The diagnosis of genetic hemochromatosis previously. Genotypic/Phenotypic Correlations in Genetic Hemochromatosis: Evolution of Diagnostic Criteria

Transcription

1 GASTROENTEROLOGY 1998;114: Genotypic/Phenotypic Correlations in Genetic Hemochromatosis: Evolution of Diagnostic Criteria PAUL C. ADAMS* and SUBRATA CHAKRABARTI Departments of *Medicine and Pathology, University of Western Ontario, London, Ontario, Canada Background & Aims: The identification of a candidate gene for hereditary hemochromatosis in 69% 100% of patients with hemochromatosis has resulted in a diagnostic genotypic test (C282Y). The aim of this study was to reassess the phenotypic diagnostic criteria for hemochromatosis in patients homozygous for the C282Y mutation of the HFE gene. Methods: saturation, ferritin, hepatic, and iron removed by venesection were studied in C282Y homozygotes and C282Y putative homozygotes. Results: Patients were homozygous for the C282Y mutation in 122 of 128 cases (95%). In C282Y homozygotes, the results were as follows: hepatic, 1.9 in 91.3%; transferrin saturation, 55% in 90%; serum ferritin, 300 g/l in 96% of men and 200 g/l in 97% of women; and iron removed, 5gin70% of men and 73% of women. There were four homozygotes for C282Y with no biochemical evidence of iron overload. Conclusions: The sensitivity of the phenotypic tests in decreasing order was as follows: serum ferritin, hepatic, transferrin saturation, and iron removed by venesection. Although the genetic test is useful in the diagnostic algorithm, this study has shown both iron-loaded patients without the mutation and homozygous patients without iron overload. The diagnosis of genetic hemochromatosis previously has depended on the phenotypic expression of iron overload. The clinical diagnosis was suspected by the signs and symptoms and an elevated transferrin saturation and/or serum ferritin level, and was confirmed by the presence of parenchymal iron overload on liver biopsy specimen. Supporting evidence was provided by the amount of iron removed by venesection therapy and by pedigree studies providing evidence of familial iron overload. 1,2 With the discovery of a candidate gene for hemochromatosis, 3 it is now possible to make a genotypic diagnosis of hemochromatosis. In this study, we review the diagnostic precision of the classical phenotypic methods in a group of well-defined hemochromatosis homozygotes that have now been classified by genotyping. Materials and Methods This hospital is a tertiary referral center for genetic hemochromatosis. Probands referred for possible hemochromatosis underwent serum iron studies, liver biopsy, HLA typing, and pedigree studies. The diagnosis was based on the clinical judgment of a single physician based on the clinical, pathological, and familial evidence for hemochromatosis. Liver biopsy was performed on all probands but was not always performed on discovered cases that were HLA identical to the proband case. The diagnosis in the proband case depended on the presence of a hepatic of 1.9 or iron removed by venesection of 5 g in the absence of other conditions associated with iron overload. Patients with elevated liver enzymes (alanine aminotransferase and aspartate aminotransferase) were tested for hepatitis B surface antigen and antibody to hepatitis C virus. The diagnosis in the discovered sibling depended on being HLA identical at the A and B locus. Many isolated cases of mild iron overload complicated by alcoholism and chronic viral hepatitis were considered by clinical and pathological assessment (hepatic of 1.9) not to have genetic hemochromatosis. We have described previously 12 patients with secondary iron overload with chronic viral hepatitis, iron-loading anemias, and multiple transfusions who had significant iron overload (hepatic of 1.9) who were not considered to have genetic hemochromatosis. 4 DNA samples had been stored on many homozygotes in anticipation of genetic testing. After the description of a method to test for the C282Y mutation of the HFE (HLA-H) gene, genotyping was performed on 128 putative homozygotes. 5 Polymerase chain reaction amplification (forward primer, 5 TGGCAAGGG- TAAACAGATCC3, and reverse primer, 5 CTCAG- GCACTCCTCTCAACC3 ) generates a product of 387 base pair size. When digested with the restriction enzyme RsaI, normal DNA generates two fragments of 247 and 140 base pairs. In a homozygote, because of the creation of a new restriction site secondary to the mutation, the mutant DNA generates a total of three fragments with 247, 111, and 29 base pairs. In heterozygotes, because of the presence of both normal Abbreviations used in this paper: C282Y, CYS282TYR mutation of the HFE gene by the American Gastroenterological Association /98/$3.00

