Predictive value of indicators for left ventricular systolic dysfunction in HD patients

doi:10.21203/rs.3.rs-1982943/v1

Download PDF

Research Article

Predictive value of indicators for left ventricular systolic dysfunction in HD patients

https://doi.org/10.21203/rs.3.rs-1982943/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: To evaluate the role of dialysis duration, creatine, cardiac troponin T (cTnT) and N-terminal pro-brain natriuretic peptide (NT-pro-BNP) for predicting left ventricular systolic dysfunction (LVSD) in hemodialysis (HD) patients.

Methods: Participants with regular hemodialysis for more than three months between January 2017 and January 2022 were enrolled. Correlation analysis was used to assess the correlations among the clinical, serological and echocardiographic variables. The risk factors of LVSD were discussed by univariate and multivariate logistic regression analysis. Receiver-operating characteristic curve (ROC) was constructed to evaluate their predictive value.

Results: The 237 HD patients were divided into two groups, NT-pro-BNP > 35000 pg/ml vs. NT-pro-BNP ≤ 35000 pg/ml. In HD patients with NT-pro-BNP ≤ 35000 pg/ml group, cTnT (r = -0.222, P = 0.011) and NT-pro-BNP (r = -0.248, P = 0.002) were significantly correlated with EF. Factors associated with LVSD and corresponding odds ratios (ORs) with 95% CI by multivariate analysis were dialysis duration [0.158 (0.036, 0.689), P ＝ 0.010], creatine [0.255 (0.078, 0.833), P ＝ 0.020] and cTnT [10.858 (2.278, 51.758), P ＝ 0.000] in HD patients with NT-pro-BNP > 35000 pg/ml. In HD patients with NT-pro-BNP ≤ 35000 pg/ml, dialysis duration [0.266 (0.075, 0.938), P ＝ 0.034] and NT-pro-BNP [11.132 (2.904, 42.679), P ＝ 0.000] were both independent risk factors of LVSD in multivariate analysis. In HD patients with NT-pro-BNP > 35000 pg/ml group, the combination of dialysis duration and creatinine [(area under the curve (AUC): 0.664); P ＝ 0.002], dialysis duration and cTnT (AUC: 0.690 ; P ＝ 0.000), creatinine and cTnT (AUC: 0.702; P ＝ 0.000) and dialysis duration, creatinine and cTnT (AUC: 0.753; P ＝ 0.000) had the significantly statistic value to predict LVSD; In HD patients with NT-pro-BNP ≤ 35000 pg/ml group, the combination of dialysis duration and NT-pro-BNP had the best diagnostic value (AUC: 0.802; sensitivity: 89.55%; P < 0.000).

Conclusion: Our results implicated that the combination of dialysis duration, creatine cTnT and NT-pro-BNP might be used for early prediction and management of LVSD in HD patients.

hemodialysis

left ventricular systolic dysfunction

dialysis duration

creatine

cardiac troponin T

N-terminal pro-brain natriuretic peptide

It is well recognized that chronic kidney disease (CKD) is a worldwide public health problem. The cardiovascular mortality is 10 to 30 times higher in dialysis patients than that in the general population. More importantly, cardiovascular disease (CVD), instead of kidney failure, is the main cause of CKD (1). As an effective renal replacement therapy, hemodialysis (HD) can improve the prognosis of patients with end-stage renal disease (ESRD), but the incidence of cardiovascular adverse events is still high in dialysis patients (2). Systolic dysfunctions represent the main factors responsible for cardiovascular risk (3). ESRD patients with left ventricular systolic dysfunction (LVSD) presented higher cardiovascular mortality than those without LVSD (4). LVSD is the main cause of increased risk of cardiovascular death in ESRD patients, which is currently also considered the strongest predictors of all-cause mortality and cardiovascular in the dialysis population (4–7). At present, few studies have combined clinical and serological indicators to predict LVSD in HD patients. Therefore, finding simple, fast and effective indicators to predict LVSD in HD patients will be valuable for early intervention and reduction of cardiovascular mortality.

Cardiac troponin T (cTnT) is a strong predictor of adverse outcomes in myocardial infarction and reflects the severity of systolic dysfunction with heart diseases. Indeed, the correlation between cTnT and LVSD in ESRD patients had been proved (3, 8). In dialysis patients, studies had shown that cTnT was significantly higher in patients with systolic dysfunction than in those with normal left ventricular function and it had certain prediction value for LVSD (3). There is evidence that measurement of cardiac-derived natriuretic peptides and their biosynthetic precursors are useful to evaluate the patients with heart failure (9). The clinical benefits of screening cardiovascular risk with N-terminal pro-brain natriuretic peptide (NT-pro-BNP) have been well demonstrated in HD patients (7). Nitta et al reported that brain natriuretic peptide (BNP) may be a possible indicator of reduced ventricular function in HD patients. Mallamaci et al. showed that NT-pro-BNP had a high sensitivity in diagnosing LVSD in HD patients (6). Studies had examined the utility of natriuretic peptides for detecting systolic dysfunction in ESRD patients (6). Several studies indicated that cTnT and cardiac natriuretic hormones had good diagnostic potential for the exclusion of LVSD in ESRD patients (3, 6). Although an increase in concentrations of cTnT and NT-pro-BNP were both associated with CVD, guidelines didn’t provide specific diagnostic and prognostic cut-offs of these cardiac biomarkers in HD patients (10).

