#3259 Intradialytic clearances and reduction ratios for middle molecular weight uremic toxins and protein-bound uremic toxins: comparison of hemodiafiltration with asymmetric cellulose triacetate vs hemodiafiltration with helicoidal polysulfone durin

Abstract Background and Aims Polysulfone (PS) is to date the most used and versatile material for hemodialysis membranes. The features of permeability and resistance of this material allow its use in hemodiafiltration (HDF) with few alternatives. Asymmetric cellulose triacetate (ATA) is a new dialysis membrane, suitable for HDF which represents a substitute in case of hypersensitivity to PS. The CHORALS study evaluates the intradialytic performances of ATA membrane compared to the widespread PS membrane in hemodiafiltration. The filter efficacy is evaluated through the removal of middle molecular weight toxins (MMUT) and protein-bound uremic toxins (PBUT). Method Twenty anuric HD patients with AVF and blood flow >300 ml/min were transitioned from HD to post-dilutional HDF. Patients were either equally assigned to ATA arm (filter Solacea 22H, Nipro) or to PS arm (FX100, Fresenius). Assigned convective volume was calculated on body surface area. Toxin depuration was focused on β2 Microglobulin (β2M), κ light Chains (Lκ) and Lambda light Chains (Lλ) for MMUT removal and P-Cresyl (PC) and Indoxil Sulphate (IS) for PBUT removal. Kinetics of removal were assessed at the first session after transition to HDF and after 6 months of follow-up. In both cases, we performed a continuous sampling of the spent dialysate to establish the concentration of toxin removed over a treatment of at least 210 minutes. Blood toxin determinations were corrected by hemoconcentration. Blood and spent dialysate samples, collected at hourly time points, were used to calculate intradialytic clearance (K) and reduction ratio (RR). Results In both sessions and for both filters K remains stable during treatment for all tested molecules except for B2M. Little to no differences were found in the comparison between sessions. Besides B2M K and Lκ K, which are significantly higher for PS membrane with good reproducibility, other toxins do not show relevant differences in their performance. In both dialysis sessions, the whole treatment RR for Lκ in PS was significantly higher than ATA but not for the other MMUT and PBUT (Table 1). During treatment, B2M instantaneous K varies significantly with a peak between the first and the third hour of treatment. Inversely, all other molecules show a very stable K over time for both tested filters, especially for Lκ. This reflects on the standard deviation of the all-time K distribution in which all measured K were depicted pooled (Fig. 1). Alongside MMUT, all-time IS K for PC is significantly higher than ATA, although this was not confirmed in both considered sessions. Besides these few differences in performance between filters, PBUT global RR remains unaffected in both sessions. Conclusion Considering that instantaneous total K does not significantly vary over time for the majority of tested toxins in both filters tested, we can exclude an impairment of the function of both filters tested. B2M clearance rises during treatment, especially for PC in which this trend is significant. This phenomenon, observed in both sessions, couldn't be related to adsorption as K is calculated on the dialysate side. Since K remains stable, hourly RR decays in response to the decrease of the toxin gradient over time between blood and dialysate. This pattern is well represented by all considered toxins, with the slope of the RR decay seemingly dependent on the K. ATA shows significantly lower B2M and Lκ clearance, but this doesn't affect B2M RR due to high variance. Besides differences in Lκ depuration, filter RR are equal.