Comparison of the radioiodinated serum albumin (RISA) dilution technique with direct volumetric measurements in animal models of peritoneal dialysis

Background: Rat models of peritoneal dialysis (PD) are useful for studying the physiology of peritoneal transport and evaluating new osmotic agents. Intraperitoneal (IP) solute concentrations and their evolution over time are easy to measure, but IP volume (IPV) is not. Direct volumetric measurements are the "gold standard;' but they are expensive and do not allow for repetitive measurements in the same animal. The indicator dilution technique is therefore used as an alternative. However, that technique is based on assumptions that are not always valid. The present study compares direct volume measurement with the indicator dilution technique [radioiodinated serum albumin (RISA)] to determine the IPV over time curves in a rat model of PD. Methods: In series 1, 17 Sprague-Dawley rats were instilled IP with 25 mL 1.36% glucose dialysate through a Teflon catheter. In 9 animals, 0.35 mL dialysate was sampled and discarded at time points 0, 3,15, 30, 60,180, and 240 minutes. In the other 8 animals, no sampling was performed. At 240 minutes, all 12 animals were humanely killed, and direct volumetric measurements of IPV were performed. In series 2, rats were instilled IP with 25 mL 1.36% glucose dialysate containing 18.5 kBq I-131 RISA. In 9 animals, dialysate was sampled at 0, 3, 15, 30, 60, 90, 120,180, and 240 minutes for the construction of the RISA concentration-over-time curve, and to calculate the elimination constant K-e. At 30, 60, 180, and 240 minutes, dialysate was sampled in 6 different animals (total n = 24) to calculate IPV using the RISA dilution technique. Immediately afterward, the animals were humanely killed, and direct volumetric measurements of IPV were performed. Results: In series 1, after 240 minutes' dwell time, the IPV was lower in the sampled animals as compared with the non sampled animals (27.11 +/- 1.85 mL vs 30.75 +/- 0.59 mL, p = 0.001). In series 2, the evolution of RISA activity in the dialysate over time was described by piece-wise linear regression, yielding 3288 - 8.2T counts (cts) for T < 52.72 minutes and 2973 - 1.99T counts for T > 52.72 minutes. The IPV was better predicted with a K-e that took into account the disappearance of RISA by sampling than with a K-e that took into account disappearance of RISA only by absorption. Conclusions: If indicator dilution techniques are used to measure IPV, attention must be paid to the disappearance of the osmotic agent and the marker by multiple sampling. The best way to meet that goal is to use micropipettes to avoid large sample volumes.