Olivier Dereure1, Nicolas Jumez1, Didier Bessis1, Benoit Gallix2 and Bernard Guillot1

Departments of 1Dermatology and 2Medical Imaging, University of Montpellier I, Hôpital Saint Eloi, Montpellier, France

Liver iron content was evaluated by a magnetic resonance imaging-based method in 20 consecutive patients with either sporadic or familial porphyria cutanea tarda. Serum ferritin, hepatitis C infection and the presence of the 2 main mutations of the hemochromatosis gene were also investigated. All patients showed good clinical response to phlebotomy. Initial liver iron content was normal (< 40 μmol/g) in 9 cases, slightly increased (40–59 μmol/g) in 3 cases, moderately increased (60–99 μmol/g) in 6 cases or markedly increased (100–199 μmol/g) in 2 cases). The ferritin level was raised (> 400 ng/ml) in 14/20 patients and there was no obvious relationship with liver iron. Increased liver iron content was observed more frequently in patients with hemochromatosis mutation and less frequent in those with hepatitis C infection. Clinical response to phlebotomies was slightly better in patients with increased liver iron content even slightly, but patients with normal liver iron content also responded well, which suggests that iron depletion is an outstanding treatment independent of liver iron content. This study shows that increased liver iron content is not a constant finding in patients with porphyria cutanea tarda, especially in women, and that it is not a prerequisite for the efficiency of phlebotomy. Key words: porphyria cutanea tarda; liver iron content; magnetic resonance imaging; phlebotomies.

(Accepted February, 14, 2008.)

Acta Derm Venereol 2008; 88: 341–345.

Olivier Dereure, Department of Dermatology, University Hospital of Montpellier, Hôpital Saint-Eloi, 80 avenue A. Fliche, FR-34295 Montpellier Cedex 5, France. E-mail: o-dereure@chu-montpellier.fr

Porphyria cutanea tarda (PCT), the most frequent subset of porphyria in Western countries, is linked to decreased activity of uroporphyrinogen decarboxylase (URO-D). The complex pathological factors leading to overt PCT are well known and have been described elsewhere (1–6). Among these factors, the peculiar relevancy of iron overload is highly suggested by the high efficiency of iron depletion by repeated phlebotomies and by independent report of an association between genetic haemochromatosis and PCT in the same patients (7, 8). Iron overload may inhibit URO-D activity in different ways, including the generation of reactive oxygen species, which can interact directly with the enzymatic activity and/or lead to the direct oxidation of uroporphyrinogen to uroporphyrin, which is not a substrate for URO-D (9); moreover, iron may play a role in photosensitization by uroporphyrin (10). Although iron overload is usually considered to be a hallmark of the disease, present in most, if not all, patients with overt PCT (4–6, 11–18), its pathomechanisms are not fully established to date; however, abnormalities of iron metabolism-regulating genes are likely to be significantly involved, for example HFE gene mutations, as demonstrated by a number of previous studies (19, 20). Despite these widely accepted data, only a few investigations assessing iron content in the liver have been conducted in series of patients with PCT over the last 20 years, probably because its accurate measurement requires complex methods such as mass spectrometry. During the last 20 years non-invasive methods have been developed for accurate evaluation of the iron content of tissues such as the liver, methods based mainly on magnetic resonance imaging (MRI) using either transverse relaxation time or liver-to-muscle gradient-echo T2*-weighted ratio, which have both proved reliable and accurate methods that can be used for routine clinical purposes (21–24). Accordingly, a study was designed to investigate liver iron content (LIC) in 20 PCT patients using an MRI-based method. The data obtained were compared with serum ferritin concentration, hepatitis C infection, the presence of 2 main HFE gene mutations, and with the clinical effect of iron depletion by a series of phlebotomies.

