Autografting as a risk factor for persisting iron overload in long-term survivors of acute myeloid leukaemia

Transcription

1 (2003) 32, & 2003 Nature Publishing Group All rights reserved /03 $ Autografting as a risk factor for persisting iron overload in long-term survivors of acute myeloid leukaemia Department Of Haematology, Royal Liverpool University Hospital, Liverpool, UK Summary: We studied the iron status of 32 evaluable adult acute myeloid leukaemia (AML)survivors who were entered into the UK Medical Research Council acute myeloid leukaemia (AML)10 and 12 trials at our institution between 1988 and Patients were required to have been independent of all blood products for at least 3 years. As a group, the median first serum ferritin level was 1323 lg/l (NR lg/l)at a median of 1321 days from the last transfusion confirming the presence of significant iron overload persisting for some years after completion of all therapy and blood products. There was a general trend for the serum ferritin level to fall with time, but the fall was less pronounced in men and carriers of the C282Y mutation. Recipients of autologous stem cell transplantation (SCT)had a higher median first serum ferritin level (3245 lg/l)than patients who received chemotherapy alone (1148 lg/l)or allogeneic SCT (1334 lg/l)because of increased use of transfused blood. Nine of the 10 recipients of autologous SCT underwent venesection. No evidence of end organ damage was seen in any patient. Serial monitoring of serum ferritin and assessment of the C282Y status may be useful in all longterm AML survivors, especially autograft recipients. (2003) 32, doi: /sj.bmt Keywords: iron overload; acute myeloid leukaemia; autograft; ferritin Patients with acute myeloid leukaemia (AML) are often heavily transfused with red cells during the course of their treatment. The natural history of this transfused iron is unknown. As survival from AML improves, it remains unclear whether the transfusion burden associated with treatment causes clinically significant iron overload in longtermsurvivors. In assessing the iron status of patients following treatment for AML, consideration must be given to Correspondence: Dr NM Butt, Department Of Haematology, Royal Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, UK; nbutt@liv.ac.uk Received 6 March 2003; accepted 27 May 2003 potential influencing factors. Firstly, the previous treatment received may be important. Many patients undergo stem cell transplantation (SCT) as part of their management, but it is not known whether this group are any more at risk of iron overload than patients receiving chemotherapy alone, since transplantation sometimes substitutes for courses of chemotherapy. Secondly, men and postmenopausal women may be at greater risk of iron overload than premenopausal women, who could be protected from iron accumulation by menstruation. Another relevant factor may be the presence of haemochromatosis gene (HFE) mutations. 1 The mechanism of iron overload produced by an abnormal HFE protein is unknown but it is thought that the body s iron status is regulated through this protein within the intestinal crypts and reticuloendothelial system. 2 The commonest of these mutations is a guanine to tyrosine (G-to-A) transition at nucleotide 845 resulting in the substitution of cysteine for tyrosine at amino acid 282 (C282Y). The relevance of a second mutation, H63D, is unclear as it only causes hereditary haemochromatosis when in association with the C282Y mutation (compound heterozygosity). 3 Homozygotes for C282Y or compound heterozygotes for C282Y/H63D are at most significant genetic risk fromiron overload. 4 The relevance of these factors in the context of AML survivors is unknown. There are a variety of techniques available to determine tissue and organ iron content. Invasive methods, such as liver biopsy for calculation of the hepatic iron index, are potentially dangerous and impractical for screening in leukaemia survivors. Noninvasive techniques for iron assessment include hepatic imaging with computerised tomography (CT), magnetic resonance imaging (MRI) or superconducting quantuminterference device (SQUID) biomagnetometry. These investigations require special equipment or expertise and are not widely available for iron assessment. 5 Serumferritin estimations are an easy and convenient test for determining total body iron stores. 5,6 Although serumferritin levels are affected by acute inflammation 7 and vitamin C deficiency, 8 they are readily available in the clinic, and have been shown to accurately mirror body iron stores. Here, we report, to our knowledge, the first serial study of iron overload in long-termsurvivors of AML. We describe the natural history of the iron burden following completion of treatment and correlate our findings with gender, treatment, C282Y mutational status and the number of red cell transfusions received.

