Urine Ammonium and Preclinical Acidosis in CKD

The kidney maintains the serum [HCO₃⁻] at a normal value of 25 mEq/L by two distinct and highly regulated processes: (1) reabsorption of the filtered load of HCO₃⁻ and (2) production and excretion of ammonium (NH₄⁺). Production of NH₄⁺ is primarily a function of the proximal tubule, and excretion is the result of the transport of NH₄⁺ by the proximal convoluted tubule and thick ascending limb of Henle loop to facilitate the accumulation of NH₃ in the medullary interstitium and its transfer into the MCD to be trapped and excreted as NH₄⁺. Both production and transport of NH₃/NH₄⁺ are linked to changes in systemic pH and [K⁺]. Ammoniagenesis is responsive to net endogenous acid production (NEAP) and thereby, dietary acid loads. In individuals ingesting a typical Western diet, NH₄⁺ production and excretion assume major significance for maintaining systemic pH: the production of NH₄⁺ and HCO₃⁻ facilitates the stoichiometric replacement of the HCO₃⁻ lost from the ECF by the entry of acid products from protein metabolism into the ECF. The “new bicarbonate” produced in this pathway will be returned to the systemic circulation only if the NH₄⁺ generated is excreted in the urine.¹

The possibility that NH₃/NH₄⁺ accumulation in the renal interstitium could be potentially nephrotoxic per se or a representation of an inflammatory process was advanced by seminal studies from Hostetter and coworkers² in an animal model of chronic metabolic acidosis with elevated uNH₄⁺ excretion that displayed accumulation of complement C3 in the interstitium. More recent evidence suggests that chronic metabolic acidosis activates signaling pathways in kidney cells³ as well as other systems and cells⁴ and that it may be associated with cell damage and fibrosis.⁵ For example, the p53 gene, which is induced by acidosis, plays a role in the inhibition of glycolysis and increase in oxidative phosphorylation and activation of reactive oxygen species in cancer cells.⁵ It is conceivable, therefore, that acidosis-induced inflammation and ER stress may represent important factors in cellular injury.

Evidence has accumulated from several clinical trials showing that patients with later-stage CKD and metabolic acidosis progress at a more rapid rate⁶ and that high dietary acid load predicts ESRD among adults with CKD.⁷ Moreover, a more rapid rate of progression of CKD has been documented in subjects with early-stage CKD eating a diet high in protein.⁷ However, patients in this group typically have a compensatory increase in NEAP and consequently, do not develop frank metabolic acidosis. Importantly, in several studies, the rate of progression of CKD was retarded when supplementary alkali therapy was administered or additional fruits and vegetables were added to the diet.^8,9 Either maneuver can balance the generation of acids from dietary protein by adding alkali to the body balance equation as HCO₃⁻ or producing citrate from dietary fruits and vegetables. The subsequent reduction in NEAP was observed in parallel with a significant reduction in urinary NAE, even in patients with early-stage CKD without frank metabolic acidosis. On the basis of these important findings, the concept of preclinical acidosis has been advanced to designate those patients with earlier-stage CKD, in whom by convention, higher dietary NEAP is matched by a higher NAE, preventing acidosis at this juncture. Avoidance of the development of overt clinical metabolic acidosis ([HCO₃⁻]<22 mEq/L) requires, however, sufficient functional renal mass. The tradeoff for augmented ammoniagenesis and NH₄⁺ excretion to preserve acid-base balance in these patients may activate potentially harmful signaling cascades in response to an “acid milieu” in the interstitium.^3,4,10,11 In this regard, studies have also shown that urine or serum levels of endothelin and angiotensinogen were elevated in nonacidotic patients with higher uNH₄⁺ and, in addition, these levels declined with alkali administration.¹⁰

In this issue of the Journal of the American Society of Nephrology, Raphael et al.¹² have examined the association between baseline NH₄⁺ excretion and clinical outcomes in subjects participating in the African American Study of Kidney Disease and Hypertension (AASK). Patients without clinical metabolic acidosis at baseline were divided into tertiles on the basis of a single measured urine NH₄⁺. Remarkably, patients with the lowest baseline NH₄⁺ excretion (<20 mEq/d) exhibited a 46% higher risk for adverse events (death and ESRD) as well as a future risk for metabolic acidosis at 1 year (table 6 in ref. 12) compared with participants in the high-NH₄⁺ excretion group (≥20 mEq/d). The authors propose that a lower uNH₄⁺ may be an “alternative and perhaps earlier indicator of risk than the serum [tCO₂].”¹²

Although the assessment of uNH₄⁺ is of great interest and may prove beneficial, additional data are needed. For example, only baseline uNH₄⁺ values were available, and neither urine pH nor urinary biomarkers of tubule injury were reported. Moreover, none of the patients received alkali therapy. Although the authors correctly point to the study of Mahajan et al.⁸ as evidence that alkali therapy slows progression of CKD in early hypertensive nephropathy, that study did not specifically evaluate patients with a low uNH₄⁺.

