- Email: [email protected]
Bystander T-cell clones in chronic hepatitis C
1050 CORRESPONDENCE
GASTROENTEROLOGY Vol. 113, No. 3
of persistent rectal bleeding after negative BE, colonoscopy detected cancer in 6%–10% of patients.9 – 12 In the only study of colonoscopy for persistent nonbleeding symptoms, cancer was detected in 6%.13 Clinicians have to be trusted to differentiate significant persistent symptoms from irritable bowel syndrome, which in young people often does not require any colon imaging for diagnosis. Based on the reported sensitivities of BE for cancer in many of the hospitals we studied and the above cited reports, we continue to recommend colonoscopy after negative BE when symptoms compatible with cancer persist. The current perforation rate of colonoscopy is unknown, but is surely less than the 1:500 determined in older studies. In the University of Minnesota FOBT trial, there were only 4 perforations in more than 12,000 colonoscopies.14 In a prospective study by Waye et al., there were no diagnostic perforations in 2097 colonoscopies.15 In skilled hands, the complications of colonoscopy are now largely related to polypectomy. To the extent that BE identifies polyps, it leads to colonoscopy and polypectomy. Thus, to contend that the strategy of initial BE results in ‘‘25 times’’ less perforation is misleading. The growth rates of colorectal tumors are not well understood. We do not have survival data on our patients, but certainly in some the failure to diagnose did not influence the ultimate outcome. However, there is no basis to contend that a delay of 10.9 months or longer does not produce some reduction in a patient’s chance of survival. From a patient satisfaction perspective, patients with colorectal cancer who experience diagnostic delays of this length are often not convinced that the delay is insignificant. Clinicians responsible for delays of this length can find themselves trying to convince juries that the delay is not significant. Dr. Glick’s comments regarding negative predictive value are misplaced. High negative predictive value is an expected consequence in a low-prevalence disease. However, in this situation, relatively high negative predictive value can still be associated with important deficiencies in sensitivity. Colorectal cancer is a serious disease, and its early and reliable diagnosis is an issue that informed patients and the clinicians charged with their care take very seriously. Relatively high negative predictive value does not justify poor sensitivity in a disease causing 55,000 deaths per year in the United States alone. The actual gain in sensitivity of colonoscopy vs. BE in our study was not 10% but 12% for DCBE vs. colonoscopy by gastroenterologists, 15% for single-contrast vs. colonoscopy by gastroenterologists, and up to 31% in individual hospitals. Would an informed patient choose to undergo full bowel preparation and then have a test with a 10%–30% chance of missing cancer if they have it, when an alternative safe and comfortable test with a 5% or much lower chance of missing cancer is available? Our recommendation, as in our paper, is to review the relative sensitivity and other quality aspects of the two tests at a local level. Then informed physicians and patients can decide for themselves. DOUGLAS K. REX, M.D.
Division of Gastroenterology/Hepatology Department of Medicine Indiana University School of Medicine Indianapolis, Indiana
1. Office of Technology Assessment: Cost-effectiveness of colorectal cancer screening in average-risk adults: background paper for the 104th Congress. 2. Johnson DC, Ilstrup DM, Fish NM, et al. Barium enema: detection
/ 5e20$$0010
08-08-97 02:18:01
gasa
3.
4.
