Abstract
BACKGROUND AND OBJECTIVES: Peritoneal protein clearance (Pcl) is determined by both effective (small pores) membrane area and relative capillary leakiness (large pores). It is not known how these two components change with duration of peritoneal dialysis (PD) in the context of progressive membrane injury and differential attrition of patients with higher Pcl, which has been associated with increased mortality risk in several studies. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: Patients treated continuously from 2000 to 2011 for a minimum of 4 years were selected from the longitudinal prospective Stoke PD Study. Pcl, membrane area (peritoneal solute transport rate [PSTR]), dialysis prescription, and residual renal function were measured every 6 months, along with comorbidity and peritonitis events. Multilevel multivariate analysis was used to determine associations with Pcl over time, taking into account within-subject correlations. RESULTS: From 280 incident patients, 335 datasets were analyzed from 49 patients receiving treatment for 4 years. Pcl correlated with PSTR at baseline (R=0.61; P<0.01), but over time there was progressive uncoupling of this relationship (year 4, R=0.28; P=0.05) with increasing PSTR (0.66-0.74; P<0.01) and stable Pcl (78.4-81.9 ml/d; P=0.7). Multivariate analysis found that age, PSTR, daily ultrafiltration, and sodium removal were significant predictors of Pcl when adjusted for sex, comorbidity, glucose exposure, and residual renal function. Peritonitis was associated with increased PSTR but a similar pattern of uncoupling. CONCLUSION: There is a progressive dissociation of the small- and large-pore pathways with time on PD, which would be in keeping with a switch from local inflammation early on to progressive fibrosis, combined with increased vascular surface area. Measuring longitudinal changes in Pcl may complement membrane function tests used to monitor progressive injury.