PATIENTS AND METHODS

Study design, selection and description of patients

A total of 20 patients diagnosed with either familial or sporadic PCT between April 2004 and April 2006 were included in this study; 12 men and 8 women aged 37–63 years (mean age 47 years) (Table I). The diagnosis of PCT was established on the basis of the usual clinical and biochemical data, including photosensitivity, facial “metallic” hyperpigmentation, blisters and milia on the dorsum of the hands, skin fragility on sun-exposed areas and high excretion rates of urinary porphyrins with dominance of uro- and heptacarboxy-porphyrins (at least 85% of total urinary porphyrins). Triggering factors usually associated with overt PCT were systematically analysed in all patients. Erythrocyte URO-D was measured in 15 patients (Laboratoire Mérieux, Lyon, France). Decreased URO-D activity, about 50% of normal, compatible with the diagnosis of familial PCT was found in 3 patients. Those showing normal activity were classified as a sporadic form of PCT (Table I). Liver function tests were abnormal in all patients (Table I) and kidney function was normal in all cases (data not shown).

After the diagnosis was established, all patients were treated with phlebotomies (usually 400 ml per course for 5–8 courses) along with the correction of any relevant triggering factor, such as excessive alcoholic consumption, active hepatitis C virus (HCV) infection (defined by positive serological tests and a viral load above 14 IU/ml according to the polymerase chain reaction (PCR)-based test) or oestrogen therapy (Table I). The clinical results were mainly evaluated on the resolution of blistering and the disappearance of skin fragility.

Biochemical blood tests

Serum ferritin concentration was determined in all patients before the start of phlebotomy therapy.

HFE gene analysis

HFE gene analysis was performed after obtaining informed consent from all the patients. Genomic DNA was extracted from peripheral blood leukocytes and analysed for the 2 main mutations of the HFE gene (C282Y and H63D) using PCR and restriction analysis of the amplified fragments, as described previously (20).

HCV status analysis

A first-step serological test was performed using an enzyme-linked immunoassay (ELISA) (HCV 3.0 ELISA*-test, Ortho-Clinical Diagnosis, Johnson & Johnson, France) for all patients. In patients with a positive result, confirmation was obtained by a chemoluminescence-derived method (Abbot-IMX HCV 3.0* test, Abbot France) followed by an evaluation of viral load in peripheral blood (Cobas Tacqman* test, Roche Diagnostics, France).

MRI-based assessment of liver iron content

Determination of iron content in the tissue was performed with a gradient echo sequences-based method with long Echo Time (TE) sequences (T2* ponderation), allowing the detection of minor iron overload compared with spin echo-based analysis. Using these sequences, a liver with normal iron content provides a hypersignal compared with muscle and LIC is calculated in patients from the signal ratio between liver and muscle, with the hypothesis that no iron overload is present in muscle independently of the LIC. This T2*-weighted gradient echo sequence is considered as highly specific, with 89% sensitivity and 80% specificity in the validation group allowing the detection of all clinically relevant liver iron overload greater than 60 μmol/g (21). The final result was expressed in μmol of iron/g of dry liver tissue and ranked as normal LIC (< 40 μmol/g), slight increase (between 40 and 59 μmol/g), moderate increase (between 60 and 99 μmol/g), marked increase (between 100 and 199 μmol/g) or important increase in LIC (200 μmol/g or more). Measurement of LIC by MRI was performed before any treatment was implemented in all included patients and after completion of phlebotomies and disappearance of relevant clinical lesions in only 3 cases.

RESULTS

According to MRI-calculated data, the initial, pre-therapeutic LIC was normal (9 cases), slightly increased (3 cases), moderately increased (6 cases) or markedly increased (2 cases). Initial ferritin level was high (above 400 ng/ml) in 14/20 patients (ranging from 411 to 2965 ng/ml in this subset). There was no clear relationship between initial high ferritin levels and and increased LIC, since only 8/14 (57%) patients with elevated ferritin displayed elevated LIC vs. 3/6 (50%) in patients with ferritin below 400 ng/ml; conversely, 8/11 (73%) patients with increased LIC showed elevated ferritin levels vs. 6/9 patients (66%) without significant overload. Overall, a discrepancy between these 2 parameters was present in 9/20 patients. Gender appeared to be a significant factor, since the percentage of PCT patients with increased LIC was 66% in men (8/12) vs. only 37.5% (3/8) in women.

