Digestive Diseases and Sciences

, Volume 57, Issue 11, pp 2988–2994

Diagnosis of Hereditary Hemochromatosis in the Era of Genetic Testing

Authors

  • Christiane Trieß
    • Department of Internal Medicine IUniversity Hospital Ulm
  • Guido von Figura
    • Department of Internal Medicine IUniversity Hospital Ulm
  • Manfred Stuhrmann
    • Institute of Human GeneticsMedical School Hannover
  • Barbara Butzeck
    • Hemochromatosis Association Germany (HVD)
  • Pierre A. Krayenbuehl
    • Klinik und Poliklinik für Innere MedizinUniversitätsSpital Zürich
  • Pavel Strnad
    • Department of Internal Medicine IUniversity Hospital Ulm
    • Department of Internal Medicine IUniversity Hospital Ulm
    • Spital Waldshut
Original Article

DOI: 10.1007/s10620-012-2243-z

Cite this article as:
Trieß, C., von Figura, G., Stuhrmann, M. et al. Dig Dis Sci (2012) 57: 2988. doi:10.1007/s10620-012-2243-z

Abstract

Background

Homozygous C282Y mutation in HFE gene is responsible for the majority of hereditary hemochromatosis cases. Since 1996 this mutation can be identified by a simple genetic test.

Aims

To determine the clinical presentations in patients with homozygous HFE C282Y mutation and the impact of genetic testing on the time needed for diagnosis.

Methods

A total of 414 patients diagnosed with C282Y homozygous hereditary hemochromatosis before and after the introduction of genetic testing were evaluated regarding symptoms and clinical findings at diagnosis as well as first hemochromatosis-related clinical features in their past medical history.

Results

At the time of diagnosis, the predominant symptom was joint pain, in particular of the hands/wrists. Those patients presenting with hand/wrist arthralgia had significantly higher ferritin levels than patients without this joint involvement (p = 0.0005 for males and p < 0.0001 for females). After the introduction of the HFE genetic test an earlier diagnosis after first onset of hemochromatosis-associated clinical features was observed between 2006 and 2009 vs. 2000–2005 p = 0.01).

Conclusions

Arthralgia, in particular of the hands/wrists, is a hallmark of hereditary hemochromatosis and its presence is associated with higher ferritin levels. Despite the availability of a genetic test, it often takes more than 6 years from the first onset of clinical features to diagnose hereditary hemochromatosis. This underlines the importance of raising the awareness of hemochromatosis and its typical clinical presentations.

Keywords

HemochromatosisHFEGenetic testArthralgia

Introduction

Hereditary hemochromatosis (HH) is a common genetic disorder [1]. The underlying genetic defects affect genes that regulate iron uptake, which leads to an inappropriately high absorption of iron, with subsequent iron deposition in liver, heart, endocrine pancreas, and joints, among others. As a consequence, HH comprises a wide range of clinical features including liver cirrhosis, heart failure, arthralgia, impotence in males, and diabetes.

Most of the HH cases are attributable to mutations in the HFE gene. Two mutations in HFE have been identified that account for most of the HH cases, namely C282Y and H63D [2]. HH may develop in individuals who are either homozygous for C282Y or compound heterozygous for C282Y and H63D. However, the penetrance is relatively low. For instance, ~28 % of male and ~1 % of female C282Y homozygous [3] and only 1–2 % of C282Y and H63D compound heterozygous patients [4] develop clinical features of hemochromatosis.

In Germany and Switzerland, genetic testing for the C282Y and H63D mutations was first introduced into clinical practice in 1996. It is recommended for use with patients that display ferritin elevations and a transferrin saturation >45 % in women and >50 % in men. In addition, this test is used as a screening test in first-degree relatives of diagnosed HH patients.

The objective of this study was to determine the clinical features at presentation and the time required for diagnosis in patients that were newly diagnosed as having HH in outpatient liver clinics and family practices. The time required for diagnosis was determined by the period of time between the onset of initial clinical features indicative of hemochromatosis and the date of definitive diagnosis of hereditary hemochromatosis. Initial clinical features were defined as elevated transaminase levels, elevated ferritin, and unexplained arthralgia. Arthralgia was found to be a common symptom at presentation. Together with diabetes, another typical feature of the disease, it was associated with higher ferritin levels. Our results indicate that despite a slight decrease in the time needed to diagnose the disease after the introduction of genetic testing into clinical practice, it still takes around 6 years to diagnose HH.