2 910 Patients and methods Patient population All patients aged between 18 and 60 years with newly diagnosed AML at our institution between 1988 and 1998, and who were entered into the UK Medical Research Council AML 10 and 12 trials were potentially eligible for study. The central question in these studies was the role of SCT in the management of AML. Our institutional ethics committee approved these studies, and all entrants gave informed consent. Briefly, these sequential trials comprised induction with DAT 3 þ 10 (daunorubicin on days 1, 3 and 5; cytosine arabinoside (ara-c) twice daily for 10 days, and oral 6-thioguanine (6-TG) for 10 days), then further treatment with DAT 3 þ 8 (as DAT 3 þ 10 except ara-c and 6-TG given for 8 days) followed by MACE (amsacrine, ara-c and etoposide each for 5 days). Further treatment was either with MidAc (mitozantrone for 5 days and intermediate dose ara-c for 3 days), or MidAc preceded by ICE (idarubicin, ara-c and etoposide). 9 Conditioning therapy for patients who subsequently proceeded to haemopoietic SCT was with cyclophosphamide 120 mg/kg and total body irradiation 14.4 Gy. Patients were excluded fromthis study if any of the following applied: not in remission prior to MACE; relapse at any time; not independent of all blood products for at least 3 years at the time of analysis, or where insufficient follow-up information was available. The C282Y HFE gene status, blood product consumption and all serial serumferritin levels were recorded. Our unit policy recommended venesection at serum ferritin levels above 2000 mg/l at the end of treatment; the venesection history was recorded for each patient. Statistical comparisons between patient groups were made using the Mann Whitney U-test. Results A total of 32 consecutive eligible patients were evaluable, and their details are given in Table 1. There were equal numbers of male and female patients with median ages of 46.5 and 39 years, respectively, at the end of treatment. In all, 16 patients had undergone SCT, of which six were allogeneic (all HLA-identical sibling donor) and 10 were autologous. Trends in serum ferritin with time Table 2 shows the trends in serumferritin with time in AML survivors. As a group, the median first serum ferritin level for all 32 patients was 1323 mg/l (normal range mg/l) at a median of 1321 days from the last transfusion. Table 1 Clinical details of AML survivors Patient no. Age at last transfusion (years) Sex MRC trial Treatment Blood received (units) Venesection history C282Y mutational status 1 42 F 10 Chemotherapy only NV Normal 2 49 M 10 Chemotherapy only NV Normal 3 19 M 10 Chemotherapy only NV Normal 4 36 M 10 Allograft NV Heterozygous 5 56 F 10 Allograft NV Normal 6 31 F 10 Allograft NV Normal 7 41 F 10 Autograft V Normal 8 43 M 10 Autograft V Normal 9 44 F 10 Autograft V Heterozygous F 10 Chemotherapy only NV Normal M 10 Chemotherapy only NV Normal F 10 Chemotherapy only NV Normal M 10 Autograft V Heterozygous M 10 Allograft NV Heterozygous M 10 Autograft NV Normal M 10 Chemotherapy only NV Normal M 10 Autograft 22 V Normal F 10 Chemotherapy only 53 NV Normal M 10 Autograft 45 V Normal M 10 Chemotherapy only 50 NV Normal F 10 Chemotherapy only 42 NV Normal F 10 Autograft 60 V Normal F 12 Allograft 48 NV Heterozygous M 12 Chemotherapy only 37 NV Heterozygous M 12 Autograft 42 V Normal F 12 Chemotherapy only 38 NV Normal F 12 Chemotherapy only 39 NV Normal F 12 Autograft 76 V Normal M 12 Chemotherapy only 39 NV Normal M 12 Chemotherapy only 28 NV Homozygous F 12 Chemotherapy only 30 NV Normal F 12 Allograft 124 V Normal M ¼ male; F ¼ female; V ¼ venesected; NV ¼ nonvenesected; ¼ data not available.

