Hereditary Nonpolyposis Colorectal Cancer (Lynch Syndrome) in Argentina: Report from a Referral Hospital Register

Hereditary Nonpolyposis Colorectal Cancer (Lynch Syndrome) in Argentina: Report from a Referral Hospital Register Carlos A. Vaccaro, M.D., 1 Fernando Bonadeo, M.D., 1 Analía V. Roverano, M.D., 1 Paivi Peltomaki, M.D., 2 Shashi Bala, Ph.D., 3 Elise Renkonen, Ph.D., 2 Maria A. Redal, Ph.D., 4 Esteban Mocetti, M.D., 5 Eduardo Mullen, M.D., 5 Guillermo Ojea-Quintana, M.D., 1 Mario L. Benati, M.D., 1 Hernan Garcia Rivello, M.D., 5 Mary B. Clark, M.S.W., 6 Jane F. Lynch, B.S.N., 6 Henry T. Lynch, M.D. 6 1 Department of Colorectal Surgery, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina 2 Department of Medical Genetics, University of Helsinki, Helsinki, Finland 3 Division of Human Cancer Genetics, Ohio State University, Columbus, Ohio 4 Basic Science and Experimental Medicine Institute, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina 5 Service of Pathology, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina 6 Creighton_s Hereditary Cancer Institute, Department of Preventive Medicine, Creighton University School of Medicine, Omaha, Nebraska PURPOSE: The first Argentine experience with epidemiologic, molecular, and genetic counseling data is reported. METHODS: We analyzed 43 families fulfilling Amsterdam criteria identified from a prospective database with data from 779 relatives. RESULTS: Eleven families (25.6 percent) presented as Lynch I, 29 (67.4 percent) as Lynch II, and 3 (7 percent) as Muir-Torre syndrome. Among the 306 affected members, 197 cases of colorectal cancer were identified (mean age at diagnosis, 52.1 (range, 21 90) years). The most frequent extracolonic tumors were gastric adenocarcinoma in males and endometrium adenocarcinoma in females. A high incidence of breast cancer was observed (16 cases among 155 females, crude rate: 11,594.20/100,000). Twenty-seven patients (8.8 percent) Poster presentation at the meeting of the Collaborative Group of the Americas on Inherited Colorectal Cancer (CGA-ICC), Salt Lake City, Utah, October 23-24, 2005. Correspondence to: Carlos A. Vaccaro, M.D., Department of Colorectal Surgery, Hospital Italiano de Buenos Aires, Guayaquil 226, 7B (1424) Capital Federal, Buenos Aires, Argentina, e-mail: carlos.vaccaro@hospitalitaliano.org.ar Dis Colon Rectum 2007; 50: 1604 1611 DOI: 10.1007/s10350-007-9037-y * The American Society of Colon and Rectal Surgeons Published online: 11 September 2007 developed more than one tumor. These patients were younger than those with only one tumor (45 vs. 51 years; P=0.001). In 5 of 11 patients who underwent molecular sequencing, a pathologic mutation was found. A novel C deletion at 1910 nucleotide, codon 637, exon 12 of MSH2 gene was identified in a family with a strong aggregation of breast cancer with lack of MSH2 immunohistochemical staining. For 78.2 percent of counseled individuals, this session represented the first time they received information, and 73.9 percent stated that their physicians were unaware of their family background. CONCLUSIONS: Argentine families presented a high incidence of stomach cancer. The elevated incidence of breast cancer and its association with a novel hmsh2 mutation bring to consideration the inclusion of this malignancy as part of the syndrome. A lack of awareness by both physicians and persons at risk was observed. [Key words: Lynch syndrome; Register; Hereditary nonpolyposis colorectal cancer; Genetic counseling; hmsh2 - hmlh1 genes] H ereditary nonpolyposis colorectal cancer (HNPCC), or Lynch syndrome, is an autosomaldominant disease that represents the most common hereditary colorectal cancer (CRC) syndrome accounting for 3 to 5 percent of all colorectal cancer. 1 1604

Vol. 50, No. 10 LYNCH SYNDROME IN ARGENTINA 1605 Five mismatch repair genes have been found to be mutated in HNPCC patients: hmlh1, hmsh2, hmsh6, hpms2, and hmlh3. 2 By the current technology, germline mutations are found in only 40 to 60 percent of patients with HNPCC with mutations in hmlh1 and hmsh2 accounting for approximately 90 percent of them. 2 The reported frequency for hmsh6 mutations varies from 7 to 10 percent and hpms2 is < 5 percent. 2,3 Clinically, HNPCC is characterized by early onset (average age, <45 years) of a variety of cancers. Two clinical variants have been described 4 : Lynch syndrome Type I, with colorectal cancer as the unique cancer running in the family, and Lynch syndrome Type II in which a variety of extracolonic tumors (especially adenocarcinoma of the endometrium and small bowel, and transitional-cell carcinoma of the renal pelvis or ureter) are present. Additionally, some families develop the so-called Muir-Torre syndrome in which an internal malignancy (often gastrointestinal) is associated with cutaneous sebaceous tumors. 