HFE gene analysis was conducted in all patients and relevant mutations were identified in 5/20 patients: 1 homozygous C282Y mutation, 2 heterozygous C282Y mutations, 1 heterozygous compound C828Y/H63D mutation and 1 homozygous H63D mutation. Overall, an increased LIC, even slight or moderate, was more frequent in patients with significant mutation of the HFE gene (4/5 including the patient with an important overload who displayed an heterozygous composite mutation: 80%) than in patients with no mutation or heterozygous H63D mutation (7/15: 47%).

The presence of HCV infection was investigated in all patients and provided a positive result that was confirmed by second-line methods in 11/20 patients (Table I). Two patients had active infection with significant viral load and were receiving interferon alpha treatment at diagnosis. Interestingly, no patient was infected by hepatitis B, whereas only one patient was HIV positive. Increased LIC was more frequent when HCV infection was absent (7/9, 78%) than when it was present (4/11, 36%) and this difference was statistically significant using a modified χ2 test (p = 0.042).

Clinical remission after phlebotomies was obtained in all the patients with increased LIC, whereas a disappearance of active clinical lesion was observed in 7 of the 9 patients (78%) with normal LIC. LIC was re-evaluated in 3 patients in clinical remission after phlebotomies, one of them with a marked iron overload (180 μmol/g) and 2 others without significant overload; in the former case, a significant decrease in liver iron load was obtained (a reduction from 180 to 75 μmol/g), while no significant change was noted in the other cases.

MRI-related data, HCV and HFE status are summarized in Table I.

DISCUSSION

This study, using a non-invasive method based on liver MRI to investigate LIC, provided new and quite unexpected results. Indeed, according to MRI-based calculations, 45% of patients displayed normal LIC, whereas most other cases displayed only slightly or moderately increased LIC, with the exception of a single patient with a markedly increased content. This trend was even more obvious in women, of whom 62.5% displayed normal LIC. Owing to the fact that these calculations used a muscle/liver ratio of proton signals, it might be argued that these surprisingly negative results were related to an increase in muscle iron content, resulting in an artificially modified ratio. However, an elevated muscle iron content has not been reported in overt PCT to date, and is unlikely to be significant due to the rarity of iron-storing cells, such as macrophages, in muscle. Accordingly, our study is not consistent with the classically admitted theory that an increased LIC is a prominent data in PCT present in at least 80% of patients, a figure far above our results even if only men are considered (1, 3, 6, 16, 18), since women usually have lower iron levels, at least before menopause. Possible reasons for this discrepancy may be: the relatively high percentage of women in our series, since iron overload in women was clearly less frequent, which might be related partly to the lower level of alcohol consumption in women; changes in patients’ profile with time, with respect to aetiological factors; onset of new triggering factors such as HCV infection; lack of accuracy of previous studies usually relying on semi-quantitative methods, less precise than mass spectrometry. Additionally, evaluation of LIC by an MRI-based method in PCT patients has been reported only in 2 previous studies investigating, respectively, 8 and 20 patients (25, 26); in both studies, MRI-estimated LIC was high in all patients compared with controls. These results appear to be different from our data, but the methods were slightly different (echospin T2* and transverse relaxation time) and these 2 studies have been conducted more than 10 years ago, which is consistent with a shift in aetiological factors since then. Moreover, the increase in LIC was only twice the baseline level in one study, a result actually close to our data for patients with elevated LIC (26).

Regarding the relationship between MRI-based results and the iron metabolism-related parameters, there was a trend toward a correlation between an increased LIC and the presence of significant mutations of the HFE gene. Conversely, in our study, serum ferritin concentration cannot be considered as a reliable predictor of LIC, since a discrepancy between these 2 parameters was present in approximately half of the patients, a figure that contradicts previous reports (24, 27, 28). This lack of reliability of ferritin level as a predictor of LIC is perhaps related to the fact that ferritin rate is influenced by many conditions independent of iron metabolism, including inflammation and liver dysfunctioning, regardless of its origin.

