Hepatic Venous Stasis Index Reflects Hepatic Congestion and Predicts Adverse Outcomes in Patients With Heart Failure

Abstract

Background

It has been reported that the hepatic vein waveforms determined by abdominal ultrasonography can assess hepatic congestion in patients with heart failure (HF). However, the parameter that quantifies hepatic vein waveforms has not been established. We suggest the hepatic venous stasis index (HVSI) as the novel indicator to evaluate hepatic congestion quantitatively. To examine the clinical significance of HVSI in patients with HF, we aimed to clarify the associations of HVSI with the parameters of cardiac function and right heart catheterization, as well as that with prognosis, in patients with HF.

Methods and Results

We performed abdominal ultrasonography, echocardiography, and right heart catheterization in patients with HF (n=513). The patients were divided into 3 groups based on HVSI as follows: HVSI 0 (HVSI=0, n=253), low HVSI (HVSI 0.01–0.20, n=132), and high HVSI (HVSI>0.20, n=128). We examined the associations of HVSI with parameters of cardiac function and right heart catheterization and followed up for cardiac events defined as cardiac death or worsening HF. There was a significant increase in level of B‐type natriuretic peptide, inferior vena cava diameter, and mean right atrial pressure with increasing HVSI. During the follow‐up period, cardiac events occurred in 87 patients. In the Kaplan–Meier analysis, cardiac event rate increased across increasing HVSI (log‐rank, P=0.002).

Conclusions

HVSI assessed by abdominal ultrasonography reflects hepatic congestion and right‐sided HF and is associated with adverse prognosis in patients with HF.

Nonstandard Abbreviations and Acronyms

AHEAD score

long‐term risk classification in acute heart failure

CVP

central venous pressure

HV

hepatic vein

HVSI

hepatic venous stasis index

IRVF

intrarenal venous flow

RAP

right atrial pressure

RHC

right heart catheterization

RVSI

renal venous stasis index

SWD

shear wave dispersion

SWE

shear wave elastography

Clinical Perspective.

What Is New?

• Hepatic venous stasis index assessed by abdominal ultrasonography is a novel indicator that is able to evaluate hepatic congestion quantitatively.

What Are the Clinical Implications?

• Hepatic venous stasis index reflected hepatic congestion and right‐sided heart failure.

• High hepatic venous stasis index was associated with adverse prognosis in patients with heart failure.

Heart failure (HF) is a refractory clinical syndrome that originates from various types of structural or functional heart diseases and is a major public health problem that leads to frequent hospitalizations, impaired quality of life, and shortened life expectancy. ¹ , ² The number of patients with HF has been rapidly increasing to an estimated 26 million patients worldwide. ³ , ⁴ , ⁵ , ⁶ The elderly often have multiple comorbidities, including hypertension, atrial fibrillation (AF), chronic kidney disease (CKD), chronic obstructive pulmonary disease, and anemia, all of which aggravate HF and affect prognosis. ⁷ Renal dysfunction and liver dysfunction are common adverse outcome predictors in HF. ⁸ , ⁹ The cardio‐hepatic syndrome can occur because of cardio‐hepatic interactions including the development of congestive hepatopathy (passive congestion) and hypoxic hepatitis (reduced arterial perfusion) in patients with HF. ¹ , ⁷ , ⁸ , ¹⁰ , ¹¹ , ¹² , ¹³ Congestion is a key feature in HF, and its presence is associated with poor prognosis. ¹⁴ Invasive tests such as right heart catheterization (RHC) have been used to objectively evaluate the degree of congestion because sometimes it is not clinically evident. ¹⁵ , ¹⁶ Recently, abdominal ultrasonography has been suggested as a form of noninvasive image testing. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸

Regarding abdominal ultrasonography for assessing congestion, it has recently been reported that the intrarenal venous flow (IRVF) patterns assessed by renal Doppler ultrasonography and the renal venous stasis index (RVSI) quantifying IRVF reflect renal congestion and are associated with adverse prognosis in patients with HF. ¹⁷ , ¹⁸ , ²¹ , ²³ , ²⁴ Concerning hepatic congestion, it has been reported that liver stiffness measured by shear wave elastography (SWE) is an indicator of hepatic congestion due to increased right‐sided filling pressure (ie, central venous pressure [CVP]). ¹⁹ , ²⁵ , ²⁶ , ²⁷ , ²⁸ In addition, it has been reported that shear wave dispersion (SWD) may reflect remaining fibrosis after improvement of congestion. ²⁰ IRVF, RVSI, SWE, and SWD are associated with adverse prognosis in patients with HF. On the other hand, it has been reported that the hepatic vein (HV) waveforms can be used to assess hepatic congestion. ²² The parameters that quantify HV waveforms have not been established. We suggest the hepatic venous stasis index (HVSI) as a new indicator to evaluate hepatic congestion quantitatively.

