Hepcidin Levels and Their Determinants In Different Types of Myelodysplastic Syndromes

Iron overload is frequently occurring in patients with myelodysplastic syndromes (MDS), with recent data suggesting an impact on both overall and leukemia-free survival^1,2. Though prolonged RBC transfusion therapy appears the main contributor, many patients develop iron overload at an early stage of the disease, before the onset of transfusion dependency. It has been postulated that an altered production of hepcidin, the key hormone regulating iron homeostasis, may play a role at this regard. Until recently, studies have been hampered by problems in the development of reliable hepcidin assays, so that only scanty and conflicting data based on semi-quantitative measurement of urinary hepcidin have been reported^3,4. This study mainly focused on analyzing serum hepcidin levels in MDS patients by means of a recently validated and improved Mass-Spectrometry based method⁵.

One hundred and thirteen consecutive patients (mean age 72.8 ± 9.2 years; 68.1% males) with different types of MDS according to the WHO classification were included in this study. To be enrolled, patients had to be previously untreated or treated only with transfusions. Besides hepcidin, in all subjects we determined serum ferritin, transferrin saturation (TS), non-transferrin-bound-iron (NTBI), along with some putative determinants of hepcidin, like GDF-15⁶ known to be associated with ineffective erythropoiesis, and C-Reactive Protein (CRP) as a surrogate of systemic IL-6 production. Fifty-four healthy individuals (61.1% males) with rigorous definition of normal iron status were used as controls.

Biochemical markers of iron overload (ferritin and TS), but also CRP and GDF-15 were significantly higher in MDS patients than in controls, even when considering only non-transfused patients. Patients with RARS and the 5q- syndrome appeared as the most iron overloaded, having the highest levels of ferritin, TS, and NTBI. In the whole MDS population, serum hepcidin levels showed a considerable variability, with overall mean values not significantly different from controls [geometric means (gm) with 95% CIs: 5.31 (3.98-7.08) versus 4.2 (3.53-5.0) nM, P=0.28], while the hepcidin/ferritin ratio was significantly lower than in controls [10.1 (7.53-13.53) versus 52.9 (43.6-64.3), P<0.001]. After stratification according to WHO subtypes, hepcidin levels showed significant differences, with the lowest levels in patients with RARS (gm 1.43 nM) and the highest levels in patients with RAEB 1–2 (gm 11.3 nM) and with CMML (gm 10.04 nM) (P=0.003 by ANOVA). The latter groups had substantial elevation of CRP as compared to other MDS subtypes (P=0.008 by ANOVA), while GDF-15 was consistently but uniformly elevated in all MDS subtypes (P=0.97 by ANOVA). Multivariate linear regression models adjusted also for age, sex, and history of RBC transfusions, showed ferritin (β-coefficient 0.45, P=0.002), CRP (β-coefficient 0.21, P=0.02), and different MDS subtypes as the main independent predictors of hepcidin levels. The different degree of correlation between hepcidin and ferritin among the MDS subtypes were analyzed in a general linear model using the F test for slopes. Hepcidin regulation by iron appeared conserved, though relatively blunted in RA, RARS, and 5q- patients, while it was lost in RAEB 1–2 and CMML.

Hepcidin levels are consistently heterogeneous in MDS according to different subtypes, likely as the result of the relative strength of competing stimuli. Relative inhibition by ineffective erythropoiesis (but not mediated by GDF-15) seems to prevail particularly in RARS and 5q- syndrome, and is likely to increase the risk of iron overload in these subgroups. On the other hand, patients with RAEB 1–2 and CMML appears to have hepcidin induction that could be driven by cytokines. If confirmed, these results may be relevant not only for a better understanding of iron pathophysiology in MDS, but also for possible future approach with hepcidin modulators⁷.

References: 1) Sanz G, et al. Blood 2008;112: abs 640. 2) Alessandrino EP, et al. Haematologica 2010;95:476-84. 3) Winder A, et al. Br J Haematol 2008;142:669-71. 4) Murphy PT, et al. Br J Haematol 2009;144:451-2. 5) Campostrini N, et al. J Biomed Biotechnol 2010;2010:329646. 6) Tanno T, et al. Nat Med 2007;13:1096-101. 7) Sasu BJ, et al. Blood 2010;115:3616-24.

No relevant conflicts of interest to declare.