Abstract
Aims
This retrospective cohort study aimed to be the first to evaluate the association between plasma protein‐bound uremic toxins (PBUTs) concentrations, echocardiographic parameters of heart failure (HF), and incident HF events in patients with chronic kidney disease (CKD) not on dialysis.
Methods and results
Retrospective, single‐centre, cohort study at the Ghent University Hospital, Belgium. Adults with CKD stages G1–G5, not on dialysis, could be included. Exclusion criteria were ongoing pregnancy, age <18 years, active acute infection, active malignancy, history of transplantation, or a cardiovascular event within 3 months prior to inclusion. Free and total concentrations of five PBUTs were quantified at baseline: indoxyl sulfate (IxS), p‐cresyl sulfate (pCS), p‐cresyl glucuronide (pCG), indole‐3 acetic acid (IAA), and hippuric acid (HA). Patients were grouped into three echocardiographic categories: normal left ventricular ejection fraction (LVEF) and normal left ventricular end‐diastolic pressure (LVEDP), normal LVEF and increased LVEDP, and reduced LVEF, based on available echocardiographic data in a time interval of ±6 months around the plasma sample collection. A total of 523 patients were included between January 2011 and January 2014. Echocardiographic data within the predefined timeframe were available for 210 patients (40% of patients). Levels of pCG and pCS were significantly higher in patients with reduced (<50%) versus normal LVEF (P < 0.05). After a median follow‐up 5.5 years, 43 (8.4%) patients reached the composite endpoint of hospitalization or mortality due to HF. Free fractions of IxS, pCS, and pCG showed the strongest association with clinical outcome: free IxS: HR 1.71 (95% CI 1.11–2.63; P = 0.015), free pCS: HR 1.82 (95% CI 1.11–3.01; P = 0.019), and free pCG: HR 1.67 (95% CI 1.08–2.58; P = 0.020), and these results were independent of age, gender, body mass index, diabetes, and systolic blood pressure. In models that were also adjusted for serum creatinine, the free fractions of these PBUTs remained significant.
Conclusions
Elevated free concentrations of IxS, pCG, and pCS were independently associated with an increased risk of HF events in non‐dialysed CKD patients. Further research is necessary to confirm these findings and investigate the potential impact of PBUT‐lowering interventions on HF events in this patient group.
Keywords: Biomarkers, Chronic kidney disease, Heart failure, Uremic toxins
Introduction
Patients with chronic kidney disease (CKD) have a mortality rate more than double the rate in the general population and more than half of deaths are from cardiovascular disease (CVD). 1 In fact, CKD patients are at greater risk of cardiovascular events or death than of progression to kidney failure. 2 Heart failure (HF) is a leading cause of morbidity and mortality in this population and the amount of patients with concurrent CKD and HF continues to grow. 1 , 3 , 4 This leads to poor health‐related quality of life and has a serious impact on the health care system. 5
Though traditional cardiovascular risk factors, such as arterial hypertension, diabetes mellitus, and obesity, are highly prevalent in the population with CKD, they do not fully account for this marked increase in CV burden. 6 Despite a good control of these traditional risk factors, CV mortality in patients with CKD remains high. 3 In the past decades, increased uremic toxin (UT) concentrations have emerged as a non‐traditional risk factor for CVD in patients with CKD, which can at least partially explain this discrepancy. 7 They are residues of organic compounds that are cleared from the body via urine and thus accumulate in the bloodstream in patients with CKD. 7 In 2003, the European Uremic Toxin Work Group (EUTox) classified UTs into three groups according to their physicochemical characteristics: (a) small water‐soluble compounds (<500 Da), which are easily removed by haemodialysis, (b) middle molecules (>500 Da), which can only be removed by dialytic membranes, which contain large pores, and (c) protein‐bound compounds. 8 The latter group largely originates from the metabolism of dietary amino acids and removal by dialysis is limited to the unbound fraction. 9 This definition and classification has recently been further finetuned. 10
Of all identified UTs, PBUTs are highly ranked for biochemical, pathophysiological and clinical evidence of toxicity 11 and a causative relationship between these PBUTs and cardiovascular events and mortality has been increasingly demonstrated, mainly for indoxyl sulfate 12 , 13 , 14 , 15 and the p‐cresol derivatives. 16 , 17 , 18 , 19 , 20 However, other studies failed to show such an association. 21 , 22 Additionally, it should be noted that most studies included patients on dialysis, which means that the results might not be fully representative of patients with non‐dialysed CKD. Furthermore, data on the association of these toxins with echocardiographic parameters and incident HF remain scarce.
Therefore, the aim of this study was to retrospectively evaluate the relationship between echocardiographic parameters of HF, incident HF events, and plasma PBUT levels in a population with CKD, not on dialysis.
Methods
Study population and ethics statement
This retrospective, single‐centre, cohort study was carried out at the Nephrology outpatient clinic of the Ghent University Hospital in Belgium, a >1000‐beds tertiary care centre. Inclusion took place between January 2011 and January 2014. Outcome parameters were monitored until June 2017. The study was approved by the local ethical committee (2010/033; B67020107926, 11 June 2010). Written informed consent was obtained from all participants. A total of 523 patients with CKD stages G1–G5 (as defined by the Kidney Disease Outcomes Quality Initiative guidelines (KDOGI)), not on dialysis, were included. Exclusion criteria were ongoing pregnancy, age <18 years, active acute infection, active malignancy, history of transplantation, or a cardiovascular event within the past 3 months.
