Characteristics of Convective Transport of Glucose During Peritoneal Dialysis

Lee, Eun Young; Yoon, Kyun Il; Park, Min Sun; Suh, Kyo Il; Lee, Hi Bahl

Abstract

The sieving coefficient(S) representing convective transport of glucose during peritoneal dialysis(PD) with glucose containing dialysis solution has been reported to be anomalous, lower than 0 or higher than 1. During peritoneal dialysis using glucose containing dialysis solution, diffusive transport of glucose is from dialysate to blood, and convective transport in the opposite direction i.e., from blood to dialysate. Glucose intolerance and hyperinsulinemia are well known adverse effects of PD using glucose containing dialysis solutions. Insulin is required for glucose transport from extracelluar fluid to intracelluar fluid in adipocytes and muscell cells. Hyperinsulinemia in PD may alter peritoneal glucose transport. If extra to intracellular glucose transport mediated by insulin is involved in the peritoneal glucose transport during PD with conventional glucose containing dialysis solutions, the diffusive and convective transport characteristics for glucose calculated using membrane model between two well-mixed compartments may not represent true values. S can be calculated best when diffusion is minimized. Male Sprague-Dawley rats were used. To minimize the diffusive transport the glucose isochratic solutions containing approximately the same concentration as in serum were used. To maximize ultrafiltration 3.86% mannitol was used as an osmotic agent. To evaluate the effect of insulin on glucose transport two different glucose concentrations, 100mg/dl(NI) and 300mg/dl(HI), were used. During the dialysis with HI solution glucose clamp technique was performed to keep blood glucose level approximately 300mg/dl. A 2 hour peritoneal dialysis was performed in 13 rats(7 Nl and 6 Hl). Serum and dialysate insulin levels were measured in 3rats in Nl, 2 rats in Hl, and 4 rats without dialysis(NC). Intraperitoneal volume(VD) was calculated using volume marker, RISA, dilution method. The diffusive mass transport coefficient(KBD) and S for urea and glucose were calculated using the modified Babb- Randerson-Farrell model. D/P glucose in Nl was 0.61+/-0.05 due to high blood glucose level 187.2+/-17.9mg/dl vs. 114.3+/-7.6 mg/dl in dialysate and 0.99+/-0.26 in Hl(360.6+/-55.6mg/dl in blood vs. 345.0+/-55.6mg/dl in dialysate). VD did not differ between the two groups. KBD for urea and glucose, and S for urea did not differ between the two groups. S for glucose in Hl was negative value and significantly lower than that in Nl(-0.903+/-0.960 vs. 1.036+/-0.137, P<0.001). Plasma insulin level was significantly higher in Hl compared with values in Nl and NC. Dialysate insulin level was similar in Nl and Hl. Dialysate insulin level in Nl was higher than plasma insulin level. The present result that S for glucose at hyperinsulinemic condition was anomalous indicates that not only simple passive transport but also other transport mechanisms mediated by insulin such as glucose influx into cells may be involved in peritoneal glucose transport. The finding of dialysate insulin level higher than plasma concentration in Nl may suggest direct leakage of insulin from pancreas or portal vein into the peritoneal cavity.