Journal of Pediatric Hematology/Oncology 22(6): 593597, November/December 2000

2000 Lippincott Williams & Wilkins, Inc., Philadelphia

Iron Overload and Iron-Chelating Therapy in Hemoglobin E- Thalassemia

Nancy F. Olivieri, M.D., Shanthimala De Silva, M.D., Anuja Premawardena, M.D., Supriya Sharma, M.D., Adrian M. Viens, Chelsea M. Taylor, B.Sc., Gary M. Brittenham, M.D., and David J. Weatherall, F.R.S.

Abstract: Whereas hemoglobin (Hb) E- thalassemia is recognized as probably the most common serious hemoglobinopathy worldwide, its natural history remains poorly defined. The interaction of hemoglobin E and -thalassemia result in a wide spectrum of clinical disorders, some indistinguishable from thalassemia major and some milder and not transfusiondependent. Partially as a result of this wide range of phenotypes, clear guidelines for approaches to transfusion and to iron-chelating therapy for patients with Hb E- thalassemia have not been developed. By contrast, data that have accumulated during the past 10 years in patients with -thalassemia permit a quantitative approach to the management of iron overload and provide guidelines for the control of body iron burden in individual patients treated with iron-chelating therapy. These guidelines may be applicable to patients with Hb E- thalassemia. Preliminary evidence from our studies of iron loading in affected patients with Hb E- thalassemia in Sri Lanka suggest that this disorder may be associated with variable, but accelerated, gastrointestinal iron absorption, and that the iron loading associated with chronic transfusions in patients with Hb Ethalassemia is similar to that observed in patients with -thalassemia. These data, in the only cohort of patients with Hb Ethalassemia to have undergone quantitative assessment of body iron burden, suggest that the principles that guide assessment of iron loading and initiation of chelating therapy in patients with -thalassemia may be generally applicable to those with Hb E- thalassemia. Further quantitative studies in both nontransfused and transfused patients will be necessary to permit firm conclusions. Key Words: Hemoglobin E -thalassemiaIron loading Iron-chelating therapyIron quantitationLiver biopsy.

mia remains poorly defined with respect to both its natural history and optimal management (1). In particular, clear guidelines for approaches to transfusion and to iron-chelating therapy for patients with Hb E- thalassemia are lacking. Data that have accumulated during the past 10 years in patients with homozygous -thalassemia provide guidelines for the control of body iron burden in individual patients treated with chelating therapy (2) and may be applicable to patients with Hb E- thalassemia. Early findings from studies of iron loading in nontransfused patients in Sri Lanka, where high heterozygote frequencies are observed for both -thalassemia (1% 4%) and Hb E (2%) (3) suggest that this disorder may be associated with variable, and sometimes accelerated, gastrointestinal iron absorption similar to that identical to that observed in patients with -thalassemia intermedia. MANAGEMENT OF IRON OVERLOAD IN PATIENTS WITH -THALASSEMIA Tissue iron overload, which is fatal in both transfused and nontransfused patients if not prevented or adequately treated, is a primary focus of -thalassemia management (2). In nontransfused patients with severe thalassemia not associated with Hb E (those termed, imprecisely, as thalassemia intermedia), abnormally regulated iron absorption results in increases in body iron burden which may, depending on the severity of erythroid expansion, vary between 2 and 5 g per year (4,5). Programs of regular (usually monthly) transfusions, as are currently administered in patients with thalassemia, likely double this rate of iron accumulation. In general, transfused patients require iron-chelating therapy with deferoxamine, the only agent approved for widespread clinical use, when body storage iron reaches concentrations demonstrated to be associated with risks for complications of iron overload. Hence, an important strategy in the management of all transfused patients is the precise assessment of body iron burden. First, this permits some estimation 593

