Prevalence and Penetrance of HFE Mutations in 4865 Unselected Primary Care Patients

Phatak et al.

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

Prevalence and Penetrance of HFE Mutations in 4865 Unselected Primary Care Patients1 Submitted 04/25/2002; revised 05/21/2002 (Communicated by E. Beutler, M.D., 06/04/2002)

Pradyumna D. Phatak,2,3 Daniel H. Ryan,3 Joseph Cappuccio,2,3 David Oakes,3 Caroline Braggins,2 Kim Provenzano,2 Shirley Eberly,3 and Ronald L. Sham2,3 ABSTRACT: Two HFE gene mutations, C282Y and H63D, underlie the vast majority of cases of hereditary hemochromatosis. We performed a cross-sectional primary care-based study to determine the allele frequency of the C282Y and H63D mutations and the penetrance of each of the affected genotypes defined by their presence. Patients had previously undergone transferrin saturation (TS) testing. A total of 4865 unselected frozen serum samples were analyzed to determine serum ferritin (SF) levels. Genomic DNA isolated from these samples was analyzed for the C282Y and H63D HFE mutations. Homozygotes for each mutation and compound heterozygotes were evaluated to determine clinical penetrance. The allele frequency of C282Y was 0.0507 among Caucasian and 0.0067 among African Americans; that of H63D was 0.1512 and 0.0263, respectively. TS was ⱖ55% in 83% of individuals with C282Y/C282Y, 14.5% of C282Y/H63D, and 5% of H63D/H63D; SF was ⱖ300 ␮G/L in 42, 9, and 5% of these genotypes, respectively. None of the 12 C282Y homozygotes had cardiac dysfunction or hepatic cirrhosis. Only 9/129 (7%) individuals with the genotypes C282Y/H63D or H63D/H63D had a SF ⱖ300␮G/L; many had explanations other than iron overload that accounted for this increase. Thus, the prevalence of the common HFE mutations is the same in our population as previously described. TS screening would detect most C282Y homozygotes but not the other two genotypes. The penetrance of C282Y/C282Y is significant. The biochemical penetrance of H63D/H63D and C282Y/H63D is modest and the clinical penetrance is low. © 2002 Elsevier Science (USA)

INTRODUCTION

studies, the penetrance of the C282Y mutation has been postulated to be high. The penetrance of the H63D mutation is probably low, leading some authorities to question its significance in causing iron overload. A large population-based screening study from Busselton, Australia found a high biochemical penetrance rate among C282Y homozygotes (7). A recent large United States-based population screening study found that only 75% of male and 40% of female C282Y homozygotes had transferrin saturation (TS) levels ⬎50% (8). Furthermore, only 1 of 152 C282Y homozygotes identi-

Hereditary hemochromatosis (HHC) is one of the most common inherited disorders among Caucasian populations (1– 4). Two common mutations in the HFE gene, c.845G 3 A (C282Y) and c.187C 3 G (H63D), have been found in the vast majority of HHC cases (5, 6). Genetic testing of individuals with proven iron overload reveals that most such individuals are homozygotes for C282Y, although some patients are compound heterozygotes or are homozygous for H63D (5, 6). Largely based upon these data and family

Correspondence and reprint requests to: Pradyumna D. Phatak, M.D., Rochester General Hospital, 1425 Portland Avenue, Rochester, NY 14621. Telephone: (585) 922-4020. Fax: (585) 922-4622. E-mail: [email protected]. 1 Grant Support: This project was supported in part by Agency for Health Care Policy and Research (Grant No. RO1 HS07616), and by the National Heart, Lung and Blood Institute (Grant No. 1 RO1 HL61428-01A1). 2 Department of Medicine, Rochester General Hospital, and the Mary M. Gooley Hemophilia Center Inc., 1425 Portland Ave. Rochester, New York 14621. 3 The University of Rochester School of Medicine and Dentistry. 1079-9796/02 $35.00 © 2002 Elsevier Science (USA)

All rights reserved.

