Iron overload in myelodysplastic syndromes

The myelodysplastic syndromes (MDSs) comprise a heterogeneous group of clonal hematopoietic disorders characterized by ineffective hematopoiesis, cellular dysplasia, peripheral cytopenias and an increased risk of transformation to acute myeloid leukemia [1]. Despite the use of erythropoietin-stimulating agents and the development of novel therapies to enhance bone marrow function, more than 80% of MDS patients will require chronic red blood cell (RBC) transfusions [2]. In humans, no physiological mechanism exists for iron elimination and, as such, patients with MDS are prone to transfusional iron overload. Eventually, the effects of iron deposition will be seen clinically, causing damage to the heart, liver and endocrine organs [3]. It is a widely held belief, echoed in published treatment guidelines [4–14], that the use of iron-chelation therapy is important in the management of patients with MDS. It remains to be discerned whether iron-chelation therapy results in improved overall survival.

How does iron overload happen in MDS?

The normal iron content of the body is 3–4 g, of which the majority exists as hemoglobin in circulating RBCs. Iron homeostasis is maintained through the loss of iron in menses, sweat and skin cells at a rate of 1 mg/day, balanced by the dietary absorption of 1–2 mg/day of iron. No mechanism exists for iron excretion and since each unit of RBCs contains 200–250 mg of iron, approximately 100-times the normal daily iron flux, MDS patients who require chronic blood transfusions are destined to develop iron overload. However, in addition to transfusional iron loading, some MDS patients have increased intestinal absorption of iron, similar to patients with hemoglobinopathies and hereditary hemochromatosis (HH) [14,15].

Hepcidin, a peptide synthesized by the liver, is the key negative regulator of intestinal iron absorption and circulating iron, blocking cellular iron efflux by binding to ferroportin and causing its internalization and proteosomal degradation [16]. Hepcidin is upregulated in response to iron overload and inflammation and downregulated in response to hypoxia, anemia and elevated levels of erythropoietin [17]. It has been observed, however, that in certain forms of chronic anemia associated with iron loading – thalassemia, congenital dyserthyropoietic anemia type 1 (CDA-1) and some forms of MDS – hepcidin levels are unexpectedly low [18,19], thus permitting overabsorption of iron even in the face of increased iron stores. The biology underpinning this apparent paradox involves a recently characterized regulatory protein known as growth differentiation factor (GDF)15, which is produced by committed erythroid precursors in response to erythropoietin and hypoxia [20]. GDF15, a member of the TGF-β family, is a direct repressor of hepcidin mRNA expression [21]. Circulating levels of GDF15 are dramatically increased in thalassemia [21] and other congenital anemias [22,23], and in the refractory anemia with ringed sideroblasts (RARS) variant of MDS [20] and thus explain, at least in part, the low-hepcidin/high-iron state observed in these conditions.

‘Free’ iron & its role in organ dysfunction

There have been several proposed mechanisms describing the relationship between iron overload and organ dysfunction. Recent debate has focused on the concept of nontransferrin bound iron or ‘free’ iron, referred to more precisely as labile plasma iron (LPI), which mediates tissue damage through the generation of free radicals via the Fenton reaction. In healthy individuals iron circulates bound to transferrin, effectively shielding it from reactions leading to redox cycling and the formation of reactive oxygen species (ROS) [24]. When transferrin becomes fully saturated, as in transfusional iron overload, LPI is increased [25–28], but is effectively and rapidly cleared following iron-chelation therapy [29,30].

Multiple small retrospective studies have demonstrated that organ dysfunction can result from transfusional iron overload in patients with MDS [31–33]. The pattern of iron-related morbidity in MDS mirrors what is seen in thalassemia: the heart is the most commonly involved organ (cardiomyopathy/heart failure and rhythm/conduction disturbances), with other organs involved including the liver (fibrosis and hepatocellular injury) and endocrine organs (diabetes and pituitary and adrenal dysfunction). These studies also found that the degree of organ dysfunction is correlated with the number of units of RBCs transfused.

A recent administrative database study has provided corroboration for these early retrospective studies. Goldberg et al. analyzed the clinical and economic consequences of MDS in the USA, comparing an age-matched cohort from the general population to newly diagnosed MDS patients and following them over 3 years [34]. In a large Medicare database, they identified 705 patients with newly diagnosed MDS between January and March 2003. Of these, 46% received blood transfusions and 2% were chelated. Transfused MDS patients experienced a higher rate of cardiac events than nontransfused patients with MDS (80 vs 69%; p = 0.002). Furthermore, 74% of the MDS cohort suffered from cardiac disease during the 3-year study, significantly more than the 42.4% of patients from the age-matched population (p < 0.001). This study provides important epidemiological support for the contention that iron overload contributes to morbidity in MDS patients.

Does iron overload decrease survival in MDS?

Although mortality due to iron overload has been extensively documented in hemoglobinopathy and HH there is a paucity of data with regard to MDS. In a retrospective study, Malcovati et al. evaluated 467 patients with MDS and found that the development of secondary iron overload (ferritin > 1000 ng/ml) significantly affected the survival of transfusion-dependent patients (hazard ratio [HR]: 1.30; p = 0.03), an effect seen predominantly in patients with refractory anemia and RARS [35]. Leukemia-free survival (LFS) was also significantly worse in transfusion-dependent patients (HR: 2.02; p = 0.01). Garcia-Manero et al. substantiated these results in a series of 95 low/intermediate-1 (International Prognostic Scoring System [IPSS]) risk MDS patients [36]. They reported a significantly worse survival for patients with a serum ferritin level of more than 1000 ng/ml (median survival not reached vs 19.8 months; p = 0.007). Lastly, a Japanese study retrospectively evaluated 152 patients with MDS, identifying fatal cardiac or liver failure in 24 and 7% of the cohort, respectively [37]. These complications were seen almost exclusively in patients classified as low/intermediate-1 IPSS risk whose serum ferritin was greater than 1000 ng/ml. Taken together, these studies indicate that in the low/intermediate-1 IPSS risk category, transfusion-dependent patients with a ferritin of more than 1000 ng/ml have a decreased overall survival and potentially a decreased LFS.