Patients and Methods

Recruitment of Patients

In this retrospective, multicentric study, we analyzed 414 patients with hereditary hemochromatosis caused by the homozygous C282Y mutation. A total of 228 patients were recruited at the university hospitals of Ulm, Heidelberg, and Zurich, and were seen at the respective specialized liver clinics. The data needed for this study was retrieved from the patients’ medical records. If clinical data (e.g., laboratory investigations, medical history) was missing, the family doctors and patients were contacted in order to collect it.

An additional 39 patients were included through collaboration with the Institute of Human Genetics at Hannover Medical School. Only patients with a homozygous HFE C282Y mutation were included. The patients and their family doctors were contacted to gather the necessary medical records.

The remaining 147 patients were recruited in collaboration with the Hemochromatosis Association Germany (HVD), which is an association supporting affected patients in Germany. These patients and their family doctors were contacted with a questionnaire to gather the data necessary for this study. The questionnaire asked for the patients’ age, sex, and year of diagnosis alongside the result of the genetic test. The result of a liver biopsy was included in this study if performed in a patient less than a year before or after the diagnosis. Clinical features evaluated at diagnosis and at the time the questionnaire was distributed were diabetes and joint pain with localization. In addition, laboratory workup results at diagnosis were retrieved, including the levels of ferritin, AST, ALT, iron and transferrin, as well as transferrin saturation. The time to diagnosis was calculated as the gap between occurrence of first hemochromatosis-related features (elevated liver enzyme levels, elevated ferritin, and unexplained arthralgia) and the date of diagnosis. The study was approved by the local ethical committee (ethical protocol Nr. 21/09).

Genetic Testing

The genetic test was performed in the respective centers as a part of the routine diagnostic procedure. Of the 414 patients, 356 were diagnosed after the introduction of the genetic test in 1996 and 58 were diagnosed before genetic testing was available. The diagnosis of all patients in this study was confirmed by genetic testing. The diagnosis of patients diagnosed before 1996 by liver biopsy was confirmed by genetic testing. Patients undergoing screening did not have hemochromatosis-associated clinical symptoms at the time of diagnosis. These patients were identified by family screenings or because of abnormally high iron parameters noted during routine clinical examinations.

Statistical Analysis

For the statistical analysis, Microsoft Excel and GraphPad Prism 4.0 were used. p values were calculated with the non-parametric Mann–Whitney test. p values for the liver biopsy results were calculated using Fisher’s exact test. For the analysis of correlations, a non-parametric Spearman’s test was performed.

Results

Clinical Features at Diagnosis of Hereditary Hemochromatosis

We evaluated the clinical features of 414 patients exhibiting hereditary hemochromatosis as a consequence of the presence of homozygous HFE C282Y mutation. Patients that were symptomatic at the time of diagnosis were compared to screening patients. In 58 symptomatic patients (43 male, 15 female), who were diagnosed before the introduction of the genetic test in 1996, the diagnosis was established through liver biopsy and later confirmed by genetic testing. A total of 282 symptomatic patients (163 male, 119 female) were diagnosed after 1996 by genetic testing. The remaining 74 patients (41 male, 33 female) were asymptomatic at diagnosis and identified by a genetic screening.

Table 1 summarizes the characteristics of symptomatic and screening patients. The mean age at diagnosis for symptomatic patients was not significantly different for males and females. However, screening patients were significantly younger than symptomatic patients at the time of diagnosis. The iron parameters determined in this study were ferritin, iron and transferrin saturation. Ferritin levels were significantly higher in males than in females of the symptomatic and screening population, whereas for both genders, symptomatic patients had significantly higher ferritin levels than the screening patients. In contrast to the ferritin results, the transferrin saturation was only found to be significantly higher in symptomatic males compared to the screening population. Except for an increased ALT in male symptomatic versus screening patients, the transaminases (AST, ALT) did not appear to significantly differ between the screening and symptomatic patients. However, AST and ALT were higher in males than in females in both populations (not significant for AST in the screening population). In addition, in a subset of the patients, the result of a liver biopsy at the time of diagnosis was available. Whereas all liver biopsies revealed iron overload, the occurrence of cirrhosis at the time of diagnosis did not differ between symptomatic males and females.
Table 1

Characteristics of the hemochromatosis patients

 

Symptomatic patients

Screening patients

Screening versus symptomatic

Male

Female

p value

Male

Female

p value

p value male

p value female

Sex (n)

206

134

 

41

33

   

Age (years)

51.8 ± 11.6

53.3 ± 11.2

0.16

39.0 ± 16.0

39.73 ± 16.3

0.78

<0.0001

<0.0001

Symptom duration (years)

6.2 ± 5.5

5.7 ± 5.0

0.67

n.a.

n.a.