3 There was no significant difference in the first serumferritin level between men and women (1434 vs 1252 mg/l) at a median of 1321 and 1307 days, respectively (P ¼ 0.522). There was a general trend for the serumferritin level to fall with time. This fall was less pronounced in men than in women (933 vs 560 mg/l at a median of 2679 and 2074 days, respectively), although this difference did not achieve statistical significance (P ¼ 0.073). This confirms the presence of significant iron overload persisting for some years after completion of all therapy and blood products. However, no patient developed any evidence of hepatic or pancreatic dysfunction (data not shown). Effects of venesection In total, 10 patients had a serumferritin level greater than 2000 mg/l at the end of treatment and required venesection (Table 2). As expected, the first median ferritin levels were significantly higher in these patients than in those not venesected (Po0.001), and also fell more rapidly in venesected patients. Table 2 also shows that the median serumferritin level fell in the nonvenesected groups with time. Correlation of iron overload with C282Y status The C282Y status of the patients and their relationship with the serum ferritin level are summarised in Table 3. Six patients (19%: four male and two female) were heterozygous, and one male patient was homozygous for the C282Y mutation. No difference in median first serum ferritin level was seen between patients with and without the mutation. In nonvenesected patients, there was a trend for a slower fall in serumferritin in C282Y heterozygotes than in normals (median 970 vs 561 mg/l at 2245 and 2770 days, respectively), although this did not achieve statistical significance. Correlation with previous treatment Table 4 shows the trends in serumferritin according to the type of SCT received. SCT recipients had a higher first serum ferritin compared with patients receiving chemotherapy alone. Recipients of autologous SCT had a higher median first serum ferritin level than patients who received chemotherapy only (3245 vs 1148 mg/l at 1094 and 1364 days, respectively; Po0.001). Among autologous SCT recipients, no evidence of C282Y mutational bias or sex bias was seen. Nine of the 10 recipients of autologous SCT underwent venesection. In contrast, no difference in serum ferritin was seen between allogeneic SCT recipients and patients treated with chemotherapy alone. Correlation with transfusion records Transfusion records of autologous SCT recipients were compared with those of patients receiving chemotherapy only (Table 1). In view of the long period of follow-up, extending over a decade in some patients, accurate computerised transfusion records were only available for five autologous SCT recipients and nine chemotherapyonly patients. There was a trend for autologous SCT recipients to have received more units of blood than chemotherapy-only patients, although this did not reach statistical significance (47 vs 39 units, P ¼ 0.23). 911 Table 2 Trends in median serum ferritin with time AML survivors Male Female V NV No. of patients Serumferritin (mg/l) Median time (days) Serumferritin (mg/l) Median time (days) V ¼ venesected; NV ¼ nonvenesected. Table 4 Previous treatment received by AML survivors Type of treatment Chemotherapy only Autograft Allograft No. of patients Serumferritin (mg/l) Median time (days) Serumferritin (mg/l) Median time (days) Table 3 C282Y status of AML survivors C282Y status Normal Heterozygotes Homozygotes Non-V normals Non-V heterozygotes No. of patients Serumferritin (mg/l) Median time (days) Serumferritin (mg/l) Median time (days) Non-V¼ nonvenesected; ¼ data not available.

4 912 Discussion It is well established that patients with transfusion-related iron overload can suffer considerable morbidity. 10,11 Although several groups have investigated liver dysfunction and iron status of SCT recipients, 12,13 to our knowledge, there has been no serial study of iron status or HFE mutational status in long-term AML survivors. In the present study, we found that iron overload remained a significant problem in AML survivors several years after completion of treatment and cessation of transfusion. We report that autologous SCT is a significant aggravating factor in the iron overload of AML survivors. The autologous SCT group had the highest serumferritin levels at the end of treatment and nine of the 10 autograft recipients required venesection. There was a trend for autologous SCT recipients to have received more blood than chemotherapy-only recipients. In AML 10, transplant recipients (both autografts and allografts) received this as an extra fifth course of treatment. However, in AML 12, allograft and autograft were delivered as course 4 or 5, and some chemotherapy-only patients received five courses of conventional chemotherapy, but no transplant. This offsets the effect in AML 10 of giving a transplant as an additional treatment. Despite this, autografts still had higher ferritin levels than allografts and chemotherapy-only recipients. Our data are in accordance with that of Lichtman et al 14 who identified five patients who had significant iron overload at least 1 year after completion of therapy for haematological malignancy. Of their three AML