5,6 The spectrum of the tumor varies in relation to the mutated gene (i.e., hmsh6 is related to a higher rate of endometrial cancer) 3 as well as to different geographic areas. 7 These epidemiologic variables illustrate the importance of reports from different parts of the world. Although molecular and epidemiologic data from North America, Europe, and Asia are widely available in the literature, 7 limited data are available from South America. 8 12 Furthermore, no data on CRC from an Argentine register are available in the MEDLINE index literature. This study describes the clinical and molecular characteristics of families fulfilling Amsterdam criteria I (ACI) 13 and II (AC II) 14 identified by an Institutional Register in Argentina. PATIENTS AND METHODS A search of our database of 3,030 consecutive patients operated on for CRC since 1970 was made based on family background and age at cancer diagnosis. From 1996, a specific questionnaire about oncologic family background was used. Pedigrees were constructed based on family background obtained by personal interview with the proband. Confirmation by pathologic tissues, slides, or pathology reports were collected whenever possible. Data confirmation was made in a second interview with the probands. Cases with personal or family background of CRC or Lynch syndrome-related tumors were sent to the Hereditary Cancer Program (Pro.Can.He.), which is an institutional register established in 1995 at the Hospital Italiano de Buenos Aires. Currently, the register includes data on 311 familial CRC, 65 families with suspected HNPCC, and 43 families fulfilling Amsterdam Criteria I or II. For a better characterization of the disease, only the latter group is analyzed in the present study. Genetic Counseling Collaboration was established with Creighton University and The Ohio State University in 1998 during a visit of one of the authors (HTL) to South America to counsel a Uruguayan family. During this time, 11 probands could be contacted and were offered genetic testing. All of them accepted and underwent germline mutations search. From these families, 46 individuals at 25 to 50 percent risk qualified for genetic testing and were counseled by one of the authors (CAV) according to the Creighton University_s recommendations. Before obtaining the blood samples, information about all pertinent aspects of the natural history of Lynch syndrome and current surveillance strategies as well as advantages, potential disadvantages, and limitations of genetic testing were discussed. According to the family preference, this first stage was provided in an individual information session or in a group information session. After molecular testing was performed, the family members were contacted to personally receive their results in an individual session during which all implications of genetic testing were again discussed. All participants were asked to complete a form with which to evaluate their counseling and testing experience. Molecular Testing Samples. Samples of blood or paraffin-derived tissues were used for molecular testing by our collaborator (PP). All human specimens were obtained after informed consent according to the guidelines of the institutional review boards. Mutation analyses. Blood DNA samples from the most genetically informative affected representatives of each family were screened for hmsh2 and hmlh1 mutations by denaturing gradient gel electrophoresis (DGGE). Exons with aberrant DGGE patterns were sequenced. Alternatively, samples were screened for mutations in hmsh2 or hmlh1 by direct exon-specific sequencing. Multiplex ligation-dependent probe amplification (MLPA) was used to test the samples for large genomic deletions or amplifications in hmlh1 or