   

n = 186

n = 122

Ferritin (μg/l)

2,273 ± 1,765

1,247 ± 1,508

<0.0001

882 ± 659

489 ± 417

0.0004

<0.0001

<0.0001

n = 186

n = 125

n = 41

n = 32

ALT (U/l)

58.1 ± 37.0

41.3 ± 101.5

<0.0001

54.6 ± 63.5

25.4 ± 13.2

0.01

0.02

0.14

n = 141

n = 103

n = 33

n = 30

AST (U/l)

40.9 ± 24.8

33.2 ± 63.0

<0.0001

34.2 ± 25.5

25.7 ± 15.5

0.12

0.50

0.53

n = 137

n = 98

n = 33

n = 26

Iron (μmol/l)

40.7 ± 28.2

36.3 ± 18.7

0.0007

34.0 ± 8.2

35.4 ± 8.9

0.53

0.31

0.54

n = 168

n = 116

n = 33

n = 29

Transferrin saturation (%)

81.8 ± 14.9

78.5 ± 14.4

0.1

70.2 ± 19.5

74.1 ± 21.0

0.41

0.003

0.42

n = 139

n = 99

n = 33

n = 28

Liver biopsy [n (%)]

67 (100)

27 (100)

 

7 (100)

4 (100)

   

No fibrosis

25 (37)

7 (26)

0.34

4 (29)

3 (75)

   

Mild fibrosis

24 (21)

6 (22)

0.23

3 (43)

0 (0)

   

Severe fibrosis

12 (18)

10 (37)

0.06

1 (14)

1 (25)

   

Cirrhosis

16 (24)

4 (15)

0.41

1 (14)

0 (0)

   

Demographic data (sex, age at diagnosis) and clinical findings (symptom duration, iron parameters, transaminases, and result of liver biopsy) of the patients included in the study are depicted. Symptomatic patients (i.e., with symptoms at the time of diagnosis) are presented separately from screening patients (i.e., non-symptomatic subjects at diagnosis). Normal reference ranges: ferritin men 30–300 μg/l, women 30–160 μg/l; transferrin saturation 16–45 %; AST/ALT men 2–45 U/l, women 2–35 U/l. n.a.not applicable

At the time of diagnosis, the patients were evaluated regarding their hemochromatosis-associated symptoms. In particular, occurrence and localization of arthralgia was evaluated. Frequency and localization of arthralgia in symptomatic patients are depicted in Fig. 1. Evidently, arthralgia was a very common symptom in the investigated population (Fig. 1a) and the predominant localization was the hand/wrist. Of note is that the presence of arthralgia was associated with higher ferritin levels at the time of diagnosis (Fig. 1b, c). In particular, hand/wrist affection appeared to be strongly associated with higher ferritin levels. Besides foot/ankle affections, the other localizations (knee, spine, hip, shoulder, elbow) did not correlate with higher ferritin levels in the blood (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs10620-012-2243-z/MediaObjects/10620_2012_2243_Fig1_HTML.gif
Fig. 1

Analysis of clinical characteristics at diagnosis and association with ferritin levels. a Frequency of joint pain at diagnosis at the indicated locations in the symptomatic patients. b Correlation of blood ferritin levels with the respective localizations of joint pain at the time of diagnosis in male patients. c Correlation of blood ferritin levels with the respective localizations of joint pain at the time of diagnosis in female patients. d Correlation of the presence of type 2 diabetes with the ferritin levels in blood at diagnosis of symptomatic male and female patients. Ankle = foot and ankle; hand = hand and wrist. − = absent; + = present

Type 2 diabetes was present in 31 of 186 male and 16 of 125 female symptomatic patients at the time of diagnosis. The occurrence of diabetes was associated with significantly higher ferritin levels in both genders (Fig. 1d).

Time to Diagnosis Before and After Introduction of Genetic Testing

To investigate whether the introduction of the genetic HFE test had changed the time needed to diagnose clinically apparent HH after the first onset of clinical features, all patients were evaluated at diagnosis regarding the onset of the following hemochromatosis-associated clinical features: (1) unexplained arthralgia, (2) high liver enzyme levels, and (3) high serum ferritin levels. This data was obtained by a retrospective analysis of patients’ medical records from the general practitioner and/or by thoroughly reviewing the medical history available at the respective medical centers.