41

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

Phatak et al.

fied in this study had signs and symptoms of advanced hemochromatosis. The current study adds to previous descriptions of the penetrance of HFE mutations and includes detailed clinical evaluation and assessment of mobilizable iron stores among all C282Y homozygotes and compound heterozygotes identified by screening. Moreover, no study to date has systematically examined penetrance among unselected H63D homozygotes. The study outlined herein constitutes the first such description in an unselected primary care population.

nique modified for analysis of DNA isolated from serum samples. Serum samples were thawed and total DNA extracted from 250 ␮l of serum using the Bio-Rad Instagene Matrix kit (Bio-Rad Inc., Richmond, CA) according to the manufacturer’s directions. Ten microliters of this preparation was used as a template for PCR. HFE C282Y and H63D mutations were detected using primers as described by Feder et al. (6) in a PCR-RFLP assay (9) using the similar enzyme BclI in place of MboI for the H63D enzyme digestion step. All samples genotyped as C282Y homozygote were confirmed with inner primers (sense: CAAGTGCCTCCTTTGGTGAA; antisense: TCACATACCCCAGATCACAA) to rule out false assignment of homozygosity due to a previously described polymorphism at primer binding sites in either intron 3 (10) or intron 4 (11). PCR contamination was monitored by simultaneous amplification of a negative control containing all reagents except template DNA. Proper amplification and cleavage of PCR product in each run was assessed by simultaneous analysis of a heterozygote control sample. In addition, all H63D homozygous mutant results were repeated to confirm the result.

MATERIALS AND METHODS Study Population This study was approved by the Institutional Review Board of the University of Rochester. The 4865 serum samples analyzed in this study were derived from the screening study performed on 16,031 primary care patients, which has been previously reported (4). All samples were frozen during the latter part of the screening study; thus, this subset of 4865 constitutes a random sample of our total population, with the only selection criterion being date of entry into the screening study. Of the 4900 samples available, adequate amounts of DNA could not be extracted from 35, leaving 4865 available for analysis. A total of 4708 samples were present in amounts adequate for serum ferritin (SF) testing; thus the correlation of genotype with SF levels is based on this subset.

Statistical Analysis Confidence intervals of 95% for the true allele frequencies for each ethnic group were calculated using the exact method for binomial proportions (SAS PROC FREQ). An Omnibus ␹2-test followed by individual comparisons using two-sample tests for binomial proportions was used to test for significant differences in allele frequencies among ethnic groups. Analysis of variance following transformation of the response variable to normality was used to assess the dependence of TS and SF levels on genotype and gender.

Biochemical Tests TS values were determined initially as part of the screening study. SF was measured on the residual serum sample after DNA extraction. SF was measured using a Bayer ACS:180 Automated Chemiluminescence System. The assay is a twosite sandwich immunoassay using direct, chemiluminometric technology, which uses constant amounts of two anti-ferritin antibodies.

RESULTS The 4865 patients had a mean age of 52 years (range 18 –96 years). Forty-two percent were male and 66% were Caucasian. Seven percent had a C282Y mutation and 22% had an H63D mutation.

Genotyping The HFE genotyping on 4865 saved serum samples was performed using a PCR-RFLP tech42

Phatak et al.

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

H63D was significantly higher among Hispanic individuals (P ⬍ 0.0001). The distribution of TS level by genotype is shown in Fig. 1. TS levels were significantly higher in any of the groups with an HFE mutation as compared to that with the wild-type genotype (overall P ⬍ 0.0001). Figure 2 shows the distribution of SF levels by genotype. Although the groups showed a significant difference overall (P ⫽ 0.0024), pairwise comparisons showed that only the C282Y homozygous group was significantly different from the wild type. Among the population of 4865 subjects, there were 12 individuals who were C282Y homozygous, 55 who were compound heterozygous, and 78 H63D homozygous. All 12 C282Y homozygotes were of Caucasian descent. When the allele frequency was analyzed by race, there was no evidence of departure from the Hardy-Weinberg equilibrium. The number of Caucasian C282Y homozygotes was slightly greater than expected but this difference was not statistically significant (P ⫽ 0.17).

TABLE 1 Allele Frequency of C282Y and H63D by Racial Group

Mutation C282Y H63D a

Caucasian N ⫽ 3227

African N ⫽ 1123

Hispanic N ⫽ 475

0.0507 0.0067 0.0126 (0.0454–0.0563)a (0.0037–0.0110) (0.0065–0.0220) 0.1512 0.0263 0.1063 (0.1426–0.1602) (0.0201–0.0338) (0.0874–0.1277)

95% confidence limits are shown in parentheses.

