs with a low in- population, where it may account for as torical Library, Ancestry.com, and a cidence (eg, 1:1000) in a mixed popu- many as one third of all cases.25 German family genealogy database. lation, and if 10 people were to emigrate The identification of a deletion en- These databases contained informa- from this population to an uninhab- compassing exons 1 to 6 of the MSH2 tion from individual family annals as ited island, and if 1 of the 10 were to gene in 7 seemingly unrelated families well as from federal and state census re- have the mutation, then a 100-fold en- with HNPCC from Creighton Universi- cords. Dates of birth, death, and mar- richment would occur at the found- ty’s hereditary cancer resource sug- riage were recorded, as were places of ing. In the new population, genetic drift gested the existence of a founder muta- family origin, temporary or perma- will determine whether the frequency tion in the United States and prompted nent residential settlements, west-

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

Figure 1. First 3 Generations of the MSH2 Deletion Founder Family in the United States

1 2 I

1 2 3 4 5 6 7 8 9 10 11 12 13 II Progenitor Couple

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 III

Male Female 1 2 Individual No. Obligate Carrier of the MSH2 Deletion Mutation

At 50% Risk of Carrying the MSH2 Deletion Mutation

Individuals Who Immigrated From Germany to the United States Together, Arriving in1727

It is currently unclear from which parent III-2 and III-10 inherited the mutation. However, inheritance from II-1 would indicate that his siblings and their offspring are also at risk.

ward migration patterns, immigration Southern Blot Analysis scribed above were selected because they records, land ownership, personal wills, Southern blot analysis of MSH2 was per- tested negative for point mutations in and religious affiliations. formed as previously described26 with MSH2, MSH6, and MLH1 after denatur- XbaI, HindIII, NsiI, EcoRI, and BclI ge- ing gradient gel electrophoresis analy- Haplotype Analysis nomic DNA digests followed by hybrid- sis or direct sequencing.8,26,29,30 The exact breakpoints of the MSH2 ization with 3 overlapping cDNA probes deletion mutation have been de- (encompassing exons 1-7, exons 7-12, RESULTS scribed previously.26 By the detailed andexons10-16).TheAmericanfounder During the genetic testing portion of our nucleotide nomenclature the deletion deletion is characterized by an aberrant family studies process, it was discov- can be characterized as g.5330483_ EcoRI fragment of approximately 14 kb ered that 9 families (Families A-I) from 5349647del19165 (NT034483). The 5Ј and by an approximately 13-kb BclI ab- across the United States shared the iden- breakpoint is 1 kb upstream of exon 1 errant band, after hybridization with the tical deletion of exons 1 through 6 of the and the 3Ј breakpoint is in intron 6.26,27 MSH2probeencompassingexons7to12. MSH2 gene and a unique haplotype of Totally the deletion comprises 16 kb of the region. Currently, a total of 566 high- genomic DNA, including exons 1 to 6. European and Australian risk individuals from these 9 families The fact that the breakpoint was ex- HNPCC Studies have been ascertained, and DNA test- actly the same even at the nucleotide A total of 407 European and Australian ing for the mutation has been per- level in the different families is evi- families with HNPCC were screened for formed in 137 individuals. Of these, 61 dence that the mutation has a com- the presence of the American founder de- individuals, residing in 14 states, were mon origin. In addition, we used nu- letion. The Dutch HNPCC cohort (250 carriers of the mutation, while the re- merous polymorphic markers to families) was collected through the maining 76 individuals did not carry it. construct haplotypes that proved the Dutch Foundation for the Detection of In concert with genealogical studies common origin of the mutation in all Hereditary Tumors and through vari- that identified geographic and ances- 9 families.26 ous Dutch clinical genetics centers. The tral relationships, we established a 13- European cohort (99 families; 58 resid- generation lineage that was ultimately Detection Method ing in Germany and 41 residing else- traced back to a single couple, the pro- As a diagnostic tool, polymerase chain where) was made up of patients with genitors, who migrated from Hesse, Ger- reaction analysis of genomic DNA pro- HNPCC participating in the Concerted many, and settled in Pennsylvania in the duces a unique 1.7-kb band when prim- Action Polyp Prevention 2 (CAPP2) early 1700s. The first 3 generations of ers 5Ј-GCTGAATTAGGTTTTG- study.28 The remaining cohorts were this family are illustrated in FIGURE 1. GAAC-3Ј and 5Ј-AAGCATCACAG made up of families from Italy (n=12), The progenitor couple (II-1 and II-2) had TTACTGTTG-3Ј are used. The experi- Australia (n=31), and Norway (n=15). 11 children, 2 of whom (III-2 and III- mental conditions are as recommended Of the 407 European and Australian 10) have been determined to be obli- in the Expand Long Template polymer- families, 37 had large genomic rearrange- gate carriers of the MSH2 deletion based ase chain reaction system (Roche Diag- ments; none had the American founder on the direct links of Family E, F, and nostics Corp, Indianapolis, Ind). mutations. The HNPCC families de- G to their descendants.

