- Email: [email protected]
Assessment of right ventricular dysfunction in end-stage renal disease patients on maintenance haemodialysis by cardiac magnetic resonance imaging
European Journal of Radiology 102 (2018) 89–94
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Research article
Assessment of right ventricular dysfunction in end-stage renal disease patients on maintenance haemodialysis by cardiac magnetic resonance imaging
T
⁎
Wanlin Penga, Zhenlin Lia, ,1, Huayan Xua, Chunchao Xiaa, Yingkun Guob, Jinge Zhanga, Keling Liua, Yuming Lia, Jin Pua, Huapeng Zhangc, Tianlei Cuid,1 a
Department of Radiology, West China Hospital of Sichuan University, Chengdu 610041, China Department of Radiology, Key Laboratory of Obstetric & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second University Hospital, China c Siemens Healthcare Itd. Chengdu Branch, Chendu 610041, China d Department of Nephrology, West China Hospital of Sichuan University, Chengdu 610041, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: End-stage renal disease Cardiac magnetic resonance Right ventricular function RVEF
Purpose: To assess right ventricular (RV) dysfunction in end-stage renal disease (ESRD) patients on maintenance haemodialysis (HD) by cardiac magnetic resonance (CMR) imaging and determined the risk factors associated with RV dysfunction. Materials and methods: Fifty ESRD patients on maintenance HD and 16 age- and gender-matched healthy individuals were prospectively enrolled and underwent CMR imaging. Left ventricular (LV) and RV function parameters, including end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV) and ejection fraction (EF), were measured and compared. Independent sample t-test and Mann–Whitney U-test were used to compare the differences between healthy individuals and ESRD patients. Pearson correlation and multiple linear regression analyses were used to assess risk factors associated with RV dysfunction. Results: Significantly lower RVEF and LVEF were observed in ESRD patients than in the control group (all p < 0.001). RVEDV, RVESV and RVSV in ESRD patients were also lower than those in the control group (all p < 0.05). Meanwhile, higher LVESV, LV mass and interventricular septum thickness were found in ESRD patients than in the control group (all p < 0.05). RVEF was positively correlated with LVEF (r = 0.37, p = 0.008) and negatively correlated with the duration of renal insufficiency (r = −0.53, p < 0.001) and dialysis (r = −0.63, p < 0.001). Moreover, multiple linear regression analyses revealed that the duration of dialysis and LVEF were independently associated with decreased RVEF (adjusted R2 = 0.53, p < 0.001). Conclusions: In ESRD patients on maintenance HD, RV function was impaired and associated with the deterioration of LV function. More importantly, the duration of dialysis was considered as a risk factor independently associated with RV dysfunction.
1. Introduction End-stage renal disease (ESRD) is defined as the terminal stage of various chronic kidney diseases (CKDs). The prevalence and incidence of ESRD continue to increase in recent years [1]. Clinically, cardiovascular diseases (CVDs) are the most common complications of ESRDs and are believed to be the leading cause of death in ESRD patients [2]. Left ventricular (LV) impairment has been extensively studied as a prominent manifestation associated with adverse clinical outcome [3]. However, several studies have recently reported a high prevalence of ⁎
1
right ventricular (RV) dysfunction in ESRD patients [4,5]. It is well known that deteriorated RV function may lead to more severely compromised LV function. Furthermore, the presence of RV abnormality may contribute to increased risk of death, shock and arrhythmias [6]. Accordingly, RV dysfunction has been proven to be independently associated with increased morbidity and mortality caused by CVDs in ESRD patients [7]. Thus, it is extremely important for ESRD patients on maintenance haemodialysis (HD) to undergo early assessment of RV function in order to prevent the worsening of heart failure (HF). At present, transthoracic echocardiography is the most common
Corresponding author at: Department of Radiology, West China Hospital of Sichuan University, 37# Guoxue Alley, Wuhou District, Chengdu, Sichuan 610041, China. E-mail address: [email protected] (Z. Li). These authors contributed equally to this work and should be considered co-corresponding authors.
https://doi.org/10.1016/j.ejrad.2018.02.036 Received 13 June 2017; Received in revised form 23 February 2018; Accepted 28 February 2018 0720-048X/ © 2018 Published by Elsevier B.V.
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
Fig. 1. Cardiac short-axis images show an example of left and right ventricular function data obtained from the patient at end-diastolic phase (A) and end-systolic phase (B) by MRI.
2.2. CMR imaging protocol
noninvasive method used to assess cardiac function. However, it is still challenging to assess RV function accurately using conventional imaging methods because of the complex RV chamber geometry, particularly in diseases without RV wall thickening. Cardiac magnetic resonance (CMR) imaging has emerged as the reference standard for the noninvasive assessment of RV size and function because of its high temporal and spatial resolution [8]. Although attention is being increasingly paid to RV impairment in ESRD patients, there is lack of data regarding the assessment of RV function by CMR imaging. Therefore, this prospective study was designed to investigate RV functions in ESRD patients on long-term maintenance HD by CMR imaging. We also explored the risk factors associated with RV dysfunction in those patients.
All CMR imaging examinations were performed on the same day of HD, and the patients were admitted for HD immediately after the examination. All the participants were examined in the supine position using a 3.0 T MR scanner (Skyra; Siemens Medical Solutions, Erlangen, Germany) with a 32-channel phased array abdominal coil to assess cardiac function. The transverse, coronal and sagittal plane localisation images were obtained using a bSSFP sequence (repetition time [TR]/ echo time [TE]: 224.16/1.23 ms, flip angle: 60°, slice thickness: 6 mm, spacing between slices: 1.5 mm, voxel size: 1.3 × 1.3 × 6.0 mm, field of view: 340 × 255 mm, matrix size: 256 × 169 mm). Then, long-axis two- and four-chamber images obtained through the apex of the left ventricle and centre of the mitral valve were acquired with a segmented electrocardiogram (ECG)-gated bSSFP 2D cine sequence (TR/TE: 39.12/1.43 ms, flip angle: 45°, slice thickness: 8 mm, voxel size: 1.6 × 1.6 × 8.0 mm, field of view: 340 × 284 mm, matrix size: 208 × 166 mm) to determine the standard short-axis images. A bSSFP sequence (TR/TE: 39.34/1.22 ms, flip angle: 40°, acquisition matrix: 208 × 166 mm, field of view: 340 × 284 mm, slice thickness: 8 mm, spacing between slices: 0 mm) was used to acquire 8-mm-thick cine short-axis images based on aforementioned long-axis images. These sequential short-axis images were perpendicular to the interventricular septum (IVS) and parallel to the mitral valve. The range of short-axis images covered the entire length of the two ventricles. All images were acquired with breath holding in end-expiration combined with electrocardiographic or peripheral sphygmous gating.
2. Materials and methods 2.1. Study population We prospectively recruited ESRD patients on maintenance HD from the nephrology department of our hospital between October 2015 and December 2017. The inclusion criteria included patients with stage V CKD (estimated glomerular filtration rates [eGFR] < 15 ml/min/ 1.73 m2) on maintenance HD. The exclusion criteria included: (1) unsuitability for magnetic resonance imaging (MRI) examination due to severe heart failure (New York Heart Association [NYHA] class IV) or echocardiographic LV ejection fraction (LVEF) < 35% [9] (n = 3); (2) cardiovascular damage caused by uncontrolled hypertension and diabetes (n = 16); (3) contraindication for MRI due to the presence of cardiac pacemaker or defibrillators and other unsafe ferromagnetic objects (n = 1); (4) claustrophobia (n = 1); (5) arrhythmia (n = 2); (6) inability to cooperate with the examination (n = 2); (7) insufficient clinical data (n = 2) and (8) noninterpretable MR image quality (n = 3). After exclusion of 30 patients, 50 ESRD patients (31 women, 19 men; mean age 53.92 ± 14.77 years) remained. Sixteen gender-matched asymptomatic normal participants with similar age range (10 women, 6 men; mean age 45.75 ± 12.69 years) were enrolled into our study as the control group. The exclusion criteria of controls were family history of CVDs and other systemic diseases associated with cardiovascular damage. This study was approved by the institutional review board (IRB), and written informed consent was obtained from all included participants.
