2016 Volume 80 Issue 5 Pages 1187-1195
Background: Heart failure (HF) causes organ congestion, which is thought to increase organ stiffness. The virtual touch quantification (VTQ) method can be used to assess liver stiffness in patients with chronic liver diseases. This study aimed to measure liver and kidney stiffness using VTQ and to determine its value for assessing organ congestion in patients with HF.
Methods and Results: This study included 10 normal subjects and 38 HF patients (age 52.3±16.7 years, left ventricular ejection fraction 27.0±9.4%, plasma B-type natriuretic peptide [BNP] 1,297.3±1,155.1 pg/ml). We investigated the relationships between clinical characteristics and hemodynamics and liver and kidney stiffness, and assessed the effects of medical treatment on these measurements. Liver stiffness was significantly higher in HF patients (1.17±0.13 m/s vs. 2.03±0.91 m/s, P=0.004) compared with normal subjects, but kidney stiffness was similar in both groups. Central venous pressure (CVP) (P=0.021) and BNP (P=0.025) were independent predictive factors for increased liver stiffness in HF patients. Liver stiffness decreased significantly from 2.37±1.09 to 1.27±0.33 m/s (P<0.001) after treatment. Changes in liver stiffness in HF patients significantly correlated with changes in CVP (R=0.636, P=0.014) and cardiac index (R=−0.557, P=0.039) according to univariate analysis, and with changes in CVP in multivariate analysis.
Conclusions: Liver stiffness measured by noninvasive VTQ methods can be used to assess liver congestion and therapeutic effects in patients with HF. (Circ J 2016; 80: 1187–1195)
Heart failure (HF) is the main cause of morbidity and mortality in cardiac patients,1 and may cause multiple organ dysfunction as a result of various interactions.2–4 In particular, kidney and liver dysfunction are closely associated with HF, referred to as cardiorenal syndrome5–7 and cardiohepatic syndrome, respectively.8,9 The main mechanisms for these associations involve neurohormonal adaptations, reduced organ perfusion, increased organ venous pressure, and right ventricular dysfunction.10–12 Organ congestion induced by high central venous pressure (CVP) in patients with decompensated HF is one of the most important causative factors. Recent studies demonstrated a significant correlation between liver congestion and CVP.13–15 However, measurement of CVP requires invasive cardiac catheterization. In contrast, elastography is a recent, noninvasive method of measuring organ stiffness that was originally developed to evaluate fibrosis caused by liver diseases such as cirrhosis.16,17 There are 2 main techniques: transient elastography (TE) and virtual touch quantification (VTQ). VTQ is the more recent method and is able to detect precise and reproducible elastic quantification by measuring the shear-wave speed created by acoustic radiation force impulse technology.18,19
A recent study demonstrated that liver stiffness measured by TE was higher in patients with decompensated HF compared with normal subjects, and was an independent prognostic factor in HF patients.20 Although several reports have revealed significant correlations between CVP and liver stiffness measured by TE,21–23 the relationship between CVP and the stiffness of other organs, including the kidney, has not been elucidated. Furthermore, the ability of TE to assess organ stiffness accurately is limited, especially in HF patients with obesity and/or ascites.24,25 Similar measurements were performed in HF patients in a recent study, and in the present study body fluid volume (BFV) was also measured, to quantify the degree of fluid accumulation.26
The aims of this study were to measure liver and kidney stiffness using the VTQ method and to determine its value for assessing organ congestion in patients with HF.
In this prospective study, we screened 254 consecutive HF patients admitted to Hokkaido University Hospital between November 2013 and July 2015. HF was defined as New York Heart Association (NYHA) functional class II–IV, left ventricular ejection fraction (LVEF) <50% by echocardiography, plasma B-type natriuretic peptide (BNP) level ≥100 pg/ml, and requirement for hospitalization because of worsening HF. We excluded 146 patients with endstage renal disease defined as estimated glomerular filtration rate (eGFR) <15 ml/min/1.73 m2, severe proteinuria ≥0.50 g/g creatinine, or underlying renal disease including diabetic nephropathy, gouty kidney, and atrophic kidney confirmed by abdominal ultrasound imaging, or organic hepatic disorders including cirrhosis. Right heart catheterization (RHC) or VTQ could not be performed in 70 HF patients, so 10 consecutive normal subjects and 38 HF patients were finally enrolled in the present study (Figure S1).