The allele frequencies of the two HFE mutations by racial group are shown in Table 1. The frequency among Caucasians was significantly higher than that among African Americans for both C282Y (P ⬍ 0.0001) and H63D (P ⬍ 0.0001) mutations. Caucasian allele frequency was also significantly higher than Hispanic for both mutations (P ⬍ 0.0001 and P ⫽ 0.0002, respectively). The allele frequency of C282Y was not significantly different between individuals of African and those of Hispanic descent; that of

FIG. 1. Distribution of transferrin saturation level by gender and HFE genotype. The distribution of transferrin saturation levels for each genotype is shown by gender. Boxplots show the median with the limits of the boxes representing the 25th and 75th percentiles. The whiskers outside the box represent the limits of the data points excluding outliers. Outliers, shown as asterisks, are defined as being outside the box by more than 1.5 times the interquartile range. Genotypes are abbreviated as follows: W/W, wild type; H/W, H63D heterozygote; C/W, C282Y heterozygote; H/H, H63D homozygote; C/H, compound heterozygote; C/C, C282Y homozygote. The number of individuals in each group is shown. Transferrin saturation for each group with an HFE mutation was significantly higher than in the wild-type group (overall P ⬍ 0.0001). 43

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

Phatak et al.

gotes and 7/9 of the other two genotypes with SF levels ⬎300 ␮G/L, were evaluated by one of the investigators. Thus, clinical data is available on most individuals who might have been expected to have signs or symptoms of iron overload. Of the 12 C282Y homozygotes, 5/12 had SF ⱖ300 ␮G/L; phlebotomy treatments were recommended in 7/12. At the time of this report, 4 C282Y homozygotes had completed iron depletion by therapeutic phlebotomy; 3 of 4 had iron stores ⬎4G. Four individuals had liver biopsies performed; all 4 had increased hepatic iron levels and none had hepatic cirrhosis. None of these subjects had significant cardiac dysfunction or diabetes mellitus but 6/12 had arthritis or arthralgias. It is important to point out, however, that arthritis and arthralgias are common in primary care populations; thus, this finding may not represent a causal relationship. Of the 7 C282Y homozygotes with SF ⬍300 ␮G/L, 3 had reasons for blood loss that might have accounted for reduced body iron stores. One individual had significantly increased iron stores despite donating at least 40 units during his lifetime. The only homozygote that missed our TS screen had a repeat TS of 54%. Retrospective chart review showed no evidence of an inflammatory process during the time of the initial screen that may have accounted for the low screening TS level. The clinical characteristics of these 12 subjects are summarized in Table 4. Of the 133 individuals with the other two genotypes, 33 had TS levels ⱖ45% and only 12 had TS ⱖ55% (Table 2). Of these individuals with elevated TS levels, 3 also had SF levels ⱖ300 ␮G/L. Overall, 9/129 had SF ⱖ300 ␮G/L; 7/9 were males. Therapeutic phlebotomy was recommended in 2 cases; one refused and the other had increased mobilizable iron stores. As shown

FIG. 2. Distribution of serum ferritin level by gender and HFE genotype. The distribution of serum ferritin levels for each genotype is shown by gender. Boxplots show the median with the limits of the boxes representing the 25th and 75th percentiles. The whiskers outside the box represent the limits of the data points excluding outliers. Outliers, shown as asterisks, are defined as being outside the box by more than 1.5 times the interquartile range. Data points for five individuals, all wild type (three male, two female), with serum ferritin levels ranging from 1534 to 3171 ␮G/L, are omitted from the plots due to their extreme values. Genotypes are abbreviated as follows: W/W, wild type; H/W, H63D heterozygote; C/W, C282Y heterozygote; H/H, H63D homozygote; C/H, compound heterozygote; C/C, C282Y homozygote.The number of individuals in each group is shown. Only the C282Y homozygote group had significantly higher serum ferritin levels compared to the wild type.

The proportion of those individuals who would meet conventional biochemical screening criteria for HHC, by gender, is shown in Table 2. The proportion with abnormal SF levels, indicating possibly increased body iron stores, is shown in Table 3. A total of 64/145 individuals with one of the three HFE genotypes that have been implicated in iron overload, including all 12 C282Y homozy-

TABLE 2 Proportion of Homozygotes and Compound Heterozygotes with Elevated TS Levels TS ⱖ 45%

TS ⱖ 55%

Genotype

Male

Female

Total

Male

Female

Total

C282Y/C282Y C282Y/H63D H63D/H63D

3/4 (75.0%) 7/26 (26.9%) 10/35 (28.6%)

8/8 (100.0%) 8/29 (27.6%) 8/43 (18.6%)

11/12 (91.7%) 15/55 (27.3%) 18/78 (23.1%)

3/4 (75.0%) 4/26 (15.4%) 3/35 (8.6%)

7/8 (87.5%) 4/29 (13.8%) 1/43 (2.3%)

10/12 (83.3%) 8/55 (14.5%) 4/78 (5.1%)

44

Phatak et al.

