Positive correlation of serum indoxyl sulfate level with peripheral arterial disease in hemodialysis patients

Abstract

Objectives

Indoxyl sulfate, known for its cardiovascular toxicity, is associated with vascular and coronary artery diseases and increased mortality. Peripheral arterial disease, defined by low ankle–brachial index, is associated with increased mortality in patients on hemodialysis. The present study aimed to determine the relationship between the serum indoxyl sulfate level and peripheral arterial disease in patients on maintenance hemodialysis.

Methods

The present cross-sectional, single-center study included 75 patients on maintenance hemodialysis. Serum indoxyl sulfate levels were determined by high-performance liquid chromatography–mass spectrometry. Ankle–brachial index values were measured using an automated oscillometric device. Patients with ankle–brachial indexes of < 0.9 were categorized into the low ankle–brachial index group.

Results

In the study cohort, 12 of the 75 patients (16.0%) had low ankle–brachial indexes. The rates of diabetes mellitus (p = 0.010) as well as the serum levels of C-reactive protein (p < 0.001) and indoxyl sulfate (p < 0.001) were higher in the low ankle–brachial index group than the normal ankle–brachial index group. The multivariable logistic regression analysis revealed that serum levels of indoxyl sulfate (odds ratio = 1.123, 95% confidence interval 1.011–1.249, p = 0.031) and C-reactive protein (each 0.1 mg/dL increase, odds ratio = 1.169, 95% confidence interval 1.018–1.343, p = 0.027) were independently associated with peripheral arterial disease in patients on maintenance hemodialysis.

Conclusions

Serum indoxyl sulfate levels were associated with peripheral arterial disease in patients on maintenance hemodialysis.

Keywords

C-reactive protein hemodialysis indoxyl sulfate peripheral artery disease

Introduction

In patients with end-stage renal disease, the mortality remains high despite treatment with renal replacement therapy. For example, in a study based on the United States Renal Data System, the 5-year mortality was 43.0% in patients on hemodialysis.¹ The high mortality is largely attributable to cardiovascular causes in these hemodialysis patients.² Peripheral arterial disease (PAD), which is defined as narrowing or obstruction of systemic vessels other than coronary or intracranial arteries, can be diagnosed based on an ankle–brachial index (ABI) of < 0.9 in either leg; PAD usually develops secondary to atherosclerosis.³ The prevalence of PAD is higher among patients on hemodialysis than the general population.⁴ Several studies reported a strong association between PAD and all-cause and cardiovascular mortality in patients on hemodialysis, with more than twofold increase in mortality risk reported in patients with PAD compared to those without PAD in a large-scale meta-analysis involving more than 30,000 patients on hemodialysis.^5,6 Therefore, it is important to identify risk factors associated with PAD in patients on hemodialysis.

Indoxyl sulfate (IS), a uremic toxin derived from tryptophan,⁷ accumulates due to tubular dysfunction during the progression of renal failure. In patients with end-stage renal disease, the clearance of IS by hemodialysis is very low due to the high protein binding by IS.⁷ The well-documented harmful effects of IS on vasculature are mediated through oxidative stress, inflammation, and impaired repair capacity of endothelial cells.⁸ Furthermore, a study in rats has demonstrated that IS-mediated toxicity in the cardiovascular system involves the deterioration of atherosclerotic lesions via the induction of vascular smooth muscle cell proliferation.⁹ In a cohort study of 139 patients with chronic kidney disease, IS was associated with aortic calcification and vascular stiffness.¹⁰ Given the mounting evidence of vascular toxic effects of IS, we hypothesized that high IS levels would be associated with increased incidence of PAD in patients with end-stage renal disease. Therefore, in the present study, we aimed to determine the association between IS and PAD in patients on maintenance hemodialysis.