2.3. Image processing and analysis Cine MRI data were transmitted to the commercially available postprocessing software (cvi42; Circle Cardiovascular Imaging Inc., Calgary, Alberta, Canada) and analysed offline by an experienced observer who was blinded to patient information. In each short-axis cine image, the endocardial and epicardial borders for the ventricle were manually delineated at the end-diastolic and end-systolic phases to obtain LV and RV functions (end-diastolic volume [EDV], end-systolic volume [ESV], stroke volume [SV] and EF) and mass parameters. In accordance with previous research [10], papillary muscles and moderator bands were carefully assigned to the ventricular cavity (Fig. 1). The most basal section surrounded by at least 50% myocardium in all heart phases was identified as the base of LV [11]. The RV border at the basal slice was carefully modified to exclude the right atrium. IVS 90
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
thickness was measured with callipers on a cine four-chamber view.
Table 1 Demographic characteristics of ESRD patients and healthy individuals.
2.4. Clinical and laboratory data collection Demographic information of the participants (including name, age, height and weight) was recorded at the time of CMR imaging. Clinical data, including the duration of dialysis treatment, duration of chronic renal insufficiency, systolic blood pressure (SBP) and diastolic blood pressure (DBP) during HD treatment, serum phosphorus, high-density lipoprotein cholesterol (HDL-C) level, plasma creatinine level (which can be used to calculate eGFR) and various complications (including secondary hyperparathyroidism, mitral regurgitation, aortic regurgitation, tricuspid regurgitation, pulmonary arterial hypertension [PAH], central venous thrombosis, renal anemia and renal hypertension) were retrospectively retrieved and recorded from the electronic medical record (EMR) systems.
Age (year) Males gender (%) Height (cm) Weight (kg) BMI (kg/m2) BSA (m2) SBP (mmHg) DBP (mmHg) Serum Phosphorus (mg/dl) HDL-C (mg/dl) Creatinine (mg/dl) eGFR (ml/min/1.732 m2) Duration of dialysis (year) Duration of renal insufficiency (year) Secondary hyperparathyroidism Mitral regurgitation Aortic regurgitation Tricuspid regurgitation PAH Central venous thrombosis Renal anemia RH
2.5. Statistical analysis Statistical analyses were performed using SPSS version, 23.0.0 (SPSS, Inc., Chicago, IL, USA). Graphs were drawn using GraphPad Prism 6.00 (GraphPad Software, San Diego, CA). Continuous variables were presented as the mean ± standard deviation (SD). All function parameters were standardised with the body surface area (BSA), except for EF. Kolmogorov–Smirnov test was performed to test for normal distribution of the data. All data were performed power of test, which showed power of test of all the RV data were enough (all power of test > 0.8). Based on the normal distribution test of data, the differences in function parameters between the patient and control groups were analysed using independent sample t-test or Mann–Whitney Utest. Pearson correlation was used to assess the association between RVEF and other factors when appropriate. Correlation between RV and LV function parameters was evaluated using Pearson correlation if data were normally distributed. Variables with significant correlation were entered into a backward stepwise multiple linear regression model to identify independent predictors of RVEF. A two-tailed p-value < 0.05 was considered to be statistically significant.
ESRD patients (n = 50)
Healthy individuals (n = 16)
p-value
53.92 ± 14.77 19/50 159.52 ± 6.86 57.32 ± 10.67 22.47 ± 3.75 1.67 ± 0.16 138.12 ± 20.92 87.04 ± 15. 39 1.82 ± 0. 63 1.14 ± 0. 36 9.26 ± 3.14 7.62 ± 3.44 4.13 ± 3.57 6.23 ± 4.53
45.75 ± 12.69 6/16 159.75 ± 4.92 60.47 ± 7.74 23.67 ± 2.66 1.71 ± 0.11 – – – – – – – –
0.051 0.97 0.902 0.225 0.114 0.344 N/A N/A N/A N/A N/A N/A N/A N/A
4 (8.0%)
–
N/A
9 (18.0%) 5 (10.0%) 9 (18.0%) 5 (10.0%) 21 (42.0%) 49 (98.0%) 28 (56.0%)
– – – – – – –
N/A N/A N/A N/A N/A N/A N/A
Note: Data presented as mean ± SD. BMI, body mass index; BSA, body surface area; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; HD, haemodialysis; PAH, pulmonary arterial hypertension; RH, renal hypertension; N/A, not applicable. Table 2 RV function and LV function parameters of ESRD patients and healthy individuals.
3. Results 3.1. Patient characteristics After exclusion, 50 ESRD patients (31 women, 19 men; mean age 53.92 ± 14.77 years) and 16 healthy individuals (10 women, 6 men; mean age 45.75 ± 12.69 years) were considered to be eligible for our study (Table 1). There were no significant differences in age, gender, weight, height, body mass index (BMI) and BSA between the two groups (all p > 0.05). For the ESRD patients, the main cause of ESRD included chronic nephritis (n = 34), polycystic kidney (n = 6), diabetes mellitus (n = 4), lupus nephritis (n = 1) and hypertension (n = 5). Furthermore, the duration of renal insufficiency was 6.23 ± 4.53 years, and patients underwent dialysis for an average of 4.13 ± 3.57 years; their SBP and DBP during examination were 138.12 ± 20.92 mmHg and 87.04 ± 15.39 mmHg, respectively. Secondary hyperparathyroidism was found in 4 (8.0%) patients; renal anemia was found in 49 (98.0%) patients; and renal hypertension (RH) was detected in 28 (56.0%) patients; Using two dimensional echocardiography, 9 (18.0%) patients were found to have mitral regurgitation; 5 (10.0%) patients were found to have aortic regurgitation; 9 (18.0%) patients were found to have tricuspid regurgitation; and 5 (10.0%) patients were found to have PAH. The digital subtraction angiography (DSA) revealed 21 (42.0%) patients had central venous thrombosis.
Variable
ESRD patients (n = 50)
Healthy individuals (n = 16)
p-value
RVEDV (ml/ m2) RVESV (ml/ m2) RVSV (ml/m2) RVEF (%) IVS (mm) LVEDV (ml/ m2) LVESV (ml/m2) LVSV (ml/m2) LVEF (%) LVM (g/m2)
67.70 ± 16.18
78.46 ± 12.58
0.008
36.14 ± 11.87
45.66 ± 7.77
0.004
31.57 46.62 17.96 75.06
± ± ± ±
10.28 10.93 3.66 22.84
46.33 59.34 14.18 70.41
± ± ± ±
7.32 5.18 2.36 10.30
0.000 0.000 0.000 0.265
35.50 39.33 54.13 52.51
± ± ± ±
21.09 11.05 10.77 17.90
27.01 45.62 64.83 27.59
± ± ± ±
8.96 7.02 4.67 6.34
0.009 0.036 0.000 0.000
Note: Data presented as mean ± SD. RVEDV, right ventricular end-diastolic volume; RVESV, right ventricular end-systolic volume; RVSV, right ventricular stroke volume; RVEF, right ventricular ejection fraction; IVS, interventricular septum; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; LVSV, left ventricular stroke volume; LVEF, left ventricular ejection fraction.
3.2. RV and LV function assessment When considering RVEF < 0.40 as a threshold to define RV dysfunction, 26% (13/50) of the ESRD patients were identified as having RV dysfunction. As shown in Table 2, the patient group showed a significant decrease in RVEDV, RVESV, RVSV and RVEF (all p < 0.05) compared with the control group. Moreover, a significantly lower LVEF and LVSV were observed in the patient group than in the control group, while a significantly higher LVESV was observed in the patient group than in the control group (all p < 0.05). Furthermore, the IVS thickness in the patient group was significantly greater than that in the 91
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
Fig. 2. Scatter plot demonstrating correlation between (A) RVEDV and LVEDV, (B) RVESV and LVESV, (C) RVSV and LVSV and (D) RVEF and LVEF.
(Table 4).
control group (p < 0.001). However, there were no significant differences in LVEDV (p = 0.265). In ESRD patients, RVEDV closely correlated with LVEDV (r = 0.62, 95% confidence interval [CI]: 0.35–0.81; p < 0.001); RVESV also correlated well with LVESV (r = 0.52, 95% CI: 0.25–0.74; p < 0.001). Accordingly, there was a positive correlation between RVSV and LVSV (r = 0.58, 95% CI: 0.40–0.74; p < 0.001) and between RVEF and LVEF (r = 0.37, 95% CI: 0.06–0.63; p < 0.001) (Fig. 2).
