Validation by computer simulation of two indirect methods for quantification of free water transport in peritoneal...

In peritoneal dialysis, approximately 40% of the total osmotic ultrafiltration (UF) induced by glucose can be predicted to be due to "free" water transport across aquaporin-1 (APQ-1). Theoretically, it would be possible to assess the fraction of free water transport in the early phase of a hypertonic dwell, when UF rate is high and the relative contribution of Na+ diffusion is low. La Milia et al. [La Milia V. et al. Fast-fast peritoneal equilibration test (FAST-FAST-PET): a simple method for peritoneal hydraulic permeability study [Abstract]. Nephrol Dial Transplant 2002; 17 (Suppl 1):17-18] suggested a technique to assess sodium-associated water transport based on sodium removal (Na+R) divided by the plasma Na+ concentration during a "fast-fast" (60 minute) peritoneal equilibration test (PET) for 3.86% glucose, yielding an estimate of the UF passing through the small pores (UFSP). Free water transport (UF through ultrasmall pores; UFUSP) was obtained by subtracting UFSP from total UF. Although peritoneal Na+ transport is almost totally convective, this technique will slightly overestimate small-pore UF due to the presence of some small-pore Na+ diffusion from the circulation during the dwell. A way of dealing with this problem was presented recently by Smit (Smit W. et al. Quantification of free water transport in peritoneal dialysis. Kidney Int 2004; 66:849-854).

In the present study we used the three-pore model of peritoneal transport to predict the degree of overestimation of UFSP for the technique presented by La Milia et al., and any potential deviations from theory for the technique presented by Smit et at. Simulations were performed under ordinary conditions and during simulated UF failure for 3.86% glucose. The fractional UF coefficient accounted for by APQ-1 was set at 2%.

Estimating the UFSP from the sodium-associated water transport according to the method by La Milia et al. consistently overestimated UFSP and underestimated UFUSP. These errors were, however, minimal for dwells lasting between 30 and 80 minutes. The technique by Smit et al. to calculate aquaporin-mediated water flow (UFUSP), using an elaborate correction for Na+ diffusion from the circulation during the dwell, seemed accurate in most situations but, in general, tended to moderately overestimate UFUSP at early dwell times (<30 minutes) and underestimate UFUSP at long dwell times (4 hours).

The technique presented by La Milia et at. to calculate free water transport during a fast-fast PET was found to be surprisingly accurate, although the procedure would further improve by the introduction of a correction algorithm. The technique by Smit is even more accurate for dwells up to 4 hours' duration. However, since the Smit technique is elaborate, it is less practical for routine determinations of aquaporin-mediated water transport in peritoneal dialysis.