Although one of the most prevalent serious hemoglobinopathies worldwide, hemoglobin (Hb) E- thalasseSubmitted for publication November 3, 1999; accepted July 12, 2000. From the Departments of Medicine and Pediatrics, University of Toronto (N.F.O., S.S., A.V., C.M.T.), Toronto, Canada; Department of Pediatrics, the Kurunegala Teaching Hospital (S.D.S.), Kurunegala, Sri Lanka; Department of Pediatrics and Medicine, Columbia University (G.M.B.), New York, New York, U.S.A.; and the Institute of Molecular Medicine (A.P., D.J.W.), Oxford, England, U.K. Address correspondence and reprint requests to Dr. Nancy F. Olivieri, MD, F.R.C.P.(C), The Hospital for Sick Children, Division of Hematology/Oncology, 555 University Avenue, M5G 1X8 Toronto, Ontario, Canada. E-mail: noliv@sickkids.on.ca

594

N.F. OLIVIERI ET AL. concentrations of iron, the partition of iron between reticuloendothelial and parenchymal cells, ascorbate status, and noniron-related factors including alcohol and viral hepatitis (2). Iron-chelating therapy that aims to maintain a normal body iron, corresponding to a hepatic iron of approximately 0.2 to 1.6 mg iron per gram of liver, dry weight, reduces the likelihood of complications of iron overload, but greatly increases the probability of dose-related drug toxicity (2). Appropriately, a perhaps conservative goal for iron-chelating therapy in patients with thalassemia is to maintain hepatic storage iron concentrations of approximately 3.2 to 7 mg iron per gram of liver, dry weight, in the range found in heterozygotes for hereditary hemochromatosis (10). The risks of deferoxamine toxicity associated with regimens to maintain body iron within this range are minor. Patients with higher body iron burdens, up to approximately 15 mg iron per gram of liver, dry weight, are considered to be at an increased risk for hepatic fibrosis, diabetes mellitus, and other complications of iron overload and need more intensive iron-chelating therapy (2). Patients with still higher body iron burdens have a greatly increased risk for cardiac disease and early death (11), and may be candidates for continuous intravenous ambulatory deferoxamine or other special programs of management (12). Deferoxamine-induced toxicity can be avoided by regular assessment of body iron burden using measurements of hepatic iron concentration. In the absence of assessments of hepatic iron, a toxicity index, defined as the mean daily dose of deferoxamine (mg/kg) divided by the serum ferritin concentration ( g/L), may be calculated regularly for individual patients and should not exceed 0.025 (13). We have recommended that doses of deferoxamine not exceed 50 mg/kg per day (2). Higher doses that have been used in an attempt to rescue patients with severe complications of iron loading have been associated with deferoxamine toxicity (13). It is difficult to justify the use of these high doses, especially in patients in whom the body iron burden has not been quantitated using the hepatic iron concentration, rather than estimated using the serum ferritin concentration. Regular evaluation of deferoxamine toxicity has been strongly recommended in all patients maintained on any dose of deferoxamine (2). BENEFITS OF IRON-CHELATING THERAPY The beneficial effects of iron-chelating therapy with parenteral deferoxamine, the only chelating agent widely available for clinical use, on the complications of iron loading have been recently reviewed (16). As a result of programs of deferoxamine therapy, the prognosis for patients in countries able to afford this therapy has greatly improved, in contrast to that for individuals (including