Previous studies focused on the predictive value of inflammatory markers for cardiac function and cardiovascular mortality in dialysis patients (11–15). It was doubted for the diagnosis of cardiac ischemia in ESRD patients due to the elevated serum cTnT levels in ESRD population (3). It had been reported that combined diagnosis value of NT-Pro-BNP, cTnT or cTnI in predicting cardiac function in coronary heart disease (3, 6). However, there is rare information about the value of combined indicators in predicting LVSD in HD patients (3). Our aim was to evaluate the individual or combined value of dialysis duration, creatinine, cTnT and NT-Pro-BNP in predicting LVSD for HD patients with different levels of NT-Pro-BNP.

Study population

We recruited 237 patients with ESRD(126 males and 111 females)who were hospitalized at the First Affiliated Hospital of University of South China (Hengyang, China) from January 2017 to January 2022 in this retrospective study. They had received routine dialysis treatment (RDT) for at least three months. All HD patients underwent hemodialysis according to the requirements of the treating nephrologist, 2–3 times a week, 4 hours each time. HD patients did not receive any other treatment during the same period. HD patients with moderate or severe valvular disease, recent acute coronary and cerebrovascular events, and active organ or hematologic malignancies were excluded. HD patients receiving immunosuppressive therapy, those who died within 6 months of starting hemodialysis and those who returned to hemodialysis after kidney transplant failure were also excluded during the study.

Data Collection

Demographic and clinical data were collected: age, gender, dialysis duration, and history of previous illness. Laboratory data were measured from blood samples, including white blood cell, neutrophil, lymphocyte, eosinophil, basophil, blood biochemistry, parathyroid (PTH), C-reactive protein (CRP), cTnT, NT-pro-BNP.

Values for left ventricular systolic function were expressed as ejection fraction (EF). EF was obtained using a modified biplane Simpson’s method from apical two- and four-chamber views (5). LVSD was defined as EF < 50%.

Blood routine, blood biochemistry, PTH and CRP were measured in routine clinical laboratory using standard methods. cTnT and NT-pro-BNP in plasma were measured by the electrochemiluminescence immunoassay (Elecsys 2010 analyser, Roche Diagnostic GmbH, Mannheim, Germany) (5). The detection limit of the cTnT assay was 14 pg/ml. The measuring range of the NT-pro-BNP assay was 5-35000 pg/ml (5). We used limits of 35000 pg/ml of NT-pro-BNP for categorizing into two groups, ≤ 35000 pg/ml vs. > 35000 pg/ml, respectively.

Statistical analysis

Continuous data were tested for normality by the Kolmogorov-Smirnov test. According to the data distribution, the data are expressed as mean ± standard deviation or median (interquartile range, IQR). Dichotomous data were presented as proportions. Differences between groups were tested by t test, Mann-Whitney test or by Chi-squared test, as appropriate. Correlations between variables were performed using the Pearson test. The associations of the indicators with LVSD were evaluated using univariate and multiple regression analysis, respectively. Odds ratios (ORs) and their 95% confidence intervals (CI) were calculated using the estimated regression coefficients and their standard errors in logistic regression analysis. The usefulness of indicators for the diagnosis of LVSD was tested by the analysis of receiver-operating characteristic curve (ROC). The best cut-off value was considered as the maximal difference of sensitivity and 1-specificity (Youden index). Data analysis were performed by using a standard statistical package (SPSS 25). All statistical tests were two-sided and a P-value < 0.05 was considered statistically significant. In addition, GraphPad Prism 8 was used to draw relevant graphics.

Patient Characteristics

A total of 237 participants with a median age of 60 years were enrolled in this study. Systolic dysfunction (as defined by EF < 50%) was presented in 53 patients (22.36%). The median duration of dialysis was 2.5 years in all HD patients. The participants were stratified into 2 groups according to the concentrations of serum NT-pro-BNP. The main demographic, clinical, laboratory and echocardiographic characteristics of the two groups were detailed in Table 1. There were no significant differences in age among the two groups, but HD patients with EF < 50% tended to be younger. Besides, there was an approximately equal distribution of sexes in the study population. The dialysis duration was statistically significant whether in the population of NT-pro-BNP > 35000 pg/ml or NT-pro-BNP ≤ 35000 pg/ml (P < 0.050). Albumin, calcium, phosphorus, PTH, creatinine, cTnT and other indicators had no statistical difference in HD patients.