5. 6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
of colonic lesions in a community population. AJR 1996;167: 49–43. Johnson DA, Gurney MS, Volpe RJ, et al. A prospective study of the prevalence of colonic neoplasms in asymptomatic patients with age-related risk. Am J Gastroenterol 1990;85:969–974. Foutch PC, Mai H, Pardy K, et al. Flexible sigmoidoscopy may be ineffective for secondary prevention of colorectal cancer in asymptomatic, average-risk men. Dig Dis Sci 1991;36:924– 928. Lieberman DA, Smith FW. Screening for colon malignancy with colonoscopy. Am J Gastroenterol 1991;86:946–951. Rex DK, Lehman GA, Ulbright TM, et al. Colonic neoplasia in asymptomatic persons with negative fecal occult blood tests: influence of age, gender, and family history. Am J Gastroenterol 1993;88:825–831. Brady AP, Stevenson GW, Stevenson I. Colorectal cancer overlooked at barium enema examination and colonoscopy: a continuing perceptual problem. Radiology 1994;192:373–378. Kewenter J, Breginge H, Engara¨s B, et al. The value of flexible sigmoidoscopy and double-contrast barium enema in the diagnosis of neoplasms in the rectum and colon in subjects with a positive hemoccult: result of 1831 rectosigmoidoscopies and double-contrast barium enemas. Endoscopy 1995;27:159– 163. Brand EJ, Sullivan BH, Sivak MV, et al. Colonoscopy in the diagnosis of unexplained rectal bleeding. Ann Surg 1980;192:111– 113. Guillem JG, Forde KA, Treat MR, et al. The impact of colonoscopy on the early detection of colonic neoplasms in patients with rectal bleeding. Ann Surg 1987;206:606–611. Swarbrick ET, Fevre DI, Hunt RH, et al. Colonoscopy for unexplained rectal bleeding. Br Med J 1978;1:1685–1687. Teague RH, Manning AP, Thornton RJ, et al. Colonoscopy for investigation of unexplained rectal bleeding. Lancet 1978:1350– 1352. Neugut AI, Garbowski GC, Waye JD, et al. Diagnostic yield of colorectal neoplasia with the use of colonoscopy for abdominal pain, change in bowel habits, and rectal bleeding. Am J Gastroenterol 1993;88:1179–1184. Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993;328:1365–1371. Waye JD, Lewis BS, Yessayan S. Colonoscopy: a prospective report of complications. J Clin Gastroenterol 1992;15:347–351.
Bystander T-Cell Clones in Chronic Hepatitis C Dear Sir: Bertoletti et al. addressed the issue of cytokine secretion of bystander T-cell clones, derived from liver tissue, in chronic hepatitis B and chronic hepatitis C.1 They reported that T helper (Th) 1 cells prevailed in chronic hepatitis C, whereas Th0 and Th2 clones were predominant in chronic hepatitis B. A total of 291 clones were derived from patients with chronic hepatitis B, and 260 clones were from patients with chronic hepatitis C. Because these clones did not react specifically to hepatitis B or hepatitis C viral antigens, and the majority was reported to express several activation markers (major histocompatability complex class II and CD69), it would be of considerable interest to know to which extent these T-cell clones also expressed the CD30 antigen, an activation marker of the tumor necrosis factor/ nerve growth factor receptor superfamily.2 Reiser et al. failed to find significant amounts of CD30-positive T lymphocytes in liver specimens of patients with chronic hepatitis C.3 However, we noted in a recent study that, in patients with chronic hepatitis C, peripheral CD30-positive lymphocytes are readily expanded in vitro after stimu-
WBS-Gastro
September 1997
CORRESPONDENCE 1051
lation with viral antigens or mitogens.4 According to our own findings in patients with chronic hepatitis C, CD30-positive lymphocytes do not secrete Th2 cytokines, as reported for human immunodeficiency virus infection,5,6 but instead produce an intermediate Th1/Th0 cytokine pattern similar to the bystander T-cell clones reported by Bertoletti in chronic hepatitis C.4 Because in the experiments of Bertoletti nonspecific T-cell clones might have resulted from in vitro expansion of such CD30-positive T cells, additional information on CD30 expression would help to understand to what extent the reported Tcell clones actually reflect the cytokine milieu in the liver or are influenced by the cloning procedure. R. P. WOITAS, M.D. T. SAUERBRUCH, M.D. U. SPENGLER, M.D.
