Kiyomi Koike, Kei Fukami, Atsushi Kawaguchi, Kazumasa Shimamatsu, Sho-ichi Yamagishi, Seiya Okuda

Higher doses of erythropoiesis-stimulating agents (ESAs) contribute to atherothrombotic cardiovascular disease in hemodialysis (HD) patients. Thrombocytosis is associated with increased mortality in ESA-treated HD patients. We investigated variables affecting platelet count and its variability (platelet count increment [Δplatelet count]) in HD patients. This retrospective longitudinal and observational study of HD outpatients was carried out over 3 years. The outcome was independent determinants of platelet count and Δplatelet count, which were associated with iron indices, ESA dose, and C-reactive protein. In univariate regression analysis, V-shaped relationship was observed between platelet count and transferrin saturation (TSAT), ferritin, serum iron, and hemoglobin (Hb) with the bottom of 0.21, 330 ng/mL, 49 µg/dL, and 10.3 g/dL, respectively. Mixed-effect multivariate regression analysis revealed that TSAT (inversely), Hb ≤10.3 g/dL (inversely), C-reactive protein, and ESA dose were independently associated with platelet count. Δplatelet count was independently and inversely correlated with ΔTSAT and directly correlated with Δferritin. TSAT was independently and inversely associated with ESA dose. ESA dose was directly correlated with iron dose and inversely correlated with TSAT, ferritin ≤330 ng/mL, and Hb ≤10.3 g/dL. ESA dose and TSAT were correlated in determining platelet count and Δplatelet count. Targets of iron indices that reflect iron supply sufficient to avoid platelet count increment and variability may be >21% of TSAT and 300 ng/mL of serum ferritin for appropriate ESA therapy in HD patients.