Endothelin-1 and Parameters of Systolic Blood Pressure in Hemodialysis

Abstract

BACKGROUND

Hypertension is common in hemodialysis (HD) patients. Increased blood pressure (BP) variability, particularly higher and lower extremes, is associated with adverse outcomes. We explored the association of endothelin-1 (ET-1), a potent vasoconstrictor, with different BP parameters (pre-HD, intra-HD, and post-HD) during HD in a contemporary patient cohort.

METHODS

This study uses the DaVita Biorepository, a longitudinal prospective cohort study with quarterly collection of clinical data and biospecimens. Unadjusted and adjusted linear mixed effects regression models were fit to determine association of pre-HD ET-1 (log-transformed and quartiles) with HD-related systolic BP (SBP) parameters (pre-HD, nadir intra-HD, and post-HD). As ET-1 was measured at baseline, analyses were restricted to 1 year of follow-up.

RESULTS

Among 769 participants, mean age was 52 years, 42% were females, and 41% were Black. Mean pre-HD SBP was 152 (±28) mm Hg and mean ET-1 concentration was 2.3 (±1.2) ng/ml. In fully adjusted models, each unit increase in SD of log-transformed ET-1 was associated with a 2.7 (95% confidence interval [CI] 1.5, 4.0) mm Hg higher pre-SBP; 1.6 (95% CI 0.9, 2.3) mm Hg higher nadir SBP; and 2.0 (95% CI 1.1, 2.9) mm Hg higher post-SBP. Each SD increase in log-transformed ET-1 was associated with 21% higher odds of experiencing intradialytic hypertension (odds ratio 1.21; 95% CI 1.10–1.34).

CONCLUSIONS

Higher baseline ET-1 levels are independently associated with higher SBP and higher odds of intradialytic hypertension. These results highlight a potential role for ET-1 in BP control in HD patients and raise the possibility of ET-1 antagonism as a therapeutic target.

Cardiovascular disease is the leading cause of mortality in patients with end-stage kidney disease, accounting for approximately 40% of deaths.¹ While hypertension is extremely common in patients with end-stage kidney disease,² both extremes of lower and higher predialysis systolic blood pressure (SBP) are known to be associated with adverse outcomes.³ Indeed, intradialytic hypertension, defined as an increase in SBP of ≥10 mm Hg from pre- to post-hemodialysis (HD), is associated with increased morbidity and mortality in HD patients.^4–6

There are a multitude of factors that contribute to BP control in HD patients, including positive sodium balance and volume overload,⁷ activation of the renin–angiotensin–aldosterone and sympathetic nervous systems,⁸ removal of antihypertensive medications by HD,⁹ the use of erythropoiesis-stimulating agents,¹⁰ and abnormalities of mineral bone disease axes.¹¹ Of particular note, endothelial cell dysfunction may play a central role via the synthesis and release of humoral factors such as endothelin-1 (ET-1), a potent vasoconstrictor.¹² This is particularly relevant, as pharmacologic antagonists of endothelin receptors are available and have been tested previously in patients with chronic kidney disease (CKD).¹³

Although prior studies have reported that ET-1 concentrations appear to be higher in HD patients, compared with non-HD controls, these were limited in size and duration of follow-up.^5,14 Herein, we examine the association of ET-1 with HD-related BP parameters using a large contemporary and prospective patient cohort. We hypothesized that higher levels of baseline ET-1 would be associated with higher predialysis SBP and development of intradialytic hypertension.

METHODS

Study population

The current analyses were performed using a prospective cohort of anonymized samples and statistically deidentified clinical data from a biorepository assembled by DaVita Clinical Research (DCR) and made available to academic organizations through the Biospecimen Research Grant (BioReG) program. Patients who were <18 years-old, with Hgb <8.0 g/dl, pregnant, or with any physical, mental, or medical condition which limited the ability to provide written informed consent were excluded from BioReG. The present study only included patients undergoing thrice-weekly HD. The sampling protocol was approved by an Institutional Review Board (Quorum IRB, Seattle, WA) and patients provided written informed consent prior to the initiation of sample collection. All clinical and HD prescription data were collected from the electronic medical record. A randomly sampled subset of the total cohort was provided to each of 4 academic institutions by DCR in a deidentified format.

Biospecimen collection and storage

Biospecimens were collected and processed according to a standardized protocol, including shipping on refrigerated packs on the same day as collection, processing, aliquoting, and storage at −80 °C. Recollection was requested for any specimen with cause for rejection (e.g., unspun tubes, insufficient volume, or thawed specimens). Specimens received >48 hours from the time of collection were also rejected and recollected. Samples were distributed frozen at −80 °C across the 4 academic medical centers.

Exposure

The primary exposure for this study was plasma ET-1 concentration, measured at baseline (defined as first pre-HD sample collection in the observation period). Since the distribution of ET-1 is right skewed, data were log-transformed for incorporation as a continuous variable in regression models, with effect estimates reported per 1-unit increase in SD. Plasma ET-1 levels were measured in duplicate at Brigham and Women’s Hospital using a commercially available ELISA kit (Quantikine Human ET-1 Immunoassay PDET100; R&D Systems, Minneapolis, MN). The mean interassay coefficient of variation from blind-split replicates was 7%.

Outcome ascertainment

The primary outcome was defined as pre-HD treatment sitting SBP. Secondary outcomes included the nadir intradialytic SBP and post-HD SBP. We also examined the association of ET-1 with intradialytic hypertension (defined as an increase in SBP of ≥10 mm Hg from pre- to post-HD)¹⁵ and intradialytic hypotension (defined as either an absolute intradialytic nadir SBP <90 mm Hg in patients with a pre-HD SBP of <160 mm Hg or nadir SBP <100 mm Hg in patients with pre-HD SBP ≥160 mm Hg).¹⁶ BP was measured at all study sessions per standard clinical guidelines.

Assessment of other covariates

Demographic information including age, race, sex, dialysis access, and comorbid conditions including diabetes, heart failure, coronary artery disease, cerebrovascular disease, peripheral vascular disease, and chronic obstructive pulmonary disease, were recorded at baseline and updated from the medical record (via ICD-9 codes) throughout the study. Additional information such as the HD prescription, erythropoietin dose, vascular access, and laboratory data were collected at each session. Dialysis session length was categorized (≤180, 181–209, 210–239, and ≥240 min). Ultrafiltration volume was calculated by subtracting the postdialysis weight from the predialysis weight.

Statistical analysis

All analyses were performed using R version 3.3.3 and Stata MP version 16 (StataCorp LP). Continuous variables were summarized using means (SDs) or medians (25th–75th percentiles) and compared with analysis of variance or Kruskal–Wallis tests, according to data distribution. Categorical variables were summarized as percentages and compared with chi-squared tests. As ET-1 is non-normally distributed, ET-1 values were log-transformed for further analyses. Initially, unadjusted repeated measures regression models (to account for within person correlation) were fit to determine the association of ET-1 (first log-transformed and then in quartiles) with outcomes of interest. Subsequently, multivariable adjusted linear mixed effects models were fit, by adding a random intercept for subject-wise variability. Model 1 adjusted for age, sex, race, access, and pre-HD SBP (the latter was excluded from analyses where pre-HD SBP was the outcome of interest); model 2 adjusted for the same variables as model 1 with additional adjustment for categories of session length, ultrafiltration volume, diabetes, heart failure, ischemic heart disease (history of coronary artery disease, myocardial infarction, or angina), peripheral vascular disease, lung disease, erythropoietin dose, and pre-HD SBP. We assessed for evidence of effect modification according to race and gender^17,18 via the inclusion of cross-product terms. We also performed an exploratory analysis using model 2 with additional adjustment for hemoglobin. The basis for the models was chosen based on clinical and biological plausibility. A 2-sided P value <0.05 was considered to be statistically significant.

RESULTS

We examined data from 769 subjects and 110,315 HD sessions from the BioReG cohort (Figure 1). There was a higher frequency of females and a lower frequency of Black participants in those excluded, compared with those included, in the final analyses. Otherwise, baseline characteristics between these groups were similar (Supplementary Table S1 online). The mean age of included participants was 52 ± 22 years, 42% were females, and 41% Black. Mean pre-HD SBP was 152 ± 28 mm Hg and mean ET-1 concentration was 2.3 ± 1.2 ng/ml. Subjects in higher quartiles of baseline ET-1 tended to be younger, diabetic, have higher ultrafiltration volume, lower serum albumin, and higher frequency of erythropoietin use (Table 1).

Figure 1.

Consort diagram. Abbreviations: ET-1, endothelin-1; HD, hemodialysis.

Table 1.

Baseline characteristics of participants according to quartiles of plasma endothelin-1

	Quartile 1 (n = 193)	Quartile 2 (n = 192)	Quartile 3 (n = 192)	Quartile 4 (n = 192)	P
Endothelin-1, ng/ml	1.3 [1.1, 1.4]	1.8 [1.7, 1.9]	2.3 [2.1, 2.5]	3.3 [2.9, 4.0]	<0.001
Age, years	56 ± 22	52 ± 22	53 ± 21	48 ± 22	<0.001
Female, n (%)	74 (38.3%)	86 (44.8%)	79 (41.1%)	80 (41.7%)	0.69
Predialysis weight, kg	89.5 ± 23.8	91.1 ± 23.5	92.9 ± 24.9	90.2 ± 24.5	0.63
Ultrafiltration volume, l	1.9 ± 1.6	2.2 ± 1.6	2.0 ± 1.5	2.4 ± 1.5	0.004
Race, n (%)					0.84
White	81 (42.0%)	85 (44.3%)	80 (41.7%)	76 (39.6%)
Black	68 (35.2%)	77 (40.1%)	85 (44.3%)	84 (43.8%)
Other	44 (22.8%)	30 (15.6%)	27 (14.1%)	32 (16.7%)
Access, n (%)					0.93
AVF	137 (71.0%)	124 (64.6%)	129 (67.2%)	131 (68.2%)
AVG	24 (12.4%)	28 (14.6%)	29 (15.1%)	34 (17.7%)
Tunneled catheter	32 (16.6%)	40 (20.8%)	34 (17.7%)	27 (14.1%)
Session length, min	210 [180, 227]	210 [182, 240]	214 [188, 240]	210 [182, 240]	0.04
Diabetes, n (%)	67 (34.7%)	81 (42.2%)	101 (52.6%)	94 (49.0%)	<0.001
Hypertension, n (%)	67 (34.7%)	50 (26.0%)	68 (35.4%)	48 (25.0%)	0.18
Ischemic heart disease, n (%)	15 (7.8%)	12 (6.2%)	19 (9.9%)	16 (8.3%)	0.54
Serum albumin, g/dl	3.6 ± 0.6	3.6 ± 0.5	3.5 ± 0.5	3.4 ± 0.5	0.005
Erythropoietin, n (%)	91 (47%)	103 (54%)	101 (53%)	126 (66%)	<0.001
Erythropoietin dose, units per HD	5,500 [3,300, 11,000]	5,500 [3,300, 10,000]	7,700 [4,000, 11,000]	7,700 [4,400, 11,000]	0.13

Values for continuous variables are presented as mean (±SD), median (25th–75th percentile). Abbreviations: AVF, arteriovenous fistula; AVG, arteriovenous graft; HD, hemodialysis. P values reflect trend across quartiles.

ET-1 and pre-HD SBP

The baseline differences in HD-related BP parameters according to quartiles of ET-1 are presented in Table 2. In unadjusted repeated measures analyses, each SD increase in log-transformed ET-1 was associated with 3.4 (95% confidence interval [CI] 2.2–4.6) mm Hg higher pre-HD SBP. In the fully adjusted model, each SD increase in log-transformed ET-1 was associated with 2.7 (95% CI 1.5–4.0) mm Hg higher pre-HD SBP. There was no evidence for effect modification according to gender (P-interaction = 0.92) or diabetes (P-interaction = 0.51). In categorical analyses, there was a monotonic increase in the association of quartiles of ET-1 with pre-HD SBP (7.1 [95% CI 3.7–10.6] mm Hg higher pre-HD SBP for Q4, compared with Q1; Figure 2).

Table 2.

Baseline systolic blood pressure parameters of participants according to quartiles of plasma endothelin-1

	Quartile 1 (n = 193)	Quartile 2 (n = 192)	Quartile 3 (n = 192)	Quartile 4 (n = 192)	P
Predialysis SBP, mm Hg	149 ± 26	151 ± 27	151 ± 28	157 ± 30	0.003
Nadir intradialytic SBP, mm Hg	119 ± 25	119 ± 24	120 ± 24	123 ± 27	0.15
Postdialysis SBP, mm Hg	145 ± 27	144 ± 27	145 ± 24	150 ± 31	0.06

Values for continuous variables are presented as mean (±SD), median (25th–75th percentile). Abbreviation: SBP, systolic blood pressure. P values represent trend across groups.