Department of General Internal Medicine University of Bonn Bonn, Germany
1. Bertoletti A, D’Elios MM, Boni C, De Carli M, Zignego AL, Durazzo M, Missale G, Penna A, Fiaccadori F, Del Prete G, Ferrari C. Different cytokine profiles of intrahepatic T cells in chronic hepatitis B and hepatitis C virus infections. Gastroenterology 1997;112: 193–199. 2. Gruss HJ, Boiani N, Williams DE, Armitage RJ, Smith CA, Goodwin RG. Pleiotropic effects of the CD30 ligand on CD30-expressing cells and lymphoma cell lines. Blood 1994;83:2045–2056. 3. Reiser M, Marousis CG, Nelson DR, Lauer G, Gonzalez-Peralta RP, Davis GL, Lau JYN. Serum interleukin 4 and interleukin 10 levels in patients with chronic hepatitis C virus infection. J Hepatol 1997; 26:471–478. 4. Woitas RP, Lechmann M, Langhans B, Ba¨r S, Ihlenfeldt HG, Jung G, Sauerbruch T, Spengler U. Cytokine response to hepatitis C virus core peptides in patients with chronic hepatitis C and healthy anti-HCV seropositives (abstr). Hepatology 1996;24:266A. 5. Del Prete G, De Carli M, Almerigogna F, Daniel CK, D’Elios MM, Zanguoghi G, Vinante F, Pizzolo G, Romagnani S. Preferential expression of CD30 by human CD4/ T cells producing Th2-type cytokines. Faseb J 1995;9:81–86. 6. Del Prete G, Maggi E, Pizzolo G, Romagnani S. CD30, Th2 cytokines and HIV infection: a complex and fascinating link. Immunol Today 1995;16:76–80.
Childhood Hepatocellular Tumors in FAP Dear Sir: Bala et al.1 recently reported a case of childhood hepatocellular adenoma associated with familial adenomatous polyposis (FAP). In a 2-year-old child, they found (1) a germline mutation of the adenomatous polyposis coli (APC) gene at codon 1451, (2) loss of heterozygosity (LOH) for APC in tumoral liver cells, and (3) a missense mutation at codon 175 in the p53 gene. This is the first time that biallelic inactivation of the APC gene has been reported in a patient with hepatocellular tumor and a family history of FAP, thus confirming that the liver tumor, at least in this case, conformed with the twohit Knudson model.2 The investigators suggest that the two-hit inactivation has been an early step for hepatocellular tumorigenesis, maybe an initiating event, also determining genetic alteration in p53 gene, facilitating accelerated tumorigenesis. In addition, they state that ‘‘despite development of hepatoblastoma in FAP-predisposed individuals has been acknowledged, no molecular data of a mutational analysis have been reported.’’ Furthermore, they hypothesize that ‘‘APC
/ 5e20$$0010
08-08-97 02:18:01
gasa
gene mutations between codon 1445 and 1578, i.e., in the ‘desmoidassociated region,’3 increase the risk for liver tumors.’’ Recently, we described a 2-year-old girl with FAP-associated hepatoblastoma carrying a germline APC mutation at codon 1061.4–6 The patient is still living (she is 10 years old presently). She belongs to an extended kindred (23 patients in four generations). There were 7 affected patients in the last two generations, all showing the same APC mutation at codon 1061.4 Five of them already underwent colectomy for colonic polyps. Interestingly, 3 of them, all female, also had asymptomatic papillary thyroid carcinoma. Two of the 3 resected specimens also had activation of ret-PTC, a chimeric gene, which is restricted to the papillary histotype (ret-PTC1 isoform).7,8 All affected patients of this kindred had congenital hypertrophy of the retinal pigment epithelium (CHRPE).9 None had desmoids. In addition, in a recent article, Giardiello et al.10 described 8 patients with hepatoblastoma collected from the Johns Hopkins Polyposis Registry. There were 6 male and 2 female patients. All had molecular data of mutational analysis of the APC gene. In particular, 2 patients had a mutation at codon 1230, 1 patient at codon 1061 (as in our patient), 1 patient at codon 1189, and the 4 remaining patients at codons 141, 215, 302, and 541, respectively. Therefore, of 9 additional patients with hepatoblastoma, who had detection of their germline APC mutations, none had a mutation in the desmoid-associated region (codons 1445–1578). On the contrary, most of them had mutations in the CHRPE-associated region.3 In particular, in a recent review of germline APC mutations in patients with another typical extracolonic FAP-associated tumor, such as thyroid carcinoma, APC mutations were found at codon 1061 (in our kindred), 1309 (in 2 patients of different kindreds), and 848, i.e., always in the CHRPEassociated region.11 In addition, all these patients actually had CHRPE. In the last kindred with APC mutation at codon 848, reported by Kashiwagi et al.,12 1 of 2 female siblings, in addition to having thyroid carcinoma and CHRPE, also had desmoids. This is not in contrast with initial observations by Caspari et al.,3 