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

Currently, it is unclear from which truly enormous numbers of potential Figure 2 also depicts the remainder of parent individuals III-2 and III-10 in- mutation carriers irrespective of whether the genealogical record emanating from herited the mutation (Figure 1). How- II-1 or II-2 was the carrier parent. the progenitor couple. It not only illus- ever, if it was inherited from their fa- At minimum, it is clear that the 9 sib- trates the direct lineal relationships that ther (II-1), there are even more at-risk lings of the 2 obligate carriers were at 50% were established connecting families E, individuals in the United States, since the risk for inheriting the founder muta- F, and G to our putative progenitors, but progenitor couple (II-1 and II-2) immi- tion. Futhermore, as illustrated in Fig- also illustrates the extensive number of grated from Germany with his parents ure 2, the obligate carriers (III-2, III-10) additional family members potentially at and 5 siblings but none of her (II-2) rela- had 12 known children, each of whom high risk for carrying the MSH2 dele- tives. Figure 1 shows the large number were also at 50% risk; the 9 at-risk sib- tion. The remaining 6 families, while har- of possible mutation carriers in genera- lings (III-1, III-3 through III-9, III-11) had boring the founder mutation, have not tion II under the assumption that II-1 was 35 known children, each of whom had a been unequivocally linked genealogi- the carrier. FIGURE 2 gives an idea of the 25% chance of inheriting the mutation. cally. These families may descend from

Figure 2. Pedigree of the Founder HNPCC Family

1 2 II Progenitor Couple

1 2 3 4 5 6 7 8 9 10 11 III

1 2 3 4 5 6 7 8 9 10 11 12 13 IV ? 9 ? ? 3 3 2 8 3 7 3 4 2 ?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 V 7 38 23 ??3 16 2 8 ? 2 35 ? 4 14 ? 3 8 6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 VI 6 24 12 ???????30 13 5 14 3 8 36 3 14 12

1 2 3 4 5 6 7 8 9 10 11 VII 6 ??37 83 Family A 18 6 ?? ?9 ? ?5 ?? 7 Sk Ut Co (85 Ind) Co

1 2 3 4 Family B Family C Family D Family H Family I Family E ? 12 49 ?? ? ?? VIII (72 Ind) (96 Ind) (64 Ind) ? (41 Ind) (45 Ind) (321 Ind) Co/36 Co/51 1 2 3 IX 6 ? Ut Sk/44 Br Co/45 En/57

Family F Family G ? X (200 Ind) (55 Ind)

XII-XIII

Male Female 3 Cancer Sites 12Individual No. Br Breast Descendants Unknown Earliest Ancestor(s) of Unaffected Co Colon ? Family Plus Subsequent Sk Skin Family C Generations (No. of Affected With Cancer Ut Uterine 3 (96 Ind) Known Family Members) Co/36 Co/36 Cancer Site/Age at Diagnosis, y Offspring of En Endometrial Siblings Combined No. of Progeny 12 2 11 (Both Sexes)

This pedigree depicts the potential number of individuals at risk for carrying the mutation. For example, 9 of the 10 children of individual III-2 had a total of 61 children, of which 23 had 43 children. HNPCC indicates hereditary nonpolyposis colorectal cancer; ind, individual(s). Dashed lines indicate earlier generations that were unable to be traced.