of the risk for complications of iron overload in individual patients. In addition, issues regarding the appropriate age for the initiation of deferoxamine and the maintenance of balance between the effectiveness and toxicity of this agent arise frequently in the management of patients with thalassemia. These issues have become less problematic because the regular quantitative assessment of body iron has increased in large treatment centers, but assessment of the iron-loaded patients using serial liver biopsies in emerging countries, where Hb Ethalassemia is prevalent, has only recently been shown to be feasible (6). ASSESSMENT OF BODY IRON Whereas both direct and indirect means for the assessment of body iron are available, no single indicator or combination of indicators is ideal for the evaluation of iron status in all clinical circumstances. In clinical practice, the serum ferritin concentration is commonly used to assess the effectiveness of chelating therapy. It is increasingly recognized that reliance on values of serum ferritin may lead to errors in management; changes in body iron account for little more than half of the variation in serum ferritin (7). Chelator-induced urinary iron excretion, too, is vulnerable to the extraneous influences of infection, inflammation, the activity and effectiveness of erythropoiesis, extramedullary hematopoiesis, liver disease, and ascorbic acid deficiency (2). By contrast, measurement of hepatic storage iron, highly correlated with total body storage iron, provides the most quantitative, specific, and sensitive method for evaluating iron burden in thalassemia. Determination of hepatic iron through liver biopsy performed with ultrasound guidance is safe, permits rational adjustment of iron-chelating therapy, and may be considered the reference method for comparison with other techniques (2). Of all the noninvasive modalities that have been used to evaluate tissue iron, only that of magnetic susceptometry, presently available in only two centers worldwide, provides a direct measure of hepatic storage iron quantitatively equivalent to that determined by biopsy, in all ranges of iron concentrations (8). By contrast, the more widely available modality of magnetic resonance imaging does not provide accurate quantitation of hepatic iron in patients with either severe iron overload, hepatic fibrosis, or both (9). OPTIMAL BODY IRON IN PATIENTS WITH THALASSEMIA MAJOR The toxic manifestations of iron overload depend not only on the amount of excess iron but also on the rate of iron accumulation, the duration of exposure to increased
J Pediatr Hematol Oncol, Vol. 22, No. 6, November/December 2000

IRON IN HEMOGLOBIN E- THALASSEMIA many patients with Hb E- thalassemia) in emerging countries where widespread implementation of this regimen is still awaited. Adequate deferoxamine prevents early death from cardiac disease; maintenance of body iron burdens corresponding to hepatic iron concentrations less than 15 mg per gram, dry weight, greatly decreases the risk of clinical disease (11). Hepatic iron can be maintained at concentrations not exceeding 7 mg per gram, dry weight, during modern regimens of deferoxamine. Moreover, deferoxamine arrests iron-induced progression of hepatic fibrosis to cirrhosis, even when administered in regimens that stabilize, rather than reduce, body iron burden (14). The importance of this finding in the seminal study that ushered in the modern era of deferoxamine therapy is highlighted by evidence that in iron overload associated with hereditary hemochromatosis, progression of hepatic fibrosis to cirrhosis is a critical event associated with an increased risk of death (15). The favorable effect of sustained reduction in body iron is also suggested by a relatively low-prevalence of endocrine (including thyroid and parathyroid) abnormalities in the modern era (16). In parallel, early, intensive, deferoxamine therapy improves the incidence of normal sexual development (17) and is effective in the prevention of diabetes mellitus (18). In summary, modern regimens of subcutaneous deferoxamine may extend survival, free of most complications of iron overload if body iron is reduced or maintained at safe levels (2,1922). PRELIMINARY STUDIES OF IRON LOADING AND CHELATING THERAPY IN HB E- THALASSEMIA Neither the guidelines for transfusion nor those for iron-chelating therapy in Hb E- thalassemia are entirely clear. In nontransfused patients with Hb E- thalassemia, the contribution of gastrointestinal iron absorption to iron loading is uncertain. In transfused patients with Hb E- thalassemia, the rate of iron accumulation, the partition of the iron burden between reticuloendothelial and parenchymal storage iron, and other factors may be different from those in homozygous -thalassemia. Because these factors influence the toxic manifestations of iron overload (2), these data will need to be amassed in large cohorts of patients with Hb E- thalassemia before firm conclusions can be made with respect to the comparability of iron loading in either condition. One problem that prevents accumulation of these data is the perception, in many centers in both emerging and developed countries, that determination of body iron burden by regular liver biopsies is not feasible. As part of the Natural History Study of Hb E- thalassemia, we have assessed body iron burden in patients

