Assessment of right ventricular dysfunction and its association with excess risk of cardiovascular events in patients undergoing maintenance hemodialysis

Abstract

Aims

Recent accumulating evidence has recently documented a significant prevalence of right ventricular dysfunction (RVD) in end-stage renal disease (ESRD) patients. Tricuspid annular plane systolic excursion (TAPSE)/pulmonary-artery systolic pressure (PASP) ratio assessed with echocardiography might be a useful clinical index of right ventricular (RV) -pulmonary arterial (PA) coupling. The current study aimed to investigate the value of the TAPSE/PASP ratios in patients on maintenance hemodialysis (MHD).

Methods

We studied 83 times echocardiographic tests from 68 patients with MHD. The associations of TAPSE/PASP ratios with echocardiography variables, clinical characteristics, and biochemical parameters were analyzed, as well as the associations of TAPSE/PASP ratios with odds of all-cause mortality, cardiovascular disease (CVD) events and frequent intermittent dialysis hypotension (IDH).

Results

Correlation analysis showed TAPSE/PASP ratios positively correlated with LVEF and negatively correlated with E/A and E/e’ values. For clinical and biochemical parameters, TAPSE/PASP ratios negatively correlated with BNP, NT-proBNP, age, CRP, and average interdialysis weight gain (ΔBW) and positively correlated with albumin. Logistic regression analysis, which induced the TAPSE/PASP ratio as a continuous variable (per 0.1 mm/mmHg increase), identified that the TAPSE/PASP ratio was associated with decreased CVD events (OR 0.386 [95% CI 0.231–0.645], p < 0.001) and frequent IDH odds (OR 0.571 [95% CI 0.397–0.820], p = 0.002). Moreover, the TAPSE/PASP ratio independently predicted CVD events (adjusted HR 0.539 [95% CI 0.391–0.743], p < 0.001) during a follow-up period of 12 months.

Conclusions

RVD, assessed by echocardiography TAPSE/PASP ratio, was found to be associated with increased risks of CVD events and frequent IDH in patients with MHD.

Keywords:

• Right ventricular dysfunction
• echocardiography
• hemodialysis
• cardiovascular disease

1. Introduction

Hemodialysis is the most common modality of renal replacement therapy worldwide in patients with end-stage renal disease (ESRD). Recent studies have shown that the number of patients undergoing maintenance hemodialysis (MHD) is increasing annually worldwide [1]. Cardiovascular diseases (CVD) such as atherosclerosis, hypertension, heart failure (HF), and coronary artery disease are the leading cause of death in patients with MHD, which is significantly higher than that among their counterparts in the general population [2,3].

While left ventricular dysfunction (LVD) was widely investigated in the field of dialysis [4–7], increasing studies have recently documented a significant prevalence of right ventricular dysfunction (RVD) in ESRD patients, including patients with dialysis [8–10]. In addition, right ventricular (RV) function was poorer in patients with hemodialysis compared to that in patients with peritoneal dialysis, regardless of left ventricular (LV) function and pulmonary hypertension [11]. Moreover, greater risks of death, arrhythmias, and cardiogenic shock have been identified in the presence of RVD [12–14].

Tricuspid annular plane systolic excursion (TAPSE), measured employing M-mode echocardiography, is decreased with impaired RV function [14] In the absence of right ventricular outflow tract obstruction or pulmonic stenosis, RV systolic pressure equals pulmonary-artery systolic pressure (PASP). Thus, the ratio of TAPSE to PASP estimates right ventricular function and its afterload coupling. Guazzi et al. first reported that the TAPSE/PASP ratio assessed with echocardiography might be a useful clinical index of RV function [15]. Moreover, the ratio is thought to provide more comprehensive information about right heart performance than TAPSE or PASP alone [16].

Previous studies showed the prognostic significance of the TAPSE/PASP ratio in patients with HF even with preserved ejection fraction [16–21]. Currently, limited evidence about its predictive value in patients with MHD is available. The aim of this analysis was to investigate the value of noninvasive RV–pulmonary arterial (PA) coupling in predicting mortality and CV events, evaluated using the TAPSE/PASP ratio, in patients with MHD, as well as the association between TAPSE/PASP ratio and metabolic, nutritional and inflammatory parameters.

2. Methods

2.1. Study population

This study was a single-center, retrospective clinical study approved by the Ethics Committee of Peking University First Hospital and conducted according to the Declaration of Helsinki. 68 patients with dialysis vintage greater than one year at Peking University First Hospital and 83 echocardiographic tests were involved from December 2021 to March 2023. The participants were clinically stable and with a proper acoustic window for echocardiography and strain analysis. All the tests were conducted on the non-dialysis day after the first or second dialysis session of the week, and the interval between the period of each echocardiographic test in the same patient was longer than one year. The exclusion criteria were unwillingness to participate in the study, dialysis for acute kidney disease, severe pulmonary hypertension due to pulmonary diseases or pulmonary vascular diseases, other severe valvular diseases, arrhythmias (including atrial fibrillation), disabling stroke and active malignancy.

2.2. Echocardiography

Echocardiographic studies were performed according to the guidelines [11,14]. LV systolic function is mainly assessed by left ventricular ejection fraction (LVEF) [22], and the LVEF was calculated using the biplane-modified Simpson’s method. TAPSE was measured as the total displacement of the tricuspid annulus (in mm) from end diastole to end systole by two-dimensional echocardiography-guided M-mode recording from the apical 4-chamber view. The maximal peak velocity of tricuspid regurgitation was measured. PASP was calculated using the transtricuspid pressure gradient plus the estimate of the right atrial pressure by the diameter and collapsibility of the inferior vena cava.

Measurements of septal ratios of transmitral flow velocity to annular velocity (E/e’) and overall mitral inflow peak E to peak A velocity ratios (E/A), which are used as an assessment of diastolic function, and other echocardiographic parameters were obtained using the software program installed on the ultrasound machine [23,24].

2.3. Biochemical parameters

Routinely determined biochemical parameters were measured in the same month as echocardiography before the first dialysis session of the week. The following parameters were evaluated: serum potassium, calcium, phosphorus, serum creatinine (Scr), blood urea nitrogen (BUN), hemoglobin (Hb), intact parathormone (iPTH), albumin, C-reactive protein (CRP), B-type natriuretic peptide (BNP), and N-terminal pro-BNP (NT-proBNP).

Δ Body weight (BW) was measured using the formula, ΔBW = interdialytic weight gain/dry weight, and the average ΔBW defined as the average values of ΔBW in six dialysis sessions around echocardiography.

Dialysis adequacy was measured using the single-pool Daugirdas formula [25].

2.4. Outcomes

We examined the following outcomes: CVD events (stroke, nonfatal myocardial infarction, unstable angina, and HF) and all-cause mortality within one year since echocardiography. Mortality and CVD events were identified by referring to medical records and confirmed by direct contact with the patients, relatives, and in-charge physicians.

Intradialytic hypotension (IDH) was defined as having a drop of systolic blood pressure of >20mmHg during dialysis, accompanied by any clinical events attributable to hypotension, such as dizziness, nausea, muscle cramp and other manifestations of tissue hypoperfusion, and in need of nursing interventions [26,27]. Frequent IDH was defined as IDH events that had occurred at least three times in all the dialysis sessions in the following month of echocardiography.

2.5. Statistical analysis

After determining the best latent profile model, statistical analyses were performed using the SPSS 26.0 software. Continuous variables with normal distribution were described by mean and standard deviation, and non-normal distribution was described by median and interquartile range. Categorical variables were described by frequency and percentage. Univariate analysis was performed by the Kruskal–Wallis test, one-way analysis of variance (ANOVA), or the chi-square test. Correlation between quantitative variables was done using Pearson’s or Spearmann’s correlations. Logistic regression tests were performed to explore the correlations among the assessed parameters. The clinical endpoint was assessed with the Kaplan–Meier method and compared with the log-rank test. Cox regression models were used to calculate hazard ratios (HR) and 95% CI for each endpoint. The predictive value of these models was assessed by the area under the receiver operating characteristic (ROC) curve (AUC). p < 0.05 was considered statistically significant.

3. Results

3.1. Characteristics of the study population

Data from Table 1 show the characteristics of the analyzed patients. There were a total of six deaths and 21 CVD events during the one-year follow-up since echocardiographic tests.

Table 1. Baseline clinical characteristics divided with TAPSE/PASP.

	All tests	TAPSE/PASP < 0.65mm/mmHg	TAPSE/PASP ≥ 0.65mm/mmHg	P value
Numbers, tests/patients	83/68	43/35	40/33
Age, y	62.4 ± 11.7	64.5 ± 12.2	59.3 ± 10.7	0.047
Female, %	51.8	51.2	52.5	0.900
Dialysis vintage, y	5.5 ± 3.7	5.1 ± 3.1	5.6 ± 3.7	0.505
CKD etiology, n DKD/HN/CGN/PKD/others*	18/15/22/13/15	9/9/10/6/9	9/6/12/7/6	0.705
Urine volume	0 [range 0–1500]	0 [range 0–1200]	0 [range 0–1500]	0.019
Previous CVD events, %	50.6	60.5	40.0	0.033
Diabetes mellitus, %	39.8	37.2	42.5	0.624
Charlson Comorbidity index	7.2 ± 2.4	7.7 ± 2.4	6.6 ± 2.3	0.051
Serum albumin, g/L	37.1 ± 3.4	36.2 ± 3.3	38.2 ± 3.4	0.010
BNP, pg/mL	902 [IQR 240–2963]	1829 [IQR 782–4083]	240 [IQR 181–832]	0.002
NT-proBNP, pg/mL	7182 [IQR 2877–18772]	17548 [IQR 7446–35000]	2883 [IQR 1452–5259]	<0.001
Serum potassium, mmol/L	4.92 ± 0.67	4.92 ± 0.61	4.90 ± 0.75	0.889
Serum calcium, mmol/L	2.26 ± 0.20	2.30 ± 0.22	2.23 ± 0.19	0.123
Serum phosphorous, mmol/L	1.81 ± 0.63	1.86 ± 0.57	1.79 ± 0.72	0.620
Serum iPTH, pg/mL	230.2 ± 119.7	231.7 ± 119.5	235.2 ± 122.4	0.907
Hemoglobin, g/dL	10.8 ± 1.3	10.5 ± 1.4	11.1 ± 1.1	0.010
CRP, mg/L	3.39 [IQR 1.64–9.49]	4.53 [IQR 2.28–17.43]	2.71 [IQR 0.83–5.62]	0.010
Kt/V	1.45 ± 0.26	1.50 ± 0.28	1.40 ± 0.23	0.109
URR	70.1 ± 6.2	71.4 ± 6.6	69.2 ± 5.8	0.144
Average ΔBW, %	3.37 ± 1.69	3.69 ± 1.71	3.16 ± 1.71	0.183
All-cause mortality, n	6	5	1	0.203
CVD events, n	21	19	2	<0.001
Frequent IDH, n	25	17	8	0.015

Abbreviations: BNP, B-type natriuretic peptide; BW, body weight; CGN, chronic glomerulonephritis; CKD, chronic kidney disease; CVD, cardiovascular diseases; CRP, C-reactive protein; DKD, diabetic kidney disease; HN, hypertensive nephropathy; IDH, intradialytic hypotension; iPTH, intact parathormone; NT-proBNP, N-terminal pro-B-type natriuretic peptide; PKD, polycystic kidney disease.

P < 0.05 are considered statistically significant and highlighted in bold.

* Other causes of CKD included antineutrophil cytoplasmic antibody-associated vasculitis, lupus nephritis, anti-glomerular basement membrane antibody-associated disease, chronic interstitial nephritis, monoclonal gammopathy of renal significance, fibrillary glomerulopathy, and renal carcinoma after surgery.

Most studies assigned the optimal cutoff for TAPSE/PASP by Cox proportional hazards, whereas some considered the median TAPSE/PASP ratio value of the overall population as the cutoff. The current study stratified the population according to the median TAPSE/PASP value (0.65 mm/mmHg). Patients with TAPSE/APSP values 0.65 mm/mmHg or less were significantly associated with lower albumin concentrations, higher BNP or NT-proBNP, and higher CRP levels. They also had significantly higher rates of CVD events and frequent IDH.

3.2. Echocardiography characteristics and clinical parameters

Data from Table 2 show the echocardiography variables of the analyzed patients.