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

TABLE 3 Proportion of Homozygotes and Compound Heterozygotes with Elevated SF Levels SF ⱖ 200 ␮G/L

SF ⱖ 300 ␮G/L

Genotype

Male

Female

Total

Male

Female

Total

C282Y/C282Y C282Y/H63D H63D/H63D

4/4 (100.0%) 13/26 (50.0%) 6/33 (18.2%)

4/8 (50.0%) 3/28 (10.7%) 4/42 (9.5%)

8/12 (66.7%) 16/54 (29.6%) 10/75 (13.3%)

2/4 (50.0%) 5/26 (19.2%) 2/33 (6.1%)

3/8 (37.5%) 0/28 (0.0%) 2/42 (4.8%)

5/12 (41.7%) 5/54 (9.3%) 4/75 (5.3%)

in Table 5, 4/9 had other medical conditions that could be potential explanations for the high SF levels.

HFE mutations in a large African American population. As expected, the frequencies of both C282Y and H63D are substantially lower among this group. The frequency of C282Y among African Americans in our population can be explained by an admixture rate of Caucasian genes of about 0.16. Most estimates suggest a higher admixture rate of Caucasian genes among African Americans in the northeastern United States (13). Thus, our data suggest that the occurrence of the C282Y mutation among African Americans is completely explained by the admixture of Caucasian genes and would predict that native Africans would have a negligibly low frequency of this allele; this is supported by existing literature (3, 14). Other groups have reported the biochemical and clinical penetrance of the homozygous C282Y genotype. The vast majority of C282Y homozygotes in our study (11/12) had TS levels ⱖ45% and would have been detected by TS screening. This finding is in agreement with the Busselton study (7) but is in contrast to the Kaiser

DISCUSSION The allele frequencies of the two common HFE mutations, C282Y and H63D, in our study are similar to those that have been previously described. The C282Y mutation is more common in individuals of Northern European descent. Studies from the United Kingdom, France, and Australia find that over 90% of HHC patients are homozygous C282Y (1, 7). In contrast, studies from Italy find a lower prevalence of C282Y, and homozygosity for this mutation accounts for a lower proportion of Italian HHC cases (12). As expected, our study confirmed that the allele frequency of C282Y was highest among Caucasian individuals. Those of Hispanic descent had higher frequencies of the H63D mutation compared to those of African descent. Our study describes the allele frequency of

TABLE 4 Clinical Features of C282Y Homozygotes ID

Age/Sex

TS

SF

Mob Fe

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

92/F 65/M 64/M 70/F 23/M 29/F 67/F 58/M 75/F 53/F 49/F 36/F

98% 94% 100% 95% 63% 66% 70% 29% 73% 59% 52% 59%

1259 977 875 506 298 417 211 281 121 95 22 44

N/A 8.2G 6.2G 3.1G N/A 5.0G N/A In process N/A N/A N/A N/A

45

Comment Deceased Regular blood donor Lost to F/U Hx bleeding ulcers Regular blood donor Heavy menstrual losses

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

Phatak et al.

TABLE 5 Clinical Features of Compound Heterozygotes and H63D Homozygotes with High SF Levels ID