Materials and methods

Patients

This cross-sectional study was conducted between 1 March 2016 and 31 July 2016 in the hemodialysis unit of a medical center in Hualien, Taiwan. Adult patients aged ≥ 20 years undergoing a standard 4-h hemodialysis session three times per week for at least 3 months were assessed for eligibility. High-flux polysulfone disposable artificial kidneys (FX class dialyzer; Fresenius Medical Care, Bad Homburg, Germany) were used in the dialysis unit. Initially, 110 hemodialysis participants enrolled, 35 participants were excluded on account of malignancy (n = 5), acute infection (n = 4), myocardial infarction (n = 3), stroke (n = 4), amputation (n = 6), heart failure (n = 2), treatment with cilostazol or pentoxifylline during blood sampling (n = 6), and elevated ABIs > 1.3 (n = 5). Finally, a total of 75 HD patients were enrolled in this study. Data on basic demographic characteristics and medical history, including diabetes mellitus (DM), hypertension, hyperlipidemia, and coronary artery disease, were collected from the medical records. The present study was approved by the Institutional Review Board of Tzu Chi Hospital and complied with the general recommendations of the Declaration of Helsinki. Written informed consent was obtained from all patients.

Anthropometric analysis

Anthropometric variables were measured in the morning after overnight fasting. Body weight and height measurements were recorded to the nearest 0.5 kg and 0.5 cm, respectively. Body mass index was calculated as weight (kg) divided by height-squared (m²).

Biochemical investigations

Blood samples were obtained after overnight fasting. In each patient, approximately 5 mL blood was immediately centrifuged at 3000 g for 10 min and stored at 4°C within 1 h of processing for biochemical analyses. We used an autoanalyzer (Siemens Advia 1800; Siemens Healthcare, Henkestr, Germany) to measure serum levels of total calcium, phosphorus, fasting glucose, albumin, blood urea nitrogen, creatinine, total cholesterol, triglycerides, and C-reactive protein (CRP). Intact parathyroid hormone (iPTH) levels were measured by a commercially available enzyme-linked immunosorbent assay (catalog number: NM59041; IBL International, Hamburg, Germany).^11–13

High-performance liquid chromatography–mass spectrometry

Total IS levels were measured by modified liquid chromatography–mass spectrometry.^12–14 Briefly, 100 µL serum was added to a 1.5-mL centrifuge tube containing 100 µL of 50 mM sodium phosphate dibasic heptahydrate solution. The stable d₄-IS isotope was used as internal control. Extraction was performed with Novum^® simplified liquid extraction cartridges (Phenomenex, Torrance, CA, USA), with elution in 1.5 mL ethyl acetate. Samples dried in nitrogen gas were later resolved in 100 µL methanol. A Waters^® e2695 high-performance liquid chromatography device was used with a single quadrupole ACQUITY QDa^® mass detector for analysis. A Phenomenex Luna^® C18(2) column (5 µ, 250 × 4.6 mm, 100 Å) was heated to 40°C, with a flow rate of 0.8 mL/min and an injection volume of 30 µL. The mobile phase included 99.9% double-distilled water (A) and 99.9% high-performance liquid chromatography-grade methanol (B); each contained 0.1% formic acid. Linear gradient from 5% B to 70% B for 14 min, maintenance at 70% B for 2 min, and linear gradient from 70% B to 50% B for 2 min. Pretreated samples were tested by negative-ion mode electrospray ionization, set at a vaporization temperature of 400°C, a capillary voltage of 0.8 kV, and a sample cone of 15.0 V. Single-ion recording mode was used for mass-to-charge ratio (m/z) scanning (d₄-IS, 216.0 m/z and IS, 211.9 m/z). The Empower^® 3.0 software was used for data acquisition and processing. Sample peak areas were used to quantify IS levels using a calibration curve of the standard.

Ankle–brachial index measurements

Blood pressure was measured using the oscillometric method in both arms and ankles for three times, with the patient in the supine position, and readings were obtained from brachial, dorsalis pedis, and posterior tibial arteries (VaSera VS-1000; Fukuda Denshi, Tokyo, Japan).¹¹ Right or left ABI was calculated as the ratio of the highest-measured systolic blood pressure in right or left ankle (dorsalis pedis or posterior tibial artery) to the highest-measured systolic blood pressure in brachial artery of either arm. Real-time electrocardiography was documented for more than 15 min. PAD was diagnosed in patients with a low ABI of < 0.9 in left or right, as described previously.¹¹