To elucidate whether the duration of clinical features was associated with parameters of iron overload, ferritin was plotted against the duration of clinical features (supplementary figure 1). It revealed that the time needed to diagnose the disease is significantly positively correlated with ferritin levels at diagnosis.

The genetic HFE test is a simple, accurate, and non-invasive test. To determine whether it has led to earlier diagnosis, the symptomatic patients were subdivided into three groups based on the year the diagnosis of hemochromatosis was established (<2000; 2000–2005; >2005). The subgroups were evaluated regarding ferritin levels at diagnosis, duration of clinical features, result of liver biopsy, and joint pain at diagnosis (Fig. 2). For ferritin, a decrease in blood levels was noted in patients that were diagnosed between 2006 and 2009 compared to patients diagnosed before 2000 (Fig. 2a). Furthermore, the time needed to diagnose the disease after the onset of clinical features decreased in patients diagnosed 2006–2009 compared to those diagnosed 2000–2005 (Fig. 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs10620-012-2243-z/MediaObjects/10620_2012_2243_Fig2_HTML.gif
Fig. 2

Change of ferritin levels and clinical features dependent on time of diagnosis. a Blood ferritin levels of symptomatic patients at the time of diagnosis with respect to year of diagnosis (<2000; 2000–2005; >2005). b The time needed to diagnose symptomatic patients with regard to year of diagnosis. c Correlation of male (grey dots; n = 206) and female (black dots; n = 134) patients’ age at diagnosis with the year in which the diagnosis of HH was made. Non-parametric Spearman’s rank correlation coefficient (rho) indicates a positive correlation of the age at diagnosis with the year of diagnosis for male patients but not female patients. d Distribution and frequency of joint pain at diagnosis with regard to the year of diagnosis. Depicted is the percentage of patients that had the indicated joint pain at the time of diagnosis. Ankle = foot and ankle; hand = hand and wrist

The correlation of patients’ age at diagnosis with the year of diagnosis showed that males’, but not females’, age at diagnosis significantly increased after introduction of the HFE genetic test (Fig. 2c). In contrast to the shorter disease duration before diagnosis, the clinical spectrum of joint pain did not reveal any striking change in the three subgroups (Fig. 2d). Also, the results of liver biopsy did not appear to change consistently in the three subgroups (supplementary figure 2), but in all subgroups a high proportion of patients already had advanced liver fibrosis at the time of diagnosis.

Discussion

Hereditary hemochromatosis (HH) is a common genetic disorder. Most of the HH cases are caused by the homozygous mutation C282Y in HFE [2]. Since 1996, a genetic test has been available in Germany and Switzerland that detects the common HFE mutations. This genetic test allows the diagnosis of HH in patients with suspected iron overload disease, e.g., elevated ferritin levels and transferrin saturation exceeding 45 % in women and 50 % in men [2]. This test is also recommended for the screening of siblings of HH patients [2]. Before the availability of the genetic test, patients suspected of having HH had to undergo a liver biopsy to verify the iron deposition typical of HH in order to make the diagnosis.

The introduction of the HFE genetic test made it possible to diagnose HH even at early stages of the disease, before a massive iron overload and irreversible organ damage have taken place. If diagnosed early enough, HH can be treated to prevent further iron accumulation, and HH complications can be avoided [5].

In this study we investigated whether the introduction of the genetic test has led to a diagnosis of HH earlier in the course of the disease and whether it has influenced the clinical features of HH observed at the time of diagnosis. The clinical spectrum of hemochromatosis at diagnosis revealed that arthralgia, predominantly of the hand/wrist, is the commonest feature at presentation in the investigated study cohort. This is in line with previous reports that found a similar prevalence of affection of the hand/wrist in hemochromatosis [6, 7].

The significant correlation of hand/wrist and foot/ankle arthralgia with ferritin levels supports its link to hemochromatosis and could suggest that these sites of arthralgia are associated more specifically with hemochromatosis than other locations. Iron burden has been described before to be linked to hand and large joint arthropathy in hemochromatosis [8, 9]. However, it was not differentiated between the different localizations of joint involvement.

