To validate azotemic markers as an index for intradialytic changes in solute concentration, we compared eight solutes (pseudouridine, xanthine, hypoxanthine, peak 4, peak 5, p-hydroxyhippuric acid, indoxyl sulfate, and hippuric acid) with five classical azotemic markers (urea, creatinine, uric acid, phosphate, and potassium). We determined concentrations by reversed-phase HPLC coupled to ultraviolet absorption or photometrically. Seven compounds showed significant intercorrelation (P less than 10(-5)): urea, pseudouridine, uric acid, peaks 4 and 5, p-hydroxyhippuric acid, and creatinine. The hippuric acid concentration change after dialysis correlated with the change for these seven compounds and also with indoxyl sulfate, hypoxanthine, potassium, and the group of unidentified ultraviolet-absorbing HPLC peaks accumulating in uremia. We conclude that urea only partially represents the concentration changes of other retention compounds after dialysis; alternative markers, e.g., hippurate, should be considered.

Contamination with uremic solutes occurred at the dialysate inlet only near the end of the session when small quantities of fresh dialysate were left in the container. Differences in dialysate temperature did not result in a different separation between used and unused dialysate, or in differences in removal of toxins or Kt/V.

Middle molecules can be defined as compounds with a molecular weight (MW) above 500 Da. An even broader definition includes those molecules that do not cross the membranes of standard low-flux dialyzers, not only because of molecular weight, but also because of protein binding and/or multicompartmental behavior. Recently, several of these middle molecules have been linked to the increased tendency of uremic patients to develop inflammation, malnutrition, and atheromatosis. Other toxic actions can also be attributed to the middle molecules. In the present publication we will consider whether improved removal of middle molecules by large pore membranes has an impact on clinical conditions related to the uremic syndrome. The clinical benefits of large pore membranes are reduction of uremia-related amyloidosis; maintenance of residual renal function; and reduction of inflammation, malnutrition, anemia, dyslipidemia, and mortality. It is concluded that middle molecules play a role in uremic toxicity and especially in the processes related to inflammation, atherogenesis, and malnutrition. Their removal seems to be related to a better outcome, although better biocompatibility of membranes might be a confounding factor.

A convenient way to classify uremic solutes is to subdivide them according to the physicochemical characteristics influencing their dialytic removal into small water-soluble compounds (<500 Da), protein-bound compounds, and middle molecules (>500 Da). The prototype of small water-soluble solutes remains urea although the proof of its toxicity is scanty. Only a few other water-soluble compounds exert toxicity (e.g., the guanidines, the purines), but most of these are characterized by an intra-dialytic behavior, which is different from that of urea. In addition, the protein-bound compounds and the middle molecules behave in a different way from urea, due to their protein binding and their molecular weights, respectively. Because of these specific removal patterns, it is suggested that new approaches of influencing uremic solute concentration should be explored, such as specific adsorptive systems, alternative dialytic timeframes, removal by intestinal adsorption, modification of toxin, or general metabolism by drug administration. Middle molecule removal has been improved by the introduction of large pore, high-flux membranes, but this approach seems to have come close to its maximal removal capacity, whereas multicompartmental behavior might become an additional factor hampering attempts to decrease toxin concentration. Hence, further enhancement of uremic toxin removal should be pursued by the introduction of alternative concepts of elimination.