The study protocol was approved by the Ethics Committee of Hokkaido University Hospital, and complied with the Declaration of Helsinki. All patients gave written informed consent.
Study ProtocolWe measured organ stiffness by VTQ method in normal subjects (control group) and HF patients. We also assessed the hemodynamic parameters, including CVP by RHC, clinical parameters, including BFV, echocardiography, and laboratory data of the HF patients. We further assessed changes in organ stiffness, RHC parameters, physical status, BFV, and laboratory data between baseline and after optimal medical treatment, including diuretics, vasodilators, and inotropes, in 14 of the 38 HF patients. Compensated HF was defined as a reduction of >2 kg of body weight (BW) and extracellular fluid (ECF) and clinical resolution of HF symptoms (dyspnea and edema) after treatment. Organ stiffness was measured subsequent to RHC with no changes in HF treatment, including diuretics, vasodilators, and inotropes. We investigated the relationship between organ stiffness and severity of clinical presentation. Severe HF was defined as patients with NYHA III/IV, prior hospitalizations, and duration of HF >1 year. Onset of HF was defined as the first day of hospitalization, to match the inclusion criteria of the present study.
We assessed the relationship between body mass index (BMI) and organ stiffness measured by VTQ method. The 38 HF patients were divided into 2 groups according to the median BMI (22 kg/m2). We compared the linear regression lines for CVP and liver stiffness between patients with BMI ≥22 with those with BMI <22.
VTQ ElastographyLiver and kidney stiffness was measured by VTQ method using an ACUSON S2000 US system (Siemens, Erlangen, Germany) with convex probe, and was expressed as shear-wave velocity (Vs [m/s]). Imaging was obtained from the right intercostal space with the patient supine and the right arm maximally abducted for the liver, and in the left lateral decubitus position for the kidney, while holding a normal breath. The probe was held lightly against the body while observing the B-mode image. A 0.6×1.0 cm region of interest devoid of large blood vessels was located 1–2 cm below the organ surface. The anatomy of the left liver lobe, which is surrounded by the diaphragm, stomach and aorta, affects the use of VTQ, and the right lobe is thus commonly used for liver stiffness measurements using this method.27 We therefore obtained data from the right liver lobe. A preliminary study conducted in 5 healthy volunteers demonstrated no differences in liver stiffness among the anteroinferior (S5), posteroinferior (S6), posterosuperior (S7), and anterosuperior segments (S8) of the right hepatic lobe (1.11±0.16 vs. 0.98±0.20 vs. 1.05±0.14 vs. 1.04±0.23 m/s, respectively; P=0.757). All measurement data regarding liver stiffness were therefore obtained from S5 in the present study. The measurements were repeated 10 times to obtain median Vs values. Each measurement of organ stiffness was expressed as median and interquartile range.
Interobserver reproducibility was confirmed in 48 healthy volunteers, prior to the present study. We measured liver stiffness using VTQ in 28 healthy volunteers (27.5±8.3 years, 14 male) and kidney stiffness in 20 healthy volunteers (27.0±5.2 years, 16 male). Interobserver agreement was analyzed using the intraclass correlation coefficient (ICC) and revealed ICCs for liver and kidney stiffness of 0.70 (95% confidence interval (CI) 0.44–0.85, P<0.001) and 0.35 (−0.04 to 0.67, 0.033), respectively.
EchocardiographyEchocardiography was performed using either an Aplio Artida SSH-88-CV or Aplio SSA-770A (Toshiba Medical Systems, Tochigi, Japan). LVEF was calculated from apical 4- and 2-chamber views using the biplane method of disks. Images were evaluated by 2 experienced observers.