Table 2. Echocardiography variables.

	All tests	TAPSE/PASP <0.65mm/mmHg	TAPSE/PASP ≥0.65mm/mmHg	P value
TAPSE, mm	20.7 ± 3.8	20.0 ± 3.6	21.4 ± 3.8	0.084
PASP, mmHg	34.1 ± 10.9	40.6 ± 11.1	27.1 ± 5.0	<0.001
TAPSE/PASP, mm/mmHg	0.65 ± 0.19	0.51 ± 0.10	0.80 ± 0.15	<0.001
LVEF, %	61.2 ± 12.5	58.1 ± 13.9	65.2 ± 10.2	0.011
IVST, cm	1.14 ± 0.16	1.17 ± 0.16	1.11 ± 0.15	0.068
LVPW, cm	1.09 ± 0.16	1.11 ± 0.16	1.07 ± 0.15	0.252
LVESD, cm	3.18 ± 0.76	3.32 ± 0.90	2.99 ± 0.75	0.052
LVEDD, cm	4.91 ± 0.77	4.87 ± 0.88	4.89 ± 0.60	0.914
RVD, cm	2.05 ± 0.33	2.08 ± 0.36	1.99 ± 0.30	0.229
RAD, cm	5.03 ± 0.72	5.03 ± 0.75	4.83 ± 0.52	0.343
E/A	0.83 ± 0.38	0.92 ± 0.49	0.75 ± 0.20	0.054
E/e’	15.9 ± 6.1	18.4 ± 6.9	13.3 ± 3.9	<0.001

Abbreviations: E/A, increased overall mitral inflow peak E/peak A velocity flow in end-diastole period; E/e′, peak E velocity flow in early-diastole period/peak e velocity flow in end-diastole period; IVST, interventricular septum thickness; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; LVEDD, left ventricular end-diastole diameter; LVPW, left ventricular posterior wall thickness; PASP, pulmonary artery systolic pressure; RVD, right ventricular diameter; RAD, right atria diameter; TAPSE, tricuspid annular plane excursion.

P < 0.05 are considered statistically significant and highlighted in bold.

TAPSE/PASP ratios positively correlated with LVEF (r = 0.363, p = 0.001) and negatively correlated with E/A and E/e’ values (r= −0.257, p = 0.027; r=-0.514, p < 0.001, respectively) (Figure 1), as LVEF negatively correlated with E/e’ values (r= −0.398, p < 0.001) as well.

Figure 1. TAPSE/PASP ratios positively correlated with LVEF and negatively correlated with E/a and E/e’ values.

TAPSE/PASP ratios negatively correlated with BNP and NT-proBNP (r= −0.613, p < 0.001; r= −0.659, p < 0.001, respectively), as LVEF negatively correlated with BNP and NT-proBNP (r= −0.462, p = 0.020; r=-0.446, p < 0.001, respectively) as well. Figure 2 showed the logarithm of BNP or NT-proBNP and the value of TAPSE/PASP or LVEF.

Figure 2. Both TAPSE/PASP ratios and LVEF negatively correlated with BNP and NT-proBNP.

TAPSE/PASP ratios positively correlated with albumin (r = 0.359, p = 0.001), and negatively correlated with age, CRP and average ΔBW (r= −0.279, p = 0.013; r= −0.254, p = 0.026; r= −0.251, p = 0.030, respectively) (Figure 3).

Figure 3. TAPSE/PASP ratios positively correlated with albumin, and negatively correlated with age, CRP and average ΔBW.

3.3. RV-PA uncoupling and prognosis

Logistic regression analysis, inducing the TAPSE/PASP ratio as a continuous variable (per 0.1 mm/mmHg increase), identified that the TAPSE/PASP ratio was associated with CVD events odds (Odds ratio (OR) 0.386 [95% CI 0.231–0.645], p < 0.001) and frequent IDH odds (OR 0.571 [95% CI 0.397–0.820], p = 0.002).