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

female family members who married and genitor couple or with their descen- mutation, we screened 407 European whose descendants have a variety of sur- dants. Currently, the time frame when and Australian families with HNPCC. names that have been difficult to trace family H coexisted with the progeni- Southern analysis of the MSH2 gene with certainty. tor couple is unclear. among these families did not reveal the Thus, while we cannot yet provide the The map of the United States with the presence of the aberrant EcoRI and BclI established lineal or collateral relation- respective migration pattern of each of fragments characteristic of the Ameri- ships connecting families A through D, the 9 HNPCC families is depicted in can founder mutation. However, 24 H, and I to our progenitors, we do pro- FIGURE 4. This map denotes the migra- other MSH2 deletion types account- vide evidence of their ancestral interac- tion patterns of the descendants of the ing for 37 cases were identified. tions with respect to time, geographical progenitors. Note that families A, B, C, location, and genetic mutation. D, H, and I reside in regions along the COMMENT FIGURE 3 is a succinct algorithm that migratory path. Studies using genealogical, clinical, and depicts the relative time frame that 8 To further test the hypothesis that molecular genetic technology have of the 9 families coexisted with the pro- this deletion is an American founder enabled us to confirm the common heri- tage of these 9 families with HNPCC, each of which had been identified inde- Figure 3. Timeline of the Migration Pattern of an HNPCC Gene-Carrying Family pendently in diverse geographic areas of

Progenitor Couple the United States (Figure 3). They were Emigrated From 1727-1745 then eventually tracked to their progeni- Hesse, Germany to Pennsylvania tors’ origin in Germany in the early 18th century (Figures 1 and 2). The presence

3rd and 4th Generations of an identical MSH2 del exon 1 to 6 1745-1800 Migrated and Settled in mutation in each family and a shared hap- North Carolina lotype provided substantial evidence in support of the founder nature of this phe- Family A nomenon. Unlike previous reports of 5th Generation Coexistence Within 80-Mile Radius of City A 1800-1830 Migrated to City A, Kentucky founder mutations in hereditary cancer via Alabama Family I syndromes, this study involved families Coexistence Within 120-Mile Radius of City A in a genetically heterogeneous population, spread over a wide geographic area.

Family B Time of Origin and Spread Coexistence Within 110-Mile of Founder Mutations 5th Through 13th Generations 5th and 6th Generations Radius of City B 1830-1850 Family E Continued Migration Through Two MLH1 founder mutations caus- Migration to Texas City B, Missouri Family C Coexistence Within 90-Mile ing HNPCC in Finland were found to City C, Iowa Radius of City C City D, Nebraska have arisen or been introduced approxi- California (via Utah) Family D mately 1000 years ago and approxi- Coexistence Within 20-Mile mately 250 years ago, respectively.31 Radius of City D Calculations of the “age” of a muta-

6th and 7th Generations tion depend on extensive haplotype 1850-Present Permanent Residence analyses and are not very precise.13 In in California the case of the American founder mutation we describe here, all the evidence suggests that the founding (ie, the 7th Through 13th Generations 7th Through 13th Generations Family G Family F bottleneck of 1 immigrant introduc- Migration to Northern California Migration to Southern California ing it on the American continent) took place 12 or 13 generations ago, in 1727. For the sake of clarity it might be em- Direct Descendants of Progenitor Couple phasized that one cannot determine ex-