Hemodynamic MeasurementsRHC was performed using a 7 French Swan-Ganz catheter (Edwards Lifesciences, CA, USA) as a flow-directed pulmonary artery catheter. Pressure parameters were measured at the end-expiratory period, with an average of 3–5 cycles. Hemodynamic measurements included CVP, pulmonary capillary wedge pressure (PCWP), pulmonary artery systolic pressure (PASP), right ventricular systolic pressure, stroke volume index, cardiac output (CO), cardiac index (CI), and renal perfusion pressure (RPP). CO was measured using the thermodilution method.
Body Composition AnalysisBody composition analysis was performed using an MLT-550N (Toray Medical Co Ltd, Tokyo, Japan) immediately after RHC. Electrodes were attached to the right arms and legs of patients while supine and body impedance was measured. BFV parameters included ECF in all subjects.
Statistical AnalysisData are given as mean±standard deviation. Parameters were compared between the control and HF groups using unpaired t-tests or Welch’s t-test for continuous variables and Fisher’s exact tests for categorical variables. Patients with liver congestion caused by HF are usually asymptomatic in the early phase. Elevated CVP is considered to be the most sensitive indicator for early diagnosis of liver congestion among other parameters including clinical manifestation and biochemical liver function. Previous study demonstrated that CVP >10 mmHg could be clinically significant and cause liver congestion in patients with HF.28,29 Therefore, we defined liver congestion as CVP >10 mmHg by RHC in HF patients. To evaluate the liver congestion associated with HF, we performed receiver-operating characteristic (ROC) curve analysis in HF group (n=38). Pearson’s correlation coefficients were calculated for the relationships between liver stiffness, kidney stiffness and each continuous variable. Analysis of covariance (ANCOVA) was used to determine if the regression lines between CVP and liver stiffness differed between the BMI groups (≥22 and <22). Significant variables in univariate analysis were also included in multivariate analysis. P<0.05 was considered statistically significant. Statistical analysis was carried out using JMP 12.0 (SAS Institute, Cary, NC, USA).
The clinical characteristics of the control and HF subjects are shown in Table 1. Age, sex, BMI, diastolic blood pressure, and heart rate were comparable between the 2 groups, but systolic blood pressure was higher in the control group. The mean age of the HF patients was 52.3±16.7 years, 76% were male, and 23 patients (61%) had NYHA functional class III/IV; 22 patients (58%) required hospitalization for worsening HF, with a frequency of hospitalization of 2.34±2.12 times. A total of 20 patients (53%) were diagnosed with HF >1 year ago, and the mean duration after HF diagnosis was 4.72±6.97 years. The underlying causes of HF were dilated cardiomyopathy in 14 (37%), ischemic cardiomyopathy in 5 (13%), valvular heart disease in 4 (11%), hypertrophic cardiomyopathy in 3 (8%), hypertensive heart disease in 3 (8%), and other causes in 9 (24%). Medical treatment included angiotensin-converting enzyme inhibitors or angiotensin II receptor antagonists and β-blockers in 27 (71%), diuretics in 36 (95%), and inotropes in 15 (39%) patients. Hemoglobin was 12.5±1.8 g/dl, eGFR was 56.2±31.3 ml/min/1.73 m2, total bilirubin (T-bil) was 1.2±1.1 mg/dl, aspartate aminotransferase was 47.6±48.1 IU/L, alanine aminotransferase was 46.6±48.5 IU/L, γ-glutamyltransferase was 112.4±87.7 IU/L, albumin was 3.7±0.5 g/dl, BNP was 1,297.3±1,155.1 pg/ml, and LVEF was 27.0±9.4% by echocardiography. In the body composition analysis, ECF was 47.1±10.3 kg. Hemodynamic measurements revealed CVP 11.8±5.4 mmHg, PCWP 22.8±8.4 mmHg, PASP 45.4±17.8 mmHg, CI 2.0±0.4 L/min/m2, and RPP 62.7±14.7 mmHg.
