Iron Overload and Haematopoiesis in MDS: Does Blood Transfusion Promote Progression to AML?

Introduction: The biology of myelodysplastic syndrome (MDS) is poorly understood, and treatment options are limited. Thus, most MDS patients require chronic red blood cell transfusion, and many develop secondary iron overload. Although the pathophysiological consequences of iron overload to the heart, liver, and endocrine organs have been well characterized, its effects on haematopoiesis have not been studied. However, it has been observed that chelation therapy in iron-overloaded MDS patients may result in reduction of transfusion requirements, and recent studies have suggested a correlation between the use of iron chelation therapy and improvement in leukaemia-free survival in MDS. At the cellular level, iron toxicity is mediated in large part via the generation of reactive oxygen species (ROS). It has been shown in animal models that accumulation of ROS leads to senescence of haematopoietic stem cells, and that ROS cause DNA damage and promote the development of malignancy. These effects of ROS may be particularly important in MDS, in which haematopoiesis is already severely compromised and genetic instability is a striking feature. Hypothesis: We hypothesize that iron overload secondary to transfusion leads to increased levels of intracellular ROS in early haematopoeitic cells in MDS. The increase in intracellular ROS in MDS would be predicted to lead further impairment of haematopoiesis via stem cell exhaustion and while promoting accumulation of DNA damage by myelodysplastic stem cells and early progenitors, thus accelerating progression of MDS to acute leukaemia. Results: To test this hypothesis, we examined the relationship between transfusion-related iron overload and ROS content of CD34+ bone marrow cells in MDS. ROS content was measured in CD34+ cells by flow cytometry in bone marrow aspirates from 34 consecutive MDS patients (CMML=4, MDS/MPD=2, RA=4, RARS=3, RCMD=2, RAEB 1=6, RAEB 2=12, RAEB-t/AML=1). The patients represented a wide range of prior transfusion burden (0->300 units PRBC) and serum ferritin levels (11->10000 μg/L). ROS was strongly correlated with serum ferritin concentration for patients with iron overload (serum ferritin >1000 μg/L; n=14, R=0.733, p<0.005). The correlation between ROS and ferritin level was even stronger in the subset of patients with RAEB 1 or RAEB 2 and iron overload (n=11, R=0.838, p<0.005). In contrast, no correlation between ROS and ferritin level was demonstrated for patients with serum ferritin <1000 μg/L (n=20). Importantly, iron chelation therapy was associated with a reduction in CD34+ cell ROS content in one patient. To assess the effect of iron overload on normal stem cell and progenitor function, we established a mouse model of subacute bone marrow iron overload. B6D2F1 mice were loaded with iron dextran by intraperitoneal injection (150mg total iron load over 21 days), and sacrificed three days after the end of iron loading. Iron staining of tissue sections confirmed iron deposition in the bone marrow, liver, and myocardium. The development of splenomegaly was noted in iron-loaded animals. Flow cytometric analysis revealed increased apoptosis of bone marrow cells in iron loaded mice based on annexin V+/7 AAD-staining (6.26±0.96% versus 3.54±0.99% for control mice, paired student’s t-Test p<0.005). However, ROS content in CD117+ progenitors of iron loaded mice was similar to control mice. Thus, subacute iron loading in mice increases apoptosis but does not alter the ROS content of HSCs; we postulate that chronic iron overload is required to achieve this effect. Conclusions: These results establish a relationship between CD34+ cell ROS content and serum ferritin concentration in MDS patients with iron overload, and indicate that iron chelation therapy in this patient population reverses this ROS accumulation. The physiological consequences of this relationship are currently being investigated in this patient set by haematopoietic colony assays and assessment of DNA damage in CD34+ cells. Nonethelesss, these data may have key implications for the deployment of iron chelation therapy in MDS patients, and may explain the association between the use of iron chelation and improved leukaemia-free survival in MDS.