Predicting the Peritoneal Absorption of Icodextrin in Rats and Humans Including the Effect of α–Amylase Activity in Dialysate

Abstract

Contrary to ultrafiltration, the three-pore model predictions of icodextrin absorption from the peritoneal cavity have not yet been reported likely, in part, due to difficulties in estimating the degradation of glucose-polymer chains by α-amylase activity in dialysate. We incorporated this degradation process in a modified three-pore model of peritoneal transport to predict ultrafiltration and icodextrin absorption simultaneously in rats and humans. Separate three-pore models were constructed for humans and rats. The model for humans was adapted from PD Adequest 2.0 including a clearance term out of the peritoneal cavity to account for the absorption of large molecules to the peritoneal tissues, and considering patients who routinely used icodextrin by establishing steady-state plasma concentrations. The model for rats employed a standard three-pore model in which human kinetic parameters were scaled for a rat based on differences in body weight. Both models described the icodextrin molecular weight (MW) distribution as five distinct MW fractions. First order kinetics was applied using degradation rate constants obtained from previous in-vitro measurements using gel permeation chromatography. Ultrafiltration and absorption were predicted during a 4-hour exchange in rats, and 9 and 14-hour exchanges in humans with slow to fast transport characteristics with and without the effect of amylase activity. In rats, the icodextrin MW profile shifted towards the low MW fractions due to complete disappearance of the MW fractions greater than 27.5 kDa. Including the effect of amylase activity (60 U/L) resulted in 21.1% increase in ultrafiltration (UF) (7.6 mL vs 6.0 mL) and 7.1% increase in icodextrin absorption (CHO) (62.5% with vs 58.1%). In humans, the shift in MW profile was less pronounced. The fast transport (H) patient absorbed more icodextrin than the slow transport (L) patient during both 14-hour (H: 47.9% vs L: 40.2%) and 9-hour (H: 37.4% vs L: 31.7%) exchanges. While the UF was higher during the longer exchange, it varied modestly among the patient types (14-hour range: 460 - 509 mL vs 9-hour range: 382 - 389 mL). When averaged over all patients, the increases in UF and CHO during the 14-hour exchange due to amylase activity (7 U/L) were 15% and 1.5%, respectively. The icodextrin absorption values predicted by the model agreed with those measured in rats and humans to accurately show the increased absorption in rats. Also, the model confirmed the previous suggestions by predicting an increase in UF specific to amylase activity in dialysate, likely due to the added osmolality by the small molecules generated as a result of the degradation process. As expected, this increase was more pronounced in rats than in humans because of higher dialysate concentrations of amylase in rats.