Statistical analysis

The Kolmogorov–Smirnov test was performed to determine normality. Continuous data were expressed as means ± standard deviation for normally distributed variables and as medians with interquartile ranges for non-normally distributed variables; between-patient comparisons were performed by two-tailed Student’s independent t test and the Mann–Whitney U test, respectively. Categorical variables were expressed as numbers with percentages and compared using the χ² test. Variables significantly associated with PAD were further evaluated by multivariable logistic regression analysis. The receiver-operating characteristic curve was used to calculate the area under the curve and determine serum IS and CRP levels to predict PAD in patients on hemodialysis. The SPSS software for Windows (version 19.0; SPSS, Chicago, IL, USA) was used for all statistical analyses. A p value of < 0.05 was considered to indicate statistical significance.

Results

Table 1 shows the baseline characteristics of the entire cohort of 75 patients categorized into the normal (n = 63) and low (n = 12) ABI groups. In the entire cohort, 33 patients (44.0%) were female and the median dialysis vintage was 49.08 (21.60–123.84) months. The rates of DM and hypertension were 41.3% (n = 31) and 50.7% (n = 38), respectively, and 12.0% (n = 9) of the patients were current smokers. The rates of DM were significantly higher in the low ABI group than in the normal ABI group (p = 0.010). Additionally, the CRP and IS levels were significantly higher in the low ABI group than in the normal ABI group (all p < 0.001). In contrast, there were no significant differences in systolic and diastolic blood pressures, sex, hemodialysis duration, Kt/V, hypertension, and smoking status between the two groups.

Table 1.

Clinical variables of the 75 hemodialysis patients with normal or low ankle–brachial index group.

Characteristic	All patients (n = 75)	Normal ABI group (n = 63)	Low ABI group (n = 12)	p value
Age (years)	62.57 ± 12.73	62.13 ± 13.26	64.92 ± 9.65	0.490
HD duration (months)	49.08 (21.60–123.84)	55.44 (23.64–130.08)	33.18 (19.41–63.81)	0.161
Height (cm)	160.31 ± 9.03	160.65 ± 9.06	158.50 ± 9.05	0.453
Pre-HD body weight (kg)	64.56 ± 15.16	65.52 ± 115.18	59.53 ± 14.68	0.212
Body mass index (kg/m²)	25.01 ± 5.10	25.28 ± 5.08	23.60 ± 5.17	0.299
Left ankle–brachial index	1.06 ± 0.14	1.09 ± 0.09	0.86 ± 0.19	< 0.001*
Right ankle–brachial index	1.05 ± 0.15	1.09 ± 0.09	0.80 ± 0.13	< 0.001*
Systolic blood pressure (mmHg)	141.72 ± 25.26	141.40 ± 25.37	143.42 ± 25.71	0.802
Diastolic blood pressure (mmHg)	76.83 ± 15.36	76.48 ± 15.43	78.67 ± 15.47	0.654
Total cholesterol (mg/dL)	144.73 ± 39.12	147.48 ± 39.07	130.33 ± 37.76	0.166
Triglyceride (mg/dL)	152.63 ± 86.60	155.56 ± 87.74	137.25 ± 82.15	0.506
Glucose (mg/dL)	136.00 (110.00–177.00)	136.00 (111.00–172.00)	135.50 (100.25–197.50)	0.994
Albumin (g/dL)	4.16 ± 0.47	4.19 ± 0.47	4.03 ± 0.41	0.276
Blood urea nitrogen (mg/dL)	58.01 ± 13.82	58.51 ± 14.58	55.42 ± 8.85	0.481
Creatinine (mg/dL)	9.34 ± 1.60	9.49 ± 1.68	8.54 ± 0.63	0.059
Total calcium (mg/dL)	8.88 ± 0.73	8.89 ± 0.75	8.85 ± 0.67	0.892
Phosphorus (mg/dL)	4.49 ± 1.25	4.45 ± 1.17	4.68 ± 1.65	0.572
Intact parathyroid hormone (pg/mL)	216.94 ± 179.55	234.01 ± 185.56	127.32 ± 111.28	0.059
C-reactive protein (mg/dL)	0.29 (0.08–0.86)	0.25 (0.06–0.59)	1.23 (0.41–2.19)	< 0.001*
Total indoxyl sulfate (mg/L)	18.95 ± 11.03	16.78 ± 8.99	30.31 ± 14.01	< 0.001*
Urea reduction rate	0.73 ± 0.04	0.74 ± 0.04	0.72 ± 0.05	0.182
Kt/V (Gotch)	1.34 ± 0.18	1.35 ± 0.18	1.27 ± 0.18	0.183
Female, n (%)	33 (44.0)	29 (46.0)	4 (33.3)	0.417
Diabetes mellitus, n (%)	31 (41.3)	22 (34.9)	9 (75.0)	0.010*
Hypertension, n (%)	38 (50.7)	32 (50.8)	6 (50.0)	0.960
Current smoking	9 (12.0)	6 (9.5)	3 (25.0)	0.131
ACE inhibitor or ARB use, n (%)	28 (37.3)	23 (36.5)	5 (41.7)	0.735
β-blocker, n (%)	31 (41.3)	27 (42.9)	4 (33.3)	0.539
Calcium channel blocker, n (%)	33 (44.0)	29 (46.0)	4 (33.3)	0.417
Statin, n (%)	16 (21.3)	11 (17.5)	5 (41.7)	0.061
Fibrate, n (%)	10 (13.3)	9 (14.3)	1 (8.3)	0.578