Genotype

Age/Sex

TS %

SF

Comment

1 2 3 4 5 6 7 8 9

C282Y/H63D C282Y/H63D C282Y/H63D C282Y/H63D C282Y/H63D H63D/H63D H63D/H63D H63D/H63D H63D/H63D

53/M 66/M 74/M 62/M 38/M 43/M 66/M 75/F 89/F

54 27 34 66 34 32 56 44 20

692 598 341 348 424 411 677 304 302

Metastatic lung cancer DM/MI; deceased CHF-Refused follow-up

Permanente study (8) and that from Wales (1). Although the number of C282Y homozygotes identified in our study is small, it is of note that 8/8 female homozygotes in our study had TS levels ⱖ50% in comparison to only 40% of female homozygotes in the Kaiser Permanente study. This difference is statistically significant (P ⫽ 0.0012). Either iron deficiency, particularly among women of child-bearing age, or an inflammatory illness that lowered serum iron may explain the occurrence of low TS levels. The one individual in our study who had a low screening TS level has had a subsequent fasting TS of 54%; however, retrospective chart review did not find any evidence of an inflammatory process that might explain his low screening level. Five of 12 C282Y homozygotes in our study had abnormally high SF levels (ⱖ300 ␮G/L). Mobilizable iron among the 5 individuals who underwent therapeutic phlebotomy ranged from 3 to 8 G and 4 were ⬎4G, thus having levels that meet phenotypic definitions of significant iron overload (15). However, none had HHC associated end-organ failure such as hepatic cirrhosis or cardiomyopathy. Of note, two individuals in our original screening study had hepatic cirrhosis (4) but neither was part of the random subset analyzed for this study. The biochemical and clinical penetrance of compound heterozygotes (for C282Y and H63D) and H63D homozygotes appears to be low. Only 9/129 individuals had an SF ⱖ300 ␮G/L. Many had other reasons for the elevation in SF and only 1 had clinically significant iron overload. Although compound heterozygous individuals ac-

Unable to contact Mob Fe 4.1G CHF, DM; deceased

count for a small proportion of HHC patients, the clinical penetrance has been calculated to be low at ⬍1.5% (16). Our data confirm this contention directly. The penetrance of the homozygous H63D phenotype is probably even lower, leading some to question the significance of this genotype (17). Our group has reported cases of iron overload among H63D homozygotes (18) but individuals with this genotype often have other co-morbidity when they develop significant iron overload. The data presented herein confirm the low expected penetrance of this genotype when population screening identifies unselected individuals. Our study addresses the penetrance of the common HFE mutations among homozygous individuals. Penetrance was defined by biochemical iron studies, documented iron overload, and organ dysfunction that is well known to be associated with the homozygous HHC phenotype. The reason for variable penetrance among homozygous individuals is unclear. Dietary factors as well as occult blood loss and blood donation may play a role. This appeared to be the case in some C282Y homozygotes in our study. Additionally, other modifier genes involved with iron metabolism may modulate the effects of these mutations. Several such candidate genes have been proposed based on animal models (19). The role of these in defining penetrance will undoubtedly be a focus of future studies. Population screening for HHC has been advocated by many but doubts regarding disease penetrance and concerns surrounding genetic labeling and discrimination have led to cautious statements 46

Phatak et al.

Blood Cells, Molecules, and Diseases (2002) 29(1) July/Aug: 41– 47 doi:10.1006/bcmd.2002.0536

by public health authorities (20). Our study adds to existing reports on the penetrance of homozygosity for C282Y and constitutes the first such study based in United States primary care practices. Additionally, we directly confirm the low penetrance of the C282Y/H63D and H63D/H63D genotypes, confirming that population screening for these genotypes is probably not a fruitful exercise.

9.

10.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

11.

Jackson, H. A., Carter, K., Darke, C., Guttridge, M. G., Ravine, D., Hutton, R. D., Napier, J. A., and Worwood, M. (2001) HFE mutations, iron deficiency and overload in 10,500 blood donors. Br. J. Haematol. 114(2), 474 – 484. Edwards, C. Q, Griffen, L. M, Goldgar, D., Drummond, C., Skolnick, M. H, and Kushner, J. P. (1988) Prevalence of hemochromatosis among 11,065 presumably healthy blood donors. N. Engl. J. Med. 318, 1355–1362. Merryweather-Clarke, A. T., Pointon, J. J., Shearman, J. D., and Robson, K. J. (1997) Global prevalence of putative haemochromatosis mutations [see comments]. J. Med. Genet. 34, 275–278. Phatak, P. D., Sham, R. L., Raubertas, R. F., Dunnigan, K., O’Leary, M. T., Braggins, C., and Cappuccio, J. D. (1998) Prevalence of hereditary hemochromatosis in 16031 primary care patients. Ann. Intern. Med. 129, 954 –961. Bacon, B. R., Powell, L. W., Adams, P. C., Kresina, T. F., and Hoofnagle, J. H. (1999) Molecular medicine and hemochromatosis: At the crossroads. Gastroenterology 116, 193–207. Feder, J. N., Gnirke, A., Thomas, W., Tsuchihashi, Z., Ruddy, D. A., Basava, A., Dormishian, F., Domingo, R., Jr., Ellis, M. C., Fullan, A., Hinton, L. M., Jones, N. L., Kimmel, B. E., Kronmal, G. S., Lauer, P., Lee, V. K., Loeb, D. B., Mapa, F. A., McClelland, E., Meyer, N. C., Mintier, G. A., Moeller, N., Moore, T., Morikang, E., and Wolff, R. K. (1996) A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis [see comments]. Nature Genet. 13, 399 – 408. Olynyk, J. K., Cullen, D. J., Aquilia, S., Rossi, E., Summerville, L., and Powell, L. W. (1999) A population-based study of the clinical expression of the hemochromatosis gene [see comments]. N. Engl. J. Med. 341, 718 –724. Beutler, E., Felitti, V., Koziol, J, Ho, N. J., and Gelbart, T. (2002) Penetrance of 845G A (C282Y)