Data collection and definition of outcome
Demographics, pre‐existing co‐morbidities, chronic medication, and HF outcome parameters were prospectively recorded by consulting the medical records and/or contacting the patient's general practitioner whenever information on the cause of death was missing.
The concentrations of five important PBUTs were determined at baseline: indoxyl sulfate (IxS), p‐cresyl sulfate (pCS), p‐cresyl glucuronide (pCG), indole‐3 acetic acid (IAA), and hippuric acid (HA). Concentrations were determined by ultrahigh performance liquid chromatography with ultraviolet and fluorescence detection. A detailed protocol can be found elsewhere. 19
Echocardiography was not routinely performed for study purposes, though if the patient had an echocardiographic examination in a time interval of ±6 months around the plasma sample collection, these data were retrieved from the electronic cardiac health record. All images were stored in the Picture Archiving and Communication System of the hospital. Echocardiographic measurements included in the study were global left ventricular ejection fraction (LVEF) (measured using eyeballing, fractional shortening and the biplane Simpson's method whenever a reduced ejection fraction was seen by eyeballing), measures of diastolic function (mitral diastolic inflow pattern with E and A velocities, E/A ratio, E wave deceleration time and tissue Doppler myocardial velocities obtained at the medial and lateral mitral annulus (e′ medial, e′ lateral)), as well as measures of left ventricular geometry [end‐diastolic interventricular septum and posterior wall thickness (resp. IVSd and PWd), end‐diastolic and end‐systolic diameters (resp. LVEDD and LVESD)] and end‐systolic left atrial volume (measured using biplane Simpson's method). From the above‐mentioned data, left ventricular mass index (LVMI) was calculated using the Devereux formula: LVMI = (0.8 × (1.04 × (((LVEDD + IVSd + PWd)3 − LVEDD3))) + 0.6)/body surface area. An increased LVMI was defined as >115 g/m2 (men) and >95 g/m2 (female). Elevated left ventricle end‐diastolic pressure (LVEDP) was defined as E/e′ average greater than 14. Although ASE/EACVI Guidelines' algorithm for determining filling pressures also includes LA volume and peak tricuspid regurgitation velocity, we decided to use E/e′ as the only marker for elevated filling pressures as the left atrial volume is often undertraced and right ventricular systolic pressure cannot always be calculated reliably from the peak tricuspid regurgitation velocity. Lastly, the left atrial volume index (LAVI) was calculated by dividing the left atrial volume by the body surface area. An elevated LAVI was defined as a value greater than 34 mL/m2.
Outcome parameters
Based on the echocardiographic data of LVEF and LVEDP three subgroups were made: (i) normal LVEF and normal LVEDP; (ii) normal LVEF and elevated LVEDP; and (iii) reduced LVEF. The clinical outcome parameter was composed of hospitalization and mortality due to HF.
Data analysis
The statistical analysis was performed using SPSS statistics (Version 28.0, IBM Corp, Armonk, NY). Categorical variables are shown as frequencies, and continuous variables as mean (standard deviation) or median (interquartile range) depending on the distribution. Differences in PBUT plasma levels between the three echocardiographic groups were evaluated by a Kruskal–Wallis test with additional pairwise post hoc comparisons using Bonferroni correction for multiple testing. Kaplan–Meier survival curves were constructed to evaluate the univariate association between PBUTs and the composite endpoint of hospitalization or death due to HF. To investigate the independent prognostic value of UTs for the clinical outcome, Cox proportional hazards models were fitted; hazard ratios (HRs), 95% confidence intervals (CIs), and statistical significances were estimated from these models. The models were adjusted for age, gender, body mass index, diabetes, and systolic blood pressure. All tests were two‐sided, and a P value < 0.05 was considered significant.
Results
Baseline characteristics
Between January 2011 and January 2014, a total of 523 patients were included. Baseline characteristics and UT concentrations have been published previously. 19 In brief, median age was 66 years and 58% of the population was male. Mean body mass index was 27.4 kg/m2. Patients had CKD stage G1 (12%), G2 (16%), G3a (21%), G3b (28%), G4 (19%), and G5 (not on dialysis, 4%). The most prevalent co‐morbidities were arterial hypertension (89%), diabetes mellitus (34%), coronary artery disease (21%), and peripheral vascular disease (18%). Chronic cardiac medication consisted mostly of lipid lowering therapy (statin and/or ezetimibe 63%), RAAS inhibitors (angiotensin‐converting enzyme inhibitors or angiotensin receptor blockers 76% and aldosterone receptor antagonists 4%), beta‐blockers (47%), and antiplatelet therapy (46%) (Table 1).