2 320 ADAMS AND CHAKRABARTI GASTROENTEROLOGY Vol. 114, No. 2 Table 1. Clinical and Genetic Features of Phenotypic Homozygotes Clinical features Male probands (n 58) Men discovered (n 38) Female probands (n 16) Women discovered (n 16) (yr) 53 (21 78) 50 (22 72) 55 (22 74) 57 (36 75) index of 37 (92%) 19 of 22 (86%) 11 of 11 (100%) 6 of 7 (86%) Iron removed 5 g 35 of 40 (88%) 9 of 23 (39%) 8 of 10 (80%) 3 of 5 (60%) HLA identical to proband 38 of of 16 C282Y homozygotes 55 of 58 (95%) 38 of of 16 (81%) 16 of 16 Mean serum ferritin a 2520 (434 10,001) 845 ( ) 1135 ( ) 786 ( ) Elevated ferritin a 58 of 58 (100%) 34 of 38 (89%) 16 of of 13 (92%) Mean transferrin saturation (%) a 81 (20 100) 70 (23 100) 69 (39 100) 74 (47 89) Elevated transferrin saturation a 56 of 58 (97%) 33 of 38 (87%) 13 of 16 (81%) 11 of 13 (85%) Chronic alcoholism 5 of 58 0 of 51 1 of 16 0 of 16 Chronic viral hepatitis 1 of 58 0 of 51 0 of 16 0 of 16 a Data are expressed as mean with range in parentheses. Reference range for serum ferritin level is for men and g/l for women. Reference range for transferrin saturation is 20% 55%. and mutated DNA, a total of four fragments with 247, 140, 111, and 29 base pairs are produced. Patients can be assigned to be homozygous, heterozygous, or normal for this mutation. Putative homozygotes that were not homozygous for the C282Y mutation (n 6) were tested for a second mutation (H63D) as described previously. 5 concentration (in micromoles per gram of dry wt; normal range, 0 35 mol/g dry wt) was determined from paraffin blocks by atomic absorption spectrophotometry, and the hepatic was the hepatic iron concentration divided by age (normal, 1.9). 6 Normal range for serum ferritin level for women was g/l and for men was g/l. Normal range for transferrin saturation was 20% 55%. Iron removed by venesection was determined by the number of venesections required to reduce the serum ferritin to approximately 50 g/l 0.25 g. This was only estimated for patients undergoing venesections at this hospital because of the inaccuracy of external medical records. Results Genetic Studies Seventy-seven probands (30 isolated probands and 47 probands subsequently found to have iron-loaded family members) and 51 subjects were discovered through family studies (96 men and 32 women) (Table 1). Patients were homozygous for the C282Y mutation in 122 of 128 cases (95%). All subjects with an HLA-identical sibling with iron overload were homozygous for the C282Y mutation (73 of 73; 100%). The 6 proband cases that were negative for the C282Y mutation (Table 2) were all tested for the H63D mutation, which was negative in all cases. Two putative homozygotes were heterozygotes for C282Y and were clinically labeled as juvenile hemochromatosis. 7 Both had severe congestive heart failure, lifethreatening arrhythmias, and endocrine dysfunction before age 30 years. No other affected family members (n 10) were found in these 2 kindreds. One patient in this series had chronic viral hepatitis C virus by serological testing. Six patients (5 men and 1 woman) had a history of alcohol abuse (men, 80 g/day; women, 60 g/day). The clinical features of these cases with coexistent risk factors are shown in Table 3. Four family members in four different families were discovered through pedigree testing to be homozygous for hemochromatosis without iron overload. Two of these cases have been reported in complete pedigree studies, 8 Table 2. Genotypic and Phenotypic Characteristics of Putative Homozygotes That Are Not Homozygous for the C282Y Mutation Patient C282Y H63D (yr) Sex saturation (%) Ferritin Iron removed (g) 1 / / 65 F / / 25 M Stain a / / 66 F / / 74 F b / / 26 M b / / 37 M , NOTE. Reference ranges are as follows: transferrin saturation, 20% 55%, ferritin level, g/l (women) and g/l (men); hepatic iron, 0 35 mol/g; hepatic, 1.9; and iron removed, 1g. a concentration was not available; 4 stain refers to the scoring of the iron stain. 21 b These 2 patients were classified as having juvenile hemochromatosis.

