Haematologica Journal of Hematology volume 84 – number 2 – February 1999

Haematologica 1999; 84:97-98 editorial, comments and views

of iron overload in patients carrying either one or no Gilbert’s syndrome mutation. Similarly, they did not find any relation between the degree of iron overload, liver histology Over the last few years the genetic basis of Gilbert’s and HFE mutations. The authors, therefore, con- syndrome has been clarified. This issue of the jour- cluded that HFE mutations may lead to overexpres- nal contains four papers on this condition. One sion of the dysmetabolic iron overload syndrome paper describes a novel molecular mechanism together with yet unknown factors.5 responsible for decreased bilirubin-UDP-glucurono- In the last few years another disease in which iron syltransferase activity, while two articles evaluate the appears to have a role has become more important: interaction between Gilbert’s syndrome and hema- this disease, called non-alcoholic steatohepatitis tologic disorders associated with increase bilirubin (NASH), was originally described in middle aged production. Finally, a review article analyzes our pre- overweight and often diabetic women, but more sent knowledge of the molecular pathogenesis of this recently has been identified in larger series of condition. patients, often of male sex. The pathognomonic fea- Why publish papers on Gilbert’s syndrome in a ture of NASH is the finding in the liver biopsy of hematology journal? There are several reasons. First, steatosis associated with cellular inflammation in the Haematologica aims to be a journal of hematolog- presence or absence of fibrosis.6,7 ic medicine in a broad sense: borderline subjects that Recently a large spectrum of pathologies including may be of interest to our readers are therefore wel- fat alone, steatohepatitis and fibrosis have been con- come. Second, Gilbert’s syndrome is borderline sub- sidered to be different stages of NASH. Unexpected- ject. This condition is characterized by mild uncon- ly, more than 50% of patients in the different series jugated hyperbilirubinemia in the absence of overt studied had hyperferritinemia whereas increased hemolysis or evidence of liver disease. However, sev- transferrin saturation was found in a much lower pro- eral studies have shown that red cell lifespan is short- portion of patients.8,9 er than normal in half the cases of Gilbert’s syn- In the present issue of Hematologica two scientif- drome, suggesting a mild, compensated hemolytic ic letters on the relation between increased ferritin state. Third, being a common condition, Gilbert’s and dysmetabolism draw attention to this subject. syndrome frequently coexists and may interact with The first, from Orlandi et al.,10 reports the case of a hematologic disorders that involve unconjugated patient with increased ferritin, normal transferrin sat- hyperbilirubinemia, per se. uration and a very slight increase in liver iron, nega- tive for the mutations of the HFE gene as well as for point mutations in the iron-regulatory element, Dysmetabolic iron overload syndrome recently described in the hyperferritinemia-cataract syndrome. The patient, who had denied alcohol A new iron overload syndrome, characterized by intake, was overweight and had non-insulin-depen- hyperferritinemia and increased liver iron concen- dent diabetes mellitus. He had only a monoclonal tration in the presence of a normal transferrin satu- gammopathy and a bone marrow biopsy showed 8% ration, was recently described by Deugnier et al.1 of plasma cells. He also had hepatomegaly and a liv- Patients with similar characteristics have also been er biopsy showed mild focal iron deposition in hepa- described in Italy but it was hypothesized that they tocytes with steatosis and mild portal fibrosis. represented a subgroup of subjects with genetic In the second letter, from Rovati et al.,11 two mono- hemochromatosis (GH).2 zygotic twins who came to medical attention be- From the start it was observed that hyperlipidemia, cause of hyperferritinemia with non-insulin-depen- glucose intolerance, increased body mass index and dent diabetes mellitus are described. Both twins con- hypertension were frequently present in subjects with sumed less than 50 g alcohol/day, both had a trans- this unusual presentation of iron overload. More ferrin saturation within the normal range and recently analysis of the mutations of the HFE gene, increased ferritin as an acute phase protein was ruled the gene associated with GH,3 showed that in this out by appropriate investigations. No association population the frequency of mutations was signifi- with HFE mutations and no point mutation of IRE cantly higher than in normal controls.4 The French was found. Both patients refused liver biopsy. group, however, observed no difference in the degree Both letters conclude that there is a link between 98 Editorial, comments and views altered metabolism and iron overload but that the far from being understood and it is to be hoped that dysmetabolic iron overload syndrome is genetically in the near feature that the relation between meta- distinct from GH.12-14 bolic abnormalities and liver iron overload will be These two interesting letters on such a poorly elucidated. defined topic stimulate several questions. Was the dia- Silvia Fargion, M.D. betes responsible for the hyperferritinemia? Were these Department of Internal Medicine, University of