Table 1.
baseline patient characteristics
| Total cohort (N = 523) | Group 1 (N = 109) | Group 2 (N = 42) | Group 3 (N = 49) | |
|---|---|---|---|---|
| Age (y) | 66 (50–76) | 65 (51–73) | 73 (67–81) | 73 (65–81) |
| Gender | ||||
| Male | 58% (304/523) | 64% (70/109) | 45% (19/42) | 69% (34/49) |
| Female | 42% (222/523) | 36% (39/109) | 55% (23/42) | 31% (15/49) |
| BMI (kg/m2) | 27.4 (24.0–31.0) | 27.5 (24.3–31.5) | 29.4 (26.5–33.0) | 28.0 (25.0–32.3) |
| Blood pressure at baseline | ||||
| Systolic (mmHg) | 131 (119–145) | 135 (120–149) | 145 (129–164) | 130 (120–143) |
| Diastolic (mmHg) | 80 (72–87) | 83 (75–90) | 77 (71–86) | 73 (66–84) |
| Smoking status | ||||
| None‐smoker | 52% (271/523) | 49% (53/109) | 57% (24/42) | 43% (21/49) |
| Former smoker | 36% (191/523) | 40% (44/109) | 31% (13/42) | 49% (24/49) |
| Active smoker | 12% (64/523) | 11% (12/109) | 12% (5/42) | 8% (4/49) |
| KDOQI class | ||||
| 1 | 12% (64/523) | 11% (12/109) | 0% (0/42) | 4% (2/49) |
| 2 | 16% (86/523) | 17% (19/109) | 12% (5/42) | 14% (7/49) |
| 3A | 21% (111/523) | 20% (22/109) | 24% (10/42) | 10% (5/49) |
| 3B | 28% (146/523) | 30% (33/109) | 43% (18/42) | 39% (19/49) |
| 4 | 19% (101/523) | 17% (18/109) | 19% (8/42) | 33% (16/49) |
| 5 (non‐dialysis) | 4% (18/523) | 5% (5/109) | 2% (1/42) | 0% (0/49) |
| Co‐morbidities | ||||
| Arterial hypertension a | 89% (469/523) | 94% (102/109) | 98% (41/42) | 96% (47/49) |
| Diabetes mellitus | 34% (176/523) | 37% (40/109) | 69% (29/42) | 51% (25/49) |
| Peripheral vascular disease | 18% (92/523) | 20% (20/109) | 13% (5/42) | 17% (8/49) |
| Cerebrovascular disease | 13% (70/523) | 13% (14/109) | 19% (8/42) | 18% (9/49) |
| Coronary artery disease | 21% (111/523) | 15% (16/109) | 33% (14/42) | 67% (33/49) |
| Pre‐existing heart failure | 12% (62/523) | 4% (4/109) | 17% (7/42) | 45% (22/49) |
| Atrial fibrillation | 13% (69/523) | 13% (14/109) | 14% (6/42) | 33% (16/49) |
| Cardiac medications | ||||
| ACEi/ARB | 76% (396/523) | 77% (84/109) | 69% (29/42) | 73% (36/49) |
| MRA | 4% (23/523) | 4% (4/109) | 10% (4/42) | 6% (3/49) |
| Betablocker | 47% (245/523) | 49% (53/109) | 69% (29/42) | 69% (34/49) |
| Diuretics (loop diuretic & thiazides) | 48% (253/523) | 42% (46/109) | 62% (26/42) | 73% (36/49) |
| (Non)‐dihydropyridine CCB | 38% (201/523) | 36% (39/109) | 57% (24/42) | 39% (19/49) |
| Antiplatelet drugs (aspirin, P2Y12‐inhibitors) | 46% (240/523) | 48% (52/109) | 64% (27/42) | 82% (40/49) |
| Lipid lowering therapy (statin & ezetemibe) | 63% (331/523) | 64% (70/109) | 69% (29/42) | 84% (41/49) |
| Echocardiography (N = 210) | ||||
| LVEF (N = 210) | ||||
| Normal (>50%) | 161 (77%) | 100% (109/109) | 100% (42/42) | 0% (0/49) |
| Mildly reduced (40–50%) | 32 (15%) | 0% (0/109) | 0% (0/42) | 65% (32/49) |
| Reduced (<40%) | 17 (8%) | 0% (0/109) | 0% (0/42) | 35% (17/49) |
| LVEDP (N = 197) (mmHg) | ||||
| Normal | 128 (65%) | 100% (109/109) | 0% (0/42) | 42% (17/41) |
| Elevated | 69 (35%) | 0% (0/109) | 100% (42/42) | 58% (24/41) |
| LVMI (N = 197) g/m2 | ||||
| Normal | 133 (68%) | 85% (87/102) | 53% (21/40) | 40% (19/47) |
| Increased | 64 (32%) | 15% (15/102) | 47% (19/40) | 60% (28/47) |
| LAVI (N = 192) mL/m2 | ||||
| Normal | 101 (53%) | 67% (66/98) | 46% (19/41) | 33% (14/43) |
| Increased | 91 (47%) | 33% (32/98) | 54% (22/41) | 67% (29/43) |
ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; CCB, calcium channel blocker; KDOQI, Kidney Disease Outcomes Quality Initiative; LAVI, left atrial volume index; LMVI, left ventricular mass index; LVEDP, left ventricular end‐diastolic pressure; LVEF, left ventricular ejection fraction; MRA, aldosterone receptor antagonist.
Defined as taking at least 1 antihypertensive medication class.
Echocardiographic data within the predefined timeframe were available for 210 patients (40% of patients). Of this population, 77% had a normal LVEF, whereas 15% had a mildly reduced LVEF and 8% had a LVEF below 40%. Furthermore, 35% of patients had echocardiographic arguments for elevated LVEDP, 32% had an increased LVMI, and 47% had an increased LAVI.
