FP655EFFECTS OF DIALYZER MEMBRANE MATERIALS ON SURVIVAL IN CHRONIC HEMODIALYSIS PATIENTS: A 2-YEAR COHORT STUDY FROM THE ANNUAL SURVEY OF THE JAPANESE RENAL DATA REGISTRY

Abstract

Introduction and Aims: The Japanese Society for Dialysis Therapy (JSDT) guidelines recommend the use of high-performance membrane (HPM) dialyzers, including high-flux membrane dialyzers because they have the potential to improve patients' prognosis and reduce dialysis-related complications, and are thus considered for use in dialysis therapy. There are seven types of HPM dialyzers in Japan: cellulose triacetate (CTA), ethylene vinyl alcohol (EVAL), polyacrylonitrile (PAN), polyester polymer alloy (PEPA), polyethersulfone (PES), polymethylmethacrylate (PMMA), and polysulfone (PS). However, there is little information available regarding which type of dialyzer results in good prognosis. Therefore, we conducted a cohort study from a nationwide registry of hemodialysis (HD) patients in Japan to clarify the association between different dialyzers and 2-year mortality rate. Methods: Weconducted acohort study using data fromanationwideregistryofthe JSDT.Subjects were 136,676 patientsonmaintenanceHD between 2009 and 2011who underwent maintenance HD for at least 2 years and were treated with one of the following seven typesof HPM dialyzers: CTA, EVAL, PAN, PEPA, PES, PMMA, and PS. Survival analyses with Cox proportional hazards regression were used to examine whether baseline basic factors including age, sex, dialysis duration, and primary kidney disease and cardiovascular comorbidity predicted survival for up to 2 years of follow-up. Additional analyses were performed adjusted for dialysis dose and b2-microglobulin. Furthermore, additional analyses were performed with adjustment for nutrition-and inflammation-related factors, including body mass index, serum levels of C-reactive protein, hemoglobin, albumin, normalized protein catabolic rate, and % creatinine generation rate. Subsequently, to reduce potential confounding and treatment selection bias, weadjusted the significant difference in baseline covariates with the use of propensity score matching. Results: Data were adjusted using basic factors, with PS as a reference group, and the hazard ratio (HR) was significantly higher inCTA, PMMA, PAN, and EVAL groups. Further data adjustment for KT/V and b2-microglobulin yielded the same results as were obtained from data adjusted for basic factors. After further adjustment for nutrition-and inflammation-related factors, HR was significantly lowered for the PES and PMMA groups compared with the PS group (HR, 0.88; 95% confidence interval [CI], 0.82-0.94 and HR, 0.84 CI; 0.76-0.93, respectively). After propensity score matching, the HRs for the PES group (0.90 [0.85-0.96]) and PMMA group (0.87 [0.78-0.96]) were significantly lower than that for the PS group (P < 0.01). Conclusions: The useofdifferent membrane typesmay affect mortality. However, further long-term prospective studies are needed to clarify whether the PES and PMMA membranes can improve prognosis.