The prevalence of diabetes in hemochromatosis patients in this study is comparable to the prevalence in the general population in Germany [10]. We did find a correlation of diabetes occurrence and ferritin levels. However, this correlation might be attributed to diabetes per se, since it is known to be associated with higher ferritin levels [11].

Surprisingly, we could not detect an immediate shift towards earlier diagnosis of HH after introduction of the genetic test when using unexplained arthralgia, elevated enzyme levels, or elevated ferritin as a measure of disease duration before diagnosis. However, the time needed to diagnose the disease after the onset of first clinical features was significantly shorter in patients diagnosed from 2006 onwards than in patients diagnosed between 2000 and 2005. When comparing the diagnosis made after 2005 to the diagnosis before 2000, no significant difference in clinical feature duration was noted. However, this might be biased by the fact that the onset of unexplained arthralgia is not accurately remembered by the patient.

In contrast, the ferritin levels at the time of diagnosis, which should correlate—albeit not very strongly—to total body iron burden [12], were significantly lower in patients diagnosed after 2005 compared to patients diagnosed before 2000; this could potentially reflect a shorter disease duration. Indeed, our data shows that the correlation between ferritin levels and the duration of hemochromatosis-associated clinical features is evident. Besides similar clinical feature durations over the time frame investigated, we also could not observe any differences in the pattern of clinical features at diagnosis that were retrieved in this study or in the proportion of patients with advanced liver damage at the time of diagnosis.

In contrast to the lower iron burden for patients diagnosed more recently, we observed a significant increase in patients’ age at diagnosis for male, but not for female patients. This pattern of decreased iron burden but increased male patients’ age after introduction of genetic testing has been described before [13]. We assume a similar explanation for the increased male patients’ age after the availability of genetic testing: with the genetic test it became possible to diagnose patients with more subtle clinical features that would have not been diagnosed as HH before. In addition, the improved awareness of physicians regarding HH could explain this. A study evaluating early diagnoses of HH in the decades before genetic testing was available showed that over the decades early diagnoses (and lower disease severity at diagnosis) had significantly increased; this was likely due to better education and diagnostic methods for HH [14]. The reason why we could not observe a substantial difference in the clinical pattern over recent years might be explained by a too narrow time frame of the study (most diagnoses lie between 1995 and 2010). We expect that this will change in the future, assuming that physicians’ awareness and genetic testing increases.

Several reasons could contribute to the fact that an early diagnosis of hereditary hemochromatosis is apparently still challenging. One reason could still be the limited awareness among physicians about the availability of a genetic test. Consequently, a trend towards earlier diagnosis in recent years could be attributed to an increasing popularity of this test. Another reason could be that physicians still think too late of this disease, since it is characterized by many non-specific clinical features. Arthralgia especially is a very unspecific symptom that may be caused by several diseases. However, our study shows that the majority of HH patients suffer from this symptom. HH should therefore always be considered in patients with unexplained arthralgia [15, 16].

In patients with suspicious clinical features, ferritin and transferrin saturation should be determined. If transferrin saturation is >50 % in men (>45 % in women) and/or ferritin >300 μg/l for men (>200 μg/l for women), a genetic test should be performed.

Our study underlines the importance of being aware of HH. Now that a genetic test is available, it is paramount to think of this diagnosis in patients with clinical features of HH, since a diagnosis can be made much earlier with the genetic test and irreversible organ damage can be avoided.

Acknowledgments

This work was supported by the Emmy Noether Program of the German Research Foundation [grant number STR 1095/2-1 to P.S.].

Conflict of interest

None.

Supplementary material

10620_2012_2243_MOESM1_ESM.pdf (51 kb)
Correlation of ferritin levels with disease duration in symptomatic patients. Ferritin levels in blood at the time of diagnosis are plotted against the disease duration (n = 282). The disease duration was defined as the time between the diagnosis and the onset of clinical features such as elevated transaminase levels, elevated ferritin or unexplained arthralgia. Non-parametric Spearman’s rank correlation coefficient (rho) indicates a positive correlation between disease duration and patients’ ferritin levels (PDF 52 kb)
10620_2012_2243_MOESM2_ESM.pdf (41 kb)
Results of liver biopsy at the time of diagnosis. Depicted is the percentage of liver biopsy results that showed a mild fibrosis, severe fibrosis or cirrhosis with regard to the year of diagnosis (<2000; 2000–2005; >2005) (PDF 42 kb)

Copyright information

© Springer Science+Business Media, LLC 2012