The relatively lower percentage of patients with HCV infection in the group with increased LIC is consistent with previous reports in which a lower prevalence of this infection was observed in patients with significant HFE mutations (20). To date, there is no convincing explanation for this loose tendency, but it can be hypothesized that HCV infection and iron overload are independent triggering factors of PCT that could be relatively exclusive of each other.

As to the clinical implications of the presence/absence of an increased LIC regardless of its level, our results clearly show that phlebotomies can be efficient even when no significant overload is present, although therapeutic results seem slightly better in patients with a higher content. Thus, removal of iron remains an outstanding treatment of PCT, independent of the initial LIC, an unexpected finding reported previously by Lundvall in 1971 (29), which suggests that the correction of increased LIC is probably not its only mechanism of action.

The unexpected results of this study must be confirmed by more extensive reports and emphasize the growing need and use of non-invasive methods for metabolic assessment. In addition to MRI-based studies, other methods may change our current understanding of the actual incidence of iron overload in PCT, such as urinary dosage of hepcidin, which seems to be closely related to the level of iron stores in the liver (30).

REFERENCES

1. Siersema PD, Rademakers LH, Cleton MI, ten Kate FJ, de Bruijn WC, Marx JJ, et al. The difference in liver pathology between sporadic and familial forms of porphyria cutanea tarda: the role of iron. J Hepatol 1995; 23: 259–267.

2. D’Alessandro Gandolfo L, Griso D, Macri A, Biolcati G, Barlattani A, Topi GC. Iron and porphyria cutanea tarda. Cell Mol Biol (Noisy-le-grand) 1997; 43: 75–79.

3. O’Reilly FM, Darby C, Fogarty J, Tormey W, Kay EW, Leader M, Murphy GM. Screening of patients with iron overload to identify hemochromatosis and porphyria cutanea tarda. Arch Dermatol 1997; 133: 1098–1101.

4. Bygum A, Brandrup F. Iron overload in porphyria cutanea tarda. Br J Dermatol 2000; 143: 1116.

5. Bonkovsky HL, Lambrecht RW, Shan Y. Iron as a co-morbid factor in nonhemochromatotic liver disease. Alcohol 2003; 30: 137–144.

6. Bygum A, Christiansen L, Petersen NE, Horder M, Thomsen K, Brandrup F. Familial and sporadic porphyria cutanea tarda: clinical, biochemical and genetic features with emphasis on iron status. Acta Derm Venereol 2003; 83: 115–120.

7. Fargion S, Fracanzani AL, Romano R, Cappellini MD, Fare M, Mattioli M, et al. Genetic hemochromatosis in Italian patients with porphyria cutanea tarda: possible explanation for iron overload. J Hepatol 1996; 24: 564–569.

8. Mehrany K, Drage LA, Brandhagen DJ, Pittelkow MR. Association of porphyria cutanea tarda with hereditary hemochromatosis. J Am Acad Dermatol 2004; 51: 205–211.

9. Mukerji SK, Pimstone NR, Burns M. Dual mechanism of inhibition of rat liver uroporphyrinogen decarboxylase activity by ferrous iron: its potential role in the genesis of porphyria cutanea tarda. Gastroenterology 1984; 87: 1248–1254.

10. Menon IA, Becker MA, Haberman HF. Role of iron in the photosensitization by uroporphyrin. Clin Chim Acta 1991; 202: 237–242.

11. Berlin SO, Brante GO. Iron metabolism in porphyria and haemochromatosis. Lancet 1962; 2 :729.

12. Lundvall O, Weinfeld A, Lundin P. Iron storage in porphyria cutanea tarda. Acta Med Scand 1970; 188: 37–53.

13. Lefkowitch JH, Grossman ME. Hepatic pathology in porphyria cutanea tarda. Liver 1983; 3: 19–29.

14. Kushner JP, Edwards CQ, Dadone MM, Skolnick MH. Heterozygosity for HLA-linked hemochromatosis as a likely cause of the hepatic siderosis associated with sporadic porphyria cutanea tarda. Gastroenterology 1985; 88: 1232–1238.

