Evidence against hepatitis C virus trapping in dialysis membranes

Sir, Haemodialysis (HD) patients have a high prevalence of hepatitis C virus (HCV) infection due to blood transfusions or nosocomial transmission w1,2x; however, the clinical course of HCV-associated liver disease in HD patients is usually mild w3,4x. Recent studies have described a consistent reduction of plasma HCV-RNA levels after dialysis w5–7x, independently of the membranes and anticoagulants used, suggesting that HD per se could protect patients from an aggressive course of HCV infection. The mechanism by which HD induces the reduction of viraemia remains largely speculative. Two major mechanisms may be involved: (i) a mechanical or physicochemical phenomenon linked to the dialysis technique and (ii) biological changes induced by the host response. The possible passage of viral particles across the membrane into the dialysate has been investigated previously in our and other laboratories w7,8x. This probably occurs only exceptionally because of the size difference between viral particles and the membrane pores. Only the presence of microlesions in some hollow fibres may allow the transfer of viral particles from blood to dialysate. However, this should involve only few virions and is therefore probably not the cause of the massive reduction of viraemia observed during the dialysis session. Okuda et al. w5x suggested trapping of HCV on the membrane surface during dialysis. To address this question, we selected 15 non-uraemic patients with HCV-related chronic hepatitis submitted to therapeutic blood letting and used ;400 ml of blood from each patient for a series of 15 haemodialysis procedures in vitro. Blood was collected in bags with Na2 EDTA and immediately used in an in vitro dialysis session, in a closed loop, with the following schedule: blood flow 300 mlumin; dialysate flow 500 mlumin; temperature 378C; ionic composition (mmolul) Na 138, K 3.0, Cl 110, NaHCO3 31, CH3COONa 5.0, Ca 2q 2.0 and Mg 0.75, with a theoretical osmolality of 290.25 mOsmul. No haemofiltration was planned and spontaneously occurring haemofiltration was restored by saline infusion, thereby maintaining a constant blood volume of ;400 ml. Membranes employed were ethylenevinyl alcohol (EVAL) in seven experiments, Gambrane (four experiments), Hemophan (two experiments), polyacrylonitrile (PAN-AN69) and polymethylmetacrylate (PMMA) in one experiment. A 2 ml aliquot of blood was collected at baseline and every 30 min for 300 min for the determination of HCV-RNA. The latter was performed using a commercial assay based on competitive amplification with an internal standard (Amplicor HCV Monitor TM, Roche Diagnostic System, Branchburg, NJ, USA). Table 1 shows the results of HCV viraemia findings from the 15 experiments performed. Only slight reductions of viraemia were seen (PsNS). In another five experiments, all performed with an EVAL 1.2 m membrane, blood was submitted for 300 min to haemofiltration without dialysate (100 ml of ultrafiltrateuh,