1606 VACCARO ET AL Dis Colon Rectum, October 2007 hmsh2 by using the SALSA P003 MLPA kit and protocols provided by the manufacturer (MRC- Holland\, Amsterdam, The Netherlands). Single nucleotide primer extension (SNuPE). Allelespecific expression analysis of MLH1 mrna was performed as described by Renkonen et al. 15 This method is based on the incorporation of a single dntp that is selected to allow the differential extension of a fluorescently labeled primer annealed next to the polymorphic site (in this case, I219V in hmlh1 exon 8). The products of the primer extension reaction were analyzed on an automated sequencer with 9 percent denaturing polyacrylamide gels. To determine allelic ratios, the peak areas of the maternal and paternal alleles in cdna, quantified with the Genotyper 2.0 program (Applied Biosystems\, Foster City, CA) were compared to those obtained from genomic DNA of the same individuals. Immunohistochemical (IHC) and microsatellite instability (MSI) analysis. Paraffin-derived tumor samples from the probands were studied for MLH1, MSH2, and MSH6 protein expression. The primary antibodies used were anti-mlh1 (clone G168-15; PharMingen\, San Diego, CA), anti-msh2 (clone FE-11; Calbiochem\ and Oncogene Research, San Diego, CA), and anti-msh6 (clone 44; Transduction Laboratories\, Franklin Lakes, NJ). The MLH1 and MSH6 antibodies are mouse monoclonal antibodies prepared with full-length human proteins, whereas the MSH2 antibody is a mouse monoclonal antibody generated with carboxyl terminal fragment of the human MSH2 protein. The DAKO Envision\ + system (DAKO\, Carpinteria, CA) was applied for visualization of the staining products. The percentage of tumor cells staining with the antibodies was recorded for each sample, and a case was considered positive for expression if any tumor cells displayed positive nuclear staining, according to the recommendations of the International Collaborative Group on HNPCC 16 (current International Society for Gastrointestinal Hereditary Tumors). Nuclear staining of normal mucosa and stromal cells included in each tumor section were used as a reference for the evaluation of the staining results. MSI was determined with the Bethesda panel of microsatellite markers (BAT25, BAT26, D5S346, D2S123, and D17S250). 17 RESULTS A total of 43 families fulfilled the Amsterdam Criteria (35 AC I and 8 AC II). Among them, 11 (25.6 percent) presented as Lynch syndrome I, 29 (67.4 percent) as Lynch syndrome II, and 3 (7 percent) as Muir-Torre syndrome. Data from 779 relatives (51.9 percent females) at a mean age of 44.9 (range, 18 94) years were analyzed. The median number of individuals per family was 15 (range, 3 66). Families characterized as Lynch syndrome I presented a trend to have less relatives compared with families characterized as Lynch syndrome II (8 (95 percent confidence interval (CI), 5 22) and 17 (95 percent CI, 12 25), respectively, P=0.09). Table 1 depicts clinical characteristics of the 43 families regarding Amsterdam criteria, Lynch variant, and number of tumors. A total of 306 members (155 females) were affected with a median 7 (95 percent CI, 5 8; range, 3 24) cases per family. In 191 cases (62.4 percent), CRC was the first diagnosed tumor and 35 (11.4 percent) developed more than one tumor (28 patients presented 2 cancers, 4 patients 3 cancers, and 3 patients 4 separate primary cancers). A total of 197 cases of colorectal cancer were identified with a median of 3 (95 percent CI, 3 5; range, 1 21) cases per family. The mean age at diagnosis was 52.1 (range, 21 90) years with 84 cases (53.1 percent) affected before aged 50 years and no difference between genders. Patients with synchronous or metachronous tumors (n=27) developed colorectal cancer at an earlier age: 45.1 (range, 21 87) vs. 51.3 (range, 22 90) years (P=0.001). Of the 306 affected members, 171 (55.9 percent) presented exclusively with colorectal cancer: 26 (8.5 percent) in association with extracolonic tumors, and 109 (35.6 percent) presented with extracolonic cancer alone. Table 2 depicts the localization of extracolonic tumors by gender. Among affected females, breast cancer (prevalence rates: affected females, 11,594.20/ 100,000, all females, 3,960.40/100,000.) was the second most common cancer after endometrial cancer with a median age at diagnosis of 55.5 (range, 25 70) years. Among males, gastric cancer (prevalence rates: affected relatives 5,839.40/100,000, all relatives, 1,026.96/100,000) was the most common extracolonic tumor (median age, 70 (range, 48 80) years). Among the 11 cases included in the collaborative protocol, 8 had MSI high and 5 were found to carry pathologic mutations (Table 3). A deletion C at nucleotide 1910, codon 637 exon 12 of the MSH2 gene (Fig. 1), not reported previously was found in a family with an aggregation of breast cancer (Fig. 2). Immunohistochemical analysis found normal expression of hmlh1 in all analyzed tissues and lack of hmsh2 expression in adenocarcinomas of the colon, duodenum, and breast.

