Recent Evidence on the Effect of Treatment of Metabolic Acid on the Progression of Kidney Disease

Sebastian Hultin; David W. Johnson; Sunil V. Badve


Curr Opin Nephrol Hypertens. 2021;30(5):467-473. 

In This Article

Overview of Recent Trials

In 2009, de Brito-Ashurst et al.[17] published the first randomized trial with 24 month follow-up duration examining the effect of bicarbonate supplementation on the rate of decline in kidney function assessed by creatinine clearance (CrCl). In this open-label trial, 134 patients with CKD with acidosis (mean serum bicarbonate 19.9 mmol/l) were randomized to bicarbonate (n = 67) or no study treatment (n = 67), matched by sex and presence of diabetes (Table 1). Bicarbonate supplementation was titrated (median bicarbonate dosage 1.82 ± 0.80 g daily) to achieve a serum bicarbonate concentration more than 24 mmol/l to maintain intergroup separation (Table 2). Compared with the control group, the rate of decline in CrCl was significantly slower in the bicarbonate supplementation group (1.73 ml/min/1.73 m2; 95% CI −0.39 to 4.15) than in the control group (5.93 ml/min/1.73 m2; 95% CI 4.19--7.76). Furthermore, less patients in the bicarbonate group progressed to kidney failure [6.5 vs, 33%; relative risk (RR) 0.13; 95% CI 0.04--0.40] (Table 2). There were no differences in blood pressure, hospitalization for heart failure, blood pressure medication requirements or peripheral edema.

Following an initial observational, interventional pilot study of sodium citrate vs. no therapy demonstrating conflicting results with respect to the rate of kidney function decline,[28] a RCT with a 5-year follow-up duration was published in 2010 by Mahajan et al..[18] In this trial, 80 patients with an eGFR greater than 60 ml/min/1.73 m2 (mean eGFR 75.5 ml/min/1.73 m2) and macro-albuminuria [mean urinary albumin--creatinine ratio (UACR) 421 mg/g] were randomized to either sodium bicarbonate, sodium chloride or placebo. Acidosis was not a prerequisite for enrolment (mean serum bicarbonate 26.1 mmol/l) (Table 1). Sodium bicarbonate therapy abrogated kidney function decline (−1.34 ± 0.20 ml/min/1.73 m2/year) compared with placebo (−2.37 ± 0.19 ml/min/1.73 m2/year) or sodium chloride (−2.19 ± 0.20 ml/min/1.73 m2/year) (Table 2). End of study UACR was significantly lower in both bicarbonate (387.5 ± 163.1 mg/g) and sodium chloride groups (466.5 ± 179.4 mg/g) compared with placebo (507.5 ± 228.2 mg/g). Although there was no difference noted in SBP between the groups, adverse effects associated with bicarbonate therapy were not reported.

The largest RCT examining kidney outcomes with bicarbonate therapy was published in 2019 by Di Iorio et al..[19] The Use of Bicarbonate in Chronic Renal insufficiency (UBI) study recruited patients with CKD stages 3–5 (mean eGFR 33 ml/min/1.73 m2) and acidosis (serum bicarbonate >18 mmol/l, <24 mmol/l, mean bicarbonate 21.7 mmol/l) with a follow-up of 3 years (Table 1). A total of 795 participants were randomized to bicarbonate supplementation twice daily or no study medication group. Bicarbonate supplementation was adjusted to maintain serum bicarbonate levels 24–28 mmol/l. Unlike in the de Brito-Ashurst study, rescue bicarbonate therapy was permitted in the control group with 147 patients receiving bicarbonate supplementation. Despite this, in-between group separation for serum bicarbonate was achieved (serum bicarbonate 26.1 ± 1.7 vs. 21.9 ± 1.9 mmol/l) (Table 2). After 3 years follow-up, the bicarbonate treatment group eGFR had declined by 4.9 ± 4.2 ml/min/1.73 m2 compared with 10.9 ± 5.2 ml/min/1.73 m2, equating to a decline of −1.4 ml/min/1.73 m2/year vs. −3.4 ml/min/1.73 m2/year (Table 2). Bicarbonate therapy reduced the risks of all-cause mortality (3.1 vs. 6.8%; hazard ratio 0.43; 95% CI 0.22--0.87), doubling of serum creatinine (6.6 vs. 17.0%; hazard ratio 0.36; 95% CI 0.22--0.58) and progression to kidney failure (6.9 vs. 12.3%; hazard ratio 0.50; 95% CI 0.31--0.81). The authors noted no differences in end of study blood pressure between the two groups with surprisingly more patients in the bicarbonate group having had a reduction in their antihypertensive therapy. Hospitalization events were less frequent in the bicarbonate group compared with the control group (400 vs. 1160 days/year) but the causes of these were not defined.

The clinical and cost-effectiveness of oral sodium bicarbonate therapy for older patients with chronic kidney disease and low-grade acidosis (BiCARB) study was a pragmatic placebo-controlled double-blind study evaluating the impact of bicarbonate therapy in elderly CKD patients (adults >60 years, mean age 74 years) with acidosis (<22 mmol/l) (Table 1).[21] Whilst the primary outcome was between-group difference in the short Physical Performance, kidney outcomes were reported as a secondary outcome over a follow-up period of 2 years. Three hundred patients were recruited and randomized to either bicarbonate 500 mg one to two tablets thrice daily (n = 152 participants; 81 participants completed 2-year follow-up) or placebo (n = 148 participants; 80 participants completed 2-year follow-up) matched by sex, degree of kidney dysfunction and age. There was no significant difference in progression to kidney failure (33 vs. 33 patients, hazard ratio 1.22; 95% CI 0.74--2.02) or the composite end point of doubling serum creatinine, 40% reduction of eGFR or progression to kidney failure (hazard ratio 1.16; 95% CI 0.73--1.84) (Table 2). Although serum bicarbonate concentration was greater in the bicarbonate group than in the placebo group at 3 months, this separation was not maintained throughout the remainder of the trial (Table 2). Adverse events from therapy were common in both trial arms (86.1 vs. 89.1%) but there was no difference in episodes of congestive heart failure, all-cause mortality and blood pressure between the two groups.