12.

13. 14.

15.

16.

17.

18.

19.

20.

47

HFE hereditary hemochromatosis mutation in the USA. Lancet 359, 211–218. Jouanolle, A. M., Gandon, G., Jezequel, P., Blayau, M., Campion, M. L., Yaouanq, J., Mosser, J., Fergelot, P., Chauvel, B., Bouric, P., Carn, G., Andrieux, N., Gicquel, I., Le Gall, J. Y., and David, V. (1996) Haemochromatosis and HLA-H [letter; comment]. Nature Genet. 14, 251–252. Beutler, E., and Gelbart, T. (2000) A common intron 3 mutation (IVS3 -48c3g) leads to misdiagnosis of the c.845G3 A (C282Y) HFE gene mutation. Blood Cells Mol. Dis. 26, 229 –233. Jeffrey, G. P., Chakrabarti, S., Hegele, R. A., and Adams, P. C. (1999) Polymorphism in intron 4 of HFE may cause overestimation of C282Y homozygote prevalence in haemochromatosis [letter]. Nature Genet. 22, 325–326. Carella, M., D’Ambrosio, L., Totaro, A., Grifa, A., Valentino, M. A., Piperno, A., Girelli, D., Roetto, A., Franco, B., Gasparini, P., and Camaschella, C. (1997) Mutation analysis of the HLA-H gene in Italian hemochromatosis patients [see comments]. Am. J. Hum. Genet. 60, 828 – 832. Reed, T., and Edward. (1966) Caucasian Genes in American Negroes. Science 165, 762–768. Roth, M., Giraldo, P., Hariti, G., Poloni, E. S., Sanchez-Mazas, A., Stefano, G. F. D., Dugoujon, J. M., and Coppin, H. (1997) Absence of the hemochromatosis gene Cys282Tyr mutation in three ethnic groups from Algeria (Mzab), Ethiopia, and Senegal. Immunogenetics 46, 222–225. Powell, L. W., George, D. K., McDonnell, S. M., and Kowdley, K. V. (1998) Diagnosis of hemochromatosis. [Review]. Ann. Intern. Med. 129, 925–931. Beutler, E., Gelbart, T., West, C., Lee, P., Adams, M., Blackstone, R., Pockros., P., Kosty, M., Venditti, C. P., Phatak, P. D., Seese, N. K., Chorney, K. A., Ten Elshof, A. E., Gerhard, G. S., and Chorney, M: Mutation analysis in hereditary hemochromatosis. Blood Cells, Mol. Dis. 22, 187–194. Beutler, E. (1997) The significance of the 187G (H63D) mutation in hemochromatosis [letter; comment]. Am. J. Hum. Genet. 61, 762–764. Sham, R. L., Ou, C. Y., Cappuccio, J., Braggins, C., Dunnigan, K., and Phatak, P. D. (1997) Correlation between genotype and phenotype in hereditary hemochromatosis: Analysis of 61 cases. Blood Cells Mol. Dis. 23, 314 –320. Andrews, N. C. (2000) Iron homeostasis: Insights from genetics and animal models. Nature Rev. Genet. 1(3), 208 –217. Adams, P., Brissot, P., Powell, L. W. (2000) EASL International Consensus Conference on Haemochromatosis. J. Hepatol. 33, 485–504.