Table 1

Baseline demographic, clinical, laboratory and echocardiographic characteristics with different levels of NT-pro-BNP in HD patients.
	NT-pro-BNP>35000		NT-pro-BNP ≤ 35000
	(N = 82)		(N = 149)
Variables	EF ≥ 50%	EF<50%	EF ≥ 50%	EF<50%
	N = 44	N = 38	N = 134	N = 15
Age, years	60.68 ± 12.23	59.79 ± 11.76	58.5 (52.0, 70.0)	54 (47, 64)
Gender, %	25 (56.82%)	21 (55.26%)	72 (53.73%)	8 (53.33%)
CAD, %	12 (27.27%)	12 (31.58%)	29 (21.64%)	4 (26.67%)
Hypertension, %	41 (29.55%)	33 (86.46%)	113 (84.33%)	11 (73.33%)
Diabetes, %	13 (29.55%)	15 (39.47%)	44 (32.84%)	4 (26.67%)
Clinical symptoms of HF, %	22 (50%)	19 (50%)	38 (28.36%)	6 (40%)
Dialysis duration, years	3.00 (1.00, 6.00)	1.17 (0.25, 3.60)^*	3.00 (1.00, 6.00)	1.17 (0.25, 3.63)^*
WBC, ×10⁹/L	6.17 ± 2.04	6.28 ± 1.82	5.90 (4.82, 7.72)	8.14 (4.86, 9.91)
Neutrophil, ×10⁹/L	3.89 (3.19, 5.48)	4.46 (3.30, 5.43)	4.49 (3.32, 5.64)	5.85 (3.57, 7.44)^*
Lymphocyte, ×10⁹/L	0.99 (0.67, 1.20)	0.97 (0.63, 1.25)	1.06 ± 0.48	1.08 ± 0.54
Eosinophil, ×10⁹/L	0.11 (0.48, 0.25)	0.11 (0.08, 0.25)	0.11 (0.05, 0.19)	0.16 (0.04, 0.24)
Hemoglobin, g/L	88.70 ± 23.19	95.00 ± 25.82	92.78 ± 24.11	89.27 ± 22.82
Albumin, g/L	40.61 ± 4.55	41.38 ± 4.93	41.60 (37.80, 44.40)	40.90 (36.43, 45.28)
Calcium, mmol/L	2.20 ± 0.31	2.09 ± 0.36	2.20 ± 0.31	2.22 ± 0.27
Phosphate, mmol/L	1.77 (1.45, 2.38)	1.71 (1.45, 2.16)	1.83 (1.39, 2.27)	1.87 (1.56, 1.99)
PTH, pg/ml	334.83 (164.35, 596.46)	260.05 (150.77,526.52)	382.56 (212.85, 793.93)	263.46 (146.96, 415.89)
Creatinine, µmol/L	842.91 ± 303.71	764.26 ± 342.66	851.10 ± 317.76	725.20 ± 225.42
UA, µmol/L	377.59 ± 123.72	358.24 ± 140.24	394.55 (300.88, 477.65)	402.80 (321.10, 521.00)
Urea, mg/dl	16.95 (13.63, 25.00)	17.55 (12.23, 23.05)	18.10 (12.75, 23.73)	15.30 (11.70, 20.25)
CRP, mg/L	7.75 (2.18, 17.50)	8.37 (2.33, 18.15)	4.20 (1.30, 16.30)	2.36 (0.41, 30.83)
CTnT, ng/L	92.60 (62.48, 146.80)	96.39 (65.68, 146.13)	57.90 (36.06, 88.18)	107.10 (35.62, 124.53)
NT-pro-BNP, pg/ml	>35000	>35000	6974.50 (2534.25, 15279.50)	17429.00 (11922.50, 35000.00)
EF, %	55 (42, 62)	40 (35.75, 43.25)	64 (61, 68)	43 (39, 47)
Data are expressed as mean ± SD or as median (interquartile range), or as percent frequency, as appropriate.
HD, hemodialysis; CAD, coronary artery disease; HF, heart failure; WBC, White blood cell; PTH, parathyroid; UA, Uric acid; CRP, C-reactive protein; CTnT, cardiac troponin T; NT-pro-BNP, N-terminal pro-brain natriuretic peptide; EF, ejection fraction.
* P<0.050

In HD patients with NT-pro-BNP > 35000 pg/ml, only the dialysis duration was different between the two groups with EF 50% as the boundary and the difference was statistically significant. Preexisting coronary artery disease (CAD), hypertension and diabetes were 24 (29.27%), 74 (90.24%) and 28 (34.15%), respectively. They had a higher prevalence of CAD, hypertensive and diabetes than that in HD patients with NT-pro-BNP ≤ 35000 pg/ml, and half of them had clinical symptoms of heart failure. The median of EF was 50.5% in this crowd. Among them, the population with LVSD accounted for 71.70%.

In HD patients with NT-pro-BNP ≤ 35000 pg/ml, preexisting CAD, hypertension and diabetes were 33 (22.15%), 124 (83.22%) and 48 (32.21%), respectively. There were statistically significant differences in NT-pro-BNP in HD patients with different LVEF levels (P = 0.000). The median of NT-pro-BNP was about 7785 pg/ml. The HD patients with NT-pro-BNP ≤ 35000 pg/ml had a significantly higher EF level compared with those HD patients with NT-pro-BNP > 35000 pg/ml (P < 0.01). In addition, we found that patients with poor cardiac function had higher neutrophil levels in HD patients with NT-pro-BNP ≤ 35000 pg/ml.

Correlation Study

There were highly significant linear associations between dialysis duration and EF whether in HD patients with NT-pro-BNP > 35000 pg/ml or NT pro BNP ≤ 35000 pg/ml. In HD patients with NT-pro-BNP ≤ 35000 pg/ml, cTnT was significantly inversely correlated with EF (r = -0.222, P= 0.011). The same relationship existed between NT-pro-BNP and EF (r= -0.248, P= 0.002). In addition, a significant positive correlation was found between cTnT and NT-pro-BNP (Table 2, Fig. 1).