4. Discussion Similar to previous studies using echocardiography [5,12,13], our study demonstrated a high prevalence (26%, 13/50) of RV dysfunction in ESRD patients undergoing HD. RV function plays a vital role in determining clinical outcomes of patients with multifarious cardiovascular and pulmonary diseases [14–17], deterioration of RV function may increase venous congestion, reduce renal blood flow, limit LV preload and increase the substrate for malignant ventricular arrhythmias, thereby leading to a poor prognosis [14,15,18]. Correspondingly, survival analysis including echocardiographic variables demonstrated that RV dysfunction is significantly associated with impaired survival in ESRD patients [7]. Therefore, it is important to pay more attention to RV function during the diagnosis, therapy and followup of ESRD patients. At present, CMR imaging has a high accuracy and reproducibility for RV size and function measurements [19] and has been able to ascertain the effects of frequent HD on ventricular volume, function and LV remodelling in ESRD patients [20]. Therefore, shifts in RV function in ESRD patients on maintenance HD were further
3.3. Determinants associated with RVEF Within the patient group, there was a positive correlation between decreased RVEF and LVEF. In addition, significant correlations were noticed between RVEF and duration of dialysis (r = −0.63, 95% CI: −0.77 to −0.43; p < 0.001), renal insufficiency (r = −0.53, 95% CI: −0.73 to −0.27; p < 0.001) (Fig. 3) or gender (r = −0.31, 95% CI: −0.57 to −0.03; p = 0.032). Factors that significantly correlated with RVEF (LVEF, duration of dialysis, renal insufficiency and gender) were entered into a multivariate linear regression model to establish the optimal regression equation (Table 3). Finally, LVEF and duration of dialysis were noted to be significantly associated with RV dysfunction
Fig. 3. Scattergrams showing results of linear regression analysis between RVEF and (A) LVEF, (B) duration of renal insufficiency and (C) duration of dialysis. The regression equation and correlation coefficient (r) are provided in each plot.
92
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
derived from intermittent and chronic volume overload. Prolonged high-volume changes may impair RV systolic function, although the right ventricle could adapt well to volume overload within limits. Further, the removal of large amounts of fluid during HD may aggravate RV dysfunction due to RV myocardial ischaemia resulting from severe hypovolemia [24,25]. Moreover, HD-induced myocardial stunning may facilitate the development of myocardial dysfunction [26]. Based on univariate correlation analysis, we also identified a negative correlation between decreased RVEF and duration of renal insufficiency, which supports the idea that RV function is impaired in ESRD patients without HD [4]. However, dialysis instead of duration of renal insufficiency was found to be the main cause of RV dysfunction in ESRD patients on maintenance HD. In addition, we found that there were no significant correlation between the common complications (e.g., mitral regurgitation) and decreased RVEF. This may be because of appropriate interventions were timely given to control these complications. Contrary to our findings, other researchers deemed that ventricular functions would not deteriorate, while BP levels, weight, hypervolemia, anemia and other metabolic disorders during HD treatment were under control [27]. BP and HDL-C levels have been associated with preclinical shifts in RV function in a Multi-Ethnic Study of Atherosclerosis (MESA) [28]. Accordingly, BP and HDL-C were included in univariate correlation analysis of our study. Contrary to the results noted by the researchers, BP and HDL-C levels did not correlate with RVEF in our study. This contradictory result may be mainly due to the clinical and laboratory indexes in ESRD patients on HD were usually kept at a certain level, except different in disease. In line with previous studies [12], our study revealed that LV function is significantly impaired in HD patients. In Chen’s three-dimensional speckle-tracking echocardiographic study on LV myocardial function, the researchers found that LV regional long-axis, short-axis and global functions are compromised in uraemia patients with normal LVEF, regardless of whether they receive HD [29]. In addition, a correlation study showed that LV function deteriorates parallel to impairments of RV function in dialysis patients [12]. It was in accordance with our findings that indices of LV function correlated well with those observed for the RV function. This finding may be explained by ventricular interdependence, which means that the size, function and load conditions of one ventricle will affect contralateral ventricular filling and contraction through direct mechanical interactions [23,30]. Paneni et al. thought that chronic pressure/volume overload caused by AVF precipitates RV dysfunction, subsequently triggering LV dysfunction via right-to-left interdependence [12]. Therefore, we considered that compromised RV function and incrassated IVS in HD patients may aggravate the deterioration of LV dysfunction, except for the aetiology of ESRD. There are several potential limitations in our study. First, our results are limited by the relatively small number of patients. This small sample size is mainly because patients were recruited with strict inclusion and exclusion criteria to minimise the effect of confounding factors on RV function. However, it was difficult to assess RV function by classified analysis on account of our small sample size. Moreover, it would be desirable to further confirm the correlation between RVEF and risk factors in a study with a larger sample size. Secondly, because of a limited amount of healthy volunteers, the age between two groups were just similar instead of exactly matching. Finally, data related to the impact of changes in RV function on cardiovascular events and long-term survival are still lacking and a substantial follow-up period would allow us to further validate the predictive value of RV function in ESRD patients.
Table 3 Correlation between RVEF and clinical and part LV parameters. Parameter
r
p-value
Age (year) Gender Weight Height BMI(kg/m2) BSA(m2) SBP (mmHg) DBP (mmHg) Serum Phosphorus (mg/dl) HDL-C (mg/dl) Creatinine (mg/dl) eGFR (ml/min/1.732 m2) LVEF (%) Duration of dialysis (year) Duration of renal insufficiency (year) Secondary hyperparathyroidism Mitral regurgitation Aortic regurgitation Tricuspid regurgitation PAH Central venous thrombosis Renal anemia RH
0.024 −0.308 0.072 −0.113 0.070 −0.032 0.086 0.093 0.029 −0.078 0.029 0.025 0.37 −0. 630 −0.526 0.010 −0.018 −0.164 0.049 −0.043 −0.035 0.153 0.046
0.871 0.032* 0.622 0.436 0.630 0.824 0.553 0.519 0.841 0.592 0.841 0.862 0.008* 0.000* 0.000* 0.944 0.901 0.255 0.737 0.77 0.811 0.287 0.751
Note: all abbreviations are as those in Tables 1 and 2. Table 4 Multivariate linear regression entering only significant correlates into the (backward stepwise) model (adjusted R2 = 0.53). Model 1 β
B Constant Duration of dialysis LVEF Duration of renal insufficiency Gender
Model 3
#
27.60 (14.02–41.19) −1.10* (−2.00 to −0.21) 0.50# (0.28–0.71) −0.24 @ (−0.96 to 0.48) −3.37@ (−7.91 to 1.16)
–
β
B #
–
0.49 −0.10
22.66 (10.75–34.57) −1.35# (−1.97 to −0.74) 0.55# (0. 35–0.76) –
0.54 –
−0.15
–
–
−0.36
−0.44
Note: Data presented as mean (95% confidence interval). LVEF, left ventricular ejection fraction; B, unstandardised regression coefficient; β, standardised coefficient. Units of measurement as before. * p < 0.05; #p < 0.001; @p > 0.05, which means that the variable was eliminated from the final linear regression model.
determined by CMR imaging in our study. At present, there have been several studies exploring the risk factors associated with LV dysfunction by echocardiography [21] or CMR imaging [22] in ESRD patients on maintenance HD. However, data involving the determinants of RV impairments in these given patients have been limited. For these patients, several pathophysiological mechanisms may precipitate the deterioration of RV function by increasing the RV preload or afterload: metabolic and hormonal derangements, sympathetic activation, renal anemia, secondary hyperparathyroidism, derangement of calcium–phosphorus homeostasis, inflammation and left-to-right shunt caused by arteriovenous fistula (AVF) [13,16]. We analysed the probable risk factors associated with RV impairment in HD patients. Consistent with previous reports [23], we identified the duration of dialysis as the strongest parameter associated with predicting RVEF and noted LVEF to be also associated with RV dysfunction. This finding further verified that HD treatment would compromise RV function. Unlike continuous ambulatory peritoneal dialysis, intermittent HD would result in the periodic accumulation of fluids, toxics and electrolytes between two consecutive HD treatment sessions. In these patients, the mechanisms underlying RV dysfunction are mainly
5. Conclusion In conclusion, our study demonstrated that both LV and RV function assessed by MRI are significantly compromised in ESRD patients on maintenance HD and RV dysfunction may contribute to the 93
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
deterioration of LV function in dialysis patients. Moreover, the duration of dialysis could be considered as a significantly independent predictor of RV dysfunction. Accordingly, CMR is a potentially prominent tool for the early detection of subclinical RV dysfunction in ESRD patients on maintenance HD.