Vol. 50, No. 10 LYNCH SYNDROME IN ARGENTINA 1607 Family No. Phenotype Table 1. Characteristics of Families Fulfilling Amsterdam Criteria No. of Relatives Generations Studied Amsterdam Criteria No. of CRC No. of Non- CRC 1 Lynch I 5 4 Amsterdam I 3 1 0 2 Lynch I 9 3 Amsterdam I 3 0 0 3 Lynch I 22 4 Amsterdam I 6 0 1 4 Lynch I 3 3 Amsterdam I 3 0 0 5 Lynch I 8 4 Amsterdam I 4 0 2 6 Lynch I 40 4 Amsterdam I 3 1 0 7 Lynch I 7 4 Amsterdam I 3 0 0 8 Lynch I 9 4 Amsterdam I 3 0 0 9 Lynch I 29 5 Amsterdam I 4 0 1 10 Lynch I 8 3 Amsterdam I 2 3 2 11 Lynch I 6 4 Amsterdam I 3 5 0 12 Lynch II 6 5 Amsterdam II 4 3 1 13 Lynch II 17 3 Amsterdam I 5 1 0 14 Lynch II 20 3 Amsterdam I 4 2 0 15 Lynch II 7 3 Amsterdam I 3 5 1 16 Lynch II 66 5 Amsterdam I 13 3 1 17 Lynch II 14 5 Amsterdam I 3 2 1 18 Lynch II 42 6 Amsterdam I 21 2 1 19 Lynch II 4 4 Amsterdam I 2 3 2 20 Lynch II 28 3 Amsterdam II 4 6 0 21 Lynch II 9 5 Amsterdam II 2 8 0 22 Lynch II 15 5 Amsterdam I 5 3 1 23 Lynch II 7 4 Amsterdam I 3 4 0 24 Lynch II 33 4 Amsterdam I 8 2 0 25 Lynch II 7 4 Amsterdam II 3 6 2 26 Lynch II 12 3 Amsterdam I 3 4 1 27 Lynch II 46 5 Amsterdam I 8 6 4 28 Lynch II 17 4 Amsterdam I 4 4 0 29 Lynch II 28 5 Amsterdam I 3 3 2 30 Lynch II 28 5 Amsterdam II 1 6 1 31 Lynch II 7 2 Amsterdam II 2 2 0 32 Lynch II 14 3 Amsterdam I 3 4 0 33 Lynch II 17 4 Amsterdam I 6 3 1 34 Lynch II 37 5 Amsterdam I 4 6 0 35 Lynch II 27 4 Amsterdam I 3 5 2 36 Lynch II 12 3 Amsterdam II 5 2 1 37 Lynch II 20 5 Amsterdam I 5 3 0 38 Lynch II 4 4 Amsterdam II 4 2 0 39 Lynch II 10 5 Amsterdam I 4 8 0 40 Lynch II 7 5 Amsterdam I 4 1 1 41 Muir-Torre 23 4 Amsterdam I 9 9 2 42 Muir-Torre 24 5 Amsterdam I 5 3 1 43 Muir-Torre 25 4 Amsterdam I 7 4 3 Total 779 197 135 35 CRC = colorectal cancer; non-crc = noncolorectal cancer. Individual with > 1 Tumor Genetic Counseling For 36 of 46 counseled people (78.3 percent), this session represented the first time that they received information about the disease. Twenty-one people (45.6 percent) realized their risk when a relative was diagnosed with cancer and 15 (32.6 percent) when a relative died from cancer. Only 22 of them had received information from their family physicians. After knowing the cancer risk, 32 (69.6 percent) expressed worry and 7 (15.2 percent) indifference. Thirty-four people (73.9 percent) stated that their personal medical doctor did not know about their family background, and only 12 (26.1 percent) of them received a screening recommendation based on their CRC syndrome. The reasons for accepting genetic counseling were: 1) to improve family care, 29 cases (63.1 percent); 2) to improve their own medical care, 8 cases (17.4 percent); 3) because of personal

1608 VACCARO ET AL Dis Colon Rectum, October 2007 Table 2. Localization of Extracolonic Cancers Related to Gender in all Affected Individuals (n=306) Site Gender Males (n=151) Endometrium j 28 Breast j 16 Ovary j 10 Stomach 8 4 Lung 7 4 Liver 5 4 Skin 3 2 Kidney 3 3 Leukemia 2 2 Pancreas 3 1 Prostate 5 j Testis 2 j Bladder 3 j Other 11 9 Total 52 83 Females (n=155) concern, 7 (15.2 percent); and 4) 2 (4.3 percent) for curiosity. Twenty-one cases (45.6 percent) reported that the results would not modify their lifestyle, and 43 (93.5 percent) estimated that these findings would expedite the recommendation for surveillance. All the individuals scored the counseling session as useful or very useful and they would recommend it to other relatives, with 41.3 percent of them feeling better than before the session and only 2.2 percent worse. Most of them (86.9 percent) were