Other Acidosis-reducing Interventions

Reduction of net acid load through dietary modification has emerged as an alternative to bicarbonate supplementation. Goraya and colleagues evaluated bicarbonate and fruit and vegetable interventions compared with standard of care in an open-label RCT.[20,25–27] One hundred and eight hypertensive CKD stage 3 patients (mean eGFR 42.6 ml/min/1.73 m2) with acidosis (plasma total CO2 <24 mmol/l and >22 mmol/l; mean 23.0 mmol/l) and macroalbuminuria (UACR >200 mg/g) were matched for age, ethnicity, eGFR and UACR and randomized to receive either sodium bicarbonate (0.3 mEq/kg/day) or fruit and vegetables calculated to reduce potential dietary kidney acid load by 50%. Patients who refused sodium bicarbonate after randomization acted as controls. Thirty-three patients in each arm completed 5-year follow-up. Despite a fairly modest increase of total plasma CO2 between intervention arms and control arm, eGFR was significantly preserved in both the fruit and vegetable and bicarbonate supplementation arms compared with controls (20.7 ± 5.8, 29.4 ± 5.8 and 27.3 ± 5.7 ml/min/1.73 m2, respectively). Proteinuria was not reported at 5 years, but at 3-year follow-up, albuminuria was lower in both the fruit and vegetable and bicarbonate groups compared with controls (UACR 242 ± 56, 262 ± 62 and 300 ± 69 mg/g, respectively). There was no difference in kidney outcomes between the bicarbonate group and the fruit and vegetables group but blood pressure (125.4 ± 4.7 135.5 ± 4.7 mmHg) and weight (BMI 26.6 ± 1.8 vs. 28.4 ± 1.8 kg/m2) were noted to be lower in the fruit and vegetable group suggestive of wider metabolic benefits.[25] There were no significant differences in medication changes between the groups. Adverse events related to interventions were not reported.

Veverimer, a novel nonabsorbable polymer selectively binds and removes hydrochloric acid in the gastrointestinal tract and has been demonstrated to increase serum bicarbonate compared with placebo.[29] A phase 3, multicentre, double-blind RCT randomized patients with CKD (eGFR 20–40 ml/min/1.73 m2) and acidosis (12–20 mmol/l) to either veverimer (n = 124 participants) or placebo (n = 93 participants). The study demonstrated a sustained increase in serum bicarbonate with good tolerability and improved patient-reported physical functioning, with benefits extended to 40 weeks of follow-up.[23,24] A large phase 3 clinical trial is currently underway aiming to recruit 1600 CKD patients (eGFR 20–40 ml/min/1.73 m2) with acidosis (serum bicarbonate 12–20 mmol/l), examining the benefit of veverimer with respect to kidney outcomes (>40% reduction in eGFR, kidney failure or kidney death) ( The absence of an unwanted cation on acid binding makes veverimer an attractive treatment modality for acidosis in patients in whom additional sodium or potassium load may preclude other alkali treatment.

Summary of Randomized Controlled Trial Evidence for Bicarbonate Therapy

We recently published a systematic review and meta-analysis evaluating the effect of bicarbonate therapy on kidney outcomes with at least 3 months follow-up.[30]

The review included 15 trials involving 2445 participants. Compared with placebo or no study medication, treatment with sodium bicarbonate increased serum bicarbonate level [weighted mean difference (WMD) 2.59 mmol/l, 95% CI 1.51--3.66 mmol/l, heterogeneity I 2 = 95%, attributed to control intervention, very low certainty evidence). Compared with placebo or no study medication, treatment with sodium bicarbonate slowed the decline in kidney function from baseline to trial completion as measured by eGFR or creatinine clearance [standardized mean difference (SMD) 0.26, 95% CI 0.13--0.40; heterogeneity I 2 = 50% attributed to study quality as assessed by Jadad scoring, low certainty evidence). Compared with the control group, bicarbonate therapy reduced the risk of progression to kidney failure [6.7 vs. 14.2%, RR 0.53; 95% CI 0.32--0.89; heterogeneity I 2 = 64%, attributed to control intervention, low certainty evidence) resulting in 67 fewer cases per 1000 patients. There was no difference in change in proteinuria (SMD −0.09, 95% CI −0.27 to 0.09, heterogeneity I 2 = 16%, very low certainty evidence), SBP (WMD −0.57 mmHg, 95% CI −2.32 to 1.18 mmHg, heterogeneity I 2 = 0%, low certainty evidence) and DBP (WMD 0.88 mmHg, 95% CI −0.61 to 2.38 mmHg, heterogeneity I 2 = 27%, low certainty evidence) between the bicarbonate and control groups. Bicarbonate treatment did not increase the risks of peripheral edema (RR 1.16; 95% CI 0.90--1.50; I 2 = 27.6%, low certainty evidence), hospitalization for heart failure (RR 1.19; 95% CI 0.30--4.67; I 2 = 0%, very low certainty of evidence) and all-cause death (RR 0.81, 95% CI 0.39--1.68, I 2 = 30%, low certainty of evidence).

Although these results appear promising, the overall certainty of the evidence was low or very low and deemed insufficient to recommend bicarbonate treatment to slow the progression of CKD.