The impact of iron supplementation for treating anemia in patients with chronic kidney disease: Results from pairwise and network meta-analyses of randomized controlled trials

Abstract

After relative erythropoietin deficiency, iron deficiency is the second most important contributing factor for anemia in chronic kidney disease (CKD) patients. Iron supplementation is a crucial part of the treatment of anemia in CKD patients, and intravenous (IV) iron supplementation is considered to be superior to per os (PO) iron supplementation. The differences between the available formulations are poorly characterized. This report presents results from pairwise and network meta-analyses carried out after a comprehensive search in sources of published and unpublished studies, according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) recommendations (International prospective register of systematic reviews PROSPERO reference ID: CRD42020148155). Meta-analytic calculations were performed for the outcome of non-response to iron supplementation (i.e., hemoglobin (Hgb) increase of <0.5–1.0 g/dL, or initiation/intensification of erythropoiesis-stimulating agent (ESA) therapy, or increase/change of iron supplement, or requirements of blood transfusion). A total of 34 randomized controlled trials (RCT) were identified, providing numerical data for analyses covering 93.7% (n = 10.097) of the total study population. At the network level, iron supplementation seems to have a more protective effect against the outcome of non-response before the start of dialysis than once dialysis is initiated, and some preparations seem to be more potent (e.g., ferumoxytol, ferric carboxymaltose), compared to the rest of iron supplements assessed (surface under the cumulative ranking area (SUCRA) > 0.8). This study provides parameters for adequately following-up patients requiring iron supplementation, by presenting the most performing preparations, and, indirectly, by making it possible to identify good responders among all patients treated with these medicines.