Background. Glucose degradation products (GDP) are generated in peritoneal dialysis (PD) fluid during heat sterilization and storage. They are thought to adversely affect the peritoneal membrane. The fate of GDP within the peritoneal cavity has not been well characterized.

Methods. A clinical study was designed to determine (1 ) whether during the dwell in the peritoneal cavity GDP concentration decreases in the PD fluid as assessed by ex vivo formation of AGE; (2) whether exposure to GDP-containing PD fluids increases plasma fluorescence (as an index of plasma AGE concentration) as well as plasma carboxymethyllysine (CML) concentration; and (3) whether exposure to GDP-containing PD fluids adversely affects glycoprotein CA 125 concentration. A two-group crossover design was adopted comprising two consecutive observation periods of eight weeks each. Stable PD patients were exposed in random order either to conventional PD fluid (heat sterilized at pH 5.5) and subsequently to PD test fluid (or the 2 fluids in reverse order). The PD test fluid was sterilized using a multicompartment bag system separating highly concentrated glucose at pH 3 from the buffer solution. Conventional and test fluids differed with respect to concentrations of GDP, that is, 3-deoxyglucosone (118 vs. 12.3 mumol/L), methylglyoxal (5.3 mumol/L vs. below detection threshold), 3, 4-dideoxyglucosone-3-ene (10 mumol/L vs. below detection threshold) and acetaldehyde (226 vs.

Results. The following results were obtained. First, methylglyoxal disappeared completely as early as two hours after intraperitoneal instillation of conventional PD fluid. Second, when spent conventional dialysate was recovered after a two hour and particularly an eight hour dwell and subsequently incubated ex vivo with 40 mg of human serum albumin, there was a continuous decrease of AGE-forming capacity, that is, less generation of fluorescence (AGE) and pyrraline (non-fluorescent Amadori product), and an increase of advanced oxidation protein products (AOPP) in the spent dialysate. Third, plasma fluorescence (exc. 350/em. 430 nm) as an index of circulating AGE compounds as well as plasma CML concentrations were significantly higher in the conventional PD fluid period versus low GDP PD fluid period. Fourth, CA 125 concentrations in spent dialysate were higher during the low GDP PD fluid period compared to the conventional PD fluid period.

Conclusions. Conventional PD fluid undergoes modifications during intraperitoneal dwell with a loss of AGE forming capacity, suggesting breakdown, precipitation or resorption of GDP in vivo. This is accompanied by an increase in plasma AGE compounds.