(2005) 112–117. [12] F. Paneni, M. Gregori, G.M. Ciavarella, et al., Relation between right and left ventricular function in patients undergoing chronic dialysis, J. Cardiovasc. Med. 14 (4) (2013) 289–295. [13] S. Yilmaz, M. Yetim, B.K. Yilmaz, T. Dogan, E. Aksoy, N. Yuksel, I. Dogan, High hemodialysis vascular access flow and impaired right ventricular function in chronic hemodialysis patients, Indian J. Nephrol. 26 (5) (2016) 352–356. [14] F.L. Dini, R.T. Demmer, A. Simioniuc, et al., Right ventricular dysfunction is associated with chronic kidney disease and predicts survival in patients with chronic systolic heart failure, Eur. J. Heart Fail. 14 (3) (2012) 287–294. [15] D. Murninkas, A.C. Alba, D. Delgado, M. McDonald, F. Billia, W.S. Chan, H.J. Ross, Right ventricular function and prognosis in stable heart failure patients, J. Card. Fail. 20 (5) (2014) 343–349. [16] K. Cenkerova, J. Dubrava, V. Pokorna, J. Kaluzay, O. Jurkovicova, Right ventricular systolic dysfunction and its prognostic value in heart failure with preserved ejection fraction, Acta Cardiol. 70 (4) (2015) 387–393. [17] L.J. Hickson, S.M. Negrotto, M. Onuigbo, et al., Echocardiography criteria for structural heart disease in patients with end-stage renal disease initiating hemodialysis, J. Am. Coll. Cardiol. 67 (10) (2016) 1173–1182. [18] R.A. Bleasdale, M.P. Frenneaux, Prognostic importance of right ventricular dysfunction, Heart 88 (4) (2002) 323–324. [19] C.F. Mooij, C.J. de Wit, D.A. Graham, A.J. Powell, T. Geva, Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles, J. Magn. Reson. Imaging 28 (1) (2008) 67–73. [20] C.T. Chan, T. Greene, G.M. Chertow, et al., Effects of frequent hemodialysis on ventricular volumes and left ventricular remodeling, Clin. J. Am. Soc. Nephrol. 8 (12) (2013) 2106–2116. [21] K. Negishi, M. Kobayashi, I. Ochiai, et al., Association between fibroblast growth factor 23 and left ventricular hypertrophy in maintenance hemodialysis patients. Comparison with B-type natriuretic peptide and cardiac troponin T, Circ. J. 74 (12) (2010) 2734–2740. [22] R. Wald, M.B. Goldstein, R.M. Wald, et al., Correlates of left ventricular mass in chronic hemodialysis recipients, Int. J. Cardiovasc. Imaging 30 (2) (2014) 349–356. [23] H. Arinc, H. Gunduz, A. Tamer, et al., Use of tissue doppler to assess right ventricle function in hemodialysis patients, Am. J. Nephrol. 25 (3) (2005) 256–261. [24] F. Haddad, R. Doyle, D.J. Murphy, S.A. Hunt, Right ventricular function in cardiovascular disease, part II: pathophysiology, clinical importance, and management of right ventricular failure, Circulation 117 (13) (2008) 1717–1731. [25] J.P. Kinet, D. Soyeur, N. Balland, et al., Hemodynamic study of hypotension during hemodialysis, Kidney Int. 21 (6) (1982) 868–876. [26] J.O. Burton, H.J. Jefferies, N.M. Selby, C.W. McIntyre, Hemodialysis-induced cardiac injury: determinants and associated outcomes, Clin. J. Am. Soc. Nephrol. 4 (5) (2009) 914–920. [27] M. Duran, A. Unal, M.T. Inanc, et al., Hemodialysis does not impair ventricular functions over 2 years, Hemodial. Int. 15 (3) (2011) 334–340. [28] H. Chahal, C. Johnson, H. Tandri, et al., Relation of cardiovascular risk factors to right ventricular structure and function as determined by magnetic resonance imaging (results from the multi-ethnic study of atherosclerosis), Am. J. Cardiol. 106 (1) (2010) 110–116. [29] R. Chen, X. Wu, L.J. Shen, B. Wang, M.M. Ma, Y. Yang, B.W. Zhao, Left ventricular myocardial function in hemodialysis and nondialysis uremia patients: a three-dimensional speckle-tracking echocardiography study, PLoS One 9 (6) (2014) e100265. [30] S. Magder, The left heart can only be as good as the right heart: determinants of function and dysfunction of the right ventricle, Crit. Care Resusc. 9 (4) (2007) 344–351.
Conflict of interest We declare that we have no conflicts of interest to this work. Funding This work was supported by the National Natural Science Foundation of China (No: 81641169, 81471722) and Program for Young Scholars And Innovative Research Team in Sichuan Province of China (No: 2017TD0005). References [1] M. Gonzalezbedat, G. Rosadiez, R. Pecoitsfilho, et al., Burden of disease: prevalence and incidence of ESRD in Latin America, Clin. Nephrol. 83 (1) (2015) S3–S6. [2] K. Kundhal, C.E. Lok, Clinical epidemiology of cardiovascular disease in chronic renal disease, Nephron Clin. Pract. 101 (2005) c47–c52. [3] A. Merchant, R. Wald, M.B. Goldstein, et al., Relationship between different blood pressure measurements and left ventricular mass by cardiac magnetic resonance imaging in end-stage renal disease, J. Am. Soc. Hypertens. 9 (4) (2015) 275–284. [4] K. Said, M. Hassan, E. Baligh, B. Zayed, K. Sorour, Ventricular function in patients with end-stage renal disease starting dialysis therapy: a tissue doppler imaging study, Echocardiography 29 (9) (2012) 1054–1059. [5] L.J. Zhao, S.M. Huang, T. Liang, H. Tang, Pulmonary hypertension and right ventricular dysfunction in hemodialysis patients, Eur. Rev. Med. Pharmacol. Sci. 18 (21) (2014) 3267–3273. [6] S.R. Mehta, J.W. Eikelboom, M.K. Natarajan, R. Diaz, C. Yi, R.J. Gibbons, S. Yusuf, Impact of right ventricular involvement on mortality and morbidity in patients with inferior myocardial infarction, J. Am. Coll. Cardiol. 37 (1) (2001) 37–43. [7] E.J. Stallworthy, H.L. Pilmore, M.W. Webster, K.K. Sidhu, E.M. Curry, P. Brown, A. Scaria, Do echocardiographic parameters predict mortality in patients with endstage renal disease? Transplantation 95 (10) (2013) 1225–1232. [8] Y.K. Guo, Z.G. Yang, H. Shao, W. Deng, G. Ning, Z.H. Dong, Right ventricular dysfunction and dilatation in patients with mitral regurgitation: analysis using ECGgated multidetector row computed tomography, Int. J. Cardiol. 167 (4) (2013) 1585–1590. [9] J.J.V. McMurray, S. Adamopoulos, S.D. Anker, et al., ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012, Eur. Heart J. 33 (2012) 1787–1847. [10] S. Zhang, Z.G. Yang, J.Y. Sun, L.Y. Wen, H.Y. Xu, G. Zhang, Y.K. Guo, Assessing right ventricular function in patients with hypertrophic cardiomyopathy with cardiac MRI: correlation with the new York heart function assessment (NYHA) classification, PLoS One 9 (9) (2014) e104312. [11] A.H. Mahnken, M. Katoh, P. Bruners, E. Spuentrup, J.E. Wildberger, R.W. Günther, A. Buecker, Acute myocardial infarction: assessment of left ventricular function with 16-detector row spiral CT versus MR imaging–study in pigs, Radiology 236 (1)
94
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Research article
Assessment of right ventricular dysfunction in end-stage renal disease patients on maintenance haemodialysis by cardiac magnetic resonance imaging
T
⁎
Wanlin Penga, Zhenlin Lia, ,1, Huayan Xua, Chunchao Xiaa, Yingkun Guob, Jinge Zhanga, Keling Liua, Yuming Lia, Jin Pua, Huapeng Zhangc, Tianlei Cuid,1 a
Department of Radiology, West China Hospital of Sichuan University, Chengdu 610041, China Department of Radiology, Key Laboratory of Obstetric & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second University Hospital, China c Siemens Healthcare Itd. Chengdu Branch, Chendu 610041, China d Department of Nephrology, West China Hospital of Sichuan University, Chengdu 610041, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: End-stage renal disease Cardiac magnetic resonance Right ventricular function RVEF
Purpose: To assess right ventricular (RV) dysfunction in end-stage renal disease (ESRD) patients on maintenance haemodialysis (HD) by cardiac magnetic resonance (CMR) imaging and determined the risk factors associated with RV dysfunction. Materials and methods: Fifty ESRD patients on maintenance HD and 16 age- and gender-matched healthy individuals were prospectively enrolled and underwent CMR imaging. Left ventricular (LV) and RV function parameters, including end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV) and ejection fraction (EF), were measured and compared. Independent sample t-test and Mann–Whitney U-test were used to compare the differences between healthy individuals and ESRD patients. Pearson correlation and multiple linear regression analyses were used to assess risk factors associated with RV dysfunction. Results: Significantly lower RVEF and LVEF were observed in ESRD patients than in the control group (all p < 0.001). RVEDV, RVESV and RVSV in ESRD patients were also lower than those in the control group (all p < 0.05). Meanwhile, higher LVESV, LV mass and interventricular septum thickness were found in ESRD patients than in the control group (all p < 0.05). RVEF was positively correlated with LVEF (r = 0.37, p = 0.008) and negatively correlated with the duration of renal insufficiency (r = −0.53, p < 0.001) and dialysis (r = −0.63, p < 0.001). Moreover, multiple linear regression analyses revealed that the duration of dialysis and LVEF were independently associated with decreased RVEF (adjusted R2 = 0.53, p < 0.001). Conclusions: In ESRD patients on maintenance HD, RV function was impaired and associated with the deterioration of LV function. More importantly, the duration of dialysis was considered as a risk factor independently associated with RV dysfunction.
