#NephJC Chat
Tuesday June 6th 9 pm EDT
Wednesday June 7th 9 pm IST
Lancet 2023 Feb 18;401(10376):557-567. doi: 10.1016/S0140-6736(22)02606-X. Epub 2023 Jan 25.
Sodium bicarbonate for kidney transplant recipients with metabolic acidosis in Switzerland: a multicentre, randomized, single-blind, placebo-controlled, phase 3 trial
Nilufar Mohebbi, Alexander Ritter, Anna Wiegand, Nicole Graf, Suzan Dahdal, Daniel Sidler, Spyridon Arampatzis, Karine Hadaya, Thomas F Mueller, Carsten A Wagner, Rudolf P Wüthrich
PMID: 36708734
Introduction
Prolonging kidney graft survival and preventing a return to dialysis is the raison d’etre of every transplant nephrologist. Retarding the progression of chronic kidney disease (CKD) through control of traditional risk factors and inhibition of the final common pathway of relentless progression has always been an integral part in the modus operandi of all nephrologists, often irrespective of the etiology. Correction of chronic metabolic acidosis potentially plays a role in slowing progression of CKD and possibly in slowing graft loss in kidney transplant recipients.The Preserve- Transplant study investigated treating metabolic acidosis to slow progression eGFR decline in kidney transplant recipients.
The implications of chronic metabolic acidosis in CKD are multitude. Cardiac health, bone health, nutritional status (muscle mass and protein energy wasting) and possible decreased mortality have been attributed to the correction of metabolic acidosis. The evidence and benefits of correcting metabolic acidosis varies from the unbelievable (600 mg of sodium bicarbonate TID slows CKD progression by a factor of 2!) to no effect at all. In CKD, prevalence and severity of acidosis increases as eGFR declines (Moranne, JASN 2009) and guidelines, including those from KDIGO, recommend keeping serum bicarbonate level above 21 mmol/l with oral supplements, unless contraindicated (guideline 3.4.1, 2B recommendation). Contraindications are quite limited in routine clinical practice, though intolerance is common. Perhaps surprisingly, the side effects of fluid overload and worsening hypertension with oral sodium bicarbonate supplements are often absent in prior studies treating metabolic acidosis in CKD (though admittedly most studies excluded participants with uncontrolled hypertension or heart failure from enrollment). Over correction is also not recommended, and serum bicarbonate levels > 32 mmol/l are associated with increased mortality (Sankar D. Navaneethan et al, CAJSN 2011).
Low cost oral supplementation with sodium bicarbonate is hypothesized to delay progression of CKD through protection of the proximal tubules and inhibition of the complement cascade activation. In a small study with 134 participants with advanced CKD stage 4 and rapid progression (control arm decline of CrCl was >3 ml/min/1.73m2), sodium bicarbonate supplements to correct acidosis retarded CKD progression and the occurrence of ESKD.
Mechanism of decline in GFR due to metabolic acidosis.
From: Łoniewski et al, KI, 2014
Sustained correction of acidosis is generally achieved with oral bicarbonate supplements and improvement in serum bicarbonate and pH levels are expected (Ione de Brito-Ashurst et al, JASN 2009). Further, the open label UBI study (Biagio R. di lorio et al, Journal of Nephrology 2019) strengthened the evidence for correction of acidosis in advanced CKD (stage 3-5), showing benefits in terms of doubling of serum creatinine (HR 0.36; (95% CI) 0.22–0.58), time to initiation of RRT (HR 0.5; 95% CI 0.31–0.81), and all cause mortality (HR 0.43; 95% CI 0.22-0.87). Whether this benefit achieved by moving serum bicarbonate from 22 to 26 mmol/L is plausible (noting an effect size much bigger than RASi or flozins, for example), is up for debate.
Conclusions from studies designed to slow CKD progression from the non-transplanted population often cannot be extrapolated to patients with transplanted kidneys. Invariably, patients with transplanted kidneys are excluded from CKD studies due to their complex nature and risk for medication interactions that might accelerate kidney graft loss. In fact, a 2019 retrospective cohort study cast doubt about the benefits of sodium bicarbonate treatment of chronic metabolic acidosis in renal transplant recipients. It showed that 46% of transplant recipients had metabolic acidosis in the first year after transplant. Unfortunately, patients who received alkali therapy had a higher rate of graft loss than patients with metabolic acidosis who did not receive alkali therapy (HR 1.52; CI 1.03-2.25) (Schulte et al, Transplant Direct 2019). Different mechanisms are postulated for the higher incidence of graft loss in patients receiving sodium bicarbonate therapy, including the possibility of impaired MMF absorption from increased gastric pH. Metabolic acidosis per se was not associated with increased frequency of death, graft failure, or bone fractures in the above study.
