Predictors of erythropoietin responsiveness in chronic hemodialysis patients

The impact of dialysis intensity on erythropoietin (EPO) requirements is unclear. Previous work suggests that increased dialysis is associated with increased erythropoietin responsiveness (ERSP), but average dialysis intensity has increased since those publications. We hypothesized that ERSP would be independent of delivered Kt/V(urea) at current intensities of hemodialysis. We prospectively studied 135 stable chronic hemodialysis patients who receive iron and subcutaneous EPO dosed according to current guidelines. We collected biochemical, hematologic, and single pool urea kinetics data. ERSP was expressed as units per kilogram per week of EPO administered. Simple and multiple linear regression were used to identify characteristics predictive of ERSP. The mean age of the patients was 62 +/- 17 years (range, 17-90 years); 68 of 135 (50.3%) were women, and 120 of 135 (88.9%) were Caucasian. Mean delivered Kt/V(urea) was 1.60 +/- 0.49, with 102 of 135 (75.6%) of patients with a delivered Kt/V(urea) > 1.3. Univariate linear regression showed seven significant independent predictors of erythropoietin requirements. Low serum albumin (p < 0.001), low serum calcium (p = 0.002), high serum phosphate (p = 0.004), and high serum iPTH (p = 0.007) were all associated with lower levels of ERSP. Lower ERSP was also correlated with lower hemoglobin and lower serum iron and transferrin saturation. Delivered dialysis (Kt/ V(urea)) was not a significant predictor of ERSP (p = 0.61). Multivariate regression confirmed low serum albumin (p < 0.01), high serum phosphate (p = 0.001), high immunoreactive parathyroid hormone (p = 0.025), and low transferrin saturation (p < 0.0005) as predictors of low ERSP, and also found high serum ferritin to be correlated with low ERSP (p = 0.016). We found no relationship between erythropoietin responsiveness and intensity of hemodialysis in this population of patients with a mean delivered Kt/V(urea) of 1.6. This may indicate a threshold effect beyond which more dialysis will not improve ERSP. However, markers of an underlying inflammatory state and of secondary hyperparathyroidism were associated with decreased response to erythropoietin.