Table 2

Correlation analysis between EF and clinical or laboratory parameters in HD patients.
Variables	NT-pro-BNP>35000	(n = 88)	NT-pro-BNP ≤ 35000	(n = 149)
Variables	r	P	r	P
Age, years	0.116	0.280	0.046	0.578
Gender, %	0.022	0.835	0.079	0.340
CAD, %	0.007	0.947	-0.169	0.041^*
Hypertension, %	0.048	0.654	0.149	0.069
Diabetes, %	-0.159	0.140	0.006	0.943
Clinical symptoms of HF, %	-0.079	0.464	-0.161	0.053
Dialysis duration, years	0.298	0.005^*	0.188	0.022^*
WBC, ×10⁹/L	0.047	0.662	-0.138	0.092
Neutrophil, ×10⁹/L	0.024	0.821	-0.095	0.248
Lymphocyte, ×10⁹/L	0.085	0.431	-0.136	0.098
Eosinophil, ×10⁹/L	-0.170	0.116	-0.128	0.120
Hemoglobin, g/L	-0.201	0.061	0.009	0.915
Albumin, g/L	-0.105	0.333	-0.008	0.925
Calcium, mmol/L	0.232	0.030^*	-0.011	0.892
Phosphate, mmol/L	-0.092	0.392	-0.060	0.475
PTH, pg/ml	0.009	0.936	0.183	0.033^*
Creatinine, µmol/L	0.076	0.480	0.080	0.331
UA, µmol/L	0.060	0.579	-0.121	0.141
Urea, mg/dl	-0.139	0.198	0.044	0.593
CRP, mg/L	0.008	0.944	-0.113	0.135
CTnT, ng/L	-0.110	0.337	-0.222	0.011^*
NT-pro-BNP, pg/ml			-0.248	0.002^*
Data are expressed as mean ± SD or as median (interquartile range), or as percent frequency, as appropriate.
HD, hemodialysis; EF, ejection function; HD, hemodialysis; CAD, coronary artery disease; HF, heart failure; WBC, White blood cell; PTH, parathyroid; UA, Uric acid; CRP, C-reactive protein; CTnT, cardiac troponin T; NT-pro-BNP, N-terminal pro-brain natriuretic peptide; EF, ejection fraction.
* P<0.050

The relationship between LVSD and related indicators were determined by univariate and multivariate logistic regression analysis. In the group of NT-pro-BNP > 35000 pg/ml, dialysis duration, creatinine and cTnT were found to be associated with LVSD (Table 3, all P < 0.050). After adjusting for confounding factors in the multivariate analysis, dialysis duration [OR = 0.158 (95% CI, 0.036 to 0.689; P = 0.010)], creatinine [OR = 0.255 (95% CI, 0.078 to 0.833; P = 0.020)] and cTnT [OR = 10.858 (95% CI, 2.278 to 51.758; P = 0.000)] were ranked as strong risk factors of LVSD. In this fully adjusted multiple logistic regression model, cTnT had an independent correlation with LVSD and cTnT concentrations above the cut-off indicated that the risk of LVSD increased by more than 4.286 times (Table 3).

Table 3

Factors associated with EF < 50%: Univariate and Multiple analysis among HD patients.
Factors	NT-pro-BNP>35000				NT-pro-BNP ≤ 35000
Factors	Univariate analysis	P	Multiple analysis	P	Univariate analysis	P	Multiple analysis	P
Age, years	1.006 (0.971, 1.043)	0.730			1.029 (0.983, 1.077)	0.220
Gender, %	1.118 (0.477, 2.618)	0.800			1.016 (0.349, 2.962)	1.000
CAD, %	0.981 (0.396, 2.427)	1.000			3.103 (0.968, 9.953)	0.060
Hypertension, %	0.733 (0.196, 2.741)	0.650			0.279 (0.090, 0.866)	0.030
Diabetes, %	1.856 (0.749, 4.597)	0.180			0.744 (0.224, 2.469)	0.630
Clinical symptoms of HF, %	1.000 (0.420, 2.382)	1.000			2.199 (0.744, 6.499)	0.150
Dialysis duration, years	0.241 (0.076, 0.760)	0.020	0.158 (0.036, 0.689)	0.010	0.295 (0.089, 0.973)	0.050	0.266 (0.075, 0.938)	0.034
WBC, ×10⁹/L	1.702 (0.704, 4.117)	0.240			8.144 (2.541, 26.098)	0.000
Neutrophil, ×10⁹/L	1.857 (0.785, 4.396)	0.160			0.203 (0.061, 0.672)	0.010
Lymphocyte, ×10⁹/L	0.652 (0.277, 1.533)	0.330			2.690 (0.835, 8.670)	0.100
Hemoglobin, g/L	3.306 (0.990, 11.041)	0.050			0.229 (0.062, 0.847)	0.030
Albumin, g/L	2.242 (0.897, 5.630)	0.080			2.242 (0.897, 5.603)	0.080
Calcium, mmol/L	0.389 (0.160, 0.943)	0.040			0.322 (0.070, 1.491)	0.150
Phosphate, mmol/L	0.520 (0.218, 1.241)	0.140			0.177 (0.022, 1.399)	0.100
PTH, pg/ml	0.458 (0.189, 1.109)	0.080			0.510 (0.169, 1.538)	0.230
Creatinine, µmol/L	0.348 (0.138, 0.879)	0.030	0.255 (0.078, 0.833)	0.020	0.245 (0.074, 0.811)	0.020
UA, µmol/L	0.500 (0.207, 1.206)	0.120			2.131 (0.723, 6.281)	0.170
Urea, mg/dl	6.783 (1.960, 23.477)	0.000			5.429 (1.096, 26.886)	0.040
CRP, mg/L	1.800 (0.708, 4.579)	0.220			0.343 (0.092, 1.283)	0.110
CTnT, ng/L	4.286 (1.273, 14.429)	0.020	10.858 (2.278, 51.758)	0.000	6.849 (1.668, 28.128)	0.010
NT-pro-BNP, pg/ml					10.486 (2.800, 39.280)	0.000	11.132 (2.904, 42.679)	0.000
Data are expressed as mean ± SD or as median (interquartile range), or as percent frequency, as appropriate.
HD, hemodialysis; CAD, coronary artery disease; HF, heart failure; WBC, White blood cell; PTH, parathyroid; UA, Uric acid; CRP, C-reactive protein; CTnT, cardiac troponin T; NT-pro-BNP, N-terminal pro-brain natriuretic peptide; EF, ejection fraction.