Families With HNPCC Gene Living in Same Area With actly when the mutation arose. It could Direct Descendants of the Progenitor Couple not have occurred later than generation II, as at least 2 members of genera- This algorithm outlines the relative time frame that the HNPCC families coexisted (ie, resided in some of the same areas) with the descendants of the progenitor couple. The time frame of coexistence of family H, which tion III were obligate mutation carriers migrated through Kentucky to its current residence in Illinois, is not known. Names of cities have been omitted (Figure 1). Conversely, it could have ex- to protect the anonymity of these families. HNPCC indicates hereditary nonpolyposis colorectal cancer. isted in the German population for un-

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

known numbers of generations before Figure 4. Map of the United States Depicting the Migration Pattern of the Founder HNPCC the American founding event. We Family screened 407 families with HNPCC by Southern hybridization of the MSH2 and MLH1 genes and found 37 with a deletion or other genomic rearrangement. Among these, there were many dele- MN ID tions of MSH2, including deletions en- WY CA NE IA compassing exons 1 to 6. More impor- NV D C PA Progenitor Couple tantly, however, by polymerase chain IL OH From Germany UT G WV VA reaction analysis, none of these were the KS MO H A same as the American founder muta- B KY I NC tion. The series of families described in F OK TN NM the “Methods” section is by far the larg- SC AL GA est series of HNPCC cases hitherto stud- E ied systematically for genomic rearrange- TX ments; therefore, the absence of the Migration Paths FL American founder mutation is signifi- First 3 Generations cant. While the existence of the Ameri- of Progenitor Family can founder mutation in Europe can- Family A Current Family Residences Family B not be excluded, it is certainly unlikely Direct Lineal Relationship to Progenitor Couple Established Family E Direct Lineal Relationship to Progenitor Couple Not Established to be common. Family F Family G States With Known Carriers of MSH2 Deletion Mutation Diagnostic Significance and Family H States With No Known Carriers of MSH2 Deletion Mutation Prevalence of Founder Mutations

In the case of Finland, the 2 founder The lines represent the migration paths of the HNPCC families and their relationship with the progenitor couple. mutations described above (ie, 3.5-kb Migration paths of families C, D, and I are currently unknown; the paths of families A and H are only partially MLH1 del exon 16; splice-site muta- known. HNPCC indicates hereditary nonpolyposis colorectal cancer. tion of MLH1 at exon 6) are so preva- lent in specific regions of Finland (50% risk whose descendants are not known Our documentation of a founder phe- and 20%, respectively, of all cases of to us. The a priori risk of having the mu- nomenon in 9 independently ascer- HNPCC) that their detection has been tation is 50% in all siblings and chil- tained American families with HNPCC used as a first step in primary screen- dren of a mutation carrier, so evi- who trace their origin to Germany, and ing for HNPCC.32,33 We argue here that dence supports the notion that the who heretofore were not known to have the American founder deletion of MSH2 affected families we have encountered been related to each other, has impor- may turn out to account for a signifi- represent but a fraction of all that ex- tant heuristic interest. Their public cant proportion of all cases of HNPCC ist. The specific mutation observed here, health impact from the cancer preven- in the white US population. This hy- being a large deletion, would not be de- tion standpoint harbors potential im- pothesis is based on 2 facts. First, we tected in a routine HNPCC mutation plications for cancer control. have already diagnosed 61 individuals analysis as currently performed in most In addition to the cancer control im- from 9 ostensibly unrelated families by laboratories. plications, knowing the molecular break- studying families with clinically diag- In our studies of HNPCC to date, this points for the founder mutation may also nosed HNPCC at just 2 Midwestern on- MSH2 mutation was the most fre- reduce the economic barriers to con- cology services. These families all re- quently observed specific alteration.34 sumers and third-party payers. Reyes et side in proximity of the historically An assay for this specific mutation al,36 in their discussion of selection strat- proven westward movement of the pro- should be added to routine MSH2 test- egies for genetic testing of patients with genitors’ descendants. In addition to ing in the United States. Previously HNPCC, already point out that molecu- these families, additional codescen- tested families with HNPCC for which lar testing for HNPCC is becoming a dants of the progenitors presumably re- no mutation was found should be re- standard of care, and is cost-effective side in locales along this path (eg, Penn- tested for this specific mutation. Should when compared with the absence of such sylvania, North Carolina, Alabama) that they test positive for the MSH2 del exon genetic testing. Thus, if the detection of remain to be explored in this regard. 1 to 6 mutation, affected patients may this founder mutation is used in the pri- Second, importantly, the earliest gen- then benefit from highly targeted mary screening for HNPCC, the cost- erations of the pedigree (Figure 1) show screening and management programs effectiveness of DNA screening may in- that there were scores of individuals at for HNPCC-associated cancers.3,35 crease even more.