595

who have received approximately monthly transfusions in the largest center of affected patients on the island of Sri Lanka at the teaching hospital of Kurunegala. Since the early 1990s, monthly transfusions have been initiated for various indications, including anemia, fever, pallor, and abdominal distension, in more than 60 patients with Hb E- thalassemia at this center. The transfusion regimen in these patients maintained pretransfusion hemoglobin levels of approximately 7 g/dL. Because of concerns with respect to accelerated iron loading, to date we have obtained liver biopsy results from 73 patients to evaluate tissue iron stores. After determination of clotting parameters, platelet counts, and conscious sedation with midazolam and meperidine, biopsies were performed using ultrasound guidance by one interventional radiologist; up to three 4-cm passes per procedure using a 18-gauge needle were undertaken. Children were scheduled for the biopsy on a day of transfusion, observed for 6 hours during which the transfusion was administered, and discharged the same day. No complications occurred. Of particular interest, the results of liver biopsies in 11 patients with Hb E- thalassemia aged (mean standard error of mean) 21.4 4.7 years, in whom (estimated) less than 10 transfusions had been administered during the clinical course, suggest that this disorder may be associated with accelerated, but variable, gastrointestinal iron absorption. The clinical details of two of the patients in this cohort are outlined in the next paragraphs. PATIENT 1 Patient 1 was a 39-year-old man who first came to medical attention at the age of 29 years and was identified as having Hb E- thalassemia at the time he presented with fever. At that time, he received a single blood transfusion because of pallor; he was never subsequently transfused. Physical examination revealed pallor, icterus, and hepatosplenomegaly, with liver spleen palpable 10 cm below the costal margin and spleen palpable 15 cm below the costal margins. Hemoglobin was 7.1 g/dL; no other laboratory data were obtained. Hepatic iron level results of a biopsy performed at age of 39years-old indicated greatly increased levels of 16.1 mg of iron per gram of liver, dry weight, placing the patient at increased risk for cardiac disease and early death. PATIENT 2 Patient 2 was an asymptomatic 42-year-old man identified as having Hb E- thalassemia by DNA screening at age 41-years-old. He had never received a blood transfusion. Physical examination revealed pallor and hepaJ Pediatr Hematol Oncol, Vol. 22, No. 6, November/December 2000

596

N.F. OLIVIERI ET AL. sorption of iron appeared to be increased with body iron burden, as estimated by serum iron concentrations. Supporting these observations of abnormally regulated absorption of iron in thalassemia intermedia was the finding of a positive iron balance of 3 to 9 mg of iron per day, or between three-times and 10-times the normal absorption of iron (23). This (the only balance study reported in patients with thalassemia intermedia) indicated that the rate of iron loading in patients with thalassemia intermedia may be approximately 2 to 5 g of iron per year. By contrast, in a 50-kg patient with thalassemia major who may receive a yearly volume of 180 mL/kg packed cells with a hematocrit of approximately 75%, iron accumulation is estimated as approximately 7 g of iron per year. It is not clear that similar disruption of the normal mechanisms regulating iron absorption is present in patients with Hb E/thalassemia because studies of iron absorption in patients with this genotype (of great potential interest) have not been obtained. Because the studies by Pippard et al. relied on measurements of transferrin saturation and serum ferritin concentrations with no direct evaluation of tissue iron, body iron burden in their patients cannot be compared with that in our patients with Hb E/thalassemia in whom body iron burden was determined by liver iron concentrations obtained at biopsy. As noted previously, direct determination of body iron burden using quantitation of hepatic iron is indicated in patients with slightly increased serum ferritin concentrations (2). Our patients with Hb E/thalassemia underwent quantitative evaluation of body iron burden and, although the numbers are small, it is clear that a strong correlation between serum ferritin and hepatic iron concentrations is not evident. From the studies cited previously, it appears that older patient with thalassemia intermedia may be considered to be at eventual risk for iron-induced hepatic, cardiac, and endocrine dysfunction that characterizes the clinical course of the poorly chelated patient with thalassemia major. Although studies of organ function were not reported in previous studies of thalassemia intermedia, three of Pippards patients were noted to be diabetic or prediabetic, suggesting that loading of tissue iron may eventually be associated with serious organ dysfunction in patients with thalassemia intermedia, as in thalassemia major. This may be true in patients with Hb E/thalassemia as well. None of our patients had clinical diabetes, but no formal testing of glucose intolerance has yet been obtained. Further, because in only two of our patients (Table 1) hepatic irons exceeded 15 mg per gram of liver, dry weight, the threshold for increased risk for cardiac disease (11) for patient numbers in this cohort are too small to determine the prevalence of this most serious complication of iron overload. Also, at the time these patients had undergone liver biopsy, the assessment of