Values for continuous variables are given as means ± standard deviation and test by Student’s t-test; variables not normally distributed are given as medians and interquartile range and test by Mann-Whitney U test; data are expressed as number of patients, and analysis was done using the chi-square test.

ABI, ankle–brachial index; HD, hemodialysis; Kt/V, fractional clearance index for urea; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; CCB, calcium-channel blocker.

*p < 0.05 was considered statistically significant.

Multivariable logistic regression analysis performed after adjusting for confounding factors (p < 0.15) associated with PAD, including DM, statin use, current smoking, serum creatinine, iPTH, CRP and IS levels, serum CRP level (each 0.1 mg/dL increase, odds ratio (OR) = 1.169, 95% confidence interval (CI) 1.018–1.343, p = 0.027), and IS level (OR = 1.123, 95% CI 1.011–1.249, p = 0.031) were independent predictors for PAD in patients on hemodialysis (Table 2). Furthermore, the receiver-operating characteristic curve for predicting PAD revealed that the areas under the curve for IS and CRP were 0.796 (95% CI, 0.688–0.881; p = 0.0003) and 0.832 (95% CI, 0.728–0.908; p < 0.001), respectively (Figure 1(a) and (b)).

Table 2.

Multivariable logistic regression analysis of the factors correlated to peripheral arterial disease among 75 hemodialysis patients.

Variable	Odds ratio	95% confidence interval	p value
Total indoxyl sulfate, 1 mg/L	1.123	1.011–1.249	0.031*
C-reactive protein, 0.1 mg/dL	1.169	1.018–1.343	0.027*
Diabetes mellitus, present	7.136	0.808–63.061	0.077
Statin, used	0.594	0.070–5.012	0.632
Creatinine, 1 mg/dL	0.573	0.280–1.170	0.126
Intact parathyroid hormone, 1 pg/mL	0.996	0.990–1.002	0.187
Smoking, present	2.202	0.205–23.593	0.514

Analysis data were done using the multivariable logistic regression analysis (adopted factors: diabetes, statin used, current smoking, creatinine, intact parathyroid hormone, C-reactive protein, and total indoxyl sulfate).

*p < 0.05 was considered statistically significant.

Figure 1.

The area under the receiver-operating characteristic curve indicates the diagnostic power of the serum indoxyl sulfate level (a) and C-reactive protein level (b) for predicting peripheral arterial disease among 75 hemodialysis patients.

Discussion

In the present cross-sectional study including 75 patients on maintenance hemodialysis, our analyses revealed that higher serum IS and CRP levels were independently associated with a higher risk of PAD.