Figure 2.

Association of quartiles of endothelin-1 with systolic blood pressure (SBP) outcomes. Abbreviation: Q1, quartile 1.

ET-1 and nadir and post-HD SBP

The association of ET-1 with other HD-related BP parameters was evaluated in unadjusted and adjusted repeated measures analysis (Table 3). Overall, in the fully adjusted model, each SD increase in log-transformed ET-1 was associated with 1.6 (95% CI 0.9–2.3) mm Hg higher nadir SBP and 2.0 (95% CI 1.1–2.9) mm Hg higher post-HD SBP. In categorical analyses, there was a monotonic association of higher quartiles of ET-1 with higher SBP in all parameters of interest (Figure 2). Similar patterns of association were observed in exploratory models that additionally adjusted for hemoglobin (Supplementary Table S2 online) and when analyses were restricted to 30 days of follow-up (Supplementary Table S3 online).

Table 3.

Unadjusted and adjusted multiple linear mixed effects regression models

SBP parameter	Model	Change in SBP per SD increase in log ET-1 (mm Hg)	95% CI	P
Pre-HD SBP	Unadjusted	3.4	[2.2, 4.6]	<0.001
	Model 1	3.3	[2.2, 4.5]	<0.001
	Model 2	2.7	[1.5, 4.0]	<0.001
Nadir SBP	Unadjusted	2.7	[1.7, 3.7]	<0.001
	Model 1	1.4	[0.6, 2.1]	<0.001
	Model 2	1.6	[0.9, 2.3]	<0.001
Post-HD SBP	Unadjusted	2.8	[1.7, 4.0]	<0.001
	Model 1	1.7	[0.8, 2.6]	<0.001
	Model 2	2.0	[1.1, 2.9]	<0.001

Models: unadjusted; model 1 adjusted for: age, gender, race, HD access, and pre-HD systolic BP; model 2: adjusted for same as model 1 plus categories of session length, ultrafiltration volume, diabetes, congestive heart failure, ischemic heart disease, peripheral vascular disease, lung disease, and erythropoietin dose. Abbreviations: CI, confidence interval; ET-1, endothelin-1; HD, hemodialysis; SBP, systolic blood pressure.

ET-1 and intradialytic hypertension and intradialytic hypotension

In adjusted analyses, each SD increase in log-transformed ET-1 was associated with 21% higher odds of experiencing intradialytic hypertension (odds ratio 1.21; 95% CI 1.10–1.34), and a nominally, but statistically nonsignificant, lower risk of intradialytic hypotension (odds ratio 0.94; 95% CI 0.85–1.04).

DISCUSSION

In this large contemporary cohort of maintenance HD patients, we found that higher baseline ET-1 concentrations are associated with higher pre-HD SBP, nadir intradialytic SBP, and higher post-HD SBP. Similarly, higher ET-1 was associated with greater odds of developing intradialytic hypertension.

ET-1 is a 21-amino-acid peptide that was initially thought to be primarily an endothelium-derived vasoconstrictor but is now known to be produced by many other cell types,¹⁹ including podocytes¹³ and glomerular endothelial cells.²⁰ There are 2 broad receptor isoforms, ET_A and ET_B. Binding to the ET_A receptor results in vasoconstriction, cell matrix accumulation, and proliferation, while binding to ET_B produces the opposite counter-regulatory responses. ET-1 concentrations are elevated in several disease states and are associated with elevations in inflammatory markers, glomerular injury, and fibrosis in animal models.¹³ Studies in non-CKD patients have reported elevated ET-1 levels in patients with atherosclerosis and highlight a central role in the development of hypertension via increased vascular resistance.²¹ Increased levels of ET-1 have also been associated with development of cardiovascular disease and progression of CKD.²²

