The sorbent cartridge could bind and retain IL-6 and/or p-cresol and its metabolites is consistent with the chemical-physical characteristics of its styrene resin that can establish strong non-covalent interactions with hydrophobic molecules and with albumin. A relevant question arises about the forms in which either pcresol or IL-6 bind to the resin. Both these molecules circulate, indeed, in the plasma both as protein-bound and as free forms. Specifically, p-Cresol and its derivatives are more than 90% bound to albumin whereas almost 70% of IL-6 circulates as a very high molecular complex with its soluble receptors. The filtering membrane of the convection stage of the HFR Supra apparatus has a low but measurable permeability to albumin with an albumin sieving coefficient of 0.02. Therefore, it lets an amount of albumin ranging around 1/40th of its plasma concentration cross the membrane and appear into the UF. This implies that uremic toxins, like p-cresol, can be transferred in the UF also in their protein bound forms. In addition, the free forms of p-cresol and of its metabolites are small enough to freely permeate through the convection stage filter. Therefore, both protein-bound and free p-cresol presumably passed through the dialysis membrane into the UF and become available for binding to the cartridge. Conversely, because of the very high molecular weight of its complex with soluble IL-6 receptors, it is likely that IL-6 passed into the UF only as a free molecule. This is consistent with the evidence that IL-6 concentration in the UF averaged 1/5th of the plasma concentration, the value expected for free IL-6 crossing through a membrane with IL-6 sieving coefficient of about 0.3. The evidence that we found low or undetectable IL-6 concentrations in the effluent from the cartridge suggests that the IL-6 that crossed the dialysis membrane of the first HFR stage to move into the UF was virtually all retained by the cartridge itself. In keeping with this result is the evidence that the ability of the UF incubation to trigger IL-6 gene expression and release in PBMC from healthy subjects was greatly reduced after its passage through the HFR cartridge sorbent bed. While the evidence discussed above suggests that HFR cartridge contributed to clear the plasma of our patients from uremic toxins, a role could have been played also by the second, diffusive stage of the HFR apparatus. As mentioned above, this portion of the HFR system operates as a conventional HD system in which diffusion represents the main mechanism of solute removal. However, the removal of p-cresol by this mechanism is expected to be small and that of IL-6 null. Indeed, less than 1% of total p-cresol circulates as free form whereas free IL-6 is still too large to cross the low flux membrane of the second HFR stage. As a whole these considerations suggest that a significant amount of p-cresol and all of IL-6 removal actually took place in the first HFR stage.