1. Introduction End-stage renal disease (ESRD) is defined as the terminal stage of various chronic kidney diseases (CKDs). The prevalence and incidence of ESRD continue to increase in recent years [1]. Clinically, cardiovascular diseases (CVDs) are the most common complications of ESRDs and are believed to be the leading cause of death in ESRD patients [2]. Left ventricular (LV) impairment has been extensively studied as a prominent manifestation associated with adverse clinical outcome [3]. However, several studies have recently reported a high prevalence of ⁎
1
right ventricular (RV) dysfunction in ESRD patients [4,5]. It is well known that deteriorated RV function may lead to more severely compromised LV function. Furthermore, the presence of RV abnormality may contribute to increased risk of death, shock and arrhythmias [6]. Accordingly, RV dysfunction has been proven to be independently associated with increased morbidity and mortality caused by CVDs in ESRD patients [7]. Thus, it is extremely important for ESRD patients on maintenance haemodialysis (HD) to undergo early assessment of RV function in order to prevent the worsening of heart failure (HF). At present, transthoracic echocardiography is the most common
Corresponding author at: Department of Radiology, West China Hospital of Sichuan University, 37# Guoxue Alley, Wuhou District, Chengdu, Sichuan 610041, China. E-mail address: [email protected] (Z. Li). These authors contributed equally to this work and should be considered co-corresponding authors.
https://doi.org/10.1016/j.ejrad.2018.02.036 Received 13 June 2017; Received in revised form 23 February 2018; Accepted 28 February 2018 0720-048X/ © 2018 Published by Elsevier B.V.
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
Fig. 1. Cardiac short-axis images show an example of left and right ventricular function data obtained from the patient at end-diastolic phase (A) and end-systolic phase (B) by MRI.
2.2. CMR imaging protocol
noninvasive method used to assess cardiac function. However, it is still challenging to assess RV function accurately using conventional imaging methods because of the complex RV chamber geometry, particularly in diseases without RV wall thickening. Cardiac magnetic resonance (CMR) imaging has emerged as the reference standard for the noninvasive assessment of RV size and function because of its high temporal and spatial resolution [8]. Although attention is being increasingly paid to RV impairment in ESRD patients, there is lack of data regarding the assessment of RV function by CMR imaging. Therefore, this prospective study was designed to investigate RV functions in ESRD patients on long-term maintenance HD by CMR imaging. We also explored the risk factors associated with RV dysfunction in those patients.
All CMR imaging examinations were performed on the same day of HD, and the patients were admitted for HD immediately after the examination. All the participants were examined in the supine position using a 3.0 T MR scanner (Skyra; Siemens Medical Solutions, Erlangen, Germany) with a 32-channel phased array abdominal coil to assess cardiac function. The transverse, coronal and sagittal plane localisation images were obtained using a bSSFP sequence (repetition time [TR]/ echo time [TE]: 224.16/1.23 ms, flip angle: 60°, slice thickness: 6 mm, spacing between slices: 1.5 mm, voxel size: 1.3 × 1.3 × 6.0 mm, field of view: 340 × 255 mm, matrix size: 256 × 169 mm). Then, long-axis two- and four-chamber images obtained through the apex of the left ventricle and centre of the mitral valve were acquired with a segmented electrocardiogram (ECG)-gated bSSFP 2D cine sequence (TR/TE: 39.12/1.43 ms, flip angle: 45°, slice thickness: 8 mm, voxel size: 1.6 × 1.6 × 8.0 mm, field of view: 340 × 284 mm, matrix size: 208 × 166 mm) to determine the standard short-axis images. A bSSFP sequence (TR/TE: 39.34/1.22 ms, flip angle: 40°, acquisition matrix: 208 × 166 mm, field of view: 340 × 284 mm, slice thickness: 8 mm, spacing between slices: 0 mm) was used to acquire 8-mm-thick cine short-axis images based on aforementioned long-axis images. These sequential short-axis images were perpendicular to the interventricular septum (IVS) and parallel to the mitral valve. The range of short-axis images covered the entire length of the two ventricles. All images were acquired with breath holding in end-expiration combined with electrocardiographic or peripheral sphygmous gating.
2. Materials and methods 2.1. Study population We prospectively recruited ESRD patients on maintenance HD from the nephrology department of our hospital between October 2015 and December 2017. The inclusion criteria included patients with stage V CKD (estimated glomerular filtration rates [eGFR] < 15 ml/min/ 1.73 m2) on maintenance HD. The exclusion criteria included: (1) unsuitability for magnetic resonance imaging (MRI) examination due to severe heart failure (New York Heart Association [NYHA] class IV) or echocardiographic LV ejection fraction (LVEF) < 35% [9] (n = 3); (2) cardiovascular damage caused by uncontrolled hypertension and diabetes (n = 16); (3) contraindication for MRI due to the presence of cardiac pacemaker or defibrillators and other unsafe ferromagnetic objects (n = 1); (4) claustrophobia (n = 1); (5) arrhythmia (n = 2); (6) inability to cooperate with the examination (n = 2); (7) insufficient clinical data (n = 2) and (8) noninterpretable MR image quality (n = 3). After exclusion of 30 patients, 50 ESRD patients (31 women, 19 men; mean age 53.92 ± 14.77 years) remained. Sixteen gender-matched asymptomatic normal participants with similar age range (10 women, 6 men; mean age 45.75 ± 12.69 years) were enrolled into our study as the control group. The exclusion criteria of controls were family history of CVDs and other systemic diseases associated with cardiovascular damage. This study was approved by the institutional review board (IRB), and written informed consent was obtained from all included participants.