15. Siersema PD, Rademakers LH, Cleton MI, ten Kate FJ, de Bruijn WC, Marx JJ, et al. The difference in liver pathology between sporadic and familial forms of porphyria cutanea tarda: the role of iron. J Hepatol 1995; 23: 259–267.

16. Elder GH. Porphyria cutanea tarda. Semin Liver Dis 1998; 18: 67–75.

17. Sampietro M, Fiorelli G, Fargion S. Iron overload in porphyria cutanea tarda. Haematologica 1999; 84: 248–253.

18. Alla V, Bonkovsky HL. Iron in nonhemochromatotic liver disorders. Semin Liver Dis 2005; 25: 461–472.

19. Bulaj ZJ, Phillips JD, Ajioka RS, Franklin MR, Griffen LM, Guinee DJ, et al. Hemochromatosis genes and other factors contributing to the pathogenesis of porphyria cutanea tarda. Blood 2000; 95: 1565–1571.

20. Dereure O, Aguilar-Martinez P, Bessis D, Perney P, Vallat C, Guillot B, et al. HFE mutations and transferrin receptor polymorphism analysis in porphyria cutanea tarda: a prospective study of 36 cases from southern France. Br J Dermatol 2001; 144: 533–539.

21. Gandon Y, Olivie D, Guyader D, Aube C, Oberti F, Sebille V, Deugnier Y. Non-invasive assessment of hepatic iron stores by MRI. Lancet 2004; 363: 357–362.

22. Carneiro AA, Fernandes JP, de Araujo DB, Elias J Jr, Martinelli AL, Covas DT, et al. Liver iron concentration evaluated by two magnetic methods: magnetic resonance imaging and magnetic susceptometry. Magn Reson Med 2005; 54: 122–128.

23. St Pierre TG, Clark PR, Chua-Anusorn W. Measurement and mapping of liver iron concentrations using magnetic resonance imaging. Ann N Y Acad Sci 2005; 1054: 379–385.

24. Papakonstantinou OG, Maris TG, Kostaridou V, Gouliamos AD, Koutoulas GK, Kalovidouris AE, et al. Assessment of liver iron overload by T2-quantitative magnetic resonance imaging: correlation of T2-QMRI measurements with serum ferritin concentration and histologic grading of siderosis. Magn Reson Imaging 1995; 13: 967–977.

25. Rocchi E, Cassanelli M, Borghi A, Paolillo F, Pradelli M, Pellizzardi S, et al. Liver iron overload and desferrioxamine treatment of porphyria cutanea tarda. Dermatologica 1991; 182: 27–31.

26. Rocchi E, Cassanelli M, Borghi A, Paolillo F, Pradelli M, Casalgrandi G, et al. Magnetic resonance imaging and different levels of iron overload in chronic liver disease. Hepatology 1993; 17: 997–1002.

27. Rocchi E, Gibertini P, Cassanelli M, Pietrangelo A, Borghi A, Ventura E. Serum ferritin in the assessment of liver iron overload and iron removal therapy in porphyria cutanea tarda. J Lab Clin Med 1986; 107: 36–42.

28. Gibertini P, Rocchi E, Cassanelli M, Pietrangelo A, Coppini M, Ventura E. Determination of serum ferritin in porphyria cutanea tarda. A reliable sign of hepatic siderosis. Minerva Med 1984; 75: 469–474.

29. Lundvall O. The effect of phlebotomy therapy in porphyria cutanea tarda. Acta Med Scand 1971; 189: 33–49.

30. Détivaud L, Nemeth E, Boudjema K, Turlin B, Troadec MG, Leroyer, et al. Hepcidin levels in humans are correlated with hepatic iron stores, hemoglobin levels and hepatic function. Blood 2005; 106: 746–748.