Concentrations of protein‐bound uremic toxins according to echocardiographic phenotype
Regarding the ultrasonographic data, 109 patients (55%) had normal LVEF and no echocardiographic evidence for elevated LVEDP (group 1), 42 patients (21%) had normal LVEF and echocardiographic evidence for elevated LVEDP (group 2), and 49 patients (23%) had reduced LVEF (<50%) (group 3). For the remaining 10 patients, not enough echocardiographic data were available to obtain LVEF and/or LVEDP. Concentrations of UTs according to ultrasonographic group are shown in Table 2 . According to the Kruskal–Wallis test, total and free concentrations of pCS and pCG were significantly different across the three groups. Pairwise comparisons revealed that patients in group 3 had significantly higher plasma levels of these PBUTs than patients in group 1 and group 2 (P < 0.05), whereas there was no significant difference between the two latter groups (P > 0.05), except for free pCS which was higher in group 2 compared with group 1 (P < 0.05). There was no significant difference across the three groups regarding the other PBUTs or regarding serum creatinine (P > 0.05).
Table 2.
Median concentrations of protein‐bound uremic toxins according to echocardiographic phenotype
| PBUTs (mg/dL) | Group 1 | Group 2 | Group 3 | P value (KW) |
|---|---|---|---|---|
|
Normal LVEF Normal LVEDP (N = 109) |
Normal LVEF Elevated LVEDP (N = 42) |
Reduced LVEF (N = 49) | ||
| IxS total | 0.117 | 0.142 | 0.155 | 0.286 |
| IxS free | 0.004 | 0.004 | 0.006 | 0.120 |
| pCS total | 0.619 | 0.867 | 1.094* / ° | 0.012* |
| pCS free | 0.016 | 0.022* | 0.025* / ° | 0.001* |
| pCG total | 0.011 | 0.015 | 0.021* / ° | 0.032* |
| pCG free | 0.008 | 0.011 | 0.017* / ° | 0.009* |
| IAA total | 0.051 | 0.051 | 0.059 | 0.502 |
| IAA free | 0.006 | 0.007 | 0.007 | 0.313 |
| HA total | 0.125 | 0.137 | 0.165 | 0.293 |
| HA free | 0.037 | 0.043 | 0.051 | 0.193 |
HA, hippuric acid; IAA, indole‐3‐acetic acid; IxS, indoxyl sulfate; KW, Kruskal–Wallis test; LVEDP, left ventricular end‐diastolic pressure; LVEF, left ventricular ejection fraction; pCG, p‐cresyl glucuronide; pCS, p‐cresyl sulfate.
P < 0.05 versus Group 1.
P < 0.05 versus Group 2.
Associations between protein‐bound uremic toxins and risk of heart failure hospitalization and death
After a median follow up of 5.5 years, 43 patients (8.4%) developed the clinical endpoint, of which 19% (8/43) was fatal. In a multivariate Cox regression model adjusted for age, gender, BMI, diabetes mellitus, and systolic blood pressure, the free fractions of three PBUTs (IxS, pCS, and pCG) were independent predictors of the clinical endpoint [IxS free: HR 1.71 (95% CI: 1.11–2.63; P = 0.015); pCS free: HR 1.82 (95% CI: 1.11–3.01; P = 0.019); pCG free: HR 1.67 (95% CI: 1.08–2.58; P = 0.020)]. Regarding the total fractions, two PBUTs remained independent predictors of the clinical endpoint: pCG total: HR 1.63 (95% CI: 1.07–2.49; P = 0.024) and pCS total: HR 1.56 (95% CI: 1.00–2.44; P = 0.049). The total fraction of IxS showed a trend towards significance (HR 1.48; 95% CI 0.97–2.26; P = 0.073). Overall, the 95% confidence intervals around the point estimates were smaller for free concentrations compared with total concentrations of each of the solutes, resulting in lower P‐values. Cox regression models are shown in Table 3 , while univariate survival analyses for the free fractions of IxS, pCS, and pCS (which most significantly correlated with the clinical endpoint) are shown in Figure 1 . Because creatinine as a measure of kidney function is part of the causal pathway leading to increased concentrations of PBUT, it can be debated whether it is appropriate to adjust multivariate models for serum creatinine. Nevertheless, as renal function/creatinine is a well‐established cardiovascular risk factor, we made an additional adjustment for serum creatinine. In this model, only the free concentrations of IxS, pCS, and pCG remained significant. These results can be seen in Table 4 .
Table 3.
Associations between protein‐bound uremic toxins and risk of heart failure hospitalization and death
| Variable | Hazard ratio per tertile change in plasma level, adjusted for age, gender, BMI, diabetes, and SBP | P value |
|---|---|---|
| IxS total | 1.48 (0.97–2.26) | 0.073 |
| IxS free | 1.71 (1.11–2.63) | 0.015* |
| pCS total | 1.56 (1.00–2.44) | 0.049* |
| pCS free | 1.82 (1.11–3.01) | 0.019* |
| pCG total | 1.63 (1.07–2.49) | 0.024* |
| pCG free | 1.67 (1.08–2.58) | 0.020* |
| HA total | 1.28 (0.85–1.92) | 0.235 |
| HA free | 1.34 (0.89–2.01) | 0.163 |
| IAA total | 1.21 (0.80–1.84) | 0.365 |
| IAA free | 1.38 (0.91–2.10) | 0.135 |
BMI, body mass index; HA, hippuric acid; IAA, indole‐3‐acetic acid; IxS, indoxyl sulfate; pCG, p‐cresyl glucuronide; pCS, p‐cresyl sulfate; SBP, systolic blood pressure.
P value below 0.05.
