In patients with Hemoglobin (Hb) E thalassemia, the most common form of severe thalassemia worldwide
(Weatherall DJ. Blood 2010;4331-6), early reports suggest that this disorder is associated with variable iron loading in the absence of frequent transfusions (Olivieri NF et al. Int J Ped Hem Onc 2000; 22:593-7), possibly related to the phenotypic variability of the disorder. At the National Thalassaemia Centre, Kurunegala, Sri Lanka, we have studied two groups of patients with Hb E thalassemia using spin density projection assisted R2-MRI (R2-MRI;
FerriScan)
(St Pierre TG et al. Blood 2005; 105:855-61). Group 1 (n=23) had received ≤ 20 transfusions (median = 9) lifelong, despite a relatively advanced age (mean ± SD, 27.5 ± 16.2 [range 7.8–57.4] years). Group 2 (n=47) had received > 20 transfusions lifelong. In Group 1, liver iron concentration (LIC) was strikingly variable (geometric mean 5.6 [range 1.0–33.0] mg Fe/g dw). LIC exceeded 7 mg Fe/g dw in 11 patients and 15 mg Fe/g dw in 4 patients. Values of serum ferritin (SF) (mean ± SD, 827 ± 611 [range 254–2484] ug/L) and serum ALT (37.6 ± 19.7 [range 12–86] U/L) did not reliably reflect the degree of elevation in LIC (R
2 = 0.38, P = 0.0017 for SF; R
2 = 0.12, P = 0.10 for ALT). These elevations of body iron burden were associated with evidence of organ dysfunction, with abnormal ALT (observed in 11 patients), hypothyroidism (6 patients), final adult height <3
rd percentile (2 patients), and no evidence of glucose intolerance. In 15 of the 23 Group 1 patients, LIC had been determined in a previous biopsy, and 9 of these 15 patients had received <1 transfusion/year since that biopsy. None of these 9 patients had significant deferoxamine (DFO) exposure between biopsy and R2-MRI. Hence, rates of non-transfusional iron accumulation (NTIA) could be estimated (over a period of 8.4 ± 2.6 [range 2.6–11.4] years between biopsy and R2-MRI). NTIA in these 9 patients was 0.40 ± 0.39 [range -0.10 to 0.96] mg Fe/g dw liver per year. In the 47 Group 2 patients, regular transfusions had been administered for years and subsequently withdrawn, as described previously
(Premawardhena A et al. Lancet 2005;366:1467-70). In 37 of these 47 patients, LIC had been determined by biopsy following the period of regular transfusions. Of these 37 patients, 25 patients had had <1 transfusion/year since biopsy. Of these 25 patients, 9 had had no significant DFO exposure between biopsy and R2-MRI, and the rate of NTIA could therefore be estimated between these two LIC measurements (over a period of 9.1 ± 1.4 [range 7.3–11.8] years). NTIA in these 9 patients was 0.12 ± 0.21 [range -0.15 to 0.41] mg Fe/g dw liver per year. The other 16 of the 25 patients had had significant DFO exposure between biopsy and R2-MRI. Of the 18 patients (9 in Group 1 and 9 in Group 2) for whom rates of NTIA could be estimated, the mean age was 32.0 ± 13.6 [range 14.0–60.3] years. Time between LIC assessments was 8.8 ± 2.1 [2.6-11.9] years. Initial and final LIC were 4.6 ± 1.9 [1.8-8.3] mg Fe/g dw liver, and 6.7 ± 3.4 [0.7-12.7] mg Fe/g dw liver, respectively. Mean NTIA
(Figure; see A) was 0.26 ± 0.33 [-0.15 to +0.96] mg Fe/g dw per year.
Rates of change of LIC in patients receiving < 1 transfusion per year (A) with no significant exposure to DFO; and (B) with significant exposure to DFO. In the 16 patients in Group 2 who had received < 1 transfusion per year and significant DFO exposure between LIC measurements, the rate of NTIA was -0.34 ± 0.66 [range -1.70 to +1.02] mg Fe/g dw liver per year (Figure; see B). These data suggest variable rates of NTIA in Hb E thalassemia, with evidence in some patients of rates up to approximately 1 mg Fe/g dw liver per year. This rate of NTIA may be associated with risk of organ dysfunction in the second decade of life. The data also suggest that DFO, administered in low doses and over short periods relative to those used in thalassemia major, may be effective in reducing body iron in Hb E thalassemia. The etiology of the observed variation in NTIA in these patients is being explored. Although hereditary hemochromatosis in Asia was believed to be rare and may often be masked by iron deficiency, recent studies (Lok CY et al. Blood 2009; 114:20-5) suggest that primary iron overload is not as rare in the Indian subcontinent and Southeast Asia as believed previously. Genetic studies, in parallel with evaluation of body iron burden, using techniques with known accuracy and precision, may elucidate the origins of the variable NTIA observed in patients with Hb E thalassemia.
