Urinary Citrate Is Associated with Kidney Outcomes in Early Polycystic Kidney Disease

Key Points

• Low urinary citrate and crystal deposition accelerated cystogenesis in an experimental model of polycystic kidney disease (PKD).

• Hypocitraturia, frequently observed in patients with autosomal dominant PKD (ADPKD) could contribute to disease progression.

• Present findings suggest lower urinary citrate in early PKD was associated with faster eGFR decline and worse kidney survival.

Introduction

Autosomal dominant polycystic kidney disease (ADPKD) is the most common hereditary cause of CKD, often leading to kidney failure. Approximately one third of patients with ADPKD develop nephrolithiasis and, among other risk factors, hypocitraturia has been observed in around 50% of patients with early ADPKD (CKD stages G1/2) (1,2). Recent data suggest hypocitraturia could contribute to disease progression because, in an animal model of polycystic kidney disease, calcium microcrystal deposition triggered tubule dilation, cyst initiation, and growth, which was prevented by citrate therapy (3). In addition, in cross-sectional studies, urinary citrate levels correlated with total kidney volume (TKV) and eGFR (4,5). To address the hypothesis that lower urinary citrate levels could predispose patients with ADPKD to a faster disease progression, we performed a single-center, retrospective study. Our findings suggest lower urinary citrate is associated with faster eGFR decline and worse kidney survival in patients with ADPKD.

Materials and Methods

We collected data from the medical records of patients followed up at the ADPKD outpatient unit from the Nephrology Division of Universidade Federal de São Paulo during the period 2002–2021, including patients from our previous prospective study (1). This study was approved by our local ethics committee. Inclusion criteria were ultrasonographic diagnosis of ADPKD, according to Pei et al. and Ravine et al. (6,7), and one 24-hour urinary citrate level determined at baseline. Exclusion criteria are presented in Figure 1. Serum and urinary creatinine were determined by isotope dilution mass spectrometry and urinary citrate by a citrate lyase enzymatic reaction method (lower limit of detection, 20 mg/L) (8). We excluded 24-hour urine collections with biologically implausible urinary creatinine excretion from further analysis. Urine citrate levels were normalized for urinary creatinine concentrations to correct for possible incomplete urine collections. Urinary urea was used to estimate protein intake through the protein equivalent of nitrogen appearance equation (9). The estimation of net rate of endogenous noncarbonic acid (net endogenous acid production), derived from diet-composition data, was calculated as described by Frassetto et al. (10). Abdominal and renal ultrasound was performed using a Siemens (Erlangen, Germany) Antares ultrasound system, and kidney length, width, and anteroposterior diameter were determined to calculate TKV using a modified ellipsoid for each kidney (11). Missing values included TKV (n=10); serum values of bicarbonate (n=13), potassium (n=20), and uric acid (n=13); and urine volume (n=2), sodium (n=1), potassium (n=5), calcium (n=2), oxalate (n=24), magnesium (n=25), phosphate (n=25), and urea (n=8).

Figure 1.

Study flowchart diagram. ADPKD, autosomal dominant polycystic kidney disease.

Statistical Analysis

Independent-sample t tests or one-way ANOVAs were used for continuous variables, and chi-squared tests were used for categoric variables. Normally distributed or log-transformed data are presented as mean±SD, and skewed distribution as median and interquartile range. We used Kaplan–Meier and Cox regression models to assess the risk rate to a kidney outcome (eGFR decline of >40%, kidney failure, or initiation of kidney replacement therapy), and linear mixed models for association of urine citrate with eGFR slope. Statistical analyses used SPSS version 28.0 (SPSS Inc., Chicago, IL) and, in all analyses, P<0.05 was considered significant.

Results

We screened a total of 736 patients that visited our clinic, and included 95 patients in our analysis. (Figure 1). Included patients were 33±14 years old, predominantly female (66%), and had an eGFR of 91±29 ml/min per 1.73 m² (Table 1). Metabolic acidosis was not observed, even in the more advanced stages of CKD (serum bicarbonate of 26±2, 26±3, and 25±3 mmol/L for CKD stage 1, 2, and 3/4, respectively; P=0.20), with none of the patients using alkali therapy. Hypocitraturia (<1.67 mmol/24 h) was highly prevalent in 52 of 95 (55%) patients, regardless of the presence of nephrolithiasis, and was more frequent in men (Table 1). This difference could not be explained by eGFR levels or use of antihypertensive treatment. Urinary citrate was lower with advanced CKD stages (1.7±0.9, 1.5±0.9, and 0.9±0.8 mmol/24 h for CKD stages 1, 2, and 3/4, respectively; P=0.003), and larger TKV (382±202, 785±489, and 1138±873 ml for CKD stages 1, 2, and 3/4, respectively; P<0.001). The urinary 24-hour citrate/creatinine ratio (uCit/Cr) correlated with eGFR (R²=0.17; P<0.001) and with TKV (R²=0.22; P<0.001; Figure 2, A and B). Corrected for eGFR, age, and sex, uCit/Cr remained significantly associated with TKV (P=0.02; data not shown).