Figure 1.
Kaplan–Meier survival curves for new heart failure events according to tertiles of indoxyl sulfate, p‐cresyl sulfate and p‐cresyl glucuronide free fraction.
Table 4.
Associations between protein‐bound uremic toxins and risk of heart failure hospitalization and death, with additional correction for serum creatinine
| Variable | Hazard ratio per tertile change in plasma level, adjusted for age, gender, BMI, diabetes, SBP and serum creatinine | P value |
|---|---|---|
| IxS total | 1.36 (0.85–2.16) | 0.196 |
| IxS free | 1.62 (1.03–2.55) | 0.038* |
| pCS total | 1.46 (0.92–2.31) | 0.111 |
| pCS free | 1.71 (1.01–2.87) | 0.044* |
| pCG total | 1.54 (1.00–2.38) | 0.050 |
| pCG free | 1.59 (1.02–2.47) | 0.042* |
| HA total | 1.19 (0.78–1.80) | 0.427 |
| HA free | 1.24 (0.81–1.88) | 0.324 |
| IAA total | 1.12 (0.72–1.73) | 0.612 |
| IAA free | 1.27 (0.81–1.98) | 0.293 |
BMI, body mass index; HA, hippuric acid; IAA, indole‐3‐acetic acid; IxS, indoxyl sulfate; pCG, p‐cresyl glucuronide; pCS, p‐cresyl sulfate; SBP, systolic blood pressure.
P value below 0.05.
Discussion
The predictive value of elevated circulating levels of PBUT for the occurrence of cardiovascular events has been demonstrated previously, mainly for indoxyl sulfate 12 , 13 , 14 , 15 and the p‐cresol derivatives. 16 , 17 , 18 , 19 , 20 However, data were sometimes conflicting and the predictive power of IxS often lost significance after adjusting for other cardiovascular risk factors. Additionally, the extent to which circulating levels of PBUT are associated with incident HF events is far less studied, 23 correlation analyses with echocardiographic data are scarce and often limited in sample size, 23 , 24 , 25 , 26 and none of these studies have included several PBUTs in their analyses. This retrospective study aimed to fill this gap by comparing concentrations of a panel of five uremic solutes in a cohort of 523 patients with CKD not on dialysis within three groups of patients with distinct echocardiographic features and by analysing the predictive value of these PBUTs for incident HF events during a median follow up of 5.5 years.
We found that circulating levels of the p‐cresol derivatives (e.g. pCG and pCS) were significantly higher in patients with reduced (<50%) versus normal LVEF. Some previous studies failed to show such a difference for IxS and pCS. 23 , 25 However, research in this field is limited and no study examined both the free and total fraction of multiple PBUTs. The exact mechanism leading to a difference in pCS and pCG levels across LVEF remains unclear, though we hypothesize that gut dysbiosis might be partly responsible. It has been previously shown that individual levels of p‐cresol derivatives are widely individually dispersed, even after adjustment for kidney function loss 17 , 27 and that interindividual variability in intestinal uptake of p‐cresol appears to be a main determinant of this dispersion. 28 This total uptake is the result of both intestinal generation and intestinal absorption of p‐cresol and both factors can be altered due to increased gut permeability, altered protein availability and change in microbial composition in patients with HFrEF. 29 , 30 However, the exact mechanism and extent to which dietary and microbial changes account for differences in intestinal formation and plasma values of different PBUTs in patients with CKD and HFrEF remain unknown and further research in this field of the so‐called ‘Gut Heart Kidney Axis’ is certainly warranted.
Furthermore, we found that the free fractions of IxS and the p‐cresol derivatives had the strongest association with the clinical endpoint of hospitalization and mortality due to HF and that these results were independent of age, gender, BMI, co‐existing diabetes mellitus, and arterial hypertension. These findings are in line with previous research, 23 which showed that a high total plasma IxS concentration was associated with higher risk of a first HF event in patients on haemodialysis. Based on our results, we can extent these findings to the non‐dialysed CKD population, both for the free fraction of IxS and free and total fraction of the p‐cresol derivatives (e.g. pCS and pCG). There are likely multiple mechanisms by which these PBUTs may lead to an increased risk of HF events. On the one hand, they may act as direct cardiotoxic metabolites as they play a role in activating leukocyte oxygen fee radical production and inducing myocardial apoptosis. 31 , 32 , 33 , 34 , 35 , 36 On the other hand, it has been demonstrated that IxS and p‐cresol derivatives can indirectly cause cardiotoxicity by inducing an inflammatory response and secretion of pro‐inflammatory cytokines with direct cardiotoxic properties, as well as inducing endothelial dysfunction and fibrosis resulting in increased ventricular stiffness. 37 In this context, one study showed diastolic dysfunction due to increased ventricular stiffness in mice treated with pCS, 36 and another study showed an increased risk for LV diastolic dysfunction in patients with high IxS levels. 38
The results of this study may have important clinical implications. The identification of PBUTs as potential biomarkers for HF events in patients with CKD may improve risk stratification and identify patients who may benefit from early initiation of HF treatment including the introduction of SLGT2i and angiotensin receptor neprilysin inhibitors. Indeed the use of conventional biomarkers of HF [e.g. troponins and natriuretic peptides (NP)] is often complicated by the presence of CKD, and usual cut‐off values cannot be used under these circumstances. 39 Furthermore, strategies aimed at reducing the generation or absorption of PBUTs in CKD patients may reduce the incidence of HF events. In this context, AST‐120, a carbonaceous oral adsorbent reduces serum PBUT levels 37 by adsorbing the precursors of IxS, pCG and pCS. AST‐120 has been shown to reduce inflammation and improve endothelial function and gut dysbiosis 37 , 40 , 41 and some small cohort studies also showed improvement in cardiac dysfunction. 42 , 43 However, further large‐scale research is needed to further examine the potential cardiovascular benefits of such a strategy.