Downloaded From: https://jamanetwork.com/ on 09/24/2021 MSH2 MUTATION AND HEREDITARY NONPOLYPOSIS COLORECTAL CANCER

In conclusion, the current strategy for Study concept and design: H. T. Lynch, Okimoto, also received from National Institutes of Health (NIH) Wagner, Fodde. grant 1U01 CA86389. Research at the Ohio State Uni- DNA screening for HNPCC-associated Acquisition of data: Coronel, Okimoto, Hampel, Sweet, versity was supported by NIH grants CA67941 and mutations has been based on the obser- J. F. Lynch, Barrows, Wijnen, van der Klift, Franken, CA16058. Wagner, Fodde, de la Chapelle. Role of the Sponsor: Neither the Nebraska Depart- vation that there are no common muta- Analysis and interpretation of data: H. T. Lynch, ment of Health and Human Services nor the National tions except in rare, relatively small and Coronel, Okimoto, Hampel, Sweet, Wijnen, Wagner, Institutes of Health had any role in the design and con- homogeneous populations with impor- Fodde, de la Chapelle. duct of the study; in the collection, analysis, and in- Drafting of the manuscript: H. T. Lynch, Coronel, terpretation of the data; in the preparation of the data; tant founder mutations. Our report high- Barrows. or in the preparation, review, or approval of the manu- lights the fact that common mutations Critical revision of the manuscript for important in- script. tellectual content: H. T. Lynch, Coronel, Okimoto, Disclaimer: The contents of this article are solely the may exist in large and diverse popula- Hampel, Sweet, J. F. Lynch, Wijnen, van der Klift, responsibility of the authors and do not necessarily rep- tions. It reinforces the continuing need Franken, Wagner, Fodde, de la Chapelle. resent the official views of the State of Nebraska or the Statistical expertise: Coronel, Wagner, Fodde. Nebraska Department of Health and Human Services. for genetic centers to share information Obtained funding: H. T. Lynch, Coronel, Wagner, Acknowledgment: We thank Trudy Shaw, MA, for on mutation frequencies via databases Fodde. providing excellent technical assistance throughout the Administrative, technical, or material support: H. T. preparation of this article. We thank Jill Griesbach, BA, and publications, and for clinical test- Lynch, Coronel, Okimoto, Hampel, Sweet, J. F. Lynch, for extensive genealogy work with Families H and I ing centers to translate new findings into Barrows, Wijnen, van der Klift, Franken, Wagner, and Ilene Comeras, RN, OCN, for holding a family in- improved DNA testing protocols. Fodde, de la Chapelle. formation service with, and coordinating blood draws Study Supervision: H. T. Lynch, Coronel, Wagner, from, members of Family I. We also thank the inves- Fodde, de la Chapelle. tigators who collected the data for the European and Author Contributions: Dr H. T. Lynch, as principal in- Funding/Support: The study and this article were sup- Australian HNPCC studies: Hans Vasen, MD (the Neth- vestigator of this study, had full access to all of the data ported by revenue from Nebraska cigarette taxes erlands); Gabriela Moslein, MD (Germany); Maija Ko- in the study and takes responsibility for the integrity of awarded to Creighton University by the Nebraska De- honen-Corish, PhD (Australia); and Pål Møller, PhD the data and the accuracy of the data analyses. partment of Health and Human Services. Support was (Norway).