Until now, there haven't been any prospective studies on the effect of sodium bicarbonate treatment to prevent progressive loss of function in kidney transplant recipients.
The Study
Methods
This was a prospective, multi-center, randomized, placebo-controlled, single-blinded trial. Kidney transplant patients >18 years of age, who had their transplant for at least a year with a stable renal function over 6 months (less than 15 % change in serum creatinine) were recruited from three university hospital centers: Zurich, Bern, and Geneva in Switzerland.
The cut off for metabolic acidosis was serum bicarbonate level less than or equal to 22 mmol/L. 240 recipients with an eGFR of 15-89 ml/min/1.7m2 were recruited and prospectively followed for a period of 2 years. Recruited patients were randomized to either sodium bicarbonate or matched placebo in a 1:1 ratio.
Participants received acid-resistant soft gelatine capsules of 500 mg sodium hydrogen carbonate (Nephrotrans) or a matching 500 mg placebo capsule. Both, placebo and study drug capsules had identical appearances and doses were titrated if bicarbonate levels remained less than 22 mmol/l, up to a dose of 4500 mg (3-9 capsules) per day.
The primary outcome was the difference in eGFR slope over a 2-year follow-up period. Secondary outcomes examined differences in albuminuria by UACR, effects on blood pressure assessed by 24 hour ABPM and differences in serum bicarbonate and pH.
Adverse events of special interest included: new onset or worsening hypertension, extremes of serum bicarbonate levels of more than 35 mmol/l and less than 16 mmol/l, serum sodium greater than 150 mmol/l with or without features of fluid overload, and hypokalemia of less than 3 mmol/l.
Statistics
The study was designed to achieve a power of 82% to detect a decline in eGFR of 1.5 ml/min/1.73m2 annually in the placebo group to a significance level of 0.05 (with an assumption of 0.0 ml/min/1.73m2 annual eGFR decline in the treatment group), which would require 240 participants. Both intention to treat and per-protocol analysis were performed.
Funding
This investigator initiated trial was funded by the Swiss National Science Foundation and the funder of the study had no part in the study from initiation to completion. Medice (Iserlohn, Germany) funded in part the study medication (sodium bicarbonate) and manufactured the matching placebo.
Results
In this Preserve-Transplant study, a total of 240 patients were included in the intention-to-treat analysis. 44 dropped out and per-protocol analysis was performed in 196 patients (99 [51%] in the placebo group and 97 [49%] in the sodium bicarbonate group).
Figure 1 Trial profile from Nilufar Mohebbi et al, Lancet, 2023
Baseline characteristics were similar and comparable in both treatment and placebo groups, except for slightly higher prevalence of a second transplant population and history of rejection in the placebo group (14 and 21% compared to 9 and 16% in the treatment group). Diabetes as the primary renal disease leading to ESKD was found to be slightly higher in the placebo group (7% and 4%). Again, the placebo arm had a higher prevalence of proteinuria, in both > 0.5.g per day (19% vs 16%) and > 1g per day (8% vs 5%) categories. Stratified multi-block randomization lead to a comparable distribution of age (55.7 vs 55.3 years), time since transplantation (9.1 vs 8.6 years) and eGFR at baseline (48.3 vs 47.7 ml/min/1.73m2), in sodium bicarbonate and placebo groups respectively.
Table 1- Baseline characteristics of study population from Nilufar Mohebbi et al, Lancet 2023
Primary Outcome
The primary outcome of eGFR slope was similar between treatment and placebo groups in both the intention-to-treat analysis and per-protocol analyses.
Table 2 from Nilufar Mohebbi et al, Lancet 2023. Median and mean eGFR slopes for intention-to-treat and per-protocol analyses, stratified by treatment group.
Secondary Outcomes
Correction of bicarbonate levels was achieved in the treatment group within three months, and mean serum bicarbonate rose from 21.3 mmol/l (SD 2.6) mmol/l to 23.0 mmol/l. Throughout the study period, serum bicarbonate remained high in the treatment group with a mean bicarbonate level of 22.5 mmol/l, whereas the mean was 20.7 mmol/l in the placebo group (P<0.001). The pH levels also had a similar statistically significant difference between groups, with a mean group difference in blood pH between month 3 and month 24 of 0.03 (95% CI 0.02–0.04).
Figure 2 from Nilufar Mohebbi et al, Lancet 2023
At 2 years of follow-up, urine albumin-to-creatinine ratio was 5.7% (95% CI - 21.8 to 42.9) higher in bicarb arm, a result without statistical significance (p=0.72).
Further, prespecified subgroup analysis did not show differences in regard to eGFR slopes in any of the subgroups analyzed: time since transplantation, type of transplant (deceased vs living donor), HLA sensitisation, type of immunosuppression, baseline serum bicarbonate, and baseline eGFR. Prespecified sensitivity analysis confirmed robustness of the primary analysis and showed no difference in eGFR slopes in subgroups.
