Is platelet-activating factor produced during hemodialysis with AN-69 polyacrylonitrile membrane?

Background: Platelet-activating factor (PAF) production during hemodialysis (HD) with cuprophane (CU) membrane has previously been demonstrated, while the results regarding PAF production during HD with AN-69 polyacrylonitrile membrane are dubious. In this study an attempt is made to show that PAF is produced during HD using AN-69 membrane while comparing this production with the corresponding one from HD with CU. Since previous studies have indicated that PAF, like the complement system, could also be implicated in HD-related leukopenia and thrombocytopenia (especially when CU membrane is used), the circulating leukocyte and platelet counts as well as the C3a-desArg and SC5b-9 (soluble, nonlytic form of the terminal complement complex) levels were measured. Methods: Ten hemodialyzed patients were subjected to HD with CU and AN-69 membranes for 2 consecutive weeks (first week with CU and second with AN-69). During the third HD session of each week and at different times (0, 2, 5, 15, 30, 60, 180 and 240 min), the PAF levels in the blood as well as the leukocyte and platelet counts were measured, while the circulating levels of the C3a-desArg and SC5b-9 were measured at 0, 5, 15, 60 and 240 min. PAF was detected by ethanol extraction, followed by purification by column chromatography and high-pressure liquid chromatography and finally quantified by bioassay. The C3a-desArg and SC5b-9 fractions of the complement were measured by immunoassay while an autoanalyzer gave the leukocyte and platelet counts. Results: Circulating PAF levels were detected at all time intervals during HD with AN-69 (PAF(AN-69)) and CU (PAF(CU)) membranes. At all time intervals PAF(AN-69) < PAF(CU), however, statistically significant differences (s) between the two membranes existed only at 15, 30, 60, 180 and 240 min. The highest PAF(AN-69) and PAF(CU) occurred at 5 and 15 min into dialysis, respectively. The same observations were made for circulating C3a-desArg levels (s existed additionally at 5 min as well). The reduction of the circulating leukocytes had almost a mirror image with the C3a-desArg as well as PAF levels while the maximal reduction of platelets was observed after 2 min into dialysis with both membranes (i.e., simultaneously with the first increase in PAF secretion). Conclusions: PAF is indeed produced during HD with AN-69 membrane, as it is during HD with CU. At all time intervals during the HD procedure, PAF(AN-69) <PAF(CU). PAF seems to contribute to HD-related leukopenia and thrombocytopenia with both membranes. Copyright (C) 2002 S. Karger AG, Basel.