because our group has shown recently that patients with desmoids may carry a germline APC mutation in a region different from codon 1445– 1548 but usually have a somatic mutation in this region.13 In conclusion, we suggest a word of caution before stating, after observing a single case, that (1) germline (or somatic) APC gene mutations between codon 1445 and 1578 increase the risk for hepatic tumors; and (2) speculation from genetic findings in patients having tumors associated with germline mutations can be extended to patients with sporadic tumors. Genotype-phenotype correlations in patients with FAP are far from being fully elucidated. Patients of our kindred with 3 siblings affected from thyroid carcinoma had up to five or seven different extracolonic manifestations per patient. The mutation site was 1061, a codon not included in the mutation region between codons 1445 and 1578, in which the highest risk of developing extraintestinal tumors has been suggested. In particular, epigenetic and environmental factors, in addition to modifier genes, are likely to be of importance to explain the wide intrafamilial variability of extracolonic manifestations, as we have recently suggested (environmental radiation, sex-related factors), while discussing the variable occurrence of thyroid carcinoma in the 7 affected siblings of our FAP kindred.14 FRANCESCO CETTA, M.D. DONATELLA CETTA
GIULIA MONTALTO, M.D. Interuniversity Center for Research in Hepatobiliary Disease Institute of Surgical Clinics University of Siena
WBS-Gastro
GASTROENTEROLOGY Vol. 113, No. 3
of persistent rectal bleeding after negative BE, colonoscopy detected cancer in 6%–10% of patients.9 – 12 In the only study of colonoscopy for persistent nonbleeding symptoms, cancer was detected in 6%.13 Clinicians have to be trusted to differentiate significant persistent symptoms from irritable bowel syndrome, which in young people often does not require any colon imaging for diagnosis. Based on the reported sensitivities of BE for cancer in many of the hospitals we studied and the above cited reports, we continue to recommend colonoscopy after negative BE when symptoms compatible with cancer persist. The current perforation rate of colonoscopy is unknown, but is surely less than the 1:500 determined in older studies. In the University of Minnesota FOBT trial, there were only 4 perforations in more than 12,000 colonoscopies.14 In a prospective study by Waye et al., there were no diagnostic perforations in 2097 colonoscopies.15 In skilled hands, the complications of colonoscopy are now largely related to polypectomy. To the extent that BE identifies polyps, it leads to colonoscopy and polypectomy. Thus, to contend that the strategy of initial BE results in ‘‘25 times’’ less perforation is misleading. The growth rates of colorectal tumors are not well understood. We do not have survival data on our patients, but certainly in some the failure to diagnose did not influence the ultimate outcome. However, there is no basis to contend that a delay of 10.9 months or longer does not produce some reduction in a patient’s chance of survival. From a patient satisfaction perspective, patients with colorectal cancer who experience diagnostic delays of this length are often not convinced that the delay is insignificant. Clinicians responsible for delays of this length can find themselves trying to convince juries that the delay is not significant. Dr. Glick’s comments regarding negative predictive value are misplaced. High negative predictive value is an expected consequence in a low-prevalence disease. However, in this situation, relatively high negative predictive value can still be associated with important deficiencies in sensitivity. Colorectal cancer is a serious disease, and its early and reliable diagnosis is an issue that informed patients and the clinicians charged with their care take very seriously. Relatively high negative predictive value does not justify poor sensitivity in a disease causing 55,000 deaths per year in the United States alone. The actual gain in sensitivity of colonoscopy vs. BE in our study was not 10% but 12% for DCBE vs. colonoscopy by gastroenterologists, 15% for single-contrast vs. colonoscopy by gastroenterologists, and up to 31% in individual hospitals. Would an informed patient choose to undergo full bowel preparation and then have a test with a 10%–30% chance of missing cancer if they have it, when an alternative safe and comfortable test with a 5% or much lower chance of missing cancer is available? Our recommendation, as in our paper, is to review the relative sensitivity and other quality aspects of the two tests at a local level. Then informed physicians and patients can decide for themselves. DOUGLAS K. REX, M.D.