3 February 1998 GENOTYPING IN HEMOCHROMATOSIS 321 Table 3. Iron Studies in C282Y Homozygotes With Alcohol Abuse or Hepatitis C Virus (yr) Sex Risk factor Ferritin level saturation (%) concentration 63 M Etoh F Etoh M Etoh M Etoh M Etoh M Etoh M Hepatitis C NOTE. Reference ranges are as follows: transferrin saturation, 20% 55%, ferritin level, g/l (women) and g/l (men); hepatic iron concentration, 0 35 mol/g; and hepatic, 1.9. Etoh, alcohol abuse. Cirrhosis and the clinical characteristics of all 4 cases are shown in Table 4. None had evidence or previous history of pathological blood loss, blood donation, or hereditary anemias. One subject had undergone an elective colonoscopy and liver biopsy. The 3 other subjects had negative stool testing for occult blood on two occasions. Patient 2 described heavy menses that ceased spontaneously at age 43 years. Phenotypic/Genotypic Correlations Of the 128 homozygotes for the C282Y mutation, 77 probands and 51 cases were discovered during familyscreening studies. Thirty-eight of 128 patients (30%) were considered to be asymptomatic for hemochromatosis at the time of diagnosis. The factors leading the patient to the physician and the physician to the diagnosis have been described previously. 1 In regard to severe phenotypic expression, 3 men diagnosed with cirrhosis were younger than 40 years (aged 26, 37, and 39 years). Serum transferrin saturation was 55% in 110 of 122 (90%) cases. The 12 cases with a normal transferrin saturation included 8 men and 4 women (median age, 57 years; age range, years). Six subjects had an elevated serum ferritin level and a transferrin saturation of 45% (mean ferritin level, 827 g/l; range, g/l). All had the transferrin saturation test repeated at least once and five times in 1 case. Five of the 12 subjects had another family member identified as homozygous for C282Y with iron overload. Serum ferritin level was 200 g/l in 97% of the women and 300 g/l in 96% of the men. Four of the subjects with a normal serum ferritin level are included in Table 4. index was available in 77 cases (48 probands and 29 discovered cases) and was 1.9 in 70 of 77 cases (91%). Iron removed by venesection was assessed in 81 patients and was 5 g in 44 of 63 patients (70%) and in 11 of 15 of the women (73%). Discussion The diagnosis of hemochromatosis has been problematic in the past for several reasons. There has been an underdiagnosis in the asymptomatic general population, as shown in population studies There has also been an overdiagnosis in patients with secondary iron overload caused by end-stage liver disease of diverse etiology. 12 The precision of the diagnostic criteria has not been validated against a gold standard. For example, it had been very difficult to make a diagnosis of genetic hemochromatosis with a normal transferrin saturation or ferritin level because these may have been the criteria for the diagnosis. Pedigree studies that only included HLAidentical siblings with iron overload had been a previous Table 4. Genotypic and Phenotypic Characteristics of C282Y Homozygotes Without Iron Overload Patient (yr) Sex saturation (%) Ferritin level concentration Other results 1 59 M Colonoscopy, antiendomysial antibody negative, and sister with iron overload 2 45 F Menorrhagia and uncle and 2 cousins with iron overload 3 69 M Two brothers with iron overload 4 71 F Sister and 2 brothers with iron overload NOTE. Reference ranges are as follows: transferrin saturation, 20% 55%; ferritin level, g/l (women) and g/l (men); hepatic iron concentration, 0 35 mol/g; and hepatic, 1.9.