2.3. Image processing and analysis Cine MRI data were transmitted to the commercially available postprocessing software (cvi42; Circle Cardiovascular Imaging Inc., Calgary, Alberta, Canada) and analysed offline by an experienced observer who was blinded to patient information. In each short-axis cine image, the endocardial and epicardial borders for the ventricle were manually delineated at the end-diastolic and end-systolic phases to obtain LV and RV functions (end-diastolic volume [EDV], end-systolic volume [ESV], stroke volume [SV] and EF) and mass parameters. In accordance with previous research [10], papillary muscles and moderator bands were carefully assigned to the ventricular cavity (Fig. 1). The most basal section surrounded by at least 50% myocardium in all heart phases was identified as the base of LV [11]. The RV border at the basal slice was carefully modified to exclude the right atrium. IVS 90
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
thickness was measured with callipers on a cine four-chamber view.
Table 1 Demographic characteristics of ESRD patients and healthy individuals.
2.4. Clinical and laboratory data collection Demographic information of the participants (including name, age, height and weight) was recorded at the time of CMR imaging. Clinical data, including the duration of dialysis treatment, duration of chronic renal insufficiency, systolic blood pressure (SBP) and diastolic blood pressure (DBP) during HD treatment, serum phosphorus, high-density lipoprotein cholesterol (HDL-C) level, plasma creatinine level (which can be used to calculate eGFR) and various complications (including secondary hyperparathyroidism, mitral regurgitation, aortic regurgitation, tricuspid regurgitation, pulmonary arterial hypertension [PAH], central venous thrombosis, renal anemia and renal hypertension) were retrospectively retrieved and recorded from the electronic medical record (EMR) systems.
Age (year) Males gender (%) Height (cm) Weight (kg) BMI (kg/m2) BSA (m2) SBP (mmHg) DBP (mmHg) Serum Phosphorus (mg/dl) HDL-C (mg/dl) Creatinine (mg/dl) eGFR (ml/min/1.732 m2) Duration of dialysis (year) Duration of renal insufficiency (year) Secondary hyperparathyroidism Mitral regurgitation Aortic regurgitation Tricuspid regurgitation PAH Central venous thrombosis Renal anemia RH
2.5. Statistical analysis Statistical analyses were performed using SPSS version, 23.0.0 (SPSS, Inc., Chicago, IL, USA). Graphs were drawn using GraphPad Prism 6.00 (GraphPad Software, San Diego, CA). Continuous variables were presented as the mean ± standard deviation (SD). All function parameters were standardised with the body surface area (BSA), except for EF. Kolmogorov–Smirnov test was performed to test for normal distribution of the data. All data were performed power of test, which showed power of test of all the RV data were enough (all power of test > 0.8). Based on the normal distribution test of data, the differences in function parameters between the patient and control groups were analysed using independent sample t-test or Mann–Whitney Utest. Pearson correlation was used to assess the association between RVEF and other factors when appropriate. Correlation between RV and LV function parameters was evaluated using Pearson correlation if data were normally distributed. Variables with significant correlation were entered into a backward stepwise multiple linear regression model to identify independent predictors of RVEF. A two-tailed p-value < 0.05 was considered to be statistically significant.
ESRD patients (n = 50)
Healthy individuals (n = 16)
p-value
53.92 ± 14.77 19/50 159.52 ± 6.86 57.32 ± 10.67 22.47 ± 3.75 1.67 ± 0.16 138.12 ± 20.92 87.04 ± 15. 39 1.82 ± 0. 63 1.14 ± 0. 36 9.26 ± 3.14 7.62 ± 3.44 4.13 ± 3.57 6.23 ± 4.53
45.75 ± 12.69 6/16 159.75 ± 4.92 60.47 ± 7.74 23.67 ± 2.66 1.71 ± 0.11 – – – – – – – –
0.051 0.97 0.902 0.225 0.114 0.344 N/A N/A N/A N/A N/A N/A N/A N/A
4 (8.0%)
–
N/A
9 (18.0%) 5 (10.0%) 9 (18.0%) 5 (10.0%) 21 (42.0%) 49 (98.0%) 28 (56.0%)
– – – – – – –
N/A N/A N/A N/A N/A N/A N/A
Note: Data presented as mean ± SD. BMI, body mass index; BSA, body surface area; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; HD, haemodialysis; PAH, pulmonary arterial hypertension; RH, renal hypertension; N/A, not applicable. Table 2 RV function and LV function parameters of ESRD patients and healthy individuals.
3. Results 3.1. Patient characteristics After exclusion, 50 ESRD patients (31 women, 19 men; mean age 53.92 ± 14.77 years) and 16 healthy individuals (10 women, 6 men; mean age 45.75 ± 12.69 years) were considered to be eligible for our study (Table 1). There were no significant differences in age, gender, weight, height, body mass index (BMI) and BSA between the two groups (all p > 0.05). For the ESRD patients, the main cause of ESRD included chronic nephritis (n = 34), polycystic kidney (n = 6), diabetes mellitus (n = 4), lupus nephritis (n = 1) and hypertension (n = 5). Furthermore, the duration of renal insufficiency was 6.23 ± 4.53 years, and patients underwent dialysis for an average of 4.13 ± 3.57 years; their SBP and DBP during examination were 138.12 ± 20.92 mmHg and 87.04 ± 15.39 mmHg, respectively. Secondary hyperparathyroidism was found in 4 (8.0%) patients; renal anemia was found in 49 (98.0%) patients; and renal hypertension (RH) was detected in 28 (56.0%) patients; Using two dimensional echocardiography, 9 (18.0%) patients were found to have mitral regurgitation; 5 (10.0%) patients were found to have aortic regurgitation; 9 (18.0%) patients were found to have tricuspid regurgitation; and 5 (10.0%) patients were found to have PAH. The digital subtraction angiography (DSA) revealed 21 (42.0%) patients had central venous thrombosis.
Variable
ESRD patients (n = 50)
Healthy individuals (n = 16)
p-value
RVEDV (ml/ m2) RVESV (ml/ m2) RVSV (ml/m2) RVEF (%) IVS (mm) LVEDV (ml/ m2) LVESV (ml/m2) LVSV (ml/m2) LVEF (%) LVM (g/m2)
67.70 ± 16.18
78.46 ± 12.58
0.008
36.14 ± 11.87
45.66 ± 7.77
0.004
31.57 46.62 17.96 75.06
± ± ± ±
10.28 10.93 3.66 22.84
46.33 59.34 14.18 70.41
± ± ± ±
7.32 5.18 2.36 10.30
0.000 0.000 0.000 0.265
35.50 39.33 54.13 52.51
± ± ± ±
21.09 11.05 10.77 17.90
27.01 45.62 64.83 27.59
± ± ± ±
8.96 7.02 4.67 6.34
0.009 0.036 0.000 0.000
Note: Data presented as mean ± SD. RVEDV, right ventricular end-diastolic volume; RVESV, right ventricular end-systolic volume; RVSV, right ventricular stroke volume; RVEF, right ventricular ejection fraction; IVS, interventricular septum; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; LVSV, left ventricular stroke volume; LVEF, left ventricular ejection fraction.
3.2. RV and LV function assessment When considering RVEF < 0.40 as a threshold to define RV dysfunction, 26% (13/50) of the ESRD patients were identified as having RV dysfunction. As shown in Table 2, the patient group showed a significant decrease in RVEDV, RVESV, RVSV and RVEF (all p < 0.05) compared with the control group. Moreover, a significantly lower LVEF and LVSV were observed in the patient group than in the control group, while a significantly higher LVESV was observed in the patient group than in the control group (all p < 0.05). Furthermore, the IVS thickness in the patient group was significantly greater than that in the 91
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
Fig. 2. Scatter plot demonstrating correlation between (A) RVEDV and LVEDV, (B) RVESV and LVESV, (C) RVSV and LVSV and (D) RVEF and LVEF.
(Table 4).
control group (p < 0.001). However, there were no significant differences in LVEDV (p = 0.265). In ESRD patients, RVEDV closely correlated with LVEDV (r = 0.62, 95% confidence interval [CI]: 0.35–0.81; p < 0.001); RVESV also correlated well with LVESV (r = 0.52, 95% CI: 0.25–0.74; p < 0.001). Accordingly, there was a positive correlation between RVSV and LVSV (r = 0.58, 95% CI: 0.40–0.74; p < 0.001) and between RVEF and LVEF (r = 0.37, 95% CI: 0.06–0.63; p < 0.001) (Fig. 2).
