Is there material hazard to treatment with intravenous iron?

We have been invited to write this article, exploring While there are several factors that contribute to the question of whether we expose our patients to suboptimal haematocrit values, the USRDS study material hazard when we treat them with intravenous noted above demonstrates the important role of iron iron. It is a reasonable question to ask with any deficiency. In fact, this study found that in 1993, 54% therapeutic agent, especially after a period in which of America haemodialysis patients were iron deficient the use of the agent has increased significantly. The as measured by transferrin saturation (TSAT ) <20%. answer to the question, to a degree, is dependent on Remarkably, fully 25% of patients had absolute iron the intended definition of the word material. We shall starvation, with TSAT <10% [7]! This is a level of seek to utilize a working definition, in which material iron deficiency at which the body’s ability to produce hazard is that which would outweigh any anticipated energy is markedly compromised. benefits that the treatment might confer. As such, we An interesting 10-year historical perspective on iron believe that there is not material hazard in the use of treatment in haemodialysis patients is emerging. In intravenous iron. Before exploring the risks of iron 1989, recombinant human erythropoietin (rHuEpo) therapy, we will begin with a brief consideration of the became clinically available in the US. This was a time benefits of intravenous iron for improving the anaemia in which iron overload was more the concern of therapy of haemodialysis patients. nephrologists than iron deficiency. Yet only 5 years Anaemia is one of the major causes of morbidity, later the aforementioned USRDS study demonstrated mortality and diminished quality of life for dialysis the emergence of an epidemic of iron deficiency. Now, patients [1–3]. Churchill and Collins have each per5 years later still, 10 years after the introduction of formed analyses that associated lower haemoglobin rHuEpo, the pendulum has swung back and iron levels with increased subsequent hospital days [1,4]. overload is again becoming a concern. The Core In another analysis by Collins, a haematocrit value Indicators Report noted a mean serum ferritin of lower than 33% was associated with an increased risk 505 ng/ml in the US in 1997 [8]. It is therefore a very of death [3]. The EPIBACDIAL multicentre study, reasonable time to reconsider the weighing of risk and found low haemoglobin concentrations to be a major reward in iron treatment. predictor of the development of bacteraemia [5]. A When haemodialysis patients are treated with synthesis of a broad literature regarding anaemia in rHuEpo, iron deficiency usually develops. The choices dialysis patients led the American NKF-DOQI anaeto supplement iron include several types of orally mia guideline development team to recommend that administered iron, or intravenous iron. Oral iron has haematocrit be maintained in the 33–36% range [6 ]. limited efficacy, and frequently causes gastrointestinal If this level is achieved, then the patient should live side effects. In contrast, the efficacy of intravenous longer, and with a better quality of life. iron for improving anaemia therapy tends to be outThe recommended haematocrit range stands in disstanding. A large series of studies have documented turbing contrast to the results actually achieved in that intravenous iron treatment leads to a significant haemodialysis patients. As recently as 1993, the improvement in rHuEpo responsiveness [9–17]. There USRDS Wave 1 Morbidity and Mortality Study found are several safety issues, however, that may lead to a that 48% of American haemodialysis patients had degree of risk to patients. I will now discuss these haematocrit values of <30% [7]. It is clear that if we issues, and consider how safety may affect treatment. seek to improve patient outcomes by raising haematoIn the US, iron dextran has been the only form of crits into the NKF-DOQI recommended range of intravenous iron available until recently. This agent is 33–36%, then we still have a long way to go. The 1998 associated with occasional severe anaphylactoid type ESRD Core Indicators Report documents that 44% of reactions. Hamstra et al. in a study of nonuraemic American haemodialysis patients had haematocrit subjects, found the rate of these reactions to be 0.6% values below this range in 1997 [8]. of patients treated [18]. Our group studied haemodialysis patients, and found the reactions to occur in 0.7% Correspondence and offprint requests to: Steven Fishbane MD, 222 Station Plaza North, Suite 510, Mineola, NY 11501, USA. of patients treated [19]. Because these types of reac-