Hypertension, smoking, DM, and dyslipidemia are risk factors for PAD, which is a manifestation of atherosclerotic vascular disease.³ The exacerbation of PAD by increased levels of inflammatory mediators, oxidative stress, overproduction of advanced glycation end-products, dyslipidemia, and atherothrombosis, which might be worse in patients with DM than those without DM.¹⁵ Inflammation and dyslipidemia, which are associated with atherosclerosis, also play a role in the initiation and progression of PAD.¹⁶ Moreover, high-circulating CRP levels are associated with major cardiovascular events in patients with PAD.¹⁷ Statin use has been shown to reduce the risk of major cardiovascular and limb events in patients with PAD.¹⁸ Our analyses also revealed that DM and higher serum CRP levels were associated with PAD in patients on hemodialysis. After multivariable logistic regression analysis, serum CRP level remained an independent predictor for PAD in patients on hemodialysis.

The direct vascular toxic effects of IS have been under extensive scrutiny in the past decade.^7,19–21 In addition to the induction of endothelial dysfunction, IS also triggers a prothrombotic state and impairs neovascularization.^19,20 In endothelial cells, IS decreases the production of nitric oxide; increases the production of reactive oxygen species; promotes pro-oxidant, pro-inflammatory, and prothrombotic processes through the activation of the aryl hydrocarbon receptor; and is involved in endothelial dysfunction.^19,21,22 IS promotes thrombosis, altered neoangiogenesis, and atherosclerosis, which may explain the risk of increased amputation rate due to PAD in patients with chronic kidney disease.^19,20 In a cohort of 139 patients with chronic kidney disease, higher IS levels were associated with significantly higher rates of vascular disease and mortality after adjustment for multiple factors, including age, sex, DM, albumin, hemoglobin, phosphate, and aortic calcification.¹⁰ In the present study, serum IS levels were positively associated with PAD after adjustment for other significant confounders. This finding highlights the potential role of IS as a marker of PAD in patients on hemodialysis. However, the mechanism underlying the association of IS with PAD in patients on hemodialysis requires further investigation. Low-protein diets or vegetarian diets have been shown to reduce serum IS levels.^23,24 Additionally, AST-120 is an oral charcoal adsorbent that could limit intestinal IS absorption and has been demonstrated to reduce serum IS levels in patients with chronic kidney disease.²⁵ Whether measurement of serum IS levels in hemodialysis patients for PAD risk evaluation or interventions to lower serum IS levels, such as dietary changes or oral adsorbents, will benefit these patients also needs further research.

The present study has several limitations that should be acknowledged. First, this was a cross-sectional study with a relatively small cohort size which was performed in a single medical center. Second, data on clinical symptoms of PAD, such as claudication and lower extremity pain, were not included. Third, in patients on hemodialysis, ABI may lead to false-negative results due to non-compressible calcified vessels.⁴ In this study, we had excluded elevated ABIs > 1.3 in our hemodialysis patients to rule out the falsely elevated ABIs due to vascular calcification. Therefore, large-scale longitudinal studies are necessary to determine the association between IS and PAD in patients on hemodialysis.

Conclusion

The findings of the present cross-sectional study suggest that serum CRP and IS levels should be considered as independent predictors of PAD in patients on hemodialysis.

Footnotes

Declaration of conflicting interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from the Buddhist Tzu Chi Medical Foundation, Taiwan (TCRD107-56, TCMF-A 109-07 and TCMF-CP 110-02).

ORCID iD

Bang-Gee Hsu

References

Yan

Norris

Xin

, et al. Facility size, race and ethnicity, and mortality for in-center hemodialysis. J Am Soc Nephrol 2013; 24: 2062–2070.

Saran

Robinson

Abbott

, et al. US renal data system 2018 annual data report: Epidemiology of kidney disease in the United States. Am J Kidney Dis 2019; 73(3 Suppl 1): A7–A8.

Fowkes

Rudan

, et al. Comparison of global estimates of prevalence and risk factors for peripheral artery disease in 2000 and 2010: a systematic review and analysis. Lancet 2013; 382: 1329–1340.