confident about the confidentiality of the data. DISCUSSION In Argentina, CRC represents the third most common tumor, which is only exceeded by breast and uterine cervical cancers in females, and lung and prostate cancers in males. Every year, 10,900 new cases are diagnosed with age-standardized incidence rates per 100,000 of 30.1 for males and 19.1 for females. Compared with other South American countries, only Uruguay with rates of 39.6 and 29.5, respectively, has a higher CRC incidence. With 5,670 deaths each year, CRC also represents a common cause of death (the third in males after lung and prostate, and the second in females after breast cancer), with age-standardized rates of 14.7 for males and 9.8 for females. Assuming that HNPCC represents 3 to 5 percent of CRC, as much as 330 to 545 Family No. Table 3. Clinical Criteria, Immunohistochemical and Sequencing Findings Clinical Criteria IHC Mutation MSI A AC I MLH1- B AC I Muir-Torre MLH1- C AC II MSH2- D AC I MLH1- MLH1 exon 16: del AAG at nt. 1846, c 616 No Change by DGGE / MLPA in MLH1 Not informative in MLH1 mrna allelic expression analysis MSH2 exon 12: del C at nt 1910, c 637 No Change by DGGE and MLPA: MSH2, MLH1, MSH6 Unbalanced mrna expression of paternal and maternal alleles of MLH1 MSH2 exon 3: del CA at nt. 388, c. 130 E AC I MSH2- F AC I No informative No mutation by seq. MLPA in MSH2/MLH1 L-MSI Balanced MLH1 mrna G AC I MLH1 75% No Change by DGGE / MLPA in MLH1 L-MSI MSH2 100% Balanced MLH1 mrna H AC I No informative No Change by DGGE/MLPA in MLH1- MSH2 N/A Balanced MLH1 mrna I AC I MLH1 10% MLH1 exon 16: del AAG at nt. 1846, c 616 MSH2 100% J AC I MLH1 90% No Change by DGGE / MLPA in MSH2 MSH2 0% K AC I MLH1 100% MSH2 del. exon 7 MSH2 <2% AC = Amsterdam criteria; IHC = immunohistochemistry; MSI = microsatellite instability; = high microsatellite instability; L-MSI = low microsatellite instability; DGGE = denaturing gradient gel electrophoresis; MLPA = multiplexdependent probe amplification; N/A = not available.

Vol. 50, No. 10 LYNCH SYNDROME IN ARGENTINA 1609 Figure 1. Sequence analysis. Novel mutation (arrow) of individual III:21 in Family C pedigree. new cases are expected each year. Unfortunately, the knowledge of the disease is limited among health professionals, and HNPCC is rarely diagnosed from specialized registers. Registers for hereditary cancer were shown to improve management of both affected and at-risk family members. Services include medical evaluation, risk assessment, and genetic counseling. A recent survey by Church et al. 18 found 18 registries in the United States. To our knowledge, just six registries are operational in South America (2 in Argentina, 2 in Brazil, 1 in Uruguay, and 1 in Chile). This is the first report from an Argentine register on HNPCC and represents the results of an international collaboration initiated in 1998. The mean age of the CRC (52.1 years) of the present series is slightly higher than those reported from Europe and North America (range, 45.6 48 years) 19,20 with no difference between families diagnosed by AC I and AC II as reported by Zhang et al. 7 from China. Gastric cancer is infrequent in the United States; however, it is a common tumor in the general population in some countries in South America (i.e., Chile and Uruguay). By 2002, the prevalence rate for this type of cancer in Argentina was 18.5 (adjusted rate, 24.3). The eight cases found in this series lead to Figure 2. Family with breast cancer aggregation carrying a novel mutation at hmsh2, exon 12, codon 637, nucleotide 1910.