Division of Gastroenterology/Hepatology Department of Medicine Indiana University School of Medicine Indianapolis, Indiana
1. Office of Technology Assessment: Cost-effectiveness of colorectal cancer screening in average-risk adults: background paper for the 104th Congress. 2. Johnson DC, Ilstrup DM, Fish NM, et al. Barium enema: detection
/ 5e20$$0010
08-08-97 02:18:01
gasa
3.
4.
5. 6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
of colonic lesions in a community population. AJR 1996;167: 49–43. Johnson DA, Gurney MS, Volpe RJ, et al. A prospective study of the prevalence of colonic neoplasms in asymptomatic patients with age-related risk. Am J Gastroenterol 1990;85:969–974. Foutch PC, Mai H, Pardy K, et al. Flexible sigmoidoscopy may be ineffective for secondary prevention of colorectal cancer in asymptomatic, average-risk men. Dig Dis Sci 1991;36:924– 928. Lieberman DA, Smith FW. Screening for colon malignancy with colonoscopy. Am J Gastroenterol 1991;86:946–951. Rex DK, Lehman GA, Ulbright TM, et al. Colonic neoplasia in asymptomatic persons with negative fecal occult blood tests: influence of age, gender, and family history. Am J Gastroenterol 1993;88:825–831. Brady AP, Stevenson GW, Stevenson I. Colorectal cancer overlooked at barium enema examination and colonoscopy: a continuing perceptual problem. Radiology 1994;192:373–378. Kewenter J, Breginge H, Engara¨s B, et al. The value of flexible sigmoidoscopy and double-contrast barium enema in the diagnosis of neoplasms in the rectum and colon in subjects with a positive hemoccult: result of 1831 rectosigmoidoscopies and double-contrast barium enemas. Endoscopy 1995;27:159– 163. Brand EJ, Sullivan BH, Sivak MV, et al. Colonoscopy in the diagnosis of unexplained rectal bleeding. Ann Surg 1980;192:111– 113. Guillem JG, Forde KA, Treat MR, et al. The impact of colonoscopy on the early detection of colonic neoplasms in patients with rectal bleeding. Ann Surg 1987;206:606–611. Swarbrick ET, Fevre DI, Hunt RH, et al. Colonoscopy for unexplained rectal bleeding. Br Med J 1978;1:1685–1687. Teague RH, Manning AP, Thornton RJ, et al. Colonoscopy for investigation of unexplained rectal bleeding. Lancet 1978:1350– 1352. Neugut AI, Garbowski GC, Waye JD, et al. Diagnostic yield of colorectal neoplasia with the use of colonoscopy for abdominal pain, change in bowel habits, and rectal bleeding. Am J Gastroenterol 1993;88:1179–1184. Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993;328:1365–1371. Waye JD, Lewis BS, Yessayan S. Colonoscopy: a prospective report of complications. J Clin Gastroenterol 1992;15:347–351.