| Control group (n=10) |
HF group (n=38) |
P value | |
|---|---|---|---|
| Age (years) | 54.2±6.9 | 52.3±16.7 | 0.735 |
| Sex (M/F) | 8/2 | 42,642 | 0.805 |
| Height (cm) | 163.8±6.6 | 166.3±8.3 | 0.377 |
| Weight (kg) | 61.5±8.7 | 65.2±17.3 | 0.517 |
| BMI (kg/m2) | 22.9±2.3 | 23.1±4.8 | 0.841 |
| SBP (mmHg) | 118.4±5.3 | 101.0±18.4 | 0.007 |
| HR (beats/min) | 74.0±8.9 | 78.5±20.1 | 0.764 |
| NYHA II, n (%) | – | 15 (39%) | – |
| NYHA III/IV, n (%) | – | 23 (61%) | – |
| No. of hospitalizations | – | 2.34±2.12 | – |
| Duration of HF (years) | – | 4.72±6.97 | – |
| Underlying heart disease | |||
| DCM, n (%) | – | 14 (37%) | – |
| ICM, n (%) | – | 5 (13%) | – |
| Valvular heart disease, n (%) | – | 4 (11%) | – |
| HCM, n (%) | – | 3 (8%) | – |
| HHD, n (%) | – | 3 (8%) | – |
| Other, n (%) | – | 9 (24%) | – |
| Medications | |||
| ACEIs/ARBs, n (%) | – | 27 (71%) | – |
| β-blockers, n (%) | – | 27 (71%) | – |
| Diuretics, n (%) | – | 36 (95%) | – |
| Inotropes, n (%) | – | 15 (39%) | – |
| Laboratory data | |||
| Hb (g/dl) | – | 12.6±1.8 | – |
| eGFR (ml/min/1.73 m2) | – | 56.2±31.3 | – |
| T-bil (mg/dl) | – | 1.2±1.1 | – |
| AST (IU/L) | – | 47.6±48.1 | – |
| ALT (IU/L) | – | 46.6±48.5 | – |
| γ-GTP (IU/L) | – | 112.4±87.7 | – |
| Albumin (g/dl) | – | 3.7±0.5 | – |
| BNP (pg/ml) | – | 1,297.3±1,155.1 | – |
| Echocardiography | |||
| LVEF (%) | – | 27.0±9.4 | – |
| Body composition analysis | |||
| ECF (kg) | – | 47.1±10.3 | – |
| Hemodynamics | |||
| CVP (mmHg) | – | 11.8±5.4 | – |
| PCWP (mmHg) | – | 22.8±8.4 | – |
| PASP (mmHg) | – | 45.4±17.8 | – |
| CI (L/min/m2) | – | 2.0±0.4 | – |
| RPP (mmHg) | – | 62.7±14.7 | – |
All values are expressed as mean±SD. P<0.05 indicates statistical significance. ACEI, angiotensin-converting enzyme inhibitor; ALT, alanine aminotransferase; ARB, angiotensin II receptor blocker; AST, aspartate aminotransferase; BMI, body mass index; BNP, plasma B-type natriuretic peptide; CI, cardiac index; CVP, central venous pressure; DCM, dilated cardiomyopathy; ECF, extracellular fluid; eGFR, estimated glomerular filtration rate; γ-GTP, γ-glutamyltransferase; Hb, hemoglobin; HCM, hypertrophic cardiomyopathy; HF, heart failure; HHD, hypertensive heart disease; HR, heart rate; ICM, ischemic cardiomyopathy; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; PASP, pulmonary artery systolic pressure; PCWP, pulmonary capillary wedge pressure; RPP, renal perfusion pressure; SBP, systolic blood pressure; T-bil, total bilirubin.