Table 1.

Baseline characteristics

Characteristics	All (n=95)	Normocitraturia (n=43)	Hypocitraturia (n=52)	P Value
Age (yr), mean±SD	33±14	31±13	35±14	0.24
Female sex, n (%)	63 (66)	35 (81)	28 (54)	<0.01
BMI (kg/m2), mean±SD	25±5	24±5	25±4	0.75
CKD stages, n (%)
G1	51 (54)	27 (63)	24 (46)
G2	30 (31)	14 (32)	16 (31)
G3/4	14 (15)	2 (5)	12 (23)	0.04
eGFR (ml/min per 1.73 m2), mean±SDa	91±29	98±24	86±31	0.04
TKV (ml), mean±SDb	546±393	435±269	667±502	<0.01
Hypertension, n (%)	39 (41)	14 (33)	25 (48)	0.15
Use of antihypertensive drugs, n (%)
ACEi	23 (24)	6 (14)	17 (33)
ARBs	8 (8)	3 (7)	5 (10)
Thiazides	20 (21)	8 (19)	12 (23)	0.73
Nephrolithiasis, n (%)	31 (33)	16 (37)	15 (29)	0.51
Serum, mean±SD
Creatinine (μmol/L)	89±25	79±18	98±28	<0.01
Bicarbonate (mmol/L)	26±2	25±2	26±3	0.06
Potassium (mmol/L)	4.2±0.4	4.1±0.3	4.3±0.5	0.27
Uric acid (mmol/dl)	0.31±0.10	0.30±0.10	0.33±0.09	0.23
Urine, mean±SD
Volume (L/d)	1.75±0.71	1.89±0.77	1.63±0.64	0.08
Creatinine (mmol/d)	12±4	12±4	11±4	0.21
Sodium (mmol/d)	196±79	214±87	180±68	0.04
Potassium (mmol/d)	48±19	50±23	47±15	0.55
Calcium (mmol/d)	2.19±1.82	3.05±1.58	1.66±1.56	<0.01
Oxalate (mmol/d)	0.31±0.11	0.33±0.11	0.30±0.12	0.30
Magnesium (mmol/d)	2.95±0.99	3.24±1.05	2.73±0.90	0.03
Phosphate (mmol/d)	22±9	23±10	21±8	0.38
Urea (mol/d)	0.60±0.29	0.59±0.30	0.61±0.28	0.66
Citrate (mmol/d)	1.48±0.97	2.54±0.91	0.95±0.45	N.T.
Citrate/Creatinine (mmol/mol)	133±92	217±98	88±52	N.T.
PNA (g/d)	62±24	61±26	62±23	0.75
NEAP (mEq/d)	64±30	60±29	67±32	0.29

BMI, body mass index; TKV, total kidney volume; ACEi, angiotensin-converting enzyme inhibitor; ARBs, angiotensin receptor blockers; N.T., not tested; PNA, protein equivalent of nitrogen appearance; NEAP, net endogenous acid production.

^aeGFR determined by Chronic Kidney Disease Epidemiology Collaboration equation.

^bTKV by ellipsoid ultrasound.

Figure 2.

Urinary 24 hr citrate/creatinine ratio correlates with eGFR and TKV and low urinary citrate is associated with faster eGFR decline and reduced kidney survival time in ADPKD patients. (A) Black dots indicate urine citrate/creatinine (uCit/Creat, mmol/mol, x axis) distribution along eGFR (y axis). Pearson correlation was used for log-transformed uCit/Creat. (B) Black dots indicate uCit/Creat (mmol/mol, x axis) distribution along total kidney volume (TKV; y axis). Pearson correlation was used for log-transformed uCit/Creat and TKV. (C) The lines display eGFR slopes with corresponding 95% CIs using univariate linear mixed models for uCit/Creat ratio divided into tertiles; tertile 1 (dashed line; range, 14–111 mmol/mol), tertile 2 (solid line; range, 112–170 mmol/mol), and tertile 3 (dashed-dotted line; range, 171–722 mmol/mol). (D) Median survival time to kidney outcome using Kaplan–Meier curve analysis according to tertiles of uCit/Creat: tertile 1, median, 9.0 (95% CI, 4.9 to 13.1) years; tertile 2, median, 16.8 (95% CI, 10.1 to 23.4) years; and tertile 3, median, 18.0 (95% CI, 14.7 to 21.3) years.