This study has several strengths. We analysed both total and free concentrations of several UT and assessed their predictive value for hard endpoints using the same statistical models, allowing head‐to‐head comparisons. The studied population included the whole range of CKD stages G1–G5 patients who were prospectively followed long enough to allow detection of hard clinical endpoints and follow‐up was complete for 99% of patients. All endpoints were identified by the same physician based on clinical reality (i.e. not only based on administrative codes using International Classification of Diseases codes), avoiding misclassification of outcomes.
This study also has some limitations. The study was conducted in a single centre, and patients were predominantly Caucasian, though results were in line with those in a predominantly Aziatic population. 23 Due to post hoc analysis, echocardiographic data were only available for a subset of patients. Still, the sample size was at least comparable, and most often larger than that of previous available research. Finally, natriuretic peptide (NP) levels were not measured although the value of these biomarkers in CKD can be questioned as conventional cut‐off values cannot be applied at the different stages of CKD and a comparison between PBUT and NP levels was beyond the scope of this article.
In conclusion, this study provides further evidence of an association between plasma levels of PBUTs, particularly the free fractions of IxS, pCG, and pCS and incident HF events in patients with CKD not on dialysis. Additionally, elevated pCG and pCS levels are associated with reduced LVEF. This study is the first to compare several PBUTs with a specific focus on HF events and inclusion of echocardiographic parameters, which are critical for the correct diagnosis and classification of HF. Further research is needed to confirm these findings and explore the potential impact of interventions/strategies aimed at reducing PBUT accumulation on HF events in this patient population.
Conflict of interest
None declared.
Funding
None.
Zwaenepoel, B. , De Backer, T. , Glorieux, G. , and Verbeke, F. (2024) Predictive value of protein‐bound uremic toxins for heart failure in patients with chronic kidney disease. ESC Heart Failure, 11: 466–474. 10.1002/ehf2.14566.
References
- 1. Tuegel C, Bansal N. Heart failure in patients with kidney disease. Heart 2017;103:1848–1853. doi: 10.1136/heartjnl-2016-310794 [DOI] [PubMed] [Google Scholar]
- 2. Lekawanvijit S. Cardiotoxicity of uremic toxins: a driver of cardiorenal syndrome. Toxins (Basel) 2018;10:352. doi: 10.3390/toxins10090352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Saran R, Robinson B, Abbott KC, Agodoa LYC, Albertus P, Ayanian J, et al. US renal data system 2016 annual data report: Epidemiology of kidney disease in the United States. Am J Kidney Dis 10:352. doi: 10.3390/toxins10090352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Kottgen A, Russell SD, Loehr LR, Crainiceanu CM, Rosamond WD, Chang PP, et al. Reduced kidney function as a risk factor for incident heart failure: The atherosclerosis risk in communities (ARIC) study. J Am Soc Nephrol 2007;18:1307–1315. doi: 10.1681/ASN.2006101159 [DOI] [PubMed] [Google Scholar]
- 5. Porter AC, Lash JP, Xie D, Pan Q, DeLuca J, Kanthety R, et al. Predictors and outcomes of health‐related quality of life in adults with CKD. Clin J Am Soc Nephrol 2016;11:1154–1162. doi: 10.2215/CJN.09990915 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jankowski J, Floege J, Fliser D, Böhm M, Marx N. Cardiovascular disease in chronic kidney disease: Pathophysiological insights and therapeutic options. Circulation 2021;143:1157–1172. doi: 10.1161/CIRCULATIONAHA.120.050686 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Falconi CA, Fogaça‐Ruiz F, Vernier ICS, da Cunha RS, Stinghen AEM, Carneiro‐Ramos MS. Uremic toxins: An alarming danger concerning the cardiovascular system. Front Physiol 2021;12:686249. doi: 10.3389/fphys.2021.686249 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Vanholder R, de Smet R, Glorieux G, Argilés A, Baurmeister U, Brunet P, et al. Review on uremic toxins: Classification, concentration, and interindividual variability. Kidney Int 2003;63:1934–1943. doi: 10.1046/j.1523-1755.2003.00924.x [DOI] [PubMed] [Google Scholar]
- 9. Lesaffer G, de Smet R, Lameire N, Dhondt A, Duym P, Vanholder R. Intradialytic removal of protein‐bound uraemic toxins: Role of solute characteristics and of dialyser membrane. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2000;15:50–57. doi: 10.1093/ndt/15.1.50 [DOI] [PubMed] [Google Scholar]