4 322 ADAMS AND CHAKRABARTI GASTROENTEROLOGY Vol. 114, No. 2 method to validate the precision of phenotypic tests such as the hepatic. 4 The identification of a single mutation (C282Y) that is present in most well-defined pedigrees with hemochromatosis has allowed for a new reevaluation of diagnostic criteria against a genotypic gold standard. An important aspect of the validation of the phenotypic tests is not so much their use in a subspecialty referral clinic such as in this report, but in population screening studies. In this study, the test with the lowest sensitivity for the diagnosis was the amount of iron removed by venesection. This may be a reflection of the heterogeneity of the disease or other unknown factors affecting iron metabolism in the homozygotes. The low sensitivity of this criteria seems to suggest that this will be an inadequate absolute criteria for the diagnosis of hemochromatosis. The transferrin saturation has also been advocated in many studies as the most reliable screening test in population studies, but with a sensitivity of 90%, it may miss up to 10% of the homozygotes at a threshold of 55%. Lowering the threshold will greatly increase the number of subjects that must return for further screening studies and, thus, the cost of a screening program. 2,13 15 The serum ferritin level had a much higher sensitivity in this study. However, many patients were identified initially because of an elevated serum ferritin level, and the specificity has been the major problem with ferritin because its level is elevated in most liver diseases and inflammatory diseases. The hepatic had a sensitivity of 91%, which is similar to predictions made in pedigree studies. 4 The specificity of the hepatic iron index is also a difficult area because of the prevalence of iron overload in cirrhosis. A recent study of 447 explanted livers showed a hepatic of 1.9 in 8.5% of cases that were not clinically judged to be hemochromatosis. 12 The phenotypic tests in this study have been validated against a genotypic gold standard. However, the C282Y test is not a perfect test for hemochromatosis. In welldefined pedigree studies, the prevalence of the mutation has been described in 93% 100% of homozygotes, 5,16 whereas in unrelated patients, the prevalence ranges from 69% to 83%. 3,17,18 There are well-documented cases of iron overload in this study (Table 2) that are not homozygous for the mutation. Whether these cases are different iron-loading diseases or different mutations will require further study. Our clinical practice is to begin venesection therapy based on the presence of iron overload rather than genetic testing. For example, the 2 patients with juvenile hemochromatosis in this study had severe iron overload (hepatic of 13.6 and 32.7), yet appear to be heterozygotes by genotyping. This strongly suggests the presence of another yet to be determined gene or mutation in these cases. It has been uncommon, except in Italy, to describe familial iron overload that is negative for the C282Y mutation. 18 Many of the Italian cases have confounding variables such as chronic viral hepatitis or alcoholism. The low prevalence of the C282Y mutation in Italian patients of only 69% cannot be explained completely on the basis of differing phenotypic assignment because the prevalence of the mutation in the general population is also much lower than in other European countries. 19 Therefore, the use of the C282Y genotyping as the gold standard in these populations seems hazardous presently. The other issue that is shown in Table 4 is the C282Y homozygote without iron overload. It has not been our clinical practice to recommend venesection therapy for these C282Y homozygotes without demonstrable iron overload. Although the absence of iron overload was confirmed by liver biopsy in the extensive investigation in case 1, we believe this case shows that liver biopsy is not necessary to exclude iron overload in the C282Y homozygote with a normal serum ferritin level and transferrin saturation. In a young adult who is homozygous for C282Y with normal iron studies, we would recommend a careful assessment for pathological blood loss and a repeat assessment with biochemical iron studies in 5 years. Venesection therapy is started if the serum ferritin level increases to above the normal range ( 200 g/l for women and 300 g/l for men). Preliminary studies in countries with European ancestry have suggested that the prevalence of the gene for hemochromatosis may be as high as 1 in 150 in Australia. The difference between the genotypic prevalence and the phenotypic prevalence could be explained by incomplete penetrance and/or an underdiagnosis of the disease. If the penetrance is much lower than suspected previously, population screening with an initial genetic test will be problematic because of the identification of a large number of people without a disease and possibly genetic discrimination. A more practical approach would be initial screening for iron overload with phenotypic testing such as the transferrin saturation or unbound iron-binding capacity as has been shown in cost-effectiveness analyses. 13,20 Genotyping could be used as a confirmatory test in a patient suspected of iron overload with an elevation in transferrin saturation or ferritin level. 15 Genotyping should also replace HLA typing in the assessment of siblings of a C282Y homozygote. In summary, the use of C282Y genotyping has broadened our horizons in regard to the diagnosis of hemochromatosis because this study has shown both iron-loaded patients without the mutation and homozygous patients

