Safety of Alkalinization in CKD

In CKD, fixed acid accumulates because the kidneys inadequately excrete protons, conserve filtered bicarbonate, and generate new bicarbonate ions to replace those consumed during systemic acid buffering. Metabolic acidosis is the culmination of this chronic process and is diagnosed when the serum bicarbonate concentration is <22 mEq/L in CKD. Metabolic acidosis has adverse effects on musculoskeletal and kidney health and is associated with higher risks of cardiovascular disease and death.¹ In the United States, approximately 15% with CKD have metabolic acidosis, corresponding to approximately 7 million individuals who are at risk for acidosis-related complications.¹ Such individuals are often treated with alkali with the hope that this therapy will reduce morbidity and mortality.

Clinical trials of varying designs, sample sizes, and durations have investigated potential benefits of systemic alkalinization in individuals with CKD. The alkalinization strategy included either nutritional therapy (fruits and vegetables) or pharmacological therapy with sodium-based alkali (almost exclusively with sodium bicarbonate) or veverimer, a novel gastrointestinal hydrochloric acid binder. Outcomes of interest in these trials were diverse and included kidney function, muscle strength, nutritional status, and skeletal health, among others. Nevertheless, it remains unclear whether alkalinization has positive effects in CKD. On the one hand, individuals in the sodium bicarbonate arm of the Use of Bicarbonate in CKD Trial had a substantially lower risk of CKD progression and death than those in the control group.² However, on the other end, the BiCARB trial reported that those in the sodium bicarbonate arm had a shorter 6-minute walk distance and reduced hand grip strength while experiencing more adverse events, suggesting that sodium bicarbonate is potentially harmful.³ Regarding nutritional therapy with fruits and vegetables, positive effects on kidney health have been reported in select individuals (i.e., hypertensive CKD in a single center) and require confirmation by other groups.⁴ Finally, the VALOR-CKD trial involving over 1400 individuals did not find that veverimer preserved kidney function despite promising results from an earlier trial.⁵ Thus, the decision to incorporate alkali therapy into the treatment plan of an individual with CKD is not straightforward and should include the potential benefits and risks to the patient. Although the potential benefits of alkalinization in general are unclear, nutritional therapy with fruits and vegetables is probably safe as long as the serum potassium is acceptable and monitored closely. In pharmacologic therapies, sodium bicarbonate is the most commonly prescribed and studied in CKD. The main concern with sodium bicarbonate is the sodium load, which can be substantial, leading to fluid retention, weight gain, and elevation of BP.

In this issue of CJASN, Beynon-Cobb and colleagues performed a meta-analysis of 14 randomized, controlled trials involving over 2100 participants to determine whether sodium bicarbonate treatment increases systolic BP in CKD.⁶ The authors reported with moderate certainty that sodium bicarbonate treatment was not associated with an increase in systolic BP. Several sensitivity analyses were performed, and the findings were consistent in each analysis. A particularly important one excluded trials in which a substantial proportion of participants did not have metabolic acidosis at baseline. Demonstrating that systolic BP did not increase in those with metabolic acidosis at baseline is reassuring because these are the individuals who receive treatment on the basis of current practice standards; including those without metabolic acidosis, who are likely healthier, could have skewed these findings toward the null. Subgroup analyses evaluating dose of sodium bicarbonate, trial duration, and CKD stage also found no substantial difference in systolic BP during follow-up. The authors also evaluated but did not find any differences with respect to increasing antihypertensive or diuretic therapy during follow-up. Therefore, the lack of an increase in systolic BP with sodium bicarbonate does not seem to be masked by an escalation in BP or diuretic therapy.

The findings from this meta-analysis suggest that, unlike sodium chloride, sodium bicarbonate does not increase BP and likely not fluid retention because the BP effects are hypothesized to be fluid-mediated. As with all meta-analyses, the results should consider the inclusion and exclusion criteria of the individual trials and any procedures that may have been incorporated before randomization. For example, several trials excluded individuals with heart failure, whereas others permitted those with class I or II heart failure symptoms. Hence, the effect of sodium bicarbonate on fluid status and BP in advanced heart failure patients with low serum bicarbonate is less certain. However, such individuals with low bicarbonate are probably better treated with diuretic therapy than alkali. Many trials required systolic and/or diastolic BP to be controlled before enrollment or included a phase in which BP was reduced below a certain threshold before randomization. Although we cannot be certain that sodium bicarbonate does not increase BP in individuals with excessive BP (i.e., systolic BP >160 mm Hg), it is probably important to reduce BP before treating metabolic acidosis in such cases. Under these assumptions in which BP is reasonably well controlled and there is not underlying fluid excess, it is likely that sodium bicarbonate does not cause excess fluid retention or increase BP in CKD.

Another cardiovascular concern with alkalinization that needs to be further investigated is vascular calcification. This concern largely stems from findings in uremic rats in which ammonium chloride induced metabolic acidosis and ameliorated calcium and phosphate deposition in the aorta⁷ and treatment of metabolic acidosis with sodium bicarbonate increased aortic calcification index.⁸ Taken together, these results suggest that metabolic acidosis prevents and treatment of metabolic acidosis with sodium bicarbonate promotes vascular calcification. Despite these concerns, there is no clear indication that sodium bicarbonate hastens vascular calcification in humans with CKD.

Although the previously mentioned and potentially positive effects of alkalinization result from chronic therapy, treatment of metabolic acidosis can be a helpful short-term treatment of modest hyperkalemia. This highlights the fact that hypokalemia is a potential side effect of treating metabolic acidosis when modest-to-severe hypokalemia is also present (i.e., serum potassium <3.5 mEq/L) if non–potassium-based alkali is used. Alkalinization with sodium bicarbonate should not be started unless the serum potassium concentration is above 3.5 mEq/L to avoid this risk. Another potential complication of alkalinization is calcium phosphate nephrolithiasis resulting from an increase in urine pH. This risk is partially offset by a concomitant increase in urinary citrate levels in response to alkali therapy.⁹ Furthermore, most with CKD have low urinary calcium levels mitigating risk of calcium stones. A rare but dreaded complication of sodium bicarbonate is rupture of the stomach due to high gastric pressure from food contents along with production of carbon dioxide. Therefore, sodium bicarbonate should preferably be taken on an empty stomach. Finally, oral citrate therapy has been linked with an increase in gastrointestinal aluminum absorption in CKD and should be kept in mind if this agent is considered. Enhanced aluminum absorption from the gastrointestinal tract has not been demonstrated with sodium bicarbonate.

Is alkalinization safe? With fruits and vegetables? Probably, as long as serum potassium is monitored. How about with sodium bicarbonate? Apart from findings in the BiCARB Trial, no other trials have reported serious safety concerns with sodium bicarbonate, including BP or weight, under the circumstances of each respective trial. Whether sodium bicarbonate is efficacious still remains to be determined.

Disclosures

M.L. Melamed reports consultancy for Guidepoint Global Consulting, an advisory or leadership role for American Board of Internal Medicine Nephrology Exam Committee, and other interests or relationships with New York Society of Nephrology. K.L. Raphael reports employment with VA Salt Lake City Health Care System.

Funding

This work was supported by research grant R01 DK132823 and R01 DK131811 from NIDDK (M.L. Melamed and K.L. Raphael) and I01 CX001695 from U.S. Department of Veterans Affairs (K.L. Raphael).

Author Contributions

Writing – original draft: Kalani L. Raphael.

Writing – review & editing: Michal L. Melamed, Kalani L. Raphael.