1610 VACCARO ET AL Dis Colon Rectum, October 2007 a crude rate of 5,839.40 per 100,000, a similar high incidence was reported by the German HNPCC Consortium. 21 Among 281 families with deleterious mutations in hmlh1 or hmsh2, gastric cancer was the second most common gastrointestinal cancer and the third most common malignancy overall, representing 5 percent of all tumors in both gene groups. Accordingly, the authors recommend upper gastrointestinal endoscopy from aged 35 years regardless of a family background of gastric cancer. In females, breast cancer was one of the most frequent extracolonic tumors. In other areas of the world, particularly in Europe, breast cancer is not considered to be an integral lesion in the Lynch syndrome. In Argentina, the incidence rate was 42.3 per 100,000 crude rate (ASR 46.0) and the rate (11,594.20/100,000) found between the studied females is ten times higher than that reported in other South American countries, such as Brazil. 8 10 The greatest aggregation of breast cancer was observed in a family with hmsh2 exon 12 mutation that had not been previously reported. The immunohistochemical analysis of the proband_s breast cancer showed lack of staining of MSH2 protein, bringing into consideration the inclusion of this lesion as a part of the Lynch syndrome cancer spectrum. In the present series, we identified three cases of Muir-Torre syndrome, which constitute, to our knowledge, the first cases reported in South America. This syndrome has been mostly related to hmsh2 mutations. 22 However, in the only case that genetic testing could be performed, although no change by DGGE nor MLPA was found in MLH1 gene, a lack of MLH1 immunohistochemical expression was observed in CRC tissue. Three of 11 cases in which genetic testing was performed (Families F, G, and H from Table 3) showed no evidence of MMR deficiency regardless fulfilling AC I. In this type of familial aggregation of CRC, the designation of Bfamilial colorectal cancer Type X^ has been suggested by Lindor et al. 23 Compared with those with MMR deficiency, these families have a lower incidence of CRC and incidence may not be increased for other cancers. This phenomenon merits more research given the known phenotypic and molecular genetic heterogeneity of Lynch syndrome. The proportion of identified mutations (45.4 percent) is similar to that reported by Rossi et al. 8 from Brazil but higher than that reported by Sarroca et al. 12 from Uruguay (2/12, 16.6 percent), who speculate that a different genetic profile may exist in the Uruguayan population. This study also underlines the lack of knowledge among physicians and families at high risk for CRC. Although almost half of counseled people knew about this risk, a majority (98 percent) had not received professional support, which points out the lack a national strategy for preventing CRC. From the evaluation of the genetic counseling sessions, we demonstrated that, following the international recommendations and findings, psychologic reactions are similar to those reported in developed countries. CONCLUSIONS Argentine families presented typical HNPCC epidemiologic characteristics showing high incidences of stomach cancer in males and breast cancer in females. The association between breast cancer and a novel hmhs2 mutation bring to consideration the inclusion of this neoplasia as a part of the syndrome. A lack of awareness by both physicians and persons at risk was observed. Genetic counseling following international recommendations allowed a successful implementation of genetic testing. REFERENCES 1. Voskuil DW, Vasen H, Kampman E, van_t Veer P. Colorectal cancer risk in HNPCC families: development during lifetime and in successive generations. Int J Cancer 1997;72:205 9.. 2. Peltomaki P, Vasen H. Mutations associated with HNPCC predisposition: Update of ICG-HNPCC-IN- SiGHT mutation database. Dis Markers 2004;20:269 76. 3. Wijnen J, Leeuw W, Vasen H, et al. Familial endometrial cancer in female carriers of MSH6 germline mutations. Nat Genet 1999;23:142 4. 4. Lynch HT, Lynch JF. The Lynch syndromes. Curr Opin Oncol 1993;5:687 96. 5. Muir EG, Bell AJ, Barlow KA. Multiple primary carcinomata of the colon, duodenum and larynx associated with kerato-acanthoma of the face. Br J Surg 1967;54:191 5. 6. Torre D. Multiple sebaceous tumors. Arch Dermatol 1968;98:549 51. 7. Zhang YZ, Sheng JQ, Li SR, Zhang H. Clinical phenotype and prevalence of hereditary nonpolyposis colorectal cancer syndrome in Chinese population. World J Gastroenterol 2005;11:1481 8.