Figure 4(A) illustrates the survival curve of CVD events according to TAPSE/PASP <0.65 mm/mmHg (Log-rank p < 0.001), and Figure 4(B) illustrates the survival curve composite outcomes including both CVD events and death according to TAPSE/PASP < 0.65 mm/mmHg (Log-rank p < 0.001). In addition, Cox regression analysis, inducing the TAPSE/PASP ratio as a continuous variable (per 0.1 mm/mmHg increase), also identified that the TAPSE/PASP ratio was associated with CVD events (HR 0.530 [95% CI 0.397–0.709], p < 0.001). The LVEF, E/e’, and BNP (HR 0.959 [95% CI 0.925–0.991], p = 0.013; HR 1.063 [95% CI 1.005–1.123], p = 0.032; HR 1.003 [95% CI 1.000–1.007], p = 0.035, respectively) were also significantly associated with the CVD events through an univariable Cox regression model, and TAPSE/PASP ratio was proven to be significantly related to CVD events even after multivariable adjustment (adjusted HR 0.539 [95% CI 0.391–0.743], p < 0.001) (Table 3). No significant association between echocardiography variables and all-cause mortality odds was found in the Logistic and Cox regression analysis.

Figure 4. Kaplan–Meier Survival curves for CVD events and composite outcome including both CVD events and death based on median TAPSE/PASP value (0.65 mm/mmHg).

Table 3. Cox regression model for prognostic prediction of CVD events.

	Unadjusted HR (95% CI)	P value	Adjusted HR (95% CI)	P value
Age, y	1.003 (0.967-1.040)	0.870	0.991 (0.959-1.025)	0.608
Female, %	0.849 (0.360-2.000)	0.708	0.619 (0.252-1.520)	0.295
BNP, ng/dL	1.003 (1.000-1.007)	0.035
TAPSE/PASP, mm/mmHg	0.530 (0.397–0.709)	<0.001	0.539 (0.391-0.743)	<0.001
LVEF, %	0.959 (0.925-0.991)	0.013	0.984 (0.951-1.018)	0.347
E/e’	1.063 (1.005-1.123)	0.032

P < 0.05 are considered statistically significant and highlighted in bold.

3.4. ROC curve analysis for TAPSE/PASP, NT-proBNP and PASP in predicting CVD events and frequent IDH

The ROC curve was plotted to analyze the application of the reciprocal of TAPSE/PASP, NT-proBNP and PASP in predicting CVD events or frequent IDH. On predicting CVD events, the reciprocal of TAPSE/PASP (AUC = 0.816, p < 0.001) exhibited a better effect than PASP (AUC = 0.769, p = 0.001), but a worse effect than NT-proBNP (AUC = 0.898, p < 0.001). On predicting frequent IDH, the reciprocal of TAPSE/PASP (AUC = 0.726, p = 0.004) also exhibited a better effect than PASP (AUC = 0.652, p = 0.049), and a similar effect to NT-proBNP (AUC = 0.726, p = 0.003) (Supplementary material). The cutoff values, sensitivity and 1-specificity were summarized in Table 4.

Table 4. ROC curve analysis for related indicators of CVD events and frequent IDH.

	AUC	P value	95% CI	Cutoff value	Sensitivity	1-specificity
CVD events
the reciprocal of TAPSE/PASP, mmHg/mm	0.816	<0.001	0.700-0.931	1/0.65	0.889	0.413
NT-proBNP, pg/ml	0.898	<0.001	0.811-0.985	11852	0.889	0.196
PASP, mmHg	0.769	0.001	0.639-0.900	33.7	0.778	0.261
frequent IDH
the reciprocal of TAPSE/PASP, mmHg/mm	0.726	0.004	0.598-0.855	1/0.62	0.752	0.372
NT-proBNP, pg/ml	0.726	0.003	0.596-0.856	5493	0.857	0.465
PASP, mmHg	0.652	0.049	0.504-0.801	35.5	0.524	0.256

4. Discussion

Combining RV systolic function with pulmonary circulation pressure, the TAPSE/PASP ratio is a noninvasive easy-to-use echocardiography-based assessment approach. This ratio may provide a novel perspective for evaluating RV function. Our results revealed significant correlations between the TAPSE/PASP ratios and LV systolic and diastolic function, BNP, NT-proBNP, albumin, age, CRP and interdialytic weight gain. Additionally, TAPSE/PASP ratios were associated with the risk of CVD events, and frequent IDH as well. It was the first report of the assessment of the noninvasive RV–PA coupling in patients with MHD.

