Acidosis and Kidney Allograft Survival
Year: 2017
Type: article
Abstract: The typical complications of CKD, such as anemia and hyperparathyroidism, are common in patients with kidney transplants. However, the mechanisms and management of each are often more complicated in the transplant setting. Metabolic acidosis, for instance, can develop because of reduced nephron number with insufficient acid excretion as in CKD, but immunologic and drug effects raise the stakes in patients with transplants. Particularly troublesome are the calcineurin inhibitors (CNIs) cyclosporin and tacrolimus, which cause renal tubular acidosis through different mechanisms. A curious effect of cyclosporin, but not tacrolimus, is that it blocks the conversion of HCO3−-secreting β-intercalated cells to H+-secreting α-intercalated cells through inhibition of peptidyl prolyl cis-trans isomerase in acidosis.1 Tacrolimus acts on several acid-base transport proteins in the proximal and distal tubules, including the Na+/H+ exchanger, the Cl−/HCO3− kidney anion exchanger (kAE1) of intercalated cells, and the Na+/HCO3− cotransporter (NBCn1).2 Nevertheless, the risk of acidosis seems to be higher with tacrolimus,3 the contemporary CNI of choice. Other contributing factors include diarrhea from mycophenolate, renal tubular acidosis from other medications (for example, renin-angiotensin system inhibitors and trimethoprin/sulfamethoxazole), and higher blood flow in the single kidney that might impair transtubular gradients and in turn, renal acid excretion.4 Metabolic acidosis can also be the presenting feature of rejection,5 possibly due to immunologically mediated injury of specific acid-base transport proteins in the kidney. These additional insults contribute to the higher prevalence of acidosis, roughly 20%–40% when defined as serum total CO2 (TCO2) <22 mEq/L, and the occurrence of acidosis at higher eGFR in kidney transplant recipients compared with CKD.6 In terms of outcomes, it is plausible that the higher prevalence and onset of acidosis at higher eGFR would translate into higher risk of acidosis-related adverse outcomes in kidney transplant recipients than in patients with CKD. Bones and muscles are affected by acidosis in kidney recipients as in CKD.7 But what about graft failure and mortality? Park et al.8 address this very important question in this issue of the Journal of the American Society of Nephrology. Different statistical techniques were used to investigate the relationship between serum TCO2 and these clinical outcomes in 2318 first-time kidney transplant recipients in South Korea over a median follow-up period of around 5 years, the largest acidosis study in kidney recipients to date. Using the 3-months post-transplant TCO2 value as the baseline, the risks of graft loss and death-censored graft failure (DCGF) were 74% and 66% higher, respectively, for those with acidosis compared with those with normal TCO2 (22–29.9 mEq/L). Mortality was 48% higher but not statistically significant, likely due to few deaths in the cohort (n=77). In analyses that treated TCO2 as a continuous variable, each 1-mEq/L lower TCO2 associated with approximately 10% higher risk of graft loss, mortality, and DCGF. Because TCO2 can vary over time, there is legitimate concern that a single value of TCO2 may not accurately reflect risk. This is particularly important in kidney transplant recipients if the index TCO2 is obtained fairly soon after transplantation, because CNI doses tend to be higher, kidney function may still be settling in, and other factors, such as acute rejection, may be occurring. Park et al.8 address this by considering TCO2 as a time-varying variable and treating eGFR and acute rejection episodes, which could influence TCO2 during follow-up, as either time-varying variables (in time-varying analyses) or time-varying confounders (in marginal structural models). Low TCO2 remained a significant risk factor for graft loss and DCGF, and an association with mortality emerged. In fact, the relationship between acidosis and these outcomes was much stronger, more than twofold higher, in these time-varying analyses than in analyses that used the 3-months post-transplant TCO2 as the index value. Thus, acidosis is a powerful predictor of poor renal and survival outcomes in kidney transplant recipients as it is in CKD. Importantly, the proportion of recipients with acidosis who were treated with alkali 3 months after transplant was quite low (5.4%) as was the proportion treated beyond 3 months (<10%). Likewise, <1% of those with normal TCO2, roughly 80% of the cohort, received treatment with alkali. Thus, the proportion of recipients treated with alkali was quite low: 1.5% at 3 months and somewhere between 1% and 3% during follow-up. This means that pharmacologic alkaline therapy probably had little effect on the outcomes, lending confidence that these results are probably not influenced by treatment. Indeed, there was no evidence of interaction between low TCO2 and alkaline therapy in the analyses. What is also noteworthy is that only about one quarter of the recipients received deceased donor kidneys, which was a risk factor for acidosis in this study. The higher proportion of living donor kidneys translated into a low occurrence of delayed graft function (1.2%). Moreover, only 9% of recipients lost the graft during follow-up. Thus, this might be considered a fairly healthy transplant cohort, which may explain the lower prevalence of acidosis in this cohort (16% at 3 months and approximately 10% beyond 6 months) compared with other transplant cohorts. Despite the low prevalence, there were substantially high risks of poor renal outcomes with acidosis. These risks might be higher in countries that transplant more deceased donor kidneys, older recipients, and/or a higher proportion of recipients with diabetes than in this cohort. This study in kidney transplant recipients highlights a need to change our mindset about what we consider as a warning TCO2. Most observational studies that have analyzed TCO2 as a categorical variable classify low TCO2 as <22 mEq/L, and most have found that low TCO2 is a risk factor for poor outcomes in CKD, non-CKD, and now, kidney recipients. However, results from this and other studies show that the risks of poor kidney outcomes and/or mortality are lowest with TCO2 values of 24–28 mEq/L.8–11 Compensatory mechanisms, including enhanced ammonia production and endothelin-1 activity,12,13 maintain systemic TCO2 in the normal range until these responses can no longer keep up with the acid load. It is conceivable that the reason why those with TCO2 in the 24–28 mEq/L range have lower risk of adverse outcomes than those with low-normal TCO2 is because they require less compensation to maintain normal TCO2 and therefore, have less kidney injury. These data from multiple cohorts support this notion and more importantly, suggest that we should seriously reconsider whether serum TCO2 of 22 mEq/L is the magic number in patients with renal disease. Finally, alkaline therapy, whether pharmacologic or nutritional, is a promising intervention in both CKD14 and renal transplant recipients. In a 1-year study of 30 kidney recipients with TCO2 of approximately 20–21 mEq/L, treatment with potassium citrate (n=19) normalized TCO2 and improved bone histology and turnover markers compared with potassium chloride (n=11).15 Whether this improves graft survival is uncertain. A clinical trial testing pleiotropic effects of alkali in kidney recipients is especially necessary to determine if this generally safe and inexpensive therapy can preserve graft function and improve patient survival. Disclosures None.
Authors: Kalani L. Raphael, Fuad S. Shihab
Institutions: University of Utah, VA Salt Lake City Healthcare System
Language: English
Subfield: Nephrology
Field: Medicine
Domain: Health Sciences
Open Access status: green
Awards: G1496914557, G4950270786
OpenAlex ID: W2600487014