Liver stiffness was significantly higher in the HF group than in the control group (1.17±0.13 vs. 2.03±0.91 m/s, P<0.001), but there were no significant differences in kidney stiffness between the groups (2.14±0.30 vs. 2.20±0.60 m/s, P=0.686) (Figure 1).
Kidney stiffness (A) and liver stiffness (B) in control and heart failure (HF) groups. Kidney stiffness was similar between groups (2.14±0.30 vs. 2.20±0.60, respectively; P=0.686), whereas liver stiffness was significantly higher in the HF patients than in the control group (1.17±0.13 vs. 2.03±0.91, P=0.004).
The relationships between organ stiffness and clinical variables are shown in Table 2. Univariate analysis found no correlation between kidney stiffness and any variable, including CVP (Figure 2A), whereas liver stiffness was significantly correlated with BW (R=0.344, P=0.040), ECF (R=0.358, P=0.032), T-bil (R=0.372, P=0.025), BNP (R=0.402, P=0.015), and CVP (R=0.578, P<0.001) (Figure 2B). We therefore performed multivariate analysis including these 5 variables by linear regression analysis. BNP and CVP were identified as independent predictive factors for increased liver stiffness.
| Kidney stiffness | Liver stiffness | |||||
|---|---|---|---|---|---|---|
| Univariate analysis | Univariate analysis | Multivariate analysis | ||||
| R | P value | R | P value | 95% CI | P value | |
| Clinical presentation | ||||||
| BW (kg) | 0.284 | 0.084 | 0.344 | 0.040 | −0.015 to 0.029 | 0.526 |
| HR (beats/min) | 0.051 | 0.765 | 0.197 | 0.257 | – | – |
| SBP (mmHg) | −0.189 | 0.256 | −0.119 | 0.491 | – | – |
| ECF (kg) | 0.152 | 0.363 | 0.358 | 0.032 | −0.077 to 0.069 | 0.905 |
| LVEF (%) | −0.031 | 0.852 | −0.185 | 0.279 | – | – |
| Laboratory data | ||||||
| Hb (g/dl) | 0.155 | 0.353 | −0.288 | 0.088 | ||
| eGFR (ml/min/1.73 m2) | 0.216 | 0.193 | −0.102 | 0.553 | – | – |
| T-bil (mg/dl) | 0.029 | 0.862 | 0.372 | 0.025 | −0.147 to 0.369 | 0.388 |
| AST (IU/L) | −0.133 | 0.427 | −0.007 | 0.968 | – | – |
| ALT (IU/L) | −0.219 | 0.187 | 0.034 | 0.846 | – | – |
| γ-GTP (IU/L) | 0.036 | 0.830 | 0.205 | 0.232 | ||
| Albumin (g/dl) | 0.16 | 0.336 | −0.169 | 0.324 | – | – |
| BNP (pg/dl) | −0.046 | 0.783 | 0.402 | 0.015 | 0.000 to 0.001 | 0.025 |
| Hemodynamics | ||||||
| CVP (mmHg) | 0.050 | 0.764 | 0.578 | <0.001 | 0.011 to 0.126 | 0.021 |
| PCWP (mmHg) | 0.286 | 0.082 | 0.281 | 0.098 | – | – |
| PASP (mmHg) | 0.111 | 0.514 | 0.194 | 0.264 | – | – |
| SVI (ml/m2) | −0.152 | 0.376 | −0.254 | 0.147 | – | – |
| RPP (mmHg) | 0.021 | 0.902 | −0.290 | 0.087 | – | – |
P<0.05 indicates statistical significance. BW, body weight; 95% CI, 95% confidence interval; SVI, stroke volume index. Other abbreviations as in Table 1.
Correlation between organ stiffness and central venous pressure (CVP). (A) Kidney stiffness shows no correlation with CVP; R=0.050, P=0.764). (B) Liver stiffness is moderately correlated with CVP (R=0.578, P<0.001).
