Metabolic Acidosis and Cardiovascular Disease Risk in CKD

Chronic metabolic acidosis is believed to contribute to several important sequelae of CKD, and the evidence base to guide its treatment has been steadily growing. Several clinical trials, all relatively small, have shown preservation of kidney function and lean body mass with alkali therapy. In select patients, the use of fruits and vegetables may have a role as a nonpharmacologic option. Nevertheless, a definitive randomized clinical trial—one with a double-blind, placebo-controlled design and preferably conducted across multiple centers in a diverse patient population—is still lacking. Given the substantial burden of cardiovascular disease in the CKD population, the relationship of chronic metabolic acidosis—and its treatment—to cardiovascular disease risk is of great interest. Paradoxically, higher levels of serum bicarbonate—but still within the normal range—have been associated with a higher risk of heart failure (1,2). Whether this reflects a causal relationship or whether serum bicarbonate is simply a useful prognostic marker is unclear. In a general population cohort, higher bicarbonate levels were associated with greater left ventricular mass and higher pulse pressure (1), but in a CKD cohort, there was no association of serum bicarbonate with cardiac structural abnormalities (3). Furthermore, in a well characterized but smaller diabetic cohort in Australia, higher serum bicarbonate was associated with a lower risk of incident coronary heart disease, and it was not associated with heart failure (4). Clearly, a better understanding is needed of the effects of acid-base balance on cardiovascular function. The report by Kendrick et al. (5) this issue of the Clinical Journal of the American Society of Nephrology addresses this important topic and adds an additional reason to consider treating metabolic acidosis. Using a randomized crossover design, these investigators examined the effect of sodium bicarbonate treatment on vascular function measured using brachial artery flow-mediated dilation (FMD). Participants received sodium bicarbonate treatment and usual care in a randomized order for 6 weeks each, with a 2-week washout period in between; FMD was measured at the beginning and end of each 6-week interval. FMD improved after treatment with sodium bicarbonate, suggesting that correction of chronic metabolic acidosis could improve vascular function. This is an intriguing result. Although acute acid-base disturbances are known to alter arterial tone, chronic metabolic acidosis has not been considered to be a cause of endothelial dysfunction in CKD. As the authors acknowledge, the mechanism underlying this effect is unknown. There was no change in inflammatory markers, and the increase in FMD was not associated with the change in serum bicarbonate. Metabolic acidosis could affect vascular function on a number of levels, including the renin-angiotensin-aldosterone system, adrenergic signaling, and insulin signaling in endothelial cells. Future investigations addressing the underlying mechanism will be informative. Several additional questions follow as well, including the following. Is the improvement in vascular function maintained over the long term? Will this translate into a reduction in cardiovascular events? Is the relationship of pH and/or serum bicarbonate with FMD linear (i.e., would additional correction of acidosis produce even greater changes in FMD)? This final question is intertwined with the concerns raised above about heart failure. Indeed, a major unresolved question is what level of serum bicarbonate should be used as the goal of treatment. Most studies have used a fixed dose of alkali, but two of the larger and more recent trials titrated therapy to achieve a goal serum bicarbonate level (≥23 mEq/L in de Brito-Ashurst et al. [6]; 24–26 mEq/L in Dubey et al. [7)]. For the majority of the 2 years in the former study, the mean serum bicarbonate in the intervention arm remained at or below 24 mEq/L; in the latter, at the end of 6 months, the intervention group achieved a mean serum bicarbonate of 23.5 mEq/L. In line with these trials and the Kidney Disease Improving Global Outcomes 2012 guidelines, Kendrick et al. (5) titrated therapy to maintain serum bicarbonate ≥23 mEq/L. Achieving such a target over a relatively brief time period is challenging: After 6 weeks of treatment with sodium bicarbonate, the mean plasma bicarbonate was 22.0 mEq/L, and 44% of participants had achieved the target. Therefore, only a minority of participants in this study and other studies of patients with CKD have maintained bicarbonate levels in the range associated with elevated heart failure risk; to adequately address the concerns stemming from those observational data, additional information is needed from interventional studies. However, one of the concerns about aggressive correction is exacerbation of vascular calcification; in this regard, the results by Kendrick et al. (5) are reassuring. Although the study did not measure vascular calcification directly, there was no change in serum calcification propensity. This was measured by the T50, which has been associated with cardiovascular and all-cause mortality in patients with nondialysis-dependent CKD and those receiving dialysis. Although encouraging, this result does not definitively allay concerns about promoting vascular calcification: this was a short-term study; the utility of the T50 as a surrogate for vascular calcification requires additional study; and understanding the long-term cardiovascular effects of alkali therapy will require investigations using definitive outcome measures. In contrast to the T50 data, another finding might raise concern about cardiovascular effects: sodium bicarbonate treatment increased levels of fibroblast growth factor 23(FGF-23), and this is the second report noting such an effect (8). The concomitant elevation of serum phosphate suggests that FGF-23 increased as a compensatory response, but additional data would be required to confirm this. Regardless of the etiology, because of the robust evidence linking FGF-23 with cardiovascular outcomes, this effect of sodium bicarbonate is a concern. However, the change in a surrogate biomarker is not equivalent to a clinically meaningful effect on patient-related outcomes: consider a similar scenario related to sodium-glucose cotransporter 2 (SGLT2) inhibitors. In patients with type 2 diabetes, this class of medications demonstrably reduces cardiovascular events and mortality (9,10), but in a short-term study, canagliflozin (a member of this class of medications) increased FGF-23 levels by 20% and serum phosphorus by 16% (11). The effect on mineral metabolism–related parameters is important for understanding the fracture risk associated with SGLT2 inhibitors, but it does not detract from their beneficial effects on hard clinical outcomes. Avoiding use of SGLT2 inhibitors because of the change in FGF-23 would deprive patients of an efficacious class of medications; for sodium bicarbonate, although comparable trial data are unavailable, the sum total of the evidence points to clinical benefit, and it would seem prudent to take a similar approach. The report by Kendrick et al. (5) is an interesting and novel study that will hopefully lead to additional investigations into this aspect of human physiology. As with all pilot studies, it consisted of a small number of participants and relied on surrogate measures. Still, it is an important step forward in clarifying the relationship between acid-base balance and cardiovascular disease in patients with CKD. It should guide future research to determine whether the treatment of metabolic acidosis meaningfully alters cardiovascular risk. Disclosures None.