Low Urinary Citrate Is Associated with Faster eGFR Decline

The median duration of follow-up was 11 years (interquartile range 5–15 years), with a mean eGFR decline of 3.0 ml/min per 1.73 m² per year (95% CI, 2.5 to 3.5 ml/min per 1.73 m² per year). uCit/Cr was analyzed in tertiles, and those in the lowest tertile (tertile 1, Figure 2C) had an eGFR decline of 3.7 ml/min per 1.73 m² per year, compared with 3.3 ml/min per 1.73 m² per year (tertile 2) and 2.3 ml/min per 1.73 m² per year (tertile 3; P=0.04 tertile 1 versus tertile 3). On the basis of the univariable linear mixed model, each log unit decrease in uCit/Cr was associated with an eGFR decline of 2.3 ml/min per 1.73 m² per year (95% CI, 0.5 to 4.2 ml/min per 1.73 m² per year; P=0.01), independent of baseline eGFR.

Low Urinary Citrate Is Associated with Worse Kidney Outcomes

We analyzed the time to kidney failure using Kaplan–Meier survival analysis. Kidney replacement therapy was initiated in 11 patients, and no death was observed during follow-up. Lower uCit/Cr was associated with significantly reduced kidney survival time (median of 9.0 versus 18.0 years for tertile 1 versus tertile 3; P=0.002; Figure 2D). Furthermore, using Cox regression analysis, each log unit decrease in uCit/Cr was significantly associated with a five-fold higher risk for the composite kidney outcome (eGFR decline >40% and/or kidney failure), independent of age, sex, body mass index, hypertension (model 2), and baseline eGFR (model 3) (Table 2). The association was lost when TKV was added to the model.

Table 2.

Risk rate to kidney outcome using Cox regression models

Urine Citratea	Model 1b		Model 2c		Model 3d		Model 4e
	Hazard Ratio (95% Confidence Interval)	P Value	Hazard Ratio (95% Confidence Interval)	P Value	Hazard Ratio (95% Confidence Interval)	P Value	Hazard Ratio (95% Confidence Interval)	P Value
Urine citrate (mmol/24 h)	0.2 (0.04 to 0.6)	0.007	0.3 (0.06 to 1.2)	0.08
Urine citrate/creatinine (mmol/mol)	0.1 (0.04 to 0.4)	<0.001	0.2 (0.04 to 0.8)	0.02	0.2 (0.04;0.9)	0.03	0.3 (0.05 to 1.6)	0.2

^aVariables were log transformed.

^bModel 1: univariate.

^cModel 2: model 1 plus adjusted for age, sex, body mass index, hypertension, use of angiotensin-converting enzyme inhibitor.

^dModel 3: model 2 plus baseline eGFR.

^eModel 4: model 3 plus baseline total kidney volume.

Discussion

Several studies identified hypocitraturia as an early and common metabolic derangement in patients with ADPKD (1,2,4). Here, we demonstrate that urine citrate excretion correlates positively with eGFR and negatively with TKV. Furthermore, we show that lower urinary citrate is associated with a more rapid disease progression and worse kidney survival in patients with ADPKD who have a relatively preserved kidney function.