4. Discussion Similar to previous studies using echocardiography [5,12,13], our study demonstrated a high prevalence (26%, 13/50) of RV dysfunction in ESRD patients undergoing HD. RV function plays a vital role in determining clinical outcomes of patients with multifarious cardiovascular and pulmonary diseases [14–17], deterioration of RV function may increase venous congestion, reduce renal blood flow, limit LV preload and increase the substrate for malignant ventricular arrhythmias, thereby leading to a poor prognosis [14,15,18]. Correspondingly, survival analysis including echocardiographic variables demonstrated that RV dysfunction is significantly associated with impaired survival in ESRD patients [7]. Therefore, it is important to pay more attention to RV function during the diagnosis, therapy and followup of ESRD patients. At present, CMR imaging has a high accuracy and reproducibility for RV size and function measurements [19] and has been able to ascertain the effects of frequent HD on ventricular volume, function and LV remodelling in ESRD patients [20]. Therefore, shifts in RV function in ESRD patients on maintenance HD were further
3.3. Determinants associated with RVEF Within the patient group, there was a positive correlation between decreased RVEF and LVEF. In addition, significant correlations were noticed between RVEF and duration of dialysis (r = −0.63, 95% CI: −0.77 to −0.43; p < 0.001), renal insufficiency (r = −0.53, 95% CI: −0.73 to −0.27; p < 0.001) (Fig. 3) or gender (r = −0.31, 95% CI: −0.57 to −0.03; p = 0.032). Factors that significantly correlated with RVEF (LVEF, duration of dialysis, renal insufficiency and gender) were entered into a multivariate linear regression model to establish the optimal regression equation (Table 3). Finally, LVEF and duration of dialysis were noted to be significantly associated with RV dysfunction
Fig. 3. Scattergrams showing results of linear regression analysis between RVEF and (A) LVEF, (B) duration of renal insufficiency and (C) duration of dialysis. The regression equation and correlation coefficient (r) are provided in each plot.
92
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
derived from intermittent and chronic volume overload. Prolonged high-volume changes may impair RV systolic function, although the right ventricle could adapt well to volume overload within limits. Further, the removal of large amounts of fluid during HD may aggravate RV dysfunction due to RV myocardial ischaemia resulting from severe hypovolemia [24,25]. Moreover, HD-induced myocardial stunning may facilitate the development of myocardial dysfunction [26]. Based on univariate correlation analysis, we also identified a negative correlation between decreased RVEF and duration of renal insufficiency, which supports the idea that RV function is impaired in ESRD patients without HD [4]. However, dialysis instead of duration of renal insufficiency was found to be the main cause of RV dysfunction in ESRD patients on maintenance HD. In addition, we found that there were no significant correlation between the common complications (e.g., mitral regurgitation) and decreased RVEF. This may be because of appropriate interventions were timely given to control these complications. Contrary to our findings, other researchers deemed that ventricular functions would not deteriorate, while BP levels, weight, hypervolemia, anemia and other metabolic disorders during HD treatment were under control [27]. BP and HDL-C levels have been associated with preclinical shifts in RV function in a Multi-Ethnic Study of Atherosclerosis (MESA) [28]. Accordingly, BP and HDL-C were included in univariate correlation analysis of our study. Contrary to the results noted by the researchers, BP and HDL-C levels did not correlate with RVEF in our study. This contradictory result may be mainly due to the clinical and laboratory indexes in ESRD patients on HD were usually kept at a certain level, except different in disease. In line with previous studies [12], our study revealed that LV function is significantly impaired in HD patients. In Chen’s three-dimensional speckle-tracking echocardiographic study on LV myocardial function, the researchers found that LV regional long-axis, short-axis and global functions are compromised in uraemia patients with normal LVEF, regardless of whether they receive HD [29]. In addition, a correlation study showed that LV function deteriorates parallel to impairments of RV function in dialysis patients [12]. It was in accordance with our findings that indices of LV function correlated well with those observed for the RV function. This finding may be explained by ventricular interdependence, which means that the size, function and load conditions of one ventricle will affect contralateral ventricular filling and contraction through direct mechanical interactions [23,30]. Paneni et al. thought that chronic pressure/volume overload caused by AVF precipitates RV dysfunction, subsequently triggering LV dysfunction via right-to-left interdependence [12]. Therefore, we considered that compromised RV function and incrassated IVS in HD patients may aggravate the deterioration of LV dysfunction, except for the aetiology of ESRD. There are several potential limitations in our study. First, our results are limited by the relatively small number of patients. This small sample size is mainly because patients were recruited with strict inclusion and exclusion criteria to minimise the effect of confounding factors on RV function. However, it was difficult to assess RV function by classified analysis on account of our small sample size. Moreover, it would be desirable to further confirm the correlation between RVEF and risk factors in a study with a larger sample size. Secondly, because of a limited amount of healthy volunteers, the age between two groups were just similar instead of exactly matching. Finally, data related to the impact of changes in RV function on cardiovascular events and long-term survival are still lacking and a substantial follow-up period would allow us to further validate the predictive value of RV function in ESRD patients.
Table 3 Correlation between RVEF and clinical and part LV parameters. Parameter
r
p-value
Age (year) Gender Weight Height BMI(kg/m2) BSA(m2) SBP (mmHg) DBP (mmHg) Serum Phosphorus (mg/dl) HDL-C (mg/dl) Creatinine (mg/dl) eGFR (ml/min/1.732 m2) LVEF (%) Duration of dialysis (year) Duration of renal insufficiency (year) Secondary hyperparathyroidism Mitral regurgitation Aortic regurgitation Tricuspid regurgitation PAH Central venous thrombosis Renal anemia RH
0.024 −0.308 0.072 −0.113 0.070 −0.032 0.086 0.093 0.029 −0.078 0.029 0.025 0.37 −0. 630 −0.526 0.010 −0.018 −0.164 0.049 −0.043 −0.035 0.153 0.046
0.871 0.032* 0.622 0.436 0.630 0.824 0.553 0.519 0.841 0.592 0.841 0.862 0.008* 0.000* 0.000* 0.944 0.901 0.255 0.737 0.77 0.811 0.287 0.751
Note: all abbreviations are as those in Tables 1 and 2. Table 4 Multivariate linear regression entering only significant correlates into the (backward stepwise) model (adjusted R2 = 0.53). Model 1 β
B Constant Duration of dialysis LVEF Duration of renal insufficiency Gender
Model 3
#
27.60 (14.02–41.19) −1.10* (−2.00 to −0.21) 0.50# (0.28–0.71) −0.24 @ (−0.96 to 0.48) −3.37@ (−7.91 to 1.16)
–
β
B #
–
0.49 −0.10
22.66 (10.75–34.57) −1.35# (−1.97 to −0.74) 0.55# (0. 35–0.76) –
0.54 –
−0.15
–
–
−0.36
−0.44
Note: Data presented as mean (95% confidence interval). LVEF, left ventricular ejection fraction; B, unstandardised regression coefficient; β, standardised coefficient. Units of measurement as before. * p < 0.05; #p < 0.001; @p > 0.05, which means that the variable was eliminated from the final linear regression model.
determined by CMR imaging in our study. At present, there have been several studies exploring the risk factors associated with LV dysfunction by echocardiography [21] or CMR imaging [22] in ESRD patients on maintenance HD. However, data involving the determinants of RV impairments in these given patients have been limited. For these patients, several pathophysiological mechanisms may precipitate the deterioration of RV function by increasing the RV preload or afterload: metabolic and hormonal derangements, sympathetic activation, renal anemia, secondary hyperparathyroidism, derangement of calcium–phosphorus homeostasis, inflammation and left-to-right shunt caused by arteriovenous fistula (AVF) [13,16]. We analysed the probable risk factors associated with RV impairment in HD patients. Consistent with previous reports [23], we identified the duration of dialysis as the strongest parameter associated with predicting RVEF and noted LVEF to be also associated with RV dysfunction. This finding further verified that HD treatment would compromise RV function. Unlike continuous ambulatory peritoneal dialysis, intermittent HD would result in the periodic accumulation of fluids, toxics and electrolytes between two consecutive HD treatment sessions. In these patients, the mechanisms underlying RV dysfunction are mainly
5. Conclusion In conclusion, our study demonstrated that both LV and RV function assessed by MRI are significantly compromised in ESRD patients on maintenance HD and RV dysfunction may contribute to the 93
European Journal of Radiology 102 (2018) 89–94
W. Peng et al.
deterioration of LV function in dialysis patients. Moreover, the duration of dialysis could be considered as a significantly independent predictor of RV dysfunction. Accordingly, CMR is a potentially prominent tool for the early detection of subclinical RV dysfunction in ESRD patients on maintenance HD.