5 February 1998 GENOTYPING IN HEMOCHROMATOSIS 323 without iron overload. Population studies will be required to quantitate the prevalence of these two clinical situations that will be important to plan ideal population screening strategies. References 1. Adams PC, Valberg LS. Evolving expression of hereditary hemochromatosis. Semin Liver Dis 1996;16: Witte DL, Crosby WH, Edwards CQ, Fairbanks VF, Mitros FA. Hereditary hemochromatosis. Clin Chim Acta 1996;245: Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, Dormishian F, Domingo R Jr, Ellis MC, Fullan A, Hinton LM, Jones NL, Kimmel BE, Kronmal GS, Lauer P, Lee VK, Loeb DB, Mapa FA, McClelland E, Meyer NC, Mintier GA, Moeller N, Moore T, Morikang E, Wolff RR, et al. A novel MHC class I like gene is mutated in patients with hereditary hemochromatosis. Nat Genet 1996;13: Adams PC, Bradley C, Henderson AR. Evaluation of the hepatic as a diagnostic criterion in hereditary hemochromatosis. J Lab Clin Med 1997;130: Jouanolle A-M, Gandon G, Jezequel P, Blayau M, Campion ML, Mosser J, Fergelot P, Chauvel B, Bouric P, Carn G, Andrieux N, Gicquel I, LeGall J-Y. Haemochromatosis and HLA-H. Nat Genet 1996;14: Bassett ML, Halliday JW, Powell LW. Value of hepatic iron measurements in early hemochromatosis and determination of the critical iron level associated with fibrosis. Hepatology 1986;6: Kaltwasser JP, Schalk KP, Werner E. Juvenile hemochromatosis. Ann NY Acad Sci 1988;526: Adams PC, Campion ML, Gwandon G, LeGall J-Y, David V, Jouanolle A-M. Clinical and family studies in genetic hemochromatosis: microsatellite and HFE studies in five atypical families. Hepatology 1997;26: Edwards CQ, Griffen LM, Goldgar D, Drummond C, Skolnick MH, Kushner JP. Prevalence of hemochromatosis among 11,065 presumably healthy blood donors. N Engl J Med 1988;318: Baer DM, Simons JL, Staples RL, Rumore GJ, Morton CJ. Hemochromatosis screening in asymptomatic ambulatory men 30 years of age and older. Am J Med 1995;98: Smith BN, Kantrowitz W, Grace ND, Greenberg MS, Patton TJ, Ookubo R, Sorger K, Semeraro J, Doyle J, Cooper A, Kamat B, Maregno L, Rand W. Prevalence of hereditary hemochromatosis in a Massachusetts corporation: is Celtic origin a risk factor? Hepatology 1997;25: Ludwig J, Hashimoto E, Porayko M, Moyer T, Baldus W. Hemosiderosis in cirrhosis: a study of 447 native livers. Gastroenterology 1997;112: Adams PC, Gregor JC, Kertesz AE, Valberg LS. Screening blood donors for hereditary hemochromatosis: decision analysis model based on a thirty-year database. Gastroenterology 1995;109: Edwards CQ, Kushner JP. Screening for hemochromatosis. N Engl J Med 1993;328: Bassett M, Leggett BA, Halliday JW, Webb S, Powell LW. Analysis of the cost of population screening for haemochromatosis using biochemical and genetic markers. J Hepatol 1997;27: Jazwinska EC, Cullen LM, Busfield F, Pyper WR, Webb SI, Powell LW, Morris CP, Walsh TP. Haemochromatosis and HLA-H. Nat Genet 1996;14: Beutler E, Gelbart T, West C, Lee P, Adams M, Blackstone R, Pockros P, Kosty M, Venditti C, Phatak P, Seese N, Chorney K, Ten Elshof A, Gerhard G, Chorney M. Mutation analysis in hereditary hemochromatosis. Blood Cells Mol Dis 1996;22: Carella M, D Ambrosio L, Totaro A, Grifa A, Valentino M, Piperno A, Girelli D, Roetto A, Franco B, Gasparini P, Camaschella C. Mutation analysis of the HLA-H gene in Italian hemochromatosis patients. Am J Hum Genet 1997;60: Merryweather-Clarke A, Pointon J, Sherman J, Robson K. Global prevalence of putative haemochromatosis mutations. J Med Genet 1997;34: Phatak PD, Guzman G, Woll JE, Robeson A, Phelps CE. Costeffectiveness of screening for hereditary hemochromatosis. Arch Intern Med 1994;154: Scheuer PJ, Williams R, Muir AR. pathology in relatives of patients with haemochromatosis. J Pathol Bacteriol 1962;84: Received July 14, Accepted October 7, Address requests for reprints to: Paul C. Adams, M.D., Department of Medicine, London Health Sciences Center University Campus, P.O. Box 5339, London, Ontario, Canada N6A 5A5. padams@julian.uwo.ca; fax: (519) The authors thank Leslie Valberg, Ann Kertesz, Bill Howson, David Mallott, and Veronique David for advice and assistance.