- 10. Rosner MH, Reis T, Husain‐Syed F, Vanholder R, Hutchison C, Stenvinkel P, et al. Classification of uremic toxins and their role in kidney failure. Clin J Am Soc Nephrol 2021;16:1918–1928. doi: 10.2215/CJN.02660221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Vanholder R, Pletinck A, Schepers E, Glorieux G. Biochemical and clinical impact of organic uremic retention solutes: A comprehensive update. Toxins (Basel) 2018;10:33. doi: 10.3390/toxins10010033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Barreto FC, Barreto DV, Liabeuf S, Meert N, Glorieux G, Temmar M, et al. Serum indoxyl sulfate is associated with vascular disease and mortality in chronic kidney disease patients. Clin J Am Soc Nephrol 2009;4:1551–1558. doi: 10.2215/CJN.03980609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lin C‐J, Liu H‐L, Pan C‐F, Chuang C‐K, Jayakumar T, Wang T‐J, et al. Indoxyl sulfate predicts cardiovascular disease and renal function deterioration in advanced chronic kidney disease. Arch Med Res 2012;43:451–456. doi: 10.1016/j.arcmed.2012.08.002 [DOI] [PubMed] [Google Scholar]
- 14. Wu C‐C, Hsieh M‐Y, Hung S‐C, Kuo K‐L, Tsai T‐H, Lai C‐L, et al. Serum Indoxyl sulfate associates with postangioplasty thrombosis of dialysis grafts. J Am Soc Nephrol 2016;27:1254–1264. doi: 10.1681/ASN.2015010068 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Imazu M, Fukuda H, Kanzaki H, Amaki M, Hasegawa T, Takahama H, et al. Plasma indoxyl sulfate levels predict cardiovascular events in patients with mild chronic heart failure. Sci Rep 2020;10:16528. doi: 10.1038/s41598-020-73633-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Wu I‐W, Hsu K‐H, Hsu H‐J, Lee C‐C, Sun C‐Y, Tsai C‐J, et al. Serum free p‐cresyl sulfate levels predict cardiovascular and all‐cause mortality in elderly hemodialysis patients—A prospective cohort study. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2012;27:1169–1175. doi: 10.2215/CJN.03980609 [DOI] [PubMed] [Google Scholar]
- 17. Liabeuf S, Barreto DV, Barreto FC, Meert N, Glorieux G, Schepers E, et al. Free p‐cresylsulphate is a predictor of mortality in patients at different stages of chronic kidney disease. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2010;25:1183–1191. doi: 10.1093/ndt/gfp592 [DOI] [PubMed] [Google Scholar]
- 18. Wu I‐W, Hsu K‐H, Lee C‐C, Sun C‐Y, Hsu H‐J, Tsai C‐J, et al. P‐Cresyl sulphate and indoxyl sulphate predict progression of chronic kidney disease. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2011;26:938–947. doi: 10.1093/ndt/gfp592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Glorieux G, Vanholder R, van Biesen W, Pletinck A, Schepers E, Neirynck N, et al. Free p‐cresyl sulfate shows the highest association with cardiovascular outcome in chronic kidney disease. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2021;36:998–1005. doi: 10.1093/ndt/gfab004 [DOI] [PubMed] [Google Scholar]
- 20. Wang C‐H, Cheng M‐L, Liu M‐H, Shiao M‐S, Hsu K‐H, Huang Y‐Y, et al. Increased p‐cresyl sulfate level is independently associated with poor outcomes in patients with heart failure. Heart Vessels 2016;31:1100–1108. doi: 10.1007/s00380-015-0702-0 [DOI] [PubMed] [Google Scholar]
- 21. wShafi T, Sirich TL, Meyer TW, Hostetter TH, Plummer NS, Hwang S, et al. Results of the HEMO study suggest that p‐cresol sulfate and indoxyl sulfate are not associated with cardiovascular outcomes. Kidney Int 2017;92:1484–1492. doi: 10.1016/j.kint.2017.05.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. van Gelder MK, Middel IR, Vernooij RWM, Bots ML, Verhaar MC, Masereeuw R, et al. Protein‐bound uremic toxins in hemodialysis patients relate to residual kidney function, are not influenced by convective transport, and do not relate to outcome. Toxins (Basel) 2020;12:234. doi: 10.3390/toxins12040234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Cao X‐S, Chen J, Zou J‐Z, Zhong Y‐H, Teng J, Ji J, et al. Association of indoxyl sulfate with heart failure among patients on hemodialysis. Clin J Am Soc Nephrol 2015;10:111–119. doi: 10.2215/CJN.04730514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Shimazu S, Hirashiki A, Okumura T, Yamada T, Okamoto R, Shinoda N, et al. Association between indoxyl sulfate and cardiac dysfunction and prognosis in patients with dilated cardiomyopathy. Circ J 2013;77:390–396. doi: 10.1253/circj.CJ-12-0715 [DOI] [PubMed] [Google Scholar]
- 25. Ni J, Zhu J, Xu T, Fang Y, Zhu Z, Lu L, et al. P‐cresyl sulfate accelerates left ventricular diastolic dysfunction associated with chronic kidney disease. Int J Clin Exp Med 2016;9:3860–3868. doi: 10.3390/toxins12040234 [DOI] [Google Scholar]
