Hydraulically-induced convective solute transport across the rabbit peritoneum.

Transport of solutes during osmotically-induced transperitoneal ultrafiltration is less than would be predicted based upon rates of transperitoneal solute diffusion. Previous workers have hypothesized that osmotically-induced convective solute transport occurs only in small pores at the arteriolar end of peritoneal capillaries, whereas solute diffusion occurs only through large venular pores. We tested this heteroporosity hypothesis in the eviscerated New Zealand White rabbit by determining sieving coefficients (S) for creatinine, p-aminohippurate (PAH) and neutral dextran during hydraulically-induced transperitoneal ultrafiltration (N = 13). A hydraulically-induced driving force directs convective solute transport through the same capillary pores employed for diffusion; therefore S for all solutes should approach unity if the heteroporosity hypothesis is valid. S for creatinine and PAH were respectively 0.72 +/- 0.03 and 0.67 +/- 0.05, values lower than unity and not different from those previously determined during osmotically induced ultrafiltration. Mean S for dextran were relatively independent of molecular size, ranging from 0.50 at 13 A to 0.40 at 50 A. Thus, dextran S were higher than those previously determined during osmotically induced ultrafiltration yet still less than unity. Control experiments (N = 6) suggested that only surface area and not transport characteristics were altered by evisceration. These observations demonstrate that the heteroporosity hypothesis fails to completely describe both diffusive and convective transport properties of the peritoneum.