Vol. 50, No. 10 LYNCH SYNDROME IN ARGENTINA 1611 8. Rossi BM, Lopes A, Oliveira Ferreira F, et al. hmlh1 and hmsh2 gene mutation in Brazilian families with suspected hereditary nonpolyposis colorectal cancer. Ann Surg Oncol 2002;9:555 61. 9. Rossi BM, Corvello CM, Epelman M, et al. Hereditary colorectal tumors: routine care and the multidisciplinary therapeutic approach. S Am J Cancer 1997;1:191 7. 10. Ferreira Oliveira F, Ferreira Napoli CC, Rossi BM, et al. Frequency of extra-colonic tumors in hereditary nonpolyposis colorectal cancer (HNPCC) and familial colorectal cancer (FCC) Brazilian families. Fam Cancer 2004;3:41 7. 11. Sarroca C, Peltomaki P, Alfano N, et al. Three new mutations in hereditary nonpolyposis colorectal cancer (Lynch II) in Uruguay. Cancer Genet Cytogenet 2003;142:13 20. 12. Sarroca C, Valle AD, Fresco R, Renkonen R, Peltomaki P, Lynch H. Frequency of hereditary non-polyposis colorectal cancer in Uruguayan patients with colorectal cancer. Clin Genet 2005;68:80 7. 13. Vasen HF, Mecklin JP, Khan PM, Lynch HT. The international collaborative group on hereditary nonpolyposis colorectal cancer (ICG-HNPCC). Dis Colon Rectum 1991;34:424 5. 14. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the international collaborative group on HNPCC. Gastroenterology 1999;116:1453 6. 15. Renkonen E, Zhang Y, Lohi H, et al. Altered expression of MLH1, MSH2 and MSH6 in predisposition to hereditary nonpolyposis colorectal cancer. J Clin Oncol 2003;21:3629 37. 16. Muller W, Burgart LJ, Krause-Paulus R, et al. The reliability of inmunohistochemistry as a prescreening method for the diagnosis of hereditary nonpolyposis colorectal cancer (HNPCC) - results of an international collaborative study. Fam Cancer 2001;1:87 92. 17. Boland CR, Thibodeau SN, Hamilton SR, et al. National Cancer Institute workshop on microsatellite instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58:5248 57. 18. Church J, Kiringoda R, LaGuardia L. Inherited colorectal cancer registries in the United States: the state of the union. Dis Colon Rectum 2004;47:674 8. 19. Rodriguez-Bigas MA, Lee PH, O_Malley L, et al. Establishment of a hereditary nonpolyposis colorectal cancer registry. Dis Colon Rectum 1996;39:649 53. 20. Fitzgibbons RJ, Jr, Lynch HT, Stanislav GV, et al. Recognition and treatment of patients with hereditary nonpolyposis colon cancer (Lynch syndromes I and II). Ann Surg 1987;206:289 95. 21. Goecke T, Schulmann K, Engel C, et al. Genotypephenotype comparison of German MLH1 and MSH2 mutation carriers clinically affected with Lynch syndrome: a report by the German HNPCC Consortium. J Clin Oncol 2006;24:4285 92. 22. Mangold E, Pagesntecher C, Leister M, et al. A genotype-phenotype correlation in HNPCC: strong predominance of MSH2 mutations in 41 patients with Muir-Torre syndrome. J Med Genet 2004;41:567 72. 23. Lindor NM, Rabe K, Petersen GM, et al. Lower cancer incidence in Amsterdam-I criteria families without mismatch repair deficiency: familial colorectal cancer type X. JAMA 2005;293:1979 85.