REFERENCES 1. International Agency for Research on Cancer. Glo- 14. Thorlacius S, Struewing JP, Hartge P, et al. Popu- cer in the United States: high mutation detection rate bocan 2000: cancer incidence, prevalence, and mor- lation-based study of risk of breast cancer in carriers among clinically selected families and characteriza- tality worldwide. Available at: http://www-dep.iarc of BRCA2 mutation. Lancet. 1998;352:1337-1339. tion of an American founder genomic deletion of the .fr/globocan/globocan.html. Accessed March 26, 2003. 15. Goodman RM, Motulsky AG, eds. Genetic Dis- MSH2 gene. Am J Hum Genet. 2003;72:1088-1100. 2. Vogelstein B, Kinzler KW, eds. The Genetic Basis eases Among Ashkenazi Jews. New York, NY: Raven 27. Nakagawa H, Hampel H, de la Chapelle A. Identi- of Human Cancer. New York, NY: McGraw-Hill; 1998. Press; 1979. fication and characterization of genomic rearrange- 3. Lynch HT, de la Chapelle A. Genetic susceptibility 16. McKusick VA, Hostetler JA, Egeland JA. Genetic ments of MSH2 and MLH1 in Lynch syndrome (HNPCC) to non-polyposis colorectal cancer. J Med Genet. 1999; studies of the Amish: background and potentialities. by novel techniques. Hum Mutat. 2003;22:258. 36:801-818. Bull Johns Hopkins Hosp. 1964;115:203-222. 28. Burn J, Chapman PD, Bishop DT, Mathers J. Diet 4. Lynch HT, de la Chapelle A. Genomic medicine: he- 17. Struewing JP, Abeliovich D, Peretz T, et al. The and cancer prevention: the Concerted Action Polyp reditary colorectal cancer. N Engl J Med. 2003;348: carrier frequency of the BRCA1 185delAG mutation Prevention (CAPP) studies. Proc Nutr Soc. 1998;57: 919-932. is approximately 1 percent in Ashkenazi Jewish indi- 183-186. 5. Aaltonen LA, Peltomaki P, Leach FS, et al. Clues viduals. Nat Genet. 1995;11:198-200. 29. Wijnen J, Khan PM, Vasen H, et al. Majority of to the pathogenesis of familial colorectal cancer. Sci- 18. Struewing JP, Hartge P, Wacholder S, et al. The hMLH1 mutations responsible for hereditary nonpol- ence. 1993;260:812-816. risk of cancer associated with specific mutations of yposis colorectal cancer cluster at the exonic region 6. Aaltonen LA, Peltomaki P, Mecklin JP, et al. Rep- BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J 15-16. Am J Hum Genet. 1996;58:300-307. lication errors in benign and malignant tumors from Med. 1997;336:1401-1408. 30. Wijnen J, Khan PM, Vasen H, et al. Hereditary non- hereditary nonpolyposis colorectal cancer patients. Can- 19. Nystrom-Lahti M, Kristo P, Nicolaides NC, et al. polyposis colorectal cancer families not complying with cer Res. 1994;54:1645-1648. Founding mutations and Alu-mediated recombina- the Amsterdam criteria show extremely low fre- 7. Risinger JI, Berchuck A, Kohler MF, Watson P, Lynch tion in hereditary colon cancer. Nat Med. 1995;1: quency of mismatch-repair-gene mutations. Am J Hum HT, Boyd J. Genetic instability of microsatellites in en- 1203-1206. Genet. 1997;61:329-335. dometrial carcinoma. Cancer Res. 1993;53:5100- 20. Peltoma¨ ki P, Aaltonen L, Sistonen P, et al. Ge- 31. Moisio A-L, Sistonen P, Weissenbach J, 5103. netic mapping of a locus predisposing to human co- de la Chapelle A, Peltomaki P. Age and origin of two 8. Wijnen J, van der Klift H, Vasen H, et al. MSH2 lorectal cancer. Science. 