Can Oral Therapy Reduce Uremic Toxins?

Removing uremic toxins has been the focus of therapeutic dialysis for >50 years, yet morbidity and mortality rates remain unacceptably high despite considerable improvements including high-efficiency dialysis, high-flux dialysis, improved biocompatibility of membranes, bicarbonate dialysate, extended treatment times, and more frequent treatments (1–3). It would seem that dialytic approaches to remove more toxins have been exhausted, and attention should be focused on novel blood purification strategies or other potential causes of adversity such as the comorbid consequences of diabetes and hypertensive cardiovascular disease. Apparently, loss of native kidney function is not completely corrected by the current intensity of solute and fluid purging, which are arguably the only contributions of dialysis. One could postulate that the kidney produces a yet-undiscovered immunologic enhancer, an inhibitor of endothelial damage, or a protector from attending comorbidities or some other adversity, analogous to uremic anemia, that cannot be corrected by dialysis. However, excretory purging of solutes is clearly the most important function of native kidneys, and the most important life-threatening solutes removed by the kidney are also removed by dialysis. Despite even the best dialysis, patients suffer from a persisting residual syndrome that includes an increased risk of hospitalization and death. One possible cause of this unfortunate clinical syndrome is incomplete removal of unspecified toxic solutes.

Solutes removed efficiently by the native kidney but poorly by dialysis include larger molecular compounds such as β2-microglobulin (B2M) and protein-bound solutes removed by the native kidney via proximal tubular transport mechanisms that are not duplicated by dialysis. Efforts to remove larger solutes by engaging more porous membranes or hemofiltration have succeeded in lowering the concentrations of B2M and other larger solutes, but clinical outcomes have benefitted little or not at all (1,4–6). By contrast, methods to remove protein-bound solutes only recently attracted investigator attention (7). In recent years, products of the colonic microbiome, such as indoxyl sulfate (IS) and p-cresyl sulfate (PCS) (8), both of which exhibit vascular toxicity, have been identified as the source of some of the major bound solutes (9–11). One can also postulate that the rich lipid environment of tissues in general could produce endogenous hydrophobic toxins in need of purging; and because blood is hydrophilic, a transport mechanism from the tissue or gut to the kidney would be required. The most prominent mechanism known for transporting lipid-soluble compounds (e.g., fatty acids) involves noncovalent binding to serum albumin. Albumin is a little-recognized component of the excretory system, which serves to deliver hydrophobic or amphiphilic compounds to the kidney. There the robust basolateral transport mechanisms in the proximal tubule act to strip these and other compounds from their relatively loose binding sites, sometimes completely in one pass through the kidney (12–14). Protein binding also serves to protect the host by minimizing the free (unbound) and presumably toxic species. Failure of dialysis to remove albumin-bound compounds is clearly demonstrated by chromatographic techniques that show high concentrations of bound solutes in well dialyzed uremic serum (15), and by the reduced potential of uremic albumin to bind marker solutes, evidence that the binding sites are highly saturated (16,17). The albumin-binding defect itself could have adverse consequences independent of the toxicity of accumulated ligands if transport of other vital but not necessarily toxic substances is impaired.

Standard thrice-weekly dialysis has limited potential to remove protein-bound substances, but increasing the dialysis membrane area and/or increasing the dialysate flow can augment removal of bound IS and PCS, especially if applied to prolonged nocturnal dialysis, which has been shown to double IS and PCS clearance without change in urea clearance (18,19). The addition of charcoal as a sorbent to the dialysate has also been shown to increase their removal (20).