In the HF group (n=38), 19 patients had CVP >10 mmHg by RHC. According to ROC curve analysis, the area under the curve of liver stiffness to identify CVP >10 mmHg was 0.81 (95% CI, 0.62–0.92), and the optimal cutoff value of liver stiffness for detection of CVP >10 mmHg was liver stiffness ≥2.05 m/s. The sensitivity and specificity of liver stiffness for differentiating between patients with CVP >10 mmHg and ≤10 mmHg were 74% and 88%, respectively.
The relationship between organ stiffness and severity of clinical presentation is shown in Figure 3. There was no correlation between kidney stiffness and clinical severity. In contrast, liver stiffness was significantly higher in HF patients with NYHA III/IV, history of prior hospitalizations, and duration of HF >1 year (P<0.001, P=0.002, and P=0.030, respectively).
Comparison of liver stiffness according to severity of heart failure (HF). Liver stiffness was significantly higher in patients with severe HF, defined as (A) NYHA (2.34±0.96 in class III/IV vs. 1.55±0.55 in class II m/s, P<0.001), (B) history of prior hospitalization (2.36±0.99 vs. 1.58±0.52 m/s, P=0.002), and (C) duration of HF (2.32±1.00 in ≥1 year vs. 1.75±0.73 in <1 year m/s, P=0.030).
The median BMI in the 38 HF patients was 22 kg/m2. CVP significantly correlated with liver stiffness in patients with BMI ≥22 (n=19; r=0.254, P=0.039) and those with BMI <22 (n=19; r=0.439, P=0.002) kg/m2. ANCOVA revealed no difference in the slopes of the regression lines between the 2 groups (P=0.457).
Effects of Treatment on Liver Stiffness in HF PatientsWe compared the clinical variables at baseline and after medical treatment in HF patients (Table 3). BW (66.7±17.5 vs. 58.8±13.1 kg, P<0.001), ECF (13.9±6.3 vs. 8.9±2.8 kg, P=0.002), BNP (1,600.8±1,341.2 vs. 379.8±334.5 pg/ml, P=0.013), CVP (13.4±5.7 vs. 7.6±2.6 mmHg, P<0.001), PCWP (26.0±6.6 vs. 12.2±6.6 mmHg, P<0.001), and PASP (51.9±12.9 vs. 30±9.6 mmHg, P<0.001) were significantly reduced and CI (1.9±0.4 vs. 2.2±0.4 L/min/m2, P=0.014) was significantly increased after treatment compared with baseline. Aspartate aminotransferase and alanine aminotransferase values in 6 patients remained slightly elevated even after HF treatment. However, abdominal ultrasound findings revealed no organic liver disease at baseline or after HF treatment in any patient.
| Baseline | After treatment | P value | |
|---|---|---|---|
| Clinical presentation | |||
| BW (kg) | 66.7±17.5 | 58.8±13.1 | <0.001 |
| HR (beats/min) | 81.9±21.4 | 73.4±16.6 | 0.214 |
| SBP (mmHg) | 99.5±14.7 | 95±14.4 | 0.287 |
| LVEF (%) by echocardiography | 27.8±10.1 | 30.2±12.4 | 0.421 |
| ECF (kg) | 13.9±6.3 | 8.9±2.8 | 0.002 |
| Laboratory data | |||
| Hb (g/dl) | 12.6±1.9 | 12.2±2.1 | 0.274 |
| eGFR (ml/min/1.73 m2) | 62.0±34.1 | 68.0±27.3 | 0.311 |
| T-bil (mg/dl) | 1.4±1.5 | 0.9±0.6 | 0.179 |
| AST (IU/L) | 38.9±38.3 | 32.1±12.8 | 0.546 |
| ALT (IU/L) | 33.4±33.9 | 33.4±25.8 | 0.994 |
| γ-GTP (IU/L) | 102.3±76.3 | 68.7±50.6 | 0.142 |
| Albumin (g/dl) | 3.5±0.5 | 3.6±0.7 | 0.711 |
| BNP (pg/ml) | 1,600.8±1,341.2 | 379.8±334.5 | 0.008 |
| Hemodynamics | |||
| CVP (mmHg) | 13.4±5.7 | 7.6±2.6 | <0.001 |
| PCWP (mmHg) | 26.0±6.6 | 12.2±6.6 | <0.001 |
| PASP (mmHg) | 51.9±12.9 | 30±9.6 | <0.001 |
| CI (L/min/m2) | 1.9±0.4 | 2.2±0.4 | 0.014 |
| RPP (mmHg) | 53.9±27.6 | 59.4±20.5 | 0.681 |
All values are expressed as mean±SD. P<0.05 indicates statistical significance. Abbreviations as in Tables 1,2.
