The current foci of renal replacement therapy with dialysis are middle molecular weight toxins, consisting of small proteins, polypeptides and products of glycosylation and lipoxygenation. Conventional high-flux dialysis is not efficient at removing these molecules, explaining the increased interest in using sorbents to supplement dialysis techniques. Prototype biocompatible sorbents have been developed and investigated for middle molecule removal; these have been shown, in man, to remove β2-microglobulin, angiogenin, leptin, cytokines and other molecules, without reducing platelets and leukocytes. Extensive clinical studies are underway to demonstrate the clinical utility and safety of adding routinely a sorbent hemoperfusion device to hemodialysis.

ß2-Microglobulin, a middle molecular weight protein (11.8 kD), has increasingly been the focus of attention in its role as the precursor of dialysis-related amyloidosis (DRA) [1]. Current evidence favors the alteration of ß2microglobulin by glycosylation [2], and subsequent deposition of glycosylated ß2-microglobulin as amyloid fibrils in many tissues, particularly in large joints (shoulders), and in the carpal tunnel. DRA in chronic dialysis patients can produce bone cysts, severe carpal tunnel syndrome and crippling arthritis [3]. While there is poor correlation between blood concentrations of ß2-microglobulin and DRA, there is direct correlation of this complication with duration of dialysis. Moreover, the disposition of ß2microglobulin, follows a 3-pool kinetic model with distribution in plasma water, extracellular fluid and a slow equilibration compartment [4]. Dialyzers constructed with cuprophan dialysis membranes (cellulosic), long the standard therapy for chronic dialysis, do not reduce ß2-microglobulin concentrations and may in fact alter the conformational structure of ß2microglobulin, promoting amyloidosis. On the other hand, noncellulosic (e.g. polyacrylonitrile and polysulfone) dialyzer membranes may reduce ß2-microglobulin concentrations during treatment, by a combination of adsorption (the majority) to the membrane and by convective removal (the minority), the latter of which is dependent on high rates of ultrafiltration. The efficiency of the latter membranes is not considered optimal, and usually adsorption is complete early in the dialysis procedure. Moreover, dialyzer reuse significantly impairs the removal of ß2-microglobulin [5]. Hemoperfusion devices containing adsorbents have been used to enhance ‘middle molecule’, amino acid, and creatinine removal, in dialysis patients, using nonspecific agents such as activated charcoal [6]. A porous resin hemoperfusion device has also been shown to reduce ß2microglobulin concentrations during extracorporeal treatment, with improvement in the clinical manifestations of DRA [7]. In this study we present a new adsorbent device with improved biocompatibility by virtue of a polymer coating [8] , designed specifically to adsorb ß2-microglobulin. In vitro and in vivo data are reported as far as efficacy and biocompatibility performances are concerned. We report the case of two long-term hemodialysis patients who volunteered to undergo combined hemodialysis/ hemoperfusion at a single session, in order to assess the

Accumulation of β2-M in ESRD correlates with the development of dialysis related amyloidosis. HFD membranes have more efficient mass transfer, mostly by convection, of β2-M than cuprophane. However HFD is only able to remove 50% of daily β2-M generation. A biocompatible polymer adsorbent hemoperfusion (HP) device (BetaSorbTM) containing 300 ml of a polystyrene-like adsorbent with efficient removal of β2-M was inserted upstream of a HFD (HF80, Fresenius, Bad Homburg) in the blood circuit in 2 ESRD patients, during one routine hemodialysis (HD). During combined HP/HD β2-M was measured at timed intervals, as were platelet and white cell counts. Blood flow rates were 405 ml/min and 380 ml/min in patients #1 and #2 respectively. Single pool (sp) Kt/Vβ-M was calculated, and compared to the high-flux arm (F80 dialyzers, Fresenius, Bad Homburg) of the HEMO study.TableCombined HP/HD was well tolerated by both pts. Neither platelets nor WBC (corrected for ultrafiltration) changed significantly. spKt/Vβ2-M was 1.7 and 1.37, while rebound equilibrium of β2-M would be achieved in 87 mins and 108 mins, in patients #1 and #2 respectively. β2-M concentrations were similar to, but spKt/Vβ-M, exceeded, that observed in the large HEMO study of HFD. This is the first resin HP device not associated with a fall in platelet count. In view of the efficient removal of β2-M a multicenter study of this device combined with HFD is planned.

Modern dialysis, used in approximately 1 million people throughout the world, carries a mortality (unadjusted for co-morbidity) in the first year of anywhere between 14 and 23% (EDTA/USRDS). Clearly, mortality is governed by co-morbid conditions, such as diabetes and heart disease, but also by the quantity of dialysis delivered (Kt/ Vurea). Even with high efficiency dialysis and consequently high Kt/Vurea, patients continue to feel unwell, have demonstrable suppression of immunity, and increased risk of the following; infection, atherosclerosis, circulating abnormalities of proteins (glycosylation) and lipids, as well as endocrine disturbances in parathyroid, pituitaryhypothalamic, and gonadal function. In addition, specific diseases associated with dialysis develop with the length of time a patient has been treated (e.g. dialysis-related amyloidosis, DRA), with hemodialysis or continuous ambulatory peritoneal dialysis (CAPD). Without the use of exogenous recombinant human erythropoietin, patients would remain anemic, and without the use of exogenous vitamin D analogues bone disease and hyperparathyroidism would be much more prevalent. The Changing Understanding of Toxins in Uremia (Small Molecular Weight Proteins)