(2005) 112–117. [12] F. Paneni, M. Gregori, G.M. Ciavarella, et al., Relation between right and left ventricular function in patients undergoing chronic dialysis, J. Cardiovasc. Med. 14 (4) (2013) 289–295. [13] S. Yilmaz, M. Yetim, B.K. Yilmaz, T. Dogan, E. Aksoy, N. Yuksel, I. Dogan, High hemodialysis vascular access flow and impaired right ventricular function in chronic hemodialysis patients, Indian J. Nephrol. 26 (5) (2016) 352–356. [14] F.L. Dini, R.T. Demmer, A. Simioniuc, et al., Right ventricular dysfunction is associated with chronic kidney disease and predicts survival in patients with chronic systolic heart failure, Eur. J. Heart Fail. 14 (3) (2012) 287–294. [15] D. Murninkas, A.C. Alba, D. Delgado, M. McDonald, F. Billia, W.S. Chan, H.J. Ross, Right ventricular function and prognosis in stable heart failure patients, J. Card. Fail. 20 (5) (2014) 343–349. [16] K. Cenkerova, J. Dubrava, V. Pokorna, J. Kaluzay, O. Jurkovicova, Right ventricular systolic dysfunction and its prognostic value in heart failure with preserved ejection fraction, Acta Cardiol. 70 (4) (2015) 387–393. [17] L.J. Hickson, S.M. Negrotto, M. Onuigbo, et al., Echocardiography criteria for structural heart disease in patients with end-stage renal disease initiating hemodialysis, J. Am. Coll. Cardiol. 67 (10) (2016) 1173–1182. [18] R.A. Bleasdale, M.P. Frenneaux, Prognostic importance of right ventricular dysfunction, Heart 88 (4) (2002) 323–324. [19] C.F. Mooij, C.J. de Wit, D.A. Graham, A.J. Powell, T. Geva, Reproducibility of MRI measurements of right ventricular size and function in patients with normal and dilated ventricles, J. Magn. Reson. Imaging 28 (1) (2008) 67–73. [20] C.T. Chan, T. Greene, G.M. Chertow, et al., Effects of frequent hemodialysis on ventricular volumes and left ventricular remodeling, Clin. J. Am. Soc. Nephrol. 8 (12) (2013) 2106–2116. [21] K. Negishi, M. Kobayashi, I. Ochiai, et al., Association between fibroblast growth factor 23 and left ventricular hypertrophy in maintenance hemodialysis patients. Comparison with B-type natriuretic peptide and cardiac troponin T, Circ. J. 74 (12) (2010) 2734–2740. [22] R. Wald, M.B. Goldstein, R.M. Wald, et al., Correlates of left ventricular mass in chronic hemodialysis recipients, Int. J. Cardiovasc. Imaging 30 (2) (2014) 349–356. [23] H. Arinc, H. Gunduz, A. Tamer, et al., Use of tissue doppler to assess right ventricle function in hemodialysis patients, Am. J. Nephrol. 25 (3) (2005) 256–261. [24] F. Haddad, R. Doyle, D.J. Murphy, S.A. Hunt, Right ventricular function in cardiovascular disease, part II: pathophysiology, clinical importance, and management of right ventricular failure, Circulation 117 (13) (2008) 1717–1731. [25] J.P. Kinet, D. Soyeur, N. Balland, et al., Hemodynamic study of hypotension during hemodialysis, Kidney Int. 21 (6) (1982) 868–876. [26] J.O. Burton, H.J. Jefferies, N.M. Selby, C.W. McIntyre, Hemodialysis-induced cardiac injury: determinants and associated outcomes, Clin. J. Am. Soc. Nephrol. 4 (5) (2009) 914–920. [27] M. Duran, A. Unal, M.T. Inanc, et al., Hemodialysis does not impair ventricular functions over 2 years, Hemodial. Int. 15 (3) (2011) 334–340. [28] H. Chahal, C. Johnson, H. Tandri, et al., Relation of cardiovascular risk factors to right ventricular structure and function as determined by magnetic resonance imaging (results from the multi-ethnic study of atherosclerosis), Am. J. Cardiol. 106 (1) (2010) 110–116. [29] R. Chen, X. Wu, L.J. Shen, B. Wang, M.M. Ma, Y. Yang, B.W. Zhao, Left ventricular myocardial function in hemodialysis and nondialysis uremia patients: a three-dimensional speckle-tracking echocardiography study, PLoS One 9 (6) (2014) e100265. [30] S. Magder, The left heart can only be as good as the right heart: determinants of function and dysfunction of the right ventricle, Crit. Care Resusc. 9 (4) (2007) 344–351.
Conflict of interest We declare that we have no conflicts of interest to this work. Funding This work was supported by the National Natural Science Foundation of China (No: 81641169, 81471722) and Program for Young Scholars And Innovative Research Team in Sichuan Province of China (No: 2017TD0005). References [1] M. Gonzalezbedat, G. Rosadiez, R. Pecoitsfilho, et al., Burden of disease: prevalence and incidence of ESRD in Latin America, Clin. Nephrol. 83 (1) (2015) S3–S6. [2] K. Kundhal, C.E. Lok, Clinical epidemiology of cardiovascular disease in chronic renal disease, Nephron Clin. Pract. 101 (2005) c47–c52. [3] A. Merchant, R. Wald, M.B. Goldstein, et al., Relationship between different blood pressure measurements and left ventricular mass by cardiac magnetic resonance imaging in end-stage renal disease, J. Am. Soc. Hypertens. 9 (4) (2015) 275–284. [4] K. Said, M. Hassan, E. Baligh, B. Zayed, K. Sorour, Ventricular function in patients with end-stage renal disease starting dialysis therapy: a tissue doppler imaging study, Echocardiography 29 (9) (2012) 1054–1059. [5] L.J. Zhao, S.M. Huang, T. Liang, H. Tang, Pulmonary hypertension and right ventricular dysfunction in hemodialysis patients, Eur. Rev. Med. Pharmacol. Sci. 18 (21) (2014) 3267–3273. [6] S.R. Mehta, J.W. Eikelboom, M.K. Natarajan, R. Diaz, C. Yi, R.J. Gibbons, S. Yusuf, Impact of right ventricular involvement on mortality and morbidity in patients with inferior myocardial infarction, J. Am. Coll. Cardiol. 37 (1) (2001) 37–43. [7] E.J. Stallworthy, H.L. Pilmore, M.W. Webster, K.K. Sidhu, E.M. Curry, P. Brown, A. Scaria, Do echocardiographic parameters predict mortality in patients with endstage renal disease? Transplantation 95 (10) (2013) 1225–1232. [8] Y.K. Guo, Z.G. Yang, H. Shao, W. Deng, G. Ning, Z.H. Dong, Right ventricular dysfunction and dilatation in patients with mitral regurgitation: analysis using ECGgated multidetector row computed tomography, Int. J. Cardiol. 167 (4) (2013) 1585–1590. [9] J.J.V. McMurray, S. Adamopoulos, S.D. Anker, et al., ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012, Eur. Heart J. 33 (2012) 1787–1847. [10] S. Zhang, Z.G. Yang, J.Y. Sun, L.Y. Wen, H.Y. Xu, G. Zhang, Y.K. Guo, Assessing right ventricular function in patients with hypertrophic cardiomyopathy with cardiac MRI: correlation with the new York heart function assessment (NYHA) classification, PLoS One 9 (9) (2014) e104312. [11] A.H. Mahnken, M. Katoh, P. Bruners, E. Spuentrup, J.E. Wildberger, R.W. Günther, A. Buecker, Acute myocardial infarction: assessment of left ventricular function with 16-detector row spiral CT versus MR imaging–study in pigs, Radiology 236 (1)
94