Bystander T-Cell Clones in Chronic Hepatitis C Dear Sir: Bertoletti et al. addressed the issue of cytokine secretion of bystander T-cell clones, derived from liver tissue, in chronic hepatitis B and chronic hepatitis C.1 They reported that T helper (Th) 1 cells prevailed in chronic hepatitis C, whereas Th0 and Th2 clones were predominant in chronic hepatitis B. A total of 291 clones were derived from patients with chronic hepatitis B, and 260 clones were from patients with chronic hepatitis C. Because these clones did not react specifically to hepatitis B or hepatitis C viral antigens, and the majority was reported to express several activation markers (major histocompatability complex class II and CD69), it would be of considerable interest to know to which extent these T-cell clones also expressed the CD30 antigen, an activation marker of the tumor necrosis factor/ nerve growth factor receptor superfamily.2 Reiser et al. failed to find significant amounts of CD30-positive T lymphocytes in liver specimens of patients with chronic hepatitis C.3 However, we noted in a recent study that, in patients with chronic hepatitis C, peripheral CD30-positive lymphocytes are readily expanded in vitro after stimu-
WBS-Gastro
September 1997
CORRESPONDENCE 1051
lation with viral antigens or mitogens.4 According to our own findings in patients with chronic hepatitis C, CD30-positive lymphocytes do not secrete Th2 cytokines, as reported for human immunodeficiency virus infection,5,6 but instead produce an intermediate Th1/Th0 cytokine pattern similar to the bystander T-cell clones reported by Bertoletti in chronic hepatitis C.4 Because in the experiments of Bertoletti nonspecific T-cell clones might have resulted from in vitro expansion of such CD30-positive T cells, additional information on CD30 expression would help to understand to what extent the reported Tcell clones actually reflect the cytokine milieu in the liver or are influenced by the cloning procedure. R. P. WOITAS, M.D. T. SAUERBRUCH, M.D. U. SPENGLER, M.D.
Department of General Internal Medicine University of Bonn Bonn, Germany
1. Bertoletti A, D’Elios MM, Boni C, De Carli M, Zignego AL, Durazzo M, Missale G, Penna A, Fiaccadori F, Del Prete G, Ferrari C. Different cytokine profiles of intrahepatic T cells in chronic hepatitis B and hepatitis C virus infections. Gastroenterology 1997;112: 193–199. 2. Gruss HJ, Boiani N, Williams DE, Armitage RJ, Smith CA, Goodwin RG. Pleiotropic effects of the CD30 ligand on CD30-expressing cells and lymphoma cell lines. Blood 1994;83:2045–2056. 3. Reiser M, Marousis CG, Nelson DR, Lauer G, Gonzalez-Peralta RP, Davis GL, Lau JYN. Serum interleukin 4 and interleukin 10 levels in patients with chronic hepatitis C virus infection. J Hepatol 1997; 26:471–478. 4. Woitas RP, Lechmann M, Langhans B, Ba¨r S, Ihlenfeldt HG, Jung G, Sauerbruch T, Spengler U. Cytokine response to hepatitis C virus core peptides in patients with chronic hepatitis C and healthy anti-HCV seropositives (abstr). Hepatology 1996;24:266A. 5. Del Prete G, De Carli M, Almerigogna F, Daniel CK, D’Elios MM, Zanguoghi G, Vinante F, Pizzolo G, Romagnani S. Preferential expression of CD30 by human CD4/ T cells producing Th2-type cytokines. Faseb J 1995;9:81–86. 6. Del Prete G, Maggi E, Pizzolo G, Romagnani S. CD30, Th2 cytokines and HIV infection: a complex and fascinating link. Immunol Today 1995;16:76–80.