Does iron overload promote progression of MDS to acute leukemia?

The observation in the Pavia study that patients with high ferritin had reduced LFS led to the conjecture that iron overload may contribute to increased mortality by increasing the risk of leukemic transformation [35]. Data presented at the most recent American Society of Hematology meeting have fueled this speculation. Sanz et al. sought to determine the independent prognostic value of transfusion dependency and iron overload in 2994 patients with MDS [38]. Their results indicated that both transfusion dependency and the development of iron overload significantly affect both overall survival (HR: 52.4; p < 0.0001 and HR: 8.8; p < 0.0001, respectively) and LFS (HR: 3.5; p = 0.003 and HR: 6.6; p < 0.0001, respectively).

How can the association between iron overload and reduced LFS be explained? The null hypothesis against which other explanations should be tested is that iron overload simply reflects severity of disease; a rather radical alternative hypothesis is that iron overload promotes progression of MDS to acute myeloid leukemia. How might this occur? As noted previously, free iron catalyzes the production of ROS, which are known to cause DNA damage, alter signaling pathways controlling growth and apoptosis and to be carcinogenic in animal models. Chan et al. measured intracellular ROS (iROS) concentrations in CD34-positive bone marrow cells from MDS patients and reported a strong correlation between iROS and serum ferritin in iron-overloaded patients [39]. Thus, a plausible model may be proposed in which iron overload leads to elaboration of ROS in hematopoietic stem cells (HSCs), resulting in accumulation of mutations and promoting clonal evolution and disease progression. It may be argued that this model is invalidated by the observation that acute leukemia is not seen in patients with thalassemia major, in which iron overload is more prolonged and severe than in MDS. However, this fails to take into account the distinct natures of these two disorders: in thalassemia HSC function is normal, while MDS is a clonal HSC disorder in which homeostatic responses to DNA damage may be impaired. In this sense it is not iron overload per se but its interaction with the preleukemic HSC in MDS that promotes the development of acute leukemia.

It must be emphasized that the studies that identified an association between iron overload and reduced LFS are retrospective in nature and thus prone to bias. Prospective studies are needed to elucidate the relationship between iron overload and LFS. Such studies have the potential for great clinical impact – if iron overload is confirmed to impact both overall survival and LFS, iron-chelation therapy will be of the utmost importance in the MDS patient population.

The importance of iron chelation

The effectiveness of iron chelation in reversing organ damage due to iron overload in MDS patients has been demonstrated in uncontrolled studies [40–42]. Iron chelation in MDS can successfully lower serum ferritin as well as liver and cardiac iron content in many patients. However patient compliance, the amount of transfusional iron overload and drug pharmacokinetics are key determinants of the success of therapy. Three iron-chelating agents are available: deferoxamine, deferasirox and deferiprone. Clinicians have had the most experience with deferoxamine, which is primarily administered by slow subcutaneous infusion. Adverse effects associated with deferoxamine are well described and patient compliance is poor. Deferasirox, an oral iron chelator, has been shown to be safe and effective in individuals with transfusional iron overload in a variety of chronic anemias, including MDS [43,44]. Recent data have demonstrated that at a dose of 20 mg/kg/day, significant reductions are seen in both serum ferritin and LPI [41,45,46]. Adverse effects, of which most are mild-to-moderate, include diarrhea (most common), gastrointestinal disturbance (nausea, abdominal pain) and rash. Further follow-up data are needed to define the long-term effectiveness and safety of deferasirox.

Not only may iron chelation be beneficial in reducing total body iron stores, it may also have an impact on mortality. Leitch et al. conducted a retrospective review of 178 patients with MDS to determine the effect of iron-chelation therapy on survival [47]. In multivariate Cox regression analysis, IPSS score (p < 0.008) and iron-chelation therapy (p < 0.02) were the only variables predictive of survival. Median overall survival for patients with low or intermediate-1 IPSS scores was 160 months with iron-chelation therapy versus 40 months with no iron-chelation therapy (p < 0.03). In a second study, the French Groupe Francophone des Myelodysplasies identified a cohort of 170 MDS patients in which they also found an improved overall survival in patients who received iron chelation therapy (median survival: 115 vs 51 months; p < 0.0001) [48]. The results of these studies must be interpreted with caution, since patients were not allocated at random into the iron-chelation therapy and non-iron-chelation therapy groups.

In conclusion, iron overload, which results from the combination of increased iron absorption and transfusion-mediated iron acquisition, appears to play a major role in the increased morbidity and mortality of patients with MDS. The toxicity of iron is mediated by LPI via the formation of free radicals, which cause cellular and tissue damage to the heart, liver and endocrine organs and may promote development of acute leukemia. While longer follow-up is needed to define the long-term efficacy of oral iron chelators in MDS, iron-chelation therapy is effective at reducing total body iron and may also prolong survival in iron-overloaded MDS patients.

Financial & competing interests disclosure

Richard Wells has received honoraria from Novartis and Celgene. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.