In our study, patients with hypocitraturia were more frequently men, and they had larger TKVs and lower eGFR levels compared with those without hypocitraturia. This raises the question whether hypocitraturia is a cause or consequence of kidney disease progression. Approximately half of our study participants with ADPKD and a preserved kidney function (CKD stage G1) had hypocitraturia. This finding is comparable to previous studies in patients with ADPKD or other forms of CKD (4,5). One proposed explanation for early onset hypocitraturia in ADPKD is an impaired ammoniagenesis (12). This tubular defect leads to eubicarbonatemic metabolic acidosis and, consequently, increased reabsorption of urinary citrate (13), in line with a recent report showing that low serum bicarbonate, even within the normal range, is associated with worse kidney outcomes in ADPKD (14). Moreover, spot uCit/Cr ratio is a surrogate marker for acid-base status in CKD (15). Low urinary citrate may, therefore, be an early consequence of disease progression in ADPKD because, despite marked enlargement and distortion of kidney anatomy, GFR remains stable until later stages of the disease. Importantly, hypocitraturia in our cohort could not be accounted for by a higher intake of acid precursors in the diet because its indirect assessment by net endogenous acid production did not differ between groups. We also demonstrate, for the first time, to the best of our knowledge, that lower urinary citrate levels were associated with rapid disease progression and with reduced kidney survival. The latter analysis was corrected for important determinants of disease progression. Although our study cannot prove causality between hypocitraturia and a faster kidney function decline, recent animal models of ADPKD support a direct link between hypocitraturia and kidney injury through microcrystal deposition and activation of the mammalian target of rapamycin pathway (3). Stimulation of the citrate carrier transporter in the inner mitochondrial membrane driven by acidosis may play a role in reduced citrate excretion but, given the changes in renal metabolism represented by enhanced aerobic glycolysis and shifts into lactate production, may reduce the availability of mitochondrial citrate content in renal cells (16). Our data suggest urinary citrate may add to prognostic models of ADPKD and, on the basis of animal studies, may even be targeted to slow disease progression (3,17). Interestingly, the percentage of patients with nephrolithiasis was not different between hypocitraturic and normocitraturic groups, suggesting hypocitraturia is a feature of ADPKD regardless of the presence of kidney stones, as previously observed (1).

The strength of this study is the inclusion of patients with ADPKD who have a relatively preserved kidney function, and the long median follow-up time of 11 years. None of our included patients used alkali therapy and we were able to correct for multiple confounders, including established risk factors for progression. However, a number of limitations should be mentioned. First, the higher risk for a kidney outcome was lost when we corrected for TKV, likely because TKV is a strong biomarker for disease progression. Alternatively, these findings suggest urinary citrate could substitute TKV as a more cost-effective biomarker. Second, the retrospective character of our study may have introduced a selection bias of patients with nephrolithiasis, although the majority (64%) of included patients were derived from our previous prospective cohort, in which urine metabolic profiling and TKV measurements were performed nonselectively (1). Third, the limited number of included patients did not allow us to perform a multivariable analysis to assess the effect of urinary citrate on eGFR decline. However, the findings from our survival analysis did allow a correction for other disease determinants. Finally, we measured TKV on the basis of the ultrasound ellipsoid formula because computed tomography/magnetic resonance imaging data were unavailable. We recognize that underestimation of TKV using ultrasound might have occurred (18).

In conclusion, low urinary citrate excretion is associated with a faster eGFR decline and worse kidney survival in patients with ADPKD, representing a potential marker of rapid disease progression. Prospective studies are warranted to confirm our findings.

Disclosures

E.J. Hoorn reports serving on the editorial boards for American Journal of Physiology – Renal Physiology, JASN, and Journal of Nephrology; receiving research funding from Aurinia; serving as a board member for the Dutch Federation of Nephrology and the European Renal Association Working Group on Inherited Kidney Diseases; and receiving honoraria from UpToDate. M. Salih reports serving on the scientific advisory board of NedMed.nl, and receiving research funding from Otsuka (unrestricted grant). All remaining authors have nothing to disclose.

Funding

This study was supported by Conselho Nacional de Desenvolvimento Científico eTecnológico—CNPq, grant 314677/2021-6 (I.P.H.) and Fundação Oswaldo Ramos—Hospital do Rim. D.R.D.R., H.M.G. are supported by a study scholarship from Coordenação deAperfeiçoamento de Pessoal de Nível Superior—CAPES.

This research was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Hospital do Rim.

Author Contributions

D.R. Rocha was responsible for investigation; D.R. Rocha, I.P. Heilberg, A.C.C. Matos, E.J. Hoorn, and M. Salih conceptualized the study; D.R. Rocha, H.M.G. Sousa, I.P. Heilberg, M. Salih, and L. Xue were responsible for formal analysis; D.R. Rocha, A.C.C. Matos, H.M.G. Sousa, and L. Xue were responsible for validation; D.R. Rocha and H.M.G. Sousa were responsible for data curation; I.P. Heilberg and M. Salih provided supervision; and all authors wrote the original draft, reviewed and edited the manuscript, and were responsible for methodology.

Data Sharing Statement

Data used for this study were previously published as “Evaluation of Nephrolithiasis in Autosomal Dominant Polycystic Kidney Disease Patients” (1). Data to support the findings in the study are available from the corresponding author on reasonable request.