ERYTHROPOETIN RESISTANCE IN PRE-DIALYSIS PERIOD AND ITS RELATION TO PROGNOSIS IN JAPANESE INCIDENT DIALYSIS PATIENTS

Abstract Background and Aims Erythropoiesis stimulating agents (ESA) are essential for better quality of life and longer life expectancy in end-stage kidney disease patients (ESKD). Resistance to ESA leads to worse prognosis in hemodialysis and peritoneal dialysis patients. Resistance to ESA exists in pre-dialysis period. We studied the clinical significance of pre-dialysis resistance to ESA by investigating the erythropoietin resistance index (ERI) in patients with ESKD prior to dialysis induction and its association with mortality and cause of death in three-year follow-up after dialysis initiation. Method Subjects included 1,420 patients in 17 centers participating in Aichi Cohort Study of Prognosis in Patients Newly Initiated into Dialysis (AICOPP) from Oct. 2011 to Sep. 2013. ESA responsiveness was estimated by using erythropoietin resistance index (ERI) (U/kg/week/g/dL), which was calculated as weekly weight-adjusted epoetin dose divided by the hemoglobin (Hb) concentration before starting dialysis. We excluded 302 patients because ERI was not available. Thus, 1,118 subjects were enrolled. Subjects were divided into four categories of ERI; ERI < 5, 5≤ ERI <10, 10≤ ERI <15 and 15≤ ERI. Cumulative survival rates from all-cause, cardiovascular disease (CVD), infection and malignancy were compared by Kaplan-Meier method according to ERI categories. Three-year mortality was analyzed by logistic regression analysis adjusted by age, gender, Charlson comorbidity index (CCI), and ERI categories. Results During three-year follow-up period, 191 (17.1%) patients died. Age and female gender share were higher in proportion to ERI. Diabetes, coronary artery disease (CAD), and values of albumin (Alb) and brain natriuretic peptide (BNP) were not correlated with ERI (Table 1). Kaplan-Meier survival curve showed that patients with the highest ERI (15≤ ERI) had a higher all-cause mortality (Log rank test p=0.002, Figure 1). Concerning the cause of death, mortality from infection (p=0.023, Figure 2) and malignancy (p=0.031) were high in the highest ERI category, however there was no significant difference in each ERI category regarding CVD mortality (p=0.33). According to multivariate analysis analyses by logistic regression model, all-cause (odd ratio [OD]: 1.65, 95% confidence interval [CI]: 1.14-2.38, p=0.008) and infection-related mortality (OR: 2.13, 95% CI: 1.08-4.11, p=0.029) were significantly correlated with the highest ERI category, but unexpectedly CVD mortality was not (OR: 1.04, 95% CI: 0.59-1.75, p=0.90). Mortality from malignancy did not reach the significant level (OR: 1.92, 95% CI: 0.96-3.73, p=0.063) and may require longer observation period for analysis. Conclusion Pre-dialysis ERI is a prognostic marker in Japanese dialysis patients and predicts the susceptibility to infection.