Childhood Hepatocellular Tumors in FAP Dear Sir: Bala et al.1 recently reported a case of childhood hepatocellular adenoma associated with familial adenomatous polyposis (FAP). In a 2-year-old child, they found (1) a germline mutation of the adenomatous polyposis coli (APC) gene at codon 1451, (2) loss of heterozygosity (LOH) for APC in tumoral liver cells, and (3) a missense mutation at codon 175 in the p53 gene. This is the first time that biallelic inactivation of the APC gene has been reported in a patient with hepatocellular tumor and a family history of FAP, thus confirming that the liver tumor, at least in this case, conformed with the twohit Knudson model.2 The investigators suggest that the two-hit inactivation has been an early step for hepatocellular tumorigenesis, maybe an initiating event, also determining genetic alteration in p53 gene, facilitating accelerated tumorigenesis. In addition, they state that ‘‘despite development of hepatoblastoma in FAP-predisposed individuals has been acknowledged, no molecular data of a mutational analysis have been reported.’’ Furthermore, they hypothesize that ‘‘APC
/ 5e20$$0010
08-08-97 02:18:01
gasa
gene mutations between codon 1445 and 1578, i.e., in the ‘desmoidassociated region,’3 increase the risk for liver tumors.’’ Recently, we described a 2-year-old girl with FAP-associated hepatoblastoma carrying a germline APC mutation at codon 1061.4–6 The patient is still living (she is 10 years old presently). She belongs to an extended kindred (23 patients in four generations). There were 7 affected patients in the last two generations, all showing the same APC mutation at codon 1061.4 Five of them already underwent colectomy for colonic polyps. Interestingly, 3 of them, all female, also had asymptomatic papillary thyroid carcinoma. Two of the 3 resected specimens also had activation of ret-PTC, a chimeric gene, which is restricted to the papillary histotype (ret-PTC1 isoform).7,8 All affected patients of this kindred had congenital hypertrophy of the retinal pigment epithelium (CHRPE).9 None had desmoids. In addition, in a recent article, Giardiello et al.10 described 8 patients with hepatoblastoma collected from the Johns Hopkins Polyposis Registry. There were 6 male and 2 female patients. All had molecular data of mutational analysis of the APC gene. In particular, 2 patients had a mutation at codon 1230, 1 patient at codon 1061 (as in our patient), 1 patient at codon 1189, and the 4 remaining patients at codons 141, 215, 302, and 541, respectively. Therefore, of 9 additional patients with hepatoblastoma, who had detection of their germline APC mutations, none had a mutation in the desmoid-associated region (codons 1445–1578). On the contrary, most of them had mutations in the CHRPE-associated region.3 In particular, in a recent review of germline APC mutations in patients with another typical extracolonic FAP-associated tumor, such as thyroid carcinoma, APC mutations were found at codon 1061 (in our kindred), 1309 (in 2 patients of different kindreds), and 848, i.e., always in the CHRPEassociated region.11 In addition, all these patients actually had CHRPE. In the last kindred with APC mutation at codon 848, reported by Kashiwagi et al.,12 1 of 2 female siblings, in addition to having thyroid carcinoma and CHRPE, also had desmoids. This is not in contrast with initial observations by Caspari et al.,3 because our group has shown recently that patients with desmoids may carry a germline APC mutation in a region different from codon 1445– 1548 but usually have a somatic mutation in this region.13 In conclusion, we suggest a word of caution before stating, after observing a single case, that (1) germline (or somatic) APC gene mutations between codon 1445 and 1578 increase the risk for hepatic tumors; and (2) speculation from genetic findings in patients having tumors associated with germline mutations can be extended to patients with sporadic tumors. Genotype-phenotype correlations in patients with FAP are far from being fully elucidated. Patients of our kindred with 3 siblings affected from thyroid carcinoma had up to five or seven different extracolonic manifestations per patient. The mutation site was 1061, a codon not included in the mutation region between codons 1445 and 1578, in which the highest risk of developing extraintestinal tumors has been suggested. In particular, epigenetic and environmental factors, in addition to modifier genes, are likely to be of importance to explain the wide intrafamilial variability of extracolonic manifestations, as we have recently suggested (environmental radiation, sex-related factors), while discussing the variable occurrence of thyroid carcinoma in the 7 affected siblings of our FAP kindred.14 FRANCESCO CETTA, M.D. DONATELLA CETTA
GIULIA MONTALTO, M.D. Interuniversity Center for Research in Hepatobiliary Disease Institute of Surgical Clinics University of Siena
WBS-Gastro