Figure 3 from Nilufar Mohebbi et al, Lancet 2023. Effect of sodium bicarbonate versus placebo on eGFR slope, overall and according to baseline subgroups.
Office blood pressure measurements and ABPM showed no difference between the two groups in the study period.
Table S8 from Nilufar Mohebbi et al, Lancet 2023. Blood pressure measurements show no difference between the groups. Data are mean (SD) in mmHg.
Safety
Adverse events of special interest occurred 92 times during the study, 57 in the placebo group (23% of patients) and 35 in the sodium bicarbonate group (15% of patients). Not surprisingly, a serum bicarbonate <16 mmol/L was twice as frequent in the placebo group. Hypokalaemia (<3 mmol/L) was infrequent but more common in the sodium bicarbonate group. There were 77 serious adverse events reported, 47 in the placebo group (27% of patients) and 30 in the sodium bicarbonate group (16% of patients). A total of 23 patients in the placebo group and 27 patients in the bicarb group prematurely ended the trial.
Table S1B from Nilufar Mohebbi et al, Lancet 2023. Detailed reasons for premature study end in selected categories. Data are number of patients.
Discussion
Phosphate, PTH, bicarbonate - outpatient nephrologists often find themselves tempted to intervene and “fix numbers”, leading to a high pill burden for patients. The theories under-pinning correction of lab values, without hard end-points, are often strongly held dogma. However, under the scrutiny of randomized trials, some of these targets have been shown to be false idols (see uric acid), or even harmful (see normalization of hemoglobin with EPO). This study adds to the list. Increasing the serum bicarbonate by ~1.7mmol/L from a baseline of 21 mmol/L for 2 years didn’t do anything for eGFR in a stable transplant population.
What did sodium bicarbonate tablets achieve? Well, they certainly corrected the serum bicarbonate levels within 3 months of initiation, and resulted in sustained improvement in serum pH (+0.03) throughout the study period. Side effects that some may expect with sodium loading did not manifest, with no change in blood pressure or fluid overload status. Pill burden needs to be balanced against evidence of measurable clinical benefits. There was no evidence for the primary endpoint of slowing graft failure in this study, and other potential beneficial effects (for example bone health) were beyond its scope. In the end there wasn’t evidence in the Preserve-Transplant Study to support routine bicarbonate supplementation in this group. On the other hand, the ill-effects on eGFR of sodium bicarbonate indicated by the Schulte retrospective study were also not bourne out in this prospective paper. It revealed an acceptable safety profile, and there was no increased rejection observed to support the hypothesis of decreased MMF absorption with treatment.
The potential for this study to show a difference in eGFR outcomes was always a tall ask. Enrollment of stable kidney transplant recipients with near normal bicarbonate levels (in this study, more than 90% percent of the study population had a bicarbonate level > 18 mmol) gives little opportunity for separation of outcomes by group. Follow up was relatively short, though separation of groups in some of the non-transplant CKD trials was seen over this time period. Benefits of bicarbonate treatment in transplant recipients with accelerated graft loss and more severe acidosis remains to be explored prospectively. Calculation of the sample size to achieve a power of 82% was based on the assumption the eGFR decline would be 0.0 ml/min/1.73m2 in the bicarbonate treatment arm, this assumption truly fell short and felt unrealistic. Moreover, the eGFR slope (used as a surrogate end point for kidney disease progression) in clinical trials, detects differences in total eGFR slope better when follow up is longer, particularly when expected differences are minimal.
Questions remain to be answered, whether bicarbonate treatment would be effective in patients with rapidly worsening graft function, with perhaps an eGFR drop more than 3 ml/min.1.73m2 per year. Moreover, correction of metabolic acidosis was targeted to a level of 22 mmol/l (correction of bicarbonate levels can not be equated to correction of acidemia as eubicarbonatemic acidosis is always possible). However, another trial with a higher target level (24-28 mmol/L, as seen in UBI), more rapid eGFR loss and more severe acidosis at baseline would be hard to justify in this population. The pill burden alone would make it impractical.
Conclusion
So should we be routinely using sodium bicarbonate supplementation to treat metabolic acidosis in renal transplant recipients? There is currently no justification for routine use of oral bicarbonate supplements for correction of mild chronic metabolic acidosis in transplant recipients to prevent graft failure. Recipients with severe acidosis and those with a rapid rate of eGFR decline have not been thoroughly examined. Perhaps inclusion of patients with rapidly deteriorating graft function, with a target of higher bicarbonate levels (as it is proven safe so far) warranted in future studies to show potential beneficial or harmful effects.
Summary by Brammah R Thangarajah
Senior Lecturer and Consultant Nephrologist
Department of Medicine, University of Jaffna, Sri Lanka
NSMC Intern 2023