- 26. Takkavatakarn K, Phannajit J, Udomkarnjananun S, Tangchitthavorngul S, Chariyavilaskul P, Sitticharoenchai P, et al. Association between indoxyl sulfate and dialysis initiation and cardiac outcomes in chronic kidney disease patients. Int J Nephrol Renovasc Dis 2022;15:115–126. doi: 10.2147/IJNRD.S354658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Eloot S, Schepers E, Barreto DV, Barreto FC, Liabeuf S, van Biesen W, et al. Estimated glomerular filtration rate is a poor predictor of concentration for a broad range of uremic toxins. Clin J Am Soc Nephrol 2011;6:1266–1273. doi: 10.2215/CJN.09981110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Poesen R, Viaene L, Verbeke K, Claes K, Bammens B, Sprangers B, et al. Renal clearance and intestinal generation of p‐cresyl sulfate and indoxyl sulfate in CKD. Clin J Am Soc Nephrol 2013;8:1508‐1514. doi: 10.2215/CJN.00300113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Rogler G, Rosano G. The heart and the gut. Eur Heart J 2014;35:426‐430. doi: 10.1093/eurheartj/eht271 [DOI] [PubMed] [Google Scholar]
- 30. Vaziri ND, Wong J, Pahl M, Piceno YM, Yuan J, DeSantis TZ, et al. Chronic kidney disease alters intestinal microbial flora. Kidney Int 2013;83:308‐315. doi: 10.1038/ki.2012.345 [DOI] [PubMed] [Google Scholar]
- 31. Koppe L, Pillon NJ, Vella RE, Croze ML, Pelletier CC, Chambert S, et al. P‐Cresyl sulfate promotes insulin resistance associated with CKD. J Am Soc Nephrol 2013;24:88‐99. doi: 10.1681/ASN.2012050503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Sun C‐Y, Hsu H‐H, Wu M‐S. P‐cresol sulfate and indoxyl sulfate induce similar cellular inflammatory gene expressions in cultured proximal renal tubular cells. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2013;28:70‐78. doi: 10.1038/ki.2012.345 [DOI] [PubMed] [Google Scholar]
- 33. Sun C‐Y, Chang S‐C, Wu M‐S. Uremic toxins induce kidney fibrosis by activating intrarenal renin‐angiotensin‐aldosterone system associated epithelial‐to‐mesenchymal transition. PLoS ONE 2012;7:e34026. doi: 10.1371/journal.pone.0052577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Watanabe H, Miyamoto Y, Honda D, Tanaka H, Wu Q, Endo M, et al. P‐Cresyl sulfate causes renal tubular cell damage by inducing oxidative stress by activation of NADPH oxidase. Kidney Int 2013;83:582‐592. doi: 10.1038/ki.2012.448 [DOI] [PubMed] [Google Scholar]
- 35. Schepers E, Meert N, Glorieux G, Goeman J, Van der Eycken J, Vanholder R. P‐cresylsulphate, the main in vivo metabolite of p‐cresol, activates leucocyte free radical production. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2007;22:592‐596. doi: 10.1093/ndt/gfl584 [DOI] [PubMed] [Google Scholar]
- 36. Han H, Zhu J, Zhu Z, Ni J, Du R, Dai Y, et al. P‐Cresyl sulfate aggravates cardiac dysfunction associated with chronic kidney disease by enhancing apoptosis of cardiomyocytes. J Am Heart Assoc 2015;4:e001852. doi: 10.1161/JAHA.115.001852 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Liu W‐C, Tomino Y, Lu K‐C. Impacts of indoxyl sulfate and p‐cresol sulfate on chronic kidney disease and mitigating effects of AST‐120. Toxins (Basel) 2018;10:367. doi: 10.3390/toxins10090367 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Sato B, Yoshikawa D, Ishii H, Suzuki S, Inoue Y, Takeshita K, et al. Relation of plasma indoxyl sulfate levels and estimated glomerular filtration rate to left ventricular diastolic dysfunction. Am J Cardiol 2013;111:712‐716. doi: 10.1016/j.amjcard.2012.11.025 [DOI] [PubMed] [Google Scholar]
- 39. Mueller C, McDonald K, de Boer RA, Maisel A, Cleland JGF, Kozhuharov N, et al. Heart Failure Association of the European Society of Cardiology practical guidance on the use of natriuretic peptide concentrations. Eur J Heart Fail 2019;21:715‐731. doi: 10.1002/ejhf.1494 [DOI] [PubMed] [Google Scholar]
- 40. Akiyama Y, Takeuchi Y, Kikuchi K, Mishima E, Yamamoto Y, Suzuki C, et al. A metabolomic approach to clarifying the effect of AST‐120 on 5/6 nephrectomized rats by capillary electrophoresis with mass spectrometry (CE‐MS). Toxins (Basel) 2012;4:1309‐1322. doi: 10.3390/toxins4111309 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Su P‐Y, Lee Y‐H, Kuo L‐N, Chen Y‐C, Chen C, Kang Y‐N, et al. Efficacy of AST‐120 for patients with chronic kidney disease: A network meta‐analysis of randomized controlled trials. Front Pharmacol 2021;12:676345. doi: 10.3389/fphar.2021.676345 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Shen W‐C, Chou Y‐H, Shi L‐S, Chen Z‐W, Tu H‐J, Lin X‐Y, et al. AST‐120 improves cardiac dysfunction in acute kidney injury mice via suppression of apoptosis and proinflammatory NF‐κB/ICAM‐1 signaling. J Inflamm Res 2021;14:505‐518. doi: 10.2147/JIR.S283378 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Fujii H, Nishijima F, Goto S, Sugano M, Yamato H, Kitazawa R, et al. Oral charcoal adsorbent (AST‐120) prevents progression of cardiac damage in chronic kidney disease through suppression of oxidative stress. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc ‐ Eur Ren Assoc 2009;24:2089‐2095. doi: 10.1093/ndt/gfp007 [DOI] [PubMed] [Google Scholar]