tosplenomegaly, with both liver and spleen palpable 10 cm below the costal margins. Hemoglobin was 7.3 g/dL; no other laboratory data were obtained. Hepatic iron levels detected in a biopsy performed at the age of 42-yearsold were modestly increased to 8.3 mg of iron per gram of liver, dry weight, exceeding the threshold for a heightened risk for hepatic fibrosis and other complications of iron overload. These data, which illustrate striking variations in two patients with Hb E- thalassemia with similar histories of transfusion, suggest that the rate and severity of iron loading in Hb E- thalassemia is highly variable and, at least in some patients, is similar to that in thalassemia intermedia. Only a few studies have attempted to define the extent of iron loading in patients with thalassemia intermedia. One such study examined the absorption and rate of accumulation of iron in 15 patients with homozygous or doubly heterozygous thalassemia, aged 4 to 42 years, who were not receiving regular transfusions (23). Some of these patients, like some of our patients with Hb E/thalassemia in the present cohort, had never been transfused. Severity of iron loading in these patients was assessed from transfusion history, determinations of serum iron, total iron binding capacity, and serum ferritin concentrations. In this study, values for total iron binding capacity in these patients with thalassemia intermedia varied from 30% to 100%; in nine of the patients, transferrin saturation exceeded 75%. By contrast, in our patients with Hb E/thalassemia, no values for total iron binding capacity or transferrin saturation are available and therefore cannot be compared with these previous studies. Concentrations of serum ferritin increased with the age of the patients with thalassemia intermedia, although increasing scatter of values was noted in the older patients. Unfortunately, comparison with the serum ferritin concentrations of our patients is not possible because, to date, so few of the values have been obtained in this cohort. The absorption of 59-labeled ferrous iron has been measured in patients with thalassemia intermedia with use of the Oxford total body counter (24). Absorption from a 5-mg dose of iron (in these patients with evidence of increased body iron burden) varied between 17% and 89%, compared with a mean of 15% in normal, ironreplete individuals. There was a highly significant correlation between the percentage of iron absorbed and the serum iron: the higher the serum iron, the greater the absorption of iron. By contrast, in normal individuals, the percentage of iron absorbed is inversely proportional to the body iron burden. Therefore, this study suggests that the mechanism or mechanisms regulating iron absorption in the presence of adequate iron stores may be abnormal in patients with thalassemia intermedia in whom the abJ Pediatr Hematol Oncol, Vol. 22, No. 6, November/December 2000

IRON IN HEMOGLOBIN E- THALASSEMIA these patients had not included studies of cardiac function, although these are now ongoing. Because the risk of complications of iron overload in selected patients with Hb E- thalassemia may be expected to be similar to that of patient with thalassemia intermedia or major, iron-chelating therapy in patients with Hb E- thalassemia, as in other hemoglobinopathies, should be initiated if the hepatic iron concentration exceeds 4 mg of iron per gram of liver, dry weight (2). Because total iron balance studies in patients with Hb E/thalassemia have not been conducted in response to administration of deferoxamine, the individual responses of these patients to iron-chelating therapy has not been proven. No quantitation of iron excretion has been determined in our patients in Sri Lanka, both because of technical limitations and difficulties in collection of these samples. Nonetheless, there appears, a priori, no rationale to suggest that the response to iron-chelating agents would be different in patients with Hb E/thalassemia than in patients with other genotypes responsible for thalassemia, but these studies have not yet been conducted. CONCLUSION The data presented here demonstrate the feasibility of accurate determinations of tissue iron in emerging countries, where such determinations may provide a rational basis for implementation of iron-chelating therapy to specific patients, and not to others, in parallel with often limited availability of such therapy in this setting. Quantitative studies of iron loading in Hb E- thalassemia should be undertaken in countries where this disorder is now a major public health problem, both to identify those at greatest possible risk for iron-induced complications and those most in need of life-saving therapy with deferoxamine. REFERENCES
1. Weatherall DJ. Hemoglobin E -thalassemia: an increasingly common disease with some diagnostic pitfalls. J Pediatr 1998;132: 7657. 2. Olivieri NF, Brittenham GM. Iron-chelating therapy and the treatment of thalassemia. Blood 1997;89:73961.