patients, two had received autografts while one had received an allograft. The highest ferritin estimations were in the autograft candidates (5000 and 9942 vs 1439 ng/ml in the allograft recipient). Transfusion data were available on only one of the autograft recipients. He had received more packed cell transfusions than the allograft recipient (66 vs 52 units). We also found that the serumferritin fell with time in all survivors of AML. However, there was a trend for male survivors to have a slower fall in ferritin than women, although this did not reach statistical significance. This trend could be due to resumption of menstrual loss in premenopausal women. There is no established link between the HFE gene and adult AML. 15 Indeed, some studies report a possible lower incidence of the C282Y allele in AML patients compared with a control group (3.8 vs 10.2%). 16 The prevalence of C282Y mutations in our cohort was higher than that reported by Hannuksela et al, with seven of our 32 patients carrying the mutation (21.9%). These carriers appeared to have a slower fall in iron stores than did patients without the C282Y mutation, although this may be mitigated by the fact that C282Y carriers may have increased intestinal absorption of iron that may slow the fall in iron stores with time. Assessment of iron status in the present study is limited to serum ferritin estimations, as this is a convenient method for assessing iron status in AML survivors. More sensitive techniques such as hepatic CT, MRI or SQUID are not widely available and are impractical for serial use in a haematology clinic setting. Another limitation is the lack of data on hepatic iron content and on liver architecture, each of which requires liver biopsy. Seven of our 32 cases were found to have mild elevations in alanine transaminase, but in each case this was transient and no patient developed any clinical evidence of hepatic or pancreatic dysfunction (data not shown). Furthermore, a recent report indicates that late cirrhosis in long-termmarrow transplant survivors is due, in most cases, to hepatitis C infection rather than iron overload, although iron may accelerate the cirrhotic process. 17 In summary, long-term AML survivors remain significantly iron overloaded for several years after completion of treatment and transfusion, with a trend of gradual decline in iron burden. The mechanism of this fall is unknown but could involve both decreased intestinal absorption of dietary iron and proportionally greater iron loss from epithelial shedding. Further study of these aspects is justified. Our principal finding is that recipients of autologous SCT may be at particular risk of iron overload. Since the incidence of iron-associated end-organ damage appears very low, more invasive assessment of the iron burden appears unjustified in otherwise uncomplicated cases. References 1 Feder JN, Gnirke A, Thomas W et al. A novel MHC class I- like gene is mutated in patients with hereditary haemochromatosis. Nat Genet 1996; 13: Townsend A, Drakesmith H. Role of HFE in iron metabolism, hereditary haemochromatosis, anaemia of chronic disease, and secondary iron overload. Lancet 2002; 359: Beutler E. The significance of the 187G (H63D) mutation in hemochromatosis. Am J Hum Genet 1997; 61: Worwood M. Pathogenesis and management of haemochromatosis. Br J Haematol 1999; 105 (Suppl. 1): Porter JB. Practical management of iron overload. Br J Haematol 2001; 115: Worwood M. Serumferritin. Clin Sci (Lond) 1986; 70: Konijn AM, Hershko C. Ferritin synthesis in inflammation. I. Pathogenesis of impaired iron release. Br J Haematol 1977; 37: Chapman RW, Hussain MA, Gorman A et al. Effect of ascorbic acid deficiency on serumferritin concentration in patients with beta-thalassaemia major and iron overload. J Clin Pathol 1982; 35: Burnett AK, Goldstone AH, Stevens RM et al. Randomised comparison of addition of autologous bone-marrow transplantation to intensive chemotherapy for acute myeloid leukaemia in first remission: results of MRC AML 10 trial. UK Medical Research Council Adult and Children s Leukaemia Working Parties. Lancet 1998; 351: Gabutti V, Borgna-Pignatti C. Clinical manifestations and therapy of transfusional haemosiderosis. Baillieres Clin Haematol 1994; 7: Hollan SR. Transfusion-associated iron overload. Curr Opin Hematol 1997; 4: McKay PJ, Murphy JA, Cameron S et al. Iron overload and liver dysfunction after allogeneic or autologous bone marrow transplantation. Bone Marrow Transplant 1996; 17: Harrison P, Neilson JR, Marwah SS et al. Role of nontransferrin bound iron in iron overload and liver dysfunction in long term survivors of acute leukaemia and bone marrow transplantation. J Clin Pathol 1996; 49:

5 14 Lichtman SM, Attivissimo L, Goldman IS et al. Secondary hemochromatosis as a long-term complication of the treatment of hematologic malignancies. Am J Hematol 1999; 61: Dorak MT, Burnett AK, Worwood M. Hemochromatosis gene in leukemia and lymphoma. Leuk Lymphoma 2002; 43: Hannuksela J, Savolainen ER, Koistinen P et al. Prevalence of HFE genotypes, C282Y and H63D, in patients with hematologic disorders. Haematologica 2002; 87: Strasser SI, Kowdley KV, Sale GE et al. Iron overload in bone marrow transplant recipients. Bone Marrow Transplant 1998; 22: