HemobloginA₂ Levels Associate with Lower ESA-Dose in African-Americans with Sickle Cell Trait and End-Stage Kidney Disease

Introduction: We have previously demonstrated that African-Americans with end-stage kidney disease on hemodialysis receive higher doses of erythropoietin-stimulating agents (ESA) if they also have sickle cell trait (SCT). In sickle cell disease, the severity of clinical manifestations is modified by the quantity of hemoglobin S (HbS) expressed and attenuated by hemoglobin F (HbF) expression. Measures of hemoglobin A₂ (HbA₂) are reported to be higher in SCT although this may be related to testing artifact. We evaluated if ESA dosing in SCT patients on hemodialysis is associated with differences in the amount of these hemoglobins.

Methods: From a cohort of 155 African-Americans on hemodialysis, we collected data on the level of hemoglobin type (HbA, HbS, HbA₂, HbF) determined by cation-exchange high-pressure liquid chromatography (HPLC) studies. Only SCT individuals were included. Demographics and medical history were reviewed for covariates including age, sex, BMI, diabetes mellitus, history of gastrointestinal bleeding, vascular access type, time on dialysis and missed treatments. All laboratory studies were reviewed and averaged from a 3-month period for ferritin, iron saturation, parathyroid hormone, serum albumin, serum phosphorus, and adequacy studies (Kt/V). We calculated delivered ESA doses and intravenous iron doses as per-treatment dosing. We compared continuous measures of the amount of each hemoglobin type with ESA dose using SpearmanÕs correlation. We also constructed multivariable linear regression models using backward elimination to identify relevant confounders from a full model. Iron dosing and iron percent saturation were retained a priori.

Results: 24(15%) of the 155 patients had SCT. Of the individuals excluded, 9 had hemoglobin C trait and 1 had β-thalassemia trait and the rest had normal hemoglobin variants. Median age of the 24 subjects was 62.1 years (IQR 48.7, 69.2) and 13 were male (54.2%). None had measurable HbF. Other relevant clinical and demographic data are presented in Table 1. In unadjusted analyses, percent of HbS weakly correlated with average ESA dose but was not statistically significant (ρ= -0.27, 95% -0.61, 0.15; p=0.2). HbA₂ however had a relatively strong negative correlation with ESA dose that was statistically significant (ρ= -0.44, 95% CI -0.72, -0.05; p=0.03). After adjusting for age, time on dialysis, iron dose, iron percent saturation, albumin and parathyroid hormone, HbA₂ still demonstrated a statistically significant negative correlation with ESA dose (p=0.04). For each tenth of percent increase in HbA₂, ESA dose was lower by approximately 440 U.

Conclusions: African-American hemodialysis patients with SCT receive higher doses of ESA to achieve similar hemoglobin levels as other patients. Among these individuals, higher HbA₂ levels are associated with lower ESA dose even after adjustment for confounders. Although measures of HbA₂ may be less accurate in SCT, these results suggest that HbA₂ content may attenuate ESA resistance in this population, and further studies are needed to understand this relationship.