Liver stiffness decreased significantly from 2.37±1.09 to 1.27±0.33 m/s (P<0.001) after HF treatment (Figure 4). Liver stiffness decreased in 13 patients (93% of the 14 patients) and was unchanged in 1 (7%). Linear regression analyses showed that a change in liver stiffness correlated with a change in CVP (R=0.636, P=0.014) and CI (R=−0.557, P=0.039) (Table 4). Multivariate analysis identified changes in CVP as an independent predictive factor for identifying a change in liver stiffness.
Changes in liver stiffness in patients with heart failure (HF) in relation to medical treatment. Liver stiffness decreased significantly from 2.37±1.09 to 1.27±0.33 m/s after HF treatment (P<0.001).
| Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|
| R | P value | 95% CI | P value | |
| Clinical presentation | ||||
| ΔBW | 0.247 | 0.394 | ||
| ΔECF | −0.004 | 0.988 | – | – |
| ΔLVEF by echocardiography | 0.076 | 0.796 | – | – |
| Laboratory data | ||||
| ΔHb | 0.210 | 0.470 | ||
| ΔeGFR | −0.267 | 0.346 | – | – |
| ΔT-bil | 0.256 | 0.378 | – | – |
| ΔAST | 0.010 | 0.973 | ||
| ΔALT | 0.112 | 0.703 | ||
| Δγ-GTP | 0.343 | 0.229 | ||
| ΔAlbumin | 0.116 | 0.692 | – | – |
| ΔBNP | 0.069 | 0.816 | – | – |
| Hemodynamics | ||||
| ΔCVP | 0.636 | 0.014 | 0.027 to 0.210 | 0.017 |
| ΔPCWP | 0.353 | 0.216 | ||
| ΔPASP | 0.078 | 0.819 | ||
| ΔCI | −0.557 | 0.039 | −2.319 to 0.005 | 0.051 |
| ΔRPP | −0.313 | 0.275 | ||
Δ value calculated as post-HF treatment value minus baseline value. P<0.05 indicates statistical significance. Abbreviations as in Tables 1,2.
This study demonstrated an association between liver stiffness and CVP, and indicated the potential value of liver stiffness measured by VTQ as a marker of liver congestion in HF patients. CVP positively correlated with liver stiffness, but not kidney stiffness in HF patients. The results of this study were consistent with the findings of Millonig et al, who also reported a correlation between liver stiffness and CVP in pigs.22 Another recent study reported that liver stiffness measured by TE correlated with CVP in HF patients.23,29 However, most prior studies have measured liver stiffness using TE, which measures the velocity of the vertebral shear wave through the organ and then converts the velocity into an organ stiffness. However, TE is unable to accurately measure the stiffness of the target lesion, because it blindly detects the shear wave generated by mechanical vibration.17 In addition, liver stiffness measurements acquired by TE may be unreliable in patients with obesity and/or ascites17 because of the increased distance between the TE probe and the liver, and the consequent attenuation of both elastic waves and ultrasound. Indeed patients with ascites or obesity were excluded from previous studies that measured liver stiffness in HF patients.23,29,30 In the present study, we measured liver stiffness using the VTQ method and included overweight patients (BMI ≥26; n=11, 29%) and those with ascites. Liver stiffness could be measured accurately and reproducibly in all subjects using VTQ, regardless of BW or the presence of ascites. CVP significantly correlated with liver stiffness in patients with BMI ≥22 and in those with BMI <22 kg/m2, based on the median BMI, and ANCOVA revealed no difference in regression lines between the 2 groups.