LVSD in patients with NT-pro-BNP ≤ 35000 pg/ml was associated with prior hypertension, dialysis duration, white blood cell, neutrophil, creatinine, cTnT and NT-pro-BNP (Table 3, all P < 0.050). No significant relationship was found between LVSD and CRP or albumin. However, multivariate analysis showed that only the dialysis duration [OR = 0.266 (95% CI, 0.075 to 0.938; P = 0.034)] and NT-pro-BNP [OR = 11.132 (95% CI, 2.904 to 42.679; P = 0.000)] were significantly associated with LVSD. NT-pro-BNP was a significant indicator which was associated with EF < 50% and the OR was as high as 11.132. In addition, in the multivariable model with NT-pro-BNP ≤ 35000 pg/ml, creatinine and cTnT showed insignificant association with LVSD (Table 3).

Roc Analysis And Combined Diagnosis

Table 4 and Table 5 showed the diagnostic value of dialysis duration, creatinine, cTnT and NT-pro-BNP for LVSD and the combined diagnostic value of these indicators in the two groups of HD patients.

Table 4

Diagnostic value of biomarkers for LVSD in HD patients with NT-pro-BNP > 35000 pg/ml.
Variable	AUC	95% CI	Sensitivity(%)	Specificity(%)	Youden index J	P
Dialysis duration	0.638	0.521 to 0.754	90.00	31.58	0.22	0.020
Creatinine	0.602	0.479 to 0.725	78.00	44.74	0.23	0.103
CTnT	0.579	0.452 to 0.706	34.88	88.89	0.24	0.223
Dialysis duration + Creatinine	0.664	0.559 to 0.769	70.00	57.89	0.28	0.002
Dialysis duration + cTnT	0.690	0.594 to 0.787	27.91	97.22	0.25	0.000
Creatinine + CTnT	0.702	0.600 to 0.804	90.70	36.11	0.27	0.000
Dialysis duration + Creatinine + CTnT	0.753	0.656 to 0.850	83.72	52.78	0.37	0.000
Data are expressed as mean ± SD or as median (interquartile range), or as percent frequency, as appropriate.
HD, hemodialysis; CTnT, cardiac troponin T; LVSD, left ventricular systolic dysfunction.

Table 5

Diagnostic value of biomarkers for LVSD in HD patients with NT-pro-BNP ≤ 35000 pg/ml.
Variable	AUC	95% CI	Sensitivity(%)	Specificity(%)	Youden index J	P
Dialysis duration	0.643	0.519 to 0.766	55.22	73.33	0.29	0.023
Creatinine	0.624	0.494 to 0.755	59.70	73.33	0.33	0.062
CTnT	0.657	0.438 to 0.876	74.59	70.00	0.45	0.159
NT-pro-BNP	0.785	0.665 to 0.905	72.39	80.00	0.52	<0.000
Dialysis duration + NT-pro-BNP	0.802	0.662 to 0.942	89.55	66.67	0.56	<0.000
Data are expressed as mean ± SD or as median (interquartile range), or as percent frequency, as appropriate.
HD, hemodalysis; CTnT, cardiac troponin T; NT-pro-BNP, N-terminal pro-brain natriuretic peptide; LVSD, left ventricular systolic dysfunction.

In HD patients with NT-pro-BNP > 35000 pg/ml, we found that only dialysis duration had certain diagnostic value for LVSD when evaluating the diagnostic value of individual indicator and the area under the corresponding ROC curves (AUC) were 0.638 (95% CI, 0.521 to 0.754; P = 0.020). However, the combined detection of dialysis duration, creatinine and cTnT had good diagnostic value and were significantly higher than that of a single indicator. The AUCs of combination of dialysis duration and creatinine, of dialysis duration and cTnT, of creatinine and cTnT in predicting LVSD were 0.664 (95% CI, 0.559 to 0.769; P= 0.002), 0.690 (95% CI, 0.594 to 0.787; P= 0.000) and 0.702 (95% CI, 0.600 to 0.804; P = 0.000), respectively. In this group, the combination of dialysis duration, creatinine and cTnT had the most significant diagnostic potential for LVSD as shown by the significantly greater AUC [0.753 (95% CI, 0.656 to 0.850; P = 0.000)] (Table 4).

In HD patients with NT-pro-BNP ≤ 35000 pg/ml, dialysis duration and NT-pro-BNP respectively had significant diagnostic potential for LVSD and their AUC were as follows [dialysis duration: 0.643 (95% CI, 0.519 to 0.766; P = 0.023); NT-pro-BNP: 0.785 (95% CI, 0.665 to 0.905; P < 0.000)]. NT-pro-BNP provided 72.39% sensitivity and 82.00% specificity with a threshold value of 13837 pg/mL for predicting LVSD. The AUC of creatinine and cTnT for LVSD were 0.624 (95% CI, 0.494 to 0.755; P = 0.062) and 0.657 (95% CI, 0.438 to 0.876; P = 0.159), respectively. They were statistically insignificant. In ROC analysis, for dialysis time, creatinine, cTnT and NT-pro-BNP, the combination of any two indicators were statistically significant. Besides, in our study, the combination of dialysis duration and NT-pro-BNP had the best diagnostic value [0.802 (95% CI, 0.662 to 0.942); P < 0.000]. The sensitivity of their combined diagnosis was 89.55% (Table 5).

CVD is highly prevalent in ESRD patients, accounting for more than half of the deaths in this population. Cardiovascular mortality is the main cause of death in HD patients. During dialysis treatment, the incidence of adverse cardiovascular events such as heart failure and myocardial infarction is also very high (16). Among them, left ventricular dysfunction is common in HD patients. This may be characterized by left ventricular dilatation or systolic insufficiency, all of which occurred frequently in HD patients. These manifestations had adverse clinical outcomes, with the worst prognosis in those with LVSD (16). Therefore, we need to pay close attention to HD patients, at the same time, through some indicators to predict their LVSD to detect “high-risk” patients early.