1993;260:810-812. common MLH1 mutations predisposing to hereditary genomic deletions are a frequent cause of HNPCC. 21. Froggatt NJ, Green J, Brassett C, et al. A com- colon cancer. Am J Hum Genet. 1996;59:1243-1251. Nat Genet. 1998;20:326-328. mon MSH2 mutation in English and North American 32. Aaltonen LA, Salovaara R, Kristo P, et al. Inci- 9. Schouten JP, McElgunn CJ, Waaijer R, Zwijnen- HNPCC families: origin, phenotypic expression, and dence of hereditary nonpolyposis colorectal cancer and burg D, Diepvens F, Pals G. Relative quantification of sex specific differences in colorectal cancer. J Med the feasibility of molecular screening for the disease. 40 nucleic acid sequences by multiplex ligation- Genet. 1999;36:97-102. N Engl J Med. 1998;338:1481-1487. dependant probe amplification. Nucleic Acids Res. 22. Froggatt NJ, Joyce JA, Davies R, et al. A frequent 33. Salovaara R, Loukola A, Kristo P, et al. Population- 2002;30:e57. hMSH2 mutation in hereditary non-polyposis colon based molecular detection of hereditary nonpolyposis 10. Yan H, Papadopoulos N, Marra G, et al. Conver- cancer syndrome. Lancet. 1995;345:727. colorectal cancer. J Clin Oncol. 2000;18:2193-2200. sion of diploidy to haploidy: individuals susceptible to 23. Desai DC, Lockman JC, Chadwick RB, et al. Re- 34. Wagner A, van der Klift H, Franken P, et al. A multigene disorders may now be spotted more eas- current germline mutation in MSH2 arises frequently 10-Mb paracentric inversion of chromosome arm 2p ily. Nature. 2000;403:723-724. de novo. J Med Genet. 2000;37:646-652. inactivates MSH2 and is responsible for HNPCC in a 11. Nakagawa H, Yan H, Lockman J, et al. Allele sepa- 24. Hutter P, Courturier A, Scott RJ, et al. Complex North-American kindred. Genes Chromosomes Can- ration facilitates interpretation of potential splicing al- genetic predisposition to cancer in an extended cer. 2002;35:49-57. terations and genomic rearrangements. Cancer Res. HNPCC family with an ancestral hMLH1 mutation. 35. Ja¨ rvinen HJ, Aarnio M, Mustonen H, et al. Con- 2002;62:4579-4582. J Med Genet. 1996;33:636-640. trolled 15-year trial on screening for colorectal can- 12. Wijnen J, de Leeuw W, Vasen H, et al. Familial 25. Foulkes WD, Thiffault I, Gruber SB, et al. The cer in families with hereditary nonpolyposis colorec- endometrial cancer in female carriers of MSH6 germ- founder mutation MSH2*1906G→Cisanimportant tal cancer. Gastroenterology. 2000;118:829-834. line mutations. Nat Genet. 1999;23:142-144. cause of hereditary nonpolyposis colorectal cancer in 36. Reyes CM, Allen BA, Terdiman JP, Wilson LS. 13. de la Chapelle A, Wright FA. Linkage disequilib- the Ashkenazi Jewish population. Am J Hum Genet. Comparison of selection strategies for genetic test- rium mapping in isolated populations: the example of 2002;71:1395-1412. ing of patients with hereditary nonpolyposis colorec- Finland revisited. Proc Natl Acad Sci U S A. 1998;95: 26. Wagner A, Barrows A, Wijnen JT, et al. Molecu- tal carcinoma: effectiveness and cost-effectiveness. 12416-12423. lar analysis of hereditary nonpolyposis colorectal can- Cancer. 2002;95:1848-1856.

Downloaded From: https://jamanetwork.com/ on 09/24/2021