In contrast with efforts at removal, in this issue of CJASN, Sirich et al. report their attempts to alter the generation of IS and PCS (21). Both are produced by gut flora in response to delivery of the amino acid substrates tryptophan and tyrosine to the colon. Both are >90% bound to serum albumin, accounting for their poor removal by dialysis and serum concentrations that are 10–20 times higher than normal. Of note, dialysis patients receiving total parenteral nutrition or following total colectomy have much lower serum levels of IS and PCS, consistent with a colonic origin. The investigators administered a prebiotic in the form of nondigestible starch to provide a substrate for colonic bacteria to grow and incorporate amino acids into structural components rather than production of toxic byproducts such as IS and PCS. Prebiotics are nondigestible compounds usually containing fiber that pass through the small bowel and into the colon where, acting as substrate, they stimulate growth and metabolic activity of certain bacterial components of the microbiome. During the 6-week study, administration of nondigestible fiber to 40 patients reduced free (unbound) IS and PCS levels nearly 25%–30% compared with digestible fiber, before and after the study. However, the effects were statistically significant only for IS. This demonstration of an effect of dietary fiber on toxin generation in dialysis patients may have implications with respect to other more toxic solutes retained in patients with life-threatening uremia. The fall in free but not total concentrations infers that the binding equilibrium shifted toward tighter binding. If this is indeed the case, then generation of other bound solutes would likely have decreased, leaving more sites for IS and PCS binding. No attempt was made to detect clinical benefits in this short-term study, but a clinical outcomes trial seems justified given the positive results and possible effect on other toxic solutes.

This study further emphasizes the potential role of the intestine as an adjunct to the dialytic management of kidney failure by either the reduced generation or facilitated clearance of toxic solutes. Although oral prebiotics or other foodstuffs aimed at shifting gut flora from proteolytic to saccharolytic are unlikely to completely reverse the uremic syndrome, they have potential to reduce its intensity or delay its onset, allowing more time to conduct crucial preparations for dialysis, including patient education and training. Perhaps more importantly, more time would be available to allow maturation of an arteriovenous access. Starting oral therapy could allow the patient to avoid exposure to indwelling dialysis catheters, which have been associated with increased morbidity and mortality, especially during the first 6 months after starting hemodialysis. In addition, intestinal strategies may provide a harbinger for nondialytic options to ameliorate uremic symptomology in patient populations without options for dialytic management.

Of note, patients managed with peritoneal dialysis (PD) apparently have lower production rates of IS and PCS, and presumably other colon-derived uremic toxins, than hemodialyzed patients (22). The cause is unknown but the well recognized reduction in appetite and food intake in PD patients might limit delivery of undigested foodstuffs including protein to the colon, providing less substrate for the microbiome. In contrast with normal people with reduced food intake, PD patients are not subject to malnutrition because they receive extra calories in the form of absorbed glucose from the peritoneal dialysate. The latter is likely the cause of the anorexia.

We currently use oral therapy to alleviate some of the consequences of kidney failure. These include dietary cation, phosphate, protein, and water restriction, phosphate binding agents, potassium binding agents, and laxatives. Unfortunately, a vegetarian diet rich in nondigestible fiber, which has been shown to reduce urinary excretion of IS and PCS in normal people (23), is relatively prohibited in dialysis patients because of the risks of potassium contained in fruits and vegetables. Protein restriction in uremic patients was the mainstay of therapy in the predialysis era, serving to decrease production of ammonia, urea, and other toxic derivatives of hepatic nitrogen metabolism. The more recent finding that dietary protein restriction may reduce colonic fermentation of protein and production of potential uremic toxins suggests an additional mechanism for the benefit.

Other methods to reduce the generation rate or increased clearance of colonic metabolites include use of antibiotics, probiotics, α-glucosidase inhibitors, and oral adsorbents such as carbon particles, or cross-linked polyelectrolytes as reviewed recently by Poesen et al. (24).

In summary, if dialysis were completely successful, we would have little need to investigate the colon in CKD or dialysis patients. Questions remain about the cause of the unacceptably high morbidity and mortality among CKD and dialysis patients, so efforts to reduce the generation or alternative clearance of known and unknown toxic products of the colonic microbiome promise to give us more insight and perhaps hope for better outcomes.

Disclosures

None.