A series of risk factors had been found to be associated with major adverse cardiovascular events in HD patients, such as age, anemia, hypoalbuminemia, hypertension, diabetes, vascular calcification, malnutrition and inflammation (16–19). LVSD was considered to be the strongest predictors of cardiovascular mortality in HD patients (16). It was suggested that some simple and reliable methods were needed to predict the LVSD in HD patients to make early intervention or prevention. Echocardiography is undoubtedly valuable in identifying LVSD, but it has a low sensitivity for the screening of LVSD (3). However, when LVSD in HD patients could be detected in echocardiography, these patients often already had clinical symptoms and the efficacy of clinical treatment may be poor. In our study, this phenomenon was more obvious in patients with NT-pro-BNP > 35000 pg/ml, 50% of whom had clinical symptoms of heart failure before LVSD occurred.

The present study demonstrated that there was a positive association between dialysis duration and LVSD in HD patients and it had certain diagnostic value. The longer the dialysis duration, the better the cardiac function. This may be the reason that dialysis could improve the cardiac function and delay the progress of heart failure within a certain period of time. Previous studies had shown that a greater dialysis duration was associated with the improvement of EF and more intensive dialysis schedules may be beneficial for patients with persistent volume overload (20). We also found that the higher the neutrophil level, the worse the cardiac function. Studies had proved that the increase of neutrophil count reflected the inflammatory state (21). Micro-inflammation was an important factor of CVD in HD patients, which may further lead to the deterioration of cardiac function (22).

Furthermore, by using multivariate logistic regression analysis, we found that creatinine was a risk factor for LVSD in HD patients. This result suggested that increasing the dialysis frequency may help HD patients to add dialysis adequacy and improve LVSD. Of note, it was proved that cTnT was a powerful predictor of cardiovascular mortality in HD patients and this phenomenon was independent of previous cardiovascular complications (23). CTnT were often elevated in patients with renal failure, even in the absence of clinically suspected ischemia(24). Although it was suggested that an elevated cTnT level might be caused by retention of proteolytic degradation products of cTnT, an increasing number of articles suggested that the increase of cTnT had a certain relationship with decreased cardiac function in HD patients (25). It was documented that the level of cTnT was higher than that of cTnI in HD patients, reflecting that cTnT might be a stronger predictor of chronic myocardial damage in HD patients (8). Our findings were consistent with previous studies that cTnT was negatively correlated with left ventricular systolic function, which can reflect LVSD. Although creatinine couldn’t be used as an independent predictor of LVSD in HD patients, it had good diagnostic value when combined with other indicators such as dialysis duration, creatinine and NT-pro-BNP.

The research showed that circulating cardiac markers (cTnT and NT-pro-BNP) could provide valuable information on HD patients to evaluate the risk of cardiac disease and LVSD (26, 27). Besides, the ability of elevated NT-pro-BNP to predict all-cause and cardiovascular mortality had been established in HD patients (26–28). NT-proBNP has been consistently shown to correlate with poor left ventricular systolic function (27). It had been proved that NT-pro-BNP had a good diagnostic performance for exclusion of LVSD and all-cause or cardiovascular mortality in HD patients (6, 10, 26). Besides, it had been proved that left ventricular systolic function had a high correlation with the NT-pro-BNP concentration (29). Our study suggested that the measurement of serum NT-pro-BNP concentration may have similar correlation and predictive value for LVSD in HD patients. Our results also showed that both NT-pro-BNP and cTnT were negatively correlated with EF in HD patients. With further elevation of serum NT-pro-BNP and cTnT levels, left ventricular systolic function may be worse in HD patients. In addition, there was a significant positive correlation between NT-pro-BNP and cTnT, which indirectly indicated that NT-pro-BNP and cTnT played the similar role in predicting LVSD in HD patients. The optimal NT-pro-BNP cut-off value to distinguish LVSD among HD patients was 13837 pg/ml. In HD patients with NT-pro-BNP ≤ 35000 pg/ml, NT-pro-BNP had a high diagnostic sensitivity for the identification of LVSD (72.39%) coupled to a high specificity (80.00%). In other words, NT-pro-BNP appeared to be useful in identifying LVSD. The measuring range of the NT-pro-BNP was 5-35000 pg/ ml analyzed by Roche Diagnostics assay (5). Due to the limitation of the detected range, we couldn’t measure the occurrence when NT-pro-BNP exceeds 35000 pg/ml. Since our study had found the value of NT-pro-BNP in predicting LVSD in HD patients with NT-pro-BNP ≤ 35000 pg/ml, it is not difficult to speculate that NT-pro-BNP might have the similar predictive value for LVSD when its concentration is greater than 35000 pg/ ml.

Previous studies had confirmed that a combination of several cardiac biomarkers may improve cardiovascular risk stratification in HD patients (30). To our knowledge, the combined power of serological biomarkers (such as creatine, cTnT and NT-pro-BNP) and clinical characteristic (such as dialysis duration) to predict LVSD has not been effectively evaluated in HD patients. Therefore, our study may contribute to stratification of cardiovascular mortality risk in regular HD patients. At the same time, we also found that the combined diagnostic value was superior to that of a single indicator. Through univariate and multivariate analysis, we found that the risk factors of the two groups of patients were different. Therefore, we conducted a ROC analysis of all risk factors to further evaluate the diagnostic value of the relevant indicators. The results showed that the combined diagnostic value of those risk factors was better. Among the HD patients with NT-pro-BNP > 35000 pg/ml, the combination of dialysis duration, creatinine and cTnT had a significant diagnostic value for LVSD [0.753 (95% CI, 0.656 to 0.850); P = 0.000], while creatinine (P = 0.103) or cTnT (P = 0.223) along had no statistically diagnostic value. Among the HD patients with NT-pro-BNP ≤ 35000 pg/ml, the combination of dialysis duration and NT-pro-BNP, which were both independent risk factors, had the highest diagnostic value [0.802 (95% CI, 0.662 to 0.942); P < 0.000]and sensitivity (83.72%). To sum up, a combination of two indicators in clinical practice might be valuable. By combining two or more indicators to predict LVSD in HD patients, better prediction effect may be achieved.