Not unexpectedly, ET-1 concentrations appear to be higher with more advanced stages of CKD.^23–25 In an ambulatory BP study of 27 patients with CKD, higher ET-1 was associated with higher mean BP and a lower frequency of nocturnal dipping, with some reversal of these findings following administration of an ET-1 antagonist.²⁶ Similar findings have been reported among patients on maintenance HD. For example, Teng et al. compared 17 stable Chinese HD patients aged <75 years with intradialytic hypertension to 17 age- and sex-matched controls, reporting that post‐HD serum ET‐1 concentrations were significantly higher in those who developed intradialytic hypertension (4.1 ± 2.1 vs. 2.8 ± 1.3 pg/ml, P-difference <0.05).²⁷ Similar findings were reported by Chou et al. from a slightly larger case–control study (n = 60), where post-HD ET-1 concentrations were almost 2-fold higher in those who developed intradialytic hypertension, compared with those who did not.²⁸ In addition, Ottosson-Seeberger et al. reported that infusions of ET-1 resulted in higher mean arterial pressure in 5 HD patients without a documented history of cardiac disease, supporting a central role for ET-1 in the regulation of BP in HD patients.²⁹ Our results from a much larger, contemporary cohort in the United States are consistent with these reports, and provide additional data related to the absolute magnitude of BP changes associated with ET-1 during the course of HD sessions. The association of higher ET-1 with hospitalization and death has been addressed by a prior analysis of the same cohort by Li et al. In fully adjusted models, they reported a 1.46-fold increased risk of death (hazard ratio [HR] 1.46, 95% CI 1.26–1.69) and 1.15-fold increased risk of hospitalization (HR 1.15, 95% CI 1.05–1.25) per unit increase in SD of log-transformed ET-1, further highlighting the potential importance of this pathway for adverse outcomes in HD patients.³⁰

Our results confirm the presence of a pronounced association of ET-1 with BP control in HD patients. While some prior studies have reported higher ET-1 concentrations in normotensive¹⁷ and hypertensive¹⁸ Black populations, when compared with white populations, we did not find evidence for effect modification according to self-reported race in our analyses. Prior studies of have also reported higher ET-1 levels in non-end-stage renal disease patients with diabetes,^31–33 compared with nondiabetics. While data regarding differences in ET-1 in patients with diabetes on maintenance HD are sparse, we did not find differences in baseline ET-1 according to diabetes, nor evidence of effect modification of the association with HD-related BP parameters. Our results are important, given the association of intradialytic hypertension with adverse outcomes in HD patients, and are particularly relevant given the availability of ET-1 receptors antagonists. Of note, the prior use of such agents in patients with CKD have been hampered by adverse effects related to hypervolemia,^34,35 providing the impetus for recent trials to have an enrichment period to identify those most likely to benefit and least likely to have adverse effects.³⁶ In theory, the risks of hypervolemia from ET receptor blockade may be lower in patients on maintenance HD therapy, who tend to be oligoanuric and less prone to renal-mediated sodium retention.

There are several strengths to our study, including the relatively large sample size, duration of follow-up, and availability of detailed HD-related hemodynamic data. However, there are several limitations to mention. These include the use of a single baseline measurement of ET-1, nonavailability of laboratory data for each HD session, limited data on the dialysate prescription, temperature, and potential for misclassification of covariables via the use of ICD-9 codes. Additionally, measures of bioimpedance, medications, and patient symptoms were not systematically recorded in this study. Furthermore, clinical BPs were used for the purposes of this study, per routine management and although these allow for a “real-word” understanding, clinical BP measurements may be unreliable and ambulatory BP data were not available. The reasons for nonavailability of blood samples for several patients were not provided and raise the possibility of some element of selection bias. Indeed, despite the performance of multivariable adjusted models, the potential for residual confounding remains. Furthermore, this was an observational study and therefore hypothesis generating. Finally, this represents a contemporary outpatient cohort from the United States, which may not be generalizable to other cohorts.

In conclusion, we observed strong association between ET-1 and higher HD-related parameters of SBP. These results support a potential role of ET-1 in BP regulation in maintenance HD patients and provide rationale for testing ET-1 antagonists in future interventional studies.

FUNDING

Dr McCausland is supported NIDDK grants U01DK096189, R03DK122240, and K23DK102511. Dr Sabbisetti is supported by NIH grants DK127587, DK085660, and CA229772.

DISCLOSURE

The authors declared no conflict of interest.

DATA AVAILABILITY

The data used in these analyses were provided by DaVita Clinical Research. Requests for access to data can be made in writing to DaVita Clinical Research.