Analysis of ultrafiltration failure in peritoneal dialysis patients by means of standard peritoneal permeability analysis.

BACKGROUND: Ultrafiltration failure (UFF) is a complication of peritoneal dialysis (PD) treatment that occurs especially in long-term patients. Etiological factors include a large effective peritoneal surface area [measured as high mass transfer area coefficient (MTAC) of creatinine], a high effective lymphatic absorption rate (ELAR), a large residual volume, or combinations.
OBJECTIVE: The prevalence and etiology of UFF were studied and the contribution of transcellular water transport (TCWT) was analyzed. A new definition of UFF and guidelines for the analysis of its etiology were derived from the results.
SETTING: Peritoneal dialysis unit in the Academic Medical Center in Amsterdam.
DESIGN: Cross-sectional study of standard peritoneal permeability analyses (4-hr dwells, dextran 70 as volume marker) with 1.36% glucose in 68 PD patients. Patients with negative net UF (change in intraperitoneal volume, dIPV < 0 mL) were analyzed further using 3.86% glucose, whenever possible.
RESULTS: Among 68 patients (duration of PD 0.3-178 months), 39 had negative net UF with 1.36% glucose. These patients had greater MTAC creatinine and glucose absorption, and higher ELAR (p < 10(-4)) than the patients with positive UF. dIPV and transcapillary UF rate (TCUFR) were lower (p < 10(-5)). Twenty of these patients could be studied using 3.86% glucose. dIPV was greater than 400 mL/4 hr in this test in 12 patients, implying that no clinically important UFF was present. Ultrafiltration failure (dIPV < 400 mL) was found in 8 patients, giving a prevalence of 23%. This last group had been treated with PD for a longer period (p = 0.03), had higher ELAR (p = 0.07), but lower residual volume (p = 0.03), and lower TCUFR (p = 0.01). Ultrafiltration failure was associated with a high MTAC creatinine in 3 patients, a high ELAR in 4 patients, and a combination of factors in one. As an additional possible cause, TCWT was studied, using the sodium gradient in the first hour of the dwell, corrected for diffusion (dNA). Five patients had dNA > 5 mmol/L, indicating normal TCWT. The 3 patients with dNA < 5 mmol/L tended to be treated longer (p = 0.19) and had lower TCUFR (p = 0.04). A smaller difference was found between dIPV 3.86% and 1.36% (p = 0.04) compared to the dNA > 5 mmol/L group, but no differences were present for MTAC creatinine, ELAR, residual volume, or glucose absorption.
CONCLUSIONS: In addition to known factors, impairment of TCWT can be a cause of UFF. A standardized dwell with 1.36% glucose overestimates UFF. Therefore, 3.86% glucose should be used for identification of patients with UFF, especially because it provides additional information on TCWT. Ultrafiltration failure can be defined as net UF < 400 mL/4 hr with 3.86% glucose during a 4-hour exchange.