This study acknowledges some limitations. First, it is a cross-sectional study which could not establish the causal relationship of LVSD among those variable factors. Secondly, this study was done based on a single center and limited to HD patients. Further longitudinal studies including more participants will require in multi-centers. Thirdly, our study only studied the predictive indicators and their predictive value for LVSD, but the effect of related indicators on the all-cause or cardiovascular mortality of those HD patients still needs further follow-up and survival analysis.

Taken together, our analysis indicated that some indicators such as dialysis duration, creatine, cTnT and NT-pro-BNP, which are easily obtained from medical history collection and routine laboratory testing, may be useful for predicting and modifying the risk of LVSD in HD patients. Early detection of relevant indicators may provide useful information for preventing the progress of LVSD in HD patients.

We demonstrated that the combination of dialysis duration, creatine and cTnT had a significant diagnostic value for LVSD in HD patients with NT-pro-BNP > 35000 pg/ml. The combination of dialysis duration and NT-pro-BNP had the best diagnostic value and the highest sensitivity in HD patients with NT-pro-BNP ≤ 35000 pg/ml. By the combination of these indicators, dialysis duration, creatine, cTnT and NT-pro-BNP all have good predictive value for LVSD in HD patients. The combination of these indicators might be used for early prediction and management of LVSD in HD patients in clinical practice.

AUC

Area under the curve

BNP

Brain natriuretic peptide

CAD

Coronary artery disease

CTnT

Cardiac troponin T

CKD

Chronic kidney disease

CVD

Cardiovascular disease

CRP

C-reactive protein

Confidence intervals

ESRD

End-stage renal disease

Ejection fraction

Hemodialysis

LVSD

Left ventricular systolic dysfunction

NT-pro-BNP

N-terminal pro-brain natriuretic peptide

ORs

Odds ratios

PTH

Parathyroid

ROC

Receiver-operating characteristic curve

RDT

Routine dialysis treatment

Acknowledgements

Not applicable.

Author contributions

Liu Ying, He Zhangxiu, Huang Hong and Tang Huifang designed the research. Liu Ying, Xiu Hao and Ding Kai collected and compiled data. Liu Ying, Xiu Hao, Ding Kai, Cao Jingsong, Hu Hengjing, Huang Hong, He Zhangxiu and Tang Huifang assisted in writing and formatting the text. All authors assisted with the data analysis and read and approved the final manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (No. 81900678), the Key Guiding Project of Hunan Provincial Health Commission (No. 20201920), Key Project of Hunan Provincial Science and Technology Innovation (No. 2020SK1013-2), Special Funds for the Construction of Innovative Provinces in Hunan (No. 2020SK4008), the Natural Science Foundation of Hunan Province (No. 2021JJ40490, 2021JJ70113).

Availability of data and materials

All data used or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All methods performed in this study were in accordance with the Declaration of Helsinki. The research protocol for this study was reviewed and approved by the Medical Ethics Committee of the First Affiliated Hospital of University of South China. Informed consent was approved to be waived for its retrospective nature by the Medical Ethics Committee of the First Affiliated Hospital of University of South China.

Consent for publication

Not applicable.

Competing interests

All authors declare no competing interests.

Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, et al. Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Hypertension. 2003;42(5):1050–65.
Wang S, Chen F, Yang S, Shi J. Interleukin-18. Int Heart J. 2018;59(4):786–90.
Mallamaci F, Zoccali C, Parlongo S, Tripepi G, Benedetto FA, Cutrupi S, et al. Diagnostic value of troponin T for alterations in left ventricular mass and function in dialysis patients. Kidney Int. 2002;62(5):1884–90.
Angkananard T, Janma J, Wannasiri T, Sangthong P, Changsirikulchai S. Associations of left ventricular systolic dysfunction with the factors among Thai patients on peritoneal dialysis: a cross-sectional study. BMC Nephrol. 2019;20(1):257.
Wang AY, Lam CW, Wang M, Chan IH, Lui SF, Zhang Y, et al. Diagnostic potential of serum biomarkers for left ventricular abnormalities in chronic peritoneal dialysis patients. Nephrol Dial Transplant. 2009;24(6):1962–9.
Mallamaci F, Zoccali C, Tripepi G, Benedetto FA, Parlongo S, Cataliotti A, et al. Diagnostic potential of cardiac natriuretic peptides in dialysis patients. Kidney Int. 2001;59(4):1559–66.
Helal I, Belhadj R, Mohseni A, Bazdeh L, Drissa H, Elyounsi F, et al. Clinical significance of N-terminal Pro-B-type natriuretic peptide (NT-proBNP) in hemodialysis patients. Saudi J Kidney Dis Transpl. 2010;21(2):262–8.
Artunc F, Mueller C, Breidthardt T, Twerenbold R, Peter A, Thamer C, et al. Sensitive troponins–which suits better for hemodialysis patients? Associated factors and prediction of mortality. PLoS One. 2012;7(10):e47610.
Alehagen U, Dahlstrom U, Rehfeld JF, Goetze JP. Prognostic assessment of elderly patients with symptoms of heart failure by combining high-sensitivity troponin T and N-terminal pro-B-type natriuretic peptide measurements. Clin Chem. 2010;56(11):1718–24.
Eriguchi M, Tsuruya K, Lopes M, Bieber B, McCullough K, Pecoits-Filho R, et al. Routinely measured cardiac troponin I and N-terminal pro-B-type natriuretic peptide as predictors of mortality in haemodialysis patients. ESC Heart Fail. 2022;9(2):1138–51.
Bazeley J, Bieber B, Li Y, Morgenstern H, de Sequera P, Combe C, et al. C-reactive protein and prediction of 1-year mortality in prevalent hemodialysis patients. Clin J Am Soc Nephrol. 2011;6(10):2452–61.
Li H, Lu X, Xiong R, Wang S. High Neutrophil-to-Lymphocyte Ratio Predicts Cardiovascular Mortality in Chronic Hemodialysis Patients. Mediators Inflamm. 2017;2017:9327136.
Xiang F, Chen R, Cao X, Shen B, Liu Z, Tan X, et al. Monocyte/lymphocyte ratio as a better predictor of cardiovascular and all-cause mortality in hemodialysis patients: A prospective cohort study. Hemodial Int. 2018;22(1):82–92.
Kaysen GA, Dubin JA, Muller HG, Mitch WE, Rosales LM, Levin NW. Relationships among inflammation nutrition and physiologic mechanisms establishing albumin levels in hemodialysis patients. Kidney Int. 2002;61(6):2240–9.
Yaprak M, Turan MN, Dayanan R, Akin S, Degirmen E, Yildirim M, et al. Platelet-to-lymphocyte ratio predicts mortality better than neutrophil-to-lymphocyte ratio in hemodialysis patients. Int Urol Nephrol. 2016;48(8):1343–8.
Yamada S, Ishii H, Takahashi H, Aoyama T, Morita Y, Kasuga H, et al. Prognostic value of reduced left ventricular ejection fraction at start of hemodialysis therapy on cardiovascular and all-cause mortality in end-stage renal disease patients. Clin J Am Soc Nephrol. 2010;5(10):1793–8.
Anavekar NS, Pfeffer MA. Cardiovascular risk in chronic kidney disease. Kidney Int Suppl. 2004(92):S11-5.
Levin A. Clinical epidemiology of cardiovascular disease in chronic kidney disease prior to dialysis. Semin Dial. 2003;16(2):101–5.
Neuen BL, Leather N, Greenwood AM, Gunnarsson R, Cho Y, Mantha ML. Neutrophil-lymphocyte ratio predicts cardiovascular and all-cause mortality in hemodialysis patients. Ren Fail. 2016;38(1):70–6.
Smyth B, Chan CT, Grieve SM, Puranik R, Zuo L, Hong D, et al. Predictors of Change in Left-Ventricular Structure and Function in a Trial of Extended Hours Hemodialysis. J Card Fail. 2020;26(6):482–91.
Arruda-Olson AM, Reeder GS, Bell MR, Weston SA, Roger VL. Neutrophilia predicts death and heart failure after myocardial infarction: a community-based study. Circ Cardiovasc Qual Outcomes. 2009;2(6):656–62.
Neirynck N, Glorieux G, Schepers E, Dhondt A, Verbeke F, Vanholder R. Pro-inflammatory cytokines and leukocyte oxidative burst in chronic kidney disease: culprits or innocent bystanders? Nephrol Dial Transplant. 2015;30(6):943–51.
Deegan PB, Lafferty ME, Blumsohn A, Henderson IS, McGregor E. Prognostic value of troponin T in hemodialysis patients is independent of comorbidity. Kidney Int. 2001;60(6):2399–405.
Freda BJ, Tang WH, Van Lente F, Peacock WF, Francis GS. Cardiac troponins in renal insufficiency: review and clinical implications. J Am Coll Cardiol. 2002;40(12):2065–71.
Havekes B, van Manen JG, Krediet RT, Boeschoten EW, Vandenbroucke JP, Dekker FW, et al. Serum troponin T concentration as a predictor of mortality in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis. 2006;47(5):823–9.
Hickman PE, McGill DA, Talaulikar G, Hiremagalur B, Bromley J, Rahman A, et al. Prognostic efficacy of cardiac biomarkers for mortality in dialysis patients. Intern Med J. 2009;39(12):812–8.
Satyan S, Light RP, Agarwal R. Relationships of N-terminal pro-B-natriuretic peptide and cardiac troponin T to left ventricular mass and function and mortality in asymptomatic hemodialysis patients. Am J Kidney Dis. 2007;50(6):1009–19.
Sommerer C, Beimler J, Schwenger V, Heckele N, Katus HA, Giannitsis E, et al. Cardiac biomarkers and survival in haemodialysis patients. Eur J Clin Invest. 2007;37(5):350–6.
Artunc F, Mueller C, Breidthardt T, Twerenbold R, Rettig I, Usta E, et al. Comparison of the diagnostic performance of three natriuretic peptides in hemodialysis patients: which is the appropriate biomarker? Kidney Blood Press Res. 2012;36(1):172–81.
Apple FS, Murakami MM, Pearce LA, Herzog CA. Multi-biomarker risk stratification of N-terminal pro-B-type natriuretic peptide, high-sensitivity C-reactive protein, and cardiac troponin T and I in end-stage renal disease for all-cause death. Clin Chem. 2004;50(12):2279–85.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Predictive value of indicators for left ventricular systolic dysfunction in HD patients

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study population

Data Collection

Statistical analysis

Results

Patient Characteristics

Correlation Study

Roc Analysis And Combined Diagnosis

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1