Milan, patients hypersensitive to modest alcohol consump- IRCCS Ospedale Maggiore tion? Did the patients have NASH? Would the ferritin via F. Sforza 35, 20122 Milan, Italy values also have been stable after a correct diet? Phone: international +39-02-55033757 The patients described in the two letters had dia- Fax: international +39-02-55180241 E-mail: [email protected] betes. In a recent paper by Turnbull et al.,15 increased values of ferritin were found in several patients with diabetes. When liver biopsy was performed, in the References hypothesis that they had GH, only one patient was 1. Moirand R, Mortaji AM, Loréal O, Paillard F, Brissot found to have liver iron overload whereas all the oth- P, Deugnier Y. A new syndrome of liver iron overload ers had steatosis. Thus, a major subject of discussion with normal transferrin saturation. Lancet 1997; 349: is whether steatosis, by itself, can cause hyperfer- 95-7. ritinemia. But if it does, why don’t all patients with 2. Fargion S, Sampietro M, Fracanzani AL, et al. Iron overload with normal transferrin saturation: a subset steatosis have increased values of ferritin? of GH? Hepatology 1997; 26:200A. It is well known that induction of ferritin synthesis 3. Feder JN, Gnirke A, Thomas W, et al. A novel MGH occurs during alcohol abuse, which in susceptible class I-like gene is mutated in patients with hereditary patients causes at least liver steatosis. Generally, com- haemochromatosis. Nat Genet 1996; 13:399-408. plete alcohol abstinence from alcohol as well as grad- 4. Mendler MH, Moirand R, Turlin B, et al. Study of HFE mutations in the “dysmetabolic iron overload syn- ual weight loss, is followed by normalization of ferritin drome”. Hepatology 1998; 28:419A. levels. One of the patients described in this issue of 5. Mendler MH, Turlin B, Moirand R, et al. The “dys- Haematologica denied alcohol abuse, while the two metabolic iron overload syndrome”: a concept which brothers consumed less than 50 g alcohol/day. Did unites hepatic iron overload associated with poly- hyperferritinemia reflect their particular susceptibilty metabolic disorders, steatosis, and NASH. Hepatol- ogy 1998; 28:501A. to small amounts of alcohol? However, the patient for 6. Sheth SG, Gordon FD, Chopra S. Non-alcoholic steato- whom a liver biopsy was available had a mildly hepatitis. Ann Intern Med 1997; 126:137-45. increased liver iron concentration, indicating that the 7. James OFW, Day CP. Non-alcoholic steatohepatitis increased ferritin reflected not only an induction by (NASH): a disease of emerging identity and impor- alcohol but, at least in part, increased iron stores. tance. J Hepatol 1998; 29:495-501. Could hepatic triglyceride accumulation interfere 8. Bacon BR, MJ Farahvash, Janney CG, Neuschwander- Tetri BA. Non-alcoholic steatohepatitis: an expanded with intrahepatocyte iron movement with reduced clinical entity. Gastroenterology 1994; 107:1103-9. iron mobilization from ferritin, causing on the one 9. George DK, Goldwurm S, Macdonald GA, et al. hand increased ferritin synthesis and on the other Increased hepatic iron concentration in non-alcoholic hand a low transferrin saturation? Increased trygli- steatohepatitis is associated with increased fibrosis. cerides in the liver provide increased substrate for lipid Gastroenterology 1998; 114:311-8. 10. Orlandi E, Arbustini E, Lazzarino M. Isolated hyper- peroxidation with consequent generation of poten- ferritinemia with normal transferrin saturation and tially reactive and cytotoxic intermediates. An imbal- dysmetabolism, in the absence of the two known ance of hepatic oxidant and antioxidant mechanisms mutations in the HFE gene of hereditary hemochro- could occur and interfere with iron mobilization from matosis. Haematologica 1999; 84:181-2. ferritin. This imbalance could become even more evi- 11. Rovati A, Bergamaschi G, Casula S, Cerani P, Grasso M, Cazzola M. The dysmetabolic iron overload syn- dent if subjects with fatty liver have mildly increased drome is clinically and genetically distinct from HFE- liver iron stores as may occur in heterozygotes for GH. related GH. Haematologica 1999; 84:182-3. Thus, the coexistence of fatty liver and heterozygosis 12. Camaschella C, Piperno A. Hereditary hemochroma- for GH could exert a synergistic effect on ferritin syn- tosis: recent advances in molecular genetics and clin- thesis. Recently, in iron overloaded mice, induction of ical management. Haematologica 1997; 82:77-84. hepatic genes involved in lipid metabolism was 13. Piperno A. Classification and diagnosis of iron over- 16 load. Haematologica 1998; 83:447-55. shown. Is this the link between iron overload and the 14. Pietrangelo A, Camaschella C. Molecular genetics and metabolic abnormalities found in the dysmetabolic control of iron metabolism in hemochromatosis. iron overload syndrome? The dysmetabolic iron over- Haematologica 1998; 83:456-61. load syndrome has, however, also been described in 15. Turnbull AJ, Mitchison HC, Peaston RTP, et al. The the absence of HFE mutations.10,11 Thus, with no evi- prevalence of hereditary haemochromatosis in a dia- betic population. Q J Med 1997; 90:271-5. dence of heterozygosis for GH, what is the trigger 16. Pigeon C, Leroyer P, Turlin B, Brissot P, Loreal O. Iden- mechanism for iron overload? tification of hepatic genes induced during chronic oral It is clear that this extremely interesting subject is iron overload in mice. Hepatology 1998; 28:716A.