In the echocardiography variables assay of the current study, TAPSE/PASP ratios positively correlated with LVEF and negatively correlated with E/A and E/e’ values, while LVEF negatively correlated with E/e’ values with a lower correlation coefficient, indicating that RV function is more relevant with diastolic function and RVD might occur before LVD in patients with MHD. Similarly, a study of patients with HF with preserved ejection fraction (HFpEF) revealed that a lower TAPSE/PASP ratio was independently associated with adverse outcomes [21]. Moreover, it has been documented that impaired RV function would result in more severely impaired LV function [28], all indicating that RV function would be of much concern in patients with MHD.

In the clinical parameters assay of the current study, TAPSE/PASP ratios positively correlated with albumin, and negatively correlated with age CRP and interdialytic weight gain, indicating worse nutritional status, aging, chronic inflammation conditions and relatively high interdialytic weight gains may impair RV function. On the one hand, malnutrition is one of the complications of chronic kidney disease (CKD). A study by Chien proved that malnutrition defined as decreased albumin concentrations, was related to ventricle remodeling [29]. Otaki also assumed that low albumin levels in patients with CKD and HF were associated with a worse prognosis [30]. On the other hand, higher interdialytic weight gains were remarkably significantly associated with higher mortality risk and cardiovascular hospitalization, as well as rapid renal urea clearance decline [31,32]. Thus, our results indicate a way to improve RV function and prognosis of patients with MHD by improving their nutritional status and extensive volume control.

In regression analysis of the current study, TAPSE/PASP ratios were identified as associated with odds of CVD events and frequent IDH. On the one hand, Nakagawa et al. reported that RV–PA uncoupling, defined as TAPSE/PASP ratio <0.48 mm/mmHg, was associated with a 2.75-fold increased risk of the composite endpoint of all-cause death or HF hospitalization in a large cohort of patients with HFpEF [21]. Similarly, patients in our study with a worse TAPSE/PASP ratio (<0.65 mm/mmHg) presented higher BNP and NT-proBNP levels, and worse LV systolic and diastolic function, suggesting a more severe cardiac involvement. On the other hand, IDH is one of the common complications of hemodialysis and is associated with all-cause and cardiovascular mortality and hospitalization odds [33]. Previous studies revealed that changes in the septal E/e’ ratio correlated with increased LV filling pressure and were the major contributing factors to IDH [34]. Taking the correlation of TAPSE/PASP and E/e’ into consideration, improving RV function may decrease the risk of IDH as well.

Several limitations of this study should be remarked upon. First, due to the assessment of RV function being a newly developed item in echocardiography in our center, the data was relatively limited, however, we still got those significant results, indicating that RVD in patients with hemodialysis would be worthwhile to investigate further. Second, the study was performed using echocardiography, whereas cardiac magnetic resonance imaging is considered the gold standard for right ventricular functional assessment.

5. Conclusion

The noninvasive RV-PA coupling, evaluated using the TAPSE/PASP ratio, in patients with MHD was found to be associated with increased CVD events and frequent IDH. Therefore, it is crucial for patients with MHD to undertake regular evaluations of RVD in routine practice.

Authors’ contributions

Chen Wang and Xu-Yang Cheng designed and planned the study. Chen wang handled the selection of suitable patients for the study, arranged the collection of clinical data, performed data analysis and drafted the manuscript. Li Meng participated the collection of clinical data. Yu-Qing Chen guided the study as a senior author. All the authors contributed to data interpretation and manuscript revision. The author(s) read and approved the final manuscript.

Ethics approval and consent to participate

This study protocol adhered to the Declaration of Helsinki and was approved by the Ethics Committee of Peking University First Hospital. Written patient-informed consent was obtained from all participants.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

Additional information

Funding

This study is supported by National Natural Science Fund [grant numbers 82270754, 81870477, 82090020], Peking University Clinical Scientist Training Program supported by “the Fundamental Research Funds for the Central Universities” (BMU2024PYJH009), Beijing Physician Scientist Training Project (BJPSTP-2024-15), Peking University Medicine Sailing Program for Young Scholars’ Scientific & Technological Innovation (BMU2023YFJHMX006).