597

3. Olivieri NF, De Silva S, Fischer C, et al. Natural history study of Hemoglobin E/ thalassemia. Blood 1998;92(S1):532a. 4. Pippard MJ, Warner GT, Callender ST, et al. Iron absorption and loading in beta-thalassaemia intermedia. Lancet 1979;2:81921. 5. Pootrakul P, Kitcharoen K, Yansukon P, et al. The effect of erythroid hyperplasia on iron balance. Blood 1988;71:11249. 6. Olivieri NF, Sharma S, DeSilva S, et al. Impact of a low transfusion regimen on growth and iron loading in hemoglobins E/ thalassemia. Blood 1999;94(S1):422a. 7. Brittenham GM, Cohen AR, McLaren CE, et al. Hepatic iron stores and plasma ferritin concentration in patients with sickle cell anemia and thalassemia major. Am J Hematol 1993;42:815. 8. Brittenham GM, Farrell DE, Harris JW, et al. Magneticsusceptibility measurement of human iron stores. N Engl J Med 1982;307:16715. 9. Angelucci E, Giovagnoni A, Valeri G, et al. Limitations of magnetic resonance imaging in measurement of hepatic iron. Blood 1997;90:473642. 10. Cartwright GE, Edwards CQ, Kravitz K, et al. Hereditary hemochromatosis: phenotypic expression of the disease. N Engl J Med 1979;301:1759. 11. Brittenham GM, Griffith PM, Nienhuis AW, et al. Efficacy of deferoxamine in preventing complications of iron overload in patients with thalassemia major. N Engl J Med 1994;331:56773. 12. Olivieri NF. Thalassaemia: Clinical Management. In: Rodgers GP, ed. Baillieres Clinical Hematology. London: Bailliere Tindall, 1998;11:14762. 13. Porter JB. A risk-benefit assessment of iron-chelation therapy. Drug Saf 1997;17:40721. 14. Barry M, Flynn D, Letsky E, et al. Long term chelation therapy in thalassemia major: effect on liver iron concentration, liver histology and clinical progress. Br Med J 1974;2:1620. 15. Niederau C, Fischer R, Purschel A, et al. Long-term survival in patients with hereditary hemochromatosis. Gastroenterology 1996; 110:110779. 16. Olivieri NF. The thalassemias. N Engl J Med 1999;341:99109. 17. Bronspeigel-Weintrob N, Olivieri NF, Tyler BJ, et al. Effect of age at the start of iron chelation therapy on gonadal function in -thalassemia major. N Engl J Med 1990;323:7139. 18. Italian Working Group on Endocrine Complications in Nonendocrine Diseases. Multicenter study on prevalence of endocrine complications in thalassaemia major. Clin Endocrinol 1995;42: 5816. 19. Olivieri NF, Nathan DG, MacMillan JH, et al. Survival of medically treated patients with homozygous thalassemia. N Engl J Med 1994;331:5748. 20. Gabutti V, Piga A. Results of long-term iron-chelating therapy. Acta Haematol 1996;95:2636. 21. Giardini C. Treatment of -thalassemia. Curr Opin Hematol 1997; 4:7987. 22. Hershko C. Iron chelators for thalassaemia. Br J Haematol 1998; 101:399406. 23. Pippard MJ, Callender ST, Warner GT, et al. Iron absorption and loading in beta-thalassaemia intermedia. Lancet 1979;ii:81921. 24. Warner GT, Oliver R. A whole-body counter for clinical measurements utilizing the shadow shield technique. Phys Med Biol 1966;11:8393.

J Pediatr Hematol Oncol, Vol. 22, No. 6, November/December 2000