Okamoto

Iida

Mano

. Current perspective on hemodialysis patients with peripheral artery disease. Ann Vasc Dis 2017; 10: 88–91.

Fishbane

Youn

Flaster

, et al. Ankle-arm blood pressure index as a predictor of mortality in hemodialysis patients. Am J Kidney Dis 1996; 27: 668–672.

Yang

Ning

Shang

, et al. Association of peripheral arterial disease with all-cause and cardiovascular mortality in hemodialysis patients: a meta-analysis. BMC Nephrol 2016; 17: 195.

Vanholder

Schepers

Pletinck

, et al. The uremic toxicity of indoxyl sulfate and p-cresyl sulfate: a systematic review. J Am Soc Nephrol 2014; 25: 1897–1907.

Jourde-Chiche

Dou

Cerini

, et al. Vascular incompetence in dialysis patients--protein-bound uremic toxins and endothelial dysfunction. Semin Dial 2011; 24: 327–337.

Yamamoto

Tsuruoka

Ioka

, et al. Indoxyl sulfate stimulates proliferation of rat vascular smooth muscle cells. Kidney Int 2006; 69: 1780–1785.

10.

Barreto

Liabeuf

, 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.

11.

Lin

Tsai

Lai

, et al. Serum osteoprotegerin level is positively associated with peripheral artery disease in patients with peritoneal dialysis. Ren Fail 2020; 42: 131–136.

12.

Wang

Lai

Kuo

, et al. Association between serum indoxyl sulfate levels and endothelial function in non-dialysis chronic kidney disease. Toxins 2019; 11: 589.

13.

Lin

Liu

Lai

, et al. Association of serum indoxyl sulfate levels with skeletal muscle mass and strength in chronic hemodialysis patients: a 2-year longitudinal analysis. Calcif Tissue Int 2020; 107: 257–265.

14.

Lin

Hsu

Wang

, et al. Serum indoxyl sulfate as a potential biomarker of aortic arterial stiffness in coronary artery disease. Nutr Metab Cardiovasc Dis 2020; 30: 2320–2327.

15.

Yang

Zhu

Han

, et al. Pathophysiology of peripheral arterial disease in diabetes mellitus. J Diabetes 2017; 9: 133–140.

16.

Signorelli

Fiore

Malaponte

. Inflammation and peripheral arterial disease: the value of circulating biomarkers. Int J Mol Med 2014; 33: 777–783.

17.

Singh

Morris

Smith

, et al. Systematic review and meta-analysis of the association between C-reactive protein and major cardiovascular events in patients with peripheral artery disease. Eur J Vasc Endovasc Surg 2017; 54: 220–233.

18.

Rymer

Swaminathan

Aday

, et al.

The current evidence for lipid management in patients with lower extremity peripheral artery disease: what is the therapeutic target?

Curr Cardiol Rep 2021; 23: 13.

19.

Lano

Burtey

Sallée

. Indoxyl sulfate, a uremic endotheliotoxin. Toxins 2020; 12: 229.

20.

Cunha

RSD

Santos

Barreto

, et al.

How do uremic toxins affect the endothelium?

Toxins 2020; 12: 412.

21.

Six

Flissi

Lenglet

, et al. Uremic toxins and vascular dysfunction. Toxins 2020; 12: 404.

22.

Gondouin

Cerini

Dou

, et al. Indolic uremic solutes increase tissue factor production in endothelial cells by the aryl hydrocarbon receptor pathway. Kidney Int 2013; 84: 733–744.

23.

Poesen

Mutsaers

Windey

, et al. The influence of dietary protein intake on mammalian tryptophan and phenolic metabolites. PLoS One. 2015; 10(10): e0140820.

24.

Patel

Luo

Plummer

, et al.

The production of p-cresol sulfate and indoxyl sulfate in vegetarians versus omnivores

Clin J Am Soc Nephrol. 2012; 7(6): 982–988.

25.

Asai

Kumakura

Kikuchi

. Review of the efficacy of AST-120 (KREMEZIN ^®) on renal function in chronic kidney disease patients. Ren Fail. 2019; 41(1): 47–56.