A sound pathophysiologic explanation for the key finding in this study is beyond the scope of the data available for analysis. The authors speculate that uNH₄⁺ is lower and that outcomes are poorer in patients with low uNH₄⁺ because of “poor excretory capacity (for NH₄⁺) and kidney function.”¹² Clearly, uNH₄⁺ fell significantly as mGFR declined (figure 3 in ref. 12) as expected. Accordingly, it might be helpful to consider the possible explanations for a decrease in uNH₄⁺ excretion. Urine NH₄⁺ excretion may be reduced because of (1) a physiologic response by the kidney to low levels of protein intake; (2) a reduction in ammoniagenesis and excretion due to metabolic or respiratory alkalosis or hyperkalemia; (3) a decrease in functional renal mass; or (4) a selective abnormality in NH₄⁺ transport and excretion, for example, transport of NH₄⁺ (or H⁺ ions) in the collecting duct or a defect in the medullary countercurrent multiplication system, either as a result of tubulointerstitial disease.^1–3 There was no evidence of metabolic alkalosis or hyperkalemia. By adjusting for NEAP and body mass index, the authors propose that the first possibility seems less likely, but it has not been eliminated unambiguously. The patients in tertile 1 had the lowest daily estimated protein intake of all three groups, about one third of the protein intake in the highest uNH₄⁺ excretion group (tertile 3). Also, those in tertile 3 had both a higher uNH₄⁺ and the highest dietary protein estimates. However, in this group, NH₄⁺ excretion was still only one third the calculated NEAP. Therefore, it remains conceivable that subjects in the low-uNH₄⁺ group had daily intakes of dietary protein that, although lower, were simply too high for the reduced GFR (i.e., explanation 3 above).

Interestingly, Wesson et al.¹³ reported recently that incremental increases in dietary acid loads in rats after 2/3 nephrectomy (model of progressive CKD) displayed evidence of cortical interstitial acid retention (compared with sham-operated controls). The renal cortical acid milieu in these animals was associated with higher excretion rates of N-acetyl-β-d-glucosaminidase, consistent with the development of tubulointerstitial disease. Biomarkers of kidney injury were not available in the study from the patients in the AASK¹² but would have enriched the analysis and should be incorporated into future clinical trials. The authors attempt to place their findings into the perspective of the observation in animal studies that complement was activated² by proposing that the uNH₄⁺ was retained and not excreted in the group with poorer outcomes and a low uNH₄⁺. However, there is no direct evidence to suggest that ammoniagenesis or NH₄⁺ excretion was playing a role in the outcomes, and speculation that uNH₄⁺ and tissue levels of NH₄⁺ may have been dissociated is not necessary.

Although the pathophysiology of the lower uNH₄⁺ cannot be discerned unequivocally, this study underscores the critical need for additional basic research using animal models to more fully explain the signal(s) linked to dietary acid loads that may initiate interstitial or tubule injury, and then to more clearly define its consequences in larger-scale multicenter prospective studies. Certainly, additional validation of the predictive utility of urine NH₄⁺ levels in patients with both early- and later-stage kidney disease is sorely needed. It seems that the time is past due for federal funding agencies to sponsor studies designed specifically to investigate the role of long-term alkali in slowing progression of CKD.

The availability of accurate measurement of uNH₄⁺ by clinical laboratories would be a welcome addition to evaluation of the integrity of kidney acid-base homeostasis in the patient with CKD. Moreover, for patients with possible renal tubular disorders, in whom surrogates for actual uNH₄⁺ measurements, such as the urine anion gap, are relied on by default, access to actual uNH₄⁺ determination is overdue.

Presently, it seems practical to provide alkali to patients with early- or later-stage CKD without metabolic acidosis who have a low or high uNH₄⁺ and recommend more universal adoption of administration of alkali to patients with CKD and clinical acidosis to correct the HCO₃⁻ to values of approximately 24 mEq/L.

Disclosures

None.