A meta-analysis that compared the diagnostic performance of VTQ and TE methods reported that VTQ had similar predictive value to TE, but VTQ was associated with significantly fewer failed or unreliable measurements compared with TE, through acoustic radiation force impulse (2.1% vs. 6.6%, P<0.0001).24
In the present study, kidney stiffness was not correlated with CVP in HF patients. This was most likely because of the heterogeneous structure of the kidneys, including the renal vasculature and tubular segments. Kidney stiffness has been shown to be affected by the cortex or medullary tissue, and by vascular or urinary pressure levels. This heterogeneity of kidney structure might also enhance the influence of the region of interest setting by examiners for kidney stiffness measurements. Indeed, interobserver variability of kidney stiffness was relatively large in the present study (ICC: 0.35, P=0.033).
This study demonstrated that liver stiffness increased with increasing severity of HF, which is consistent with the report by Alegre et al, who showed that liver stiffness was significantly increased in decompensated HF patients who required hospitalization and those with clinically severe HF.21
In the present study, liver stiffness and CVP both decreased after optimal HF treatment, and their changes were significantly correlated. Previous studies evaluated the effects of treatment on clinical improvement in parameters such as BW and edema.21,22,31 In addition to these clinical presentations, we also evaluated the effects of treatment on hemodynamic parameters, including CVP measured by RHC and ECF volume obtained from body composition analysis. This study thus provided more accurate and appropriate assessment of improvements in HF compared with previous studies.
Study LimitationsFirst, it was a single-center study with a small sample population. Analysis of a much larger data sample is thus needed to confirm our results. In addition, the patients enrolled in the current study were limited to those able to undergo emergency RHC and subsequent elastography as soon as possible. Second, increased liver stiffness could indicate fibrosis, as well as liver congestion. Third, although both CVP and liver stiffness decreased after optimal medical treatment in HF patients, it was unclear if a decrease in liver stiffness would lead to favorable outcomes in HF patients. Long-term follow-up of these patients is thus clearly needed. Fourth, the number of eligible HF patients in the present study was limited by the inclusion and exclusion criteria, including NYHA class II–IV, LVEF <50%, and BNP ≥100 pg/ml. Further studies are therefore needed to determine if the present findings can be generalized to the total HF population. Finally, no obese patients with HF (BMI >30) were included in the present study. Although the VTQ method is recognized as a useful measurement technique regardless of BW, further studies are needed to determine its ability to evaluate organ stiffness in obese patients.
In conclusion, liver stiffness assessed by VTQ can predict liver congestion in patients with decompensated HF. Compared with CVP measurements by Swan-Ganz catheterization, this method is noninvasive, convenient, and an efficient technique for evaluating liver congestion and for monitoring treatment-related improvements in patients with HF.
H.T. has received speakers’ bureau/honoraria from MSD K.K., Ono Pharmaceutical Co, Ltd, Takeda Pharmaceutical Co, Ltd, Mitsubishi Tanabe Pharma Corp, Daiichi-Sankyo Co, Ltd, Teijin Ltd, Nippon Boehringer Ingelheim Co, Ltd, Bayer Yakuhin, Ltd, Pfizer Co, Ltd, Bristol-Myers Squibb Co, and research funds from Takeda Pharmaceutical Co, Ltd, Nippon Boehringer Ingelheim Co, Ltd, Mitsubishi Tanabe Pharma Corp, Daiichi-Sankyo Co, Ltd, and consultation fees from Novartis Pharma K.K., Ono Pharmaceutical Co, Ltd, Pfizer Co, Ltd. Other authors declare no conflicts of interest in relation to this article.
Supplementary File 1
Figure S1. Flow chart of patient enrollment in the present study.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-15-1200