To examine the clinical significance of HVSI in patients with HF, we aimed to clarify the associations of HVSI with laboratory tests, echocardiography, and RHC, as well as its prognostic impact in patients with HF.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Subjects and Protocol

Figure 1 shows a patient flow diagram. This was a prospective observational study of 556 patients who were classified as either stage C or stage D of HF stage classification in the American College of Cardiology Foundation/ American Heart Association guidelines and were hospitalized and underwent abdominal ultrasonography in Fukushima Medical University Hospital between April 2018 and September 2021. ²⁹ Treatment of decompensated HF was provided by each patient's attending cardiologist based on the established HF guidelines. ²⁹ , ³⁰ , ³¹ , ³² Blood samples, abdominal ultrasonography, echocardiography, and RHC were obtained during hospitalization from patients in a stable condition before discharge. A total of 556 patients underwent abdominal ultrasonography, all of whom underwent testing for hepatitis B surface antigen and hepatitis C antibodies, and their medical histories were checked for chronic liver disease (eg, cirrhosis, hepatic tumors, bile duct disease) and alcohol abuse (≥30 g/day for men, ≥20 g/day for women). Any patients who had the aforementioned liver diseases (n=32) or were undergoing dialysis (n=11) were excluded from the study. Finally, 513 patients were enrolled, among whom 310 had undergone RHC. Patients' baseline data were collected at discharge. Abdominal ultrasonography was performed during hospitalization in patients in a stable condition without changes in medications and the duration from baseline data collection to abdominal ultrasonography was median 4 days. Echocardiography and RHC were performed within 3 days of abdominal ultrasonography. Laboratory test was performed on the same day as RHC. Of these patients, those with HVSI of 0 were classified as the HVSI 0 group (HVSI=0, n=253, 49%). Because HVSI is our original and novel indicator, there is no common cutoff value. Therefore, patients with HVSI above 0 were divided into 2 groups based on the median HVSI value (0.20) in the present population: the low HVSI group (HVSI 0.01–0.20, n=132, 26%) and the high HVSI group (HVSI >0.20, n=128, 25%). These patients were also classified into 3 categories based on their HV waveforms: those with a continuous pattern (C group), those whose V wave ran under the baseline (U group), and those with a reversed V wave (R group). ²²

Figure 1. Patient flow diagram.

A total of 556 patients who were classified as either stage C or D of HF stage classification in the ACCF/AHA guidelines and were hospitalized at Fukushima Medical University Hospital between April 2018 and September 2021 underwent abdominal ultrasonography. Patients who had chronic liver disease (eg, viral hepatitis, cirrhosis, hepatic tumors, bile duct disease, alcohol abuse) or were undergoing dialysis were excluded from the study. Finally, 513 patients were enrolled, among whom 310 had undergone right heart catheterization. Of these patients, those with HVSI of 0 were classified as the HVSI 0 group. Patients with HVSI above 0 were divided into 2 groups based on the median HVSI value (0.20) in the present population: the low HVSI group and the high HVSI group. ACCF indicates American College of Cardiology Foundation; AHA, American Heart Association; HF, heart failure; and HVSI, hepatic venous stasis index.

First, we compared the clinical features and the results from laboratory tests, echocardiography, and RHC among the 3 groups. Second, the patients were followed up until March 2022 for cardiac events as composites of cardiac death or worsening HF. Cardiac death was defined as death from lethal arrhythmia such as ventricular fibrillation, acute coronary syndromes, or worsening HF. Worsening HF was defined as unplanned rehospitalization for HF treatment. For patients who experienced ≥2 events, only the first event was included in the analysis. We were able to follow up with all patients because they all visited the hospital monthly or every other month. Status and dates of death were obtained from the patients' medical records. Third, we evaluated prognostic value of HVSI. We calculated the AHEAD (A–atrial fibrillation, H–hemoglobin levels of <12.0 g/dL in women and <13.0 g/dL in men [anemia], E–elderly [age >70 years], A–abnormal renal parameters [creatinine>130 μmol/L], D–diabetes) score, which was known as predictive model for HF. Each risk factor of the AHEAD score was counted as 1 point. ³³ Then, we created a modified AHEAD score by adding HVSI to the AHEAD score and tested the improvement in prediction. Those administering the survey were blind to the analyses, and written informed consent was obtained from all study subjects. The study protocol was approved by the Ethics Committee of Fukushima Medical University (No. 823) and was carried out in accordance with the principles outlined in the Declaration of Helsinki. Reporting of the study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines and the Enhancing the Quality and Transparency of Health Research guidelines.

Ischemic coronary artery disease was confirmed by either past medical history or by myocardial scintigraphy, coronary computed tomography angiography, or coronary angiography. AF was confirmed by ECG performed during hospitalization or from medical records including past medical history. Hypertension was defined as a systolic blood pressure of ≥140 mm Hg, diastolic blood pressure of ≥90 mm Hg, or use of antihypertensive drugs. ³⁴ Dyslipidemia was defined as triglyceride levels of ≥150 mg/dL, low‐density lipoprotein cholesterol levels of ≥140 mg/dL, high‐density lipoprotein cholesterol levels of <40 mg/dL, or use of cholesterol‐lowering drugs. ³⁵ Diabetes was defined as the recent use of antidiabetic drugs, fasting glucose levels of ≥126 mg/dL, casual glucose levels of ≥200 mg/dL, or glycated hemoglobin levels of ≥6.5% (National Glycohemoglobin Standardization Program). ³⁶ CKD was defined as estimated glomerular filtration rate of <60 mL/min per 1.73 m². ³⁷ , ³⁸ , ³⁹ Anemia was defined as hemoglobin levels of <12.0 g/dL in women and <13.0 g/dL in men. ⁴⁰ , ⁴¹

Abdominal Ultrasonography

The actual methods of acquisition were as follows. Two experienced sonographers (M.M., with 30 years of experience in abdominal ultrasonography, and S.I., with 20 years of experience) performed abdominal ultrasonography, using an Aplio i800 system (Canon Medical Systems Corporation, Tochigi, Japan) with a 1.8 to 6.4 MHz convex transducer. The velocity range of the color Doppler was set to approximately 10 to 20 cm/s. The 2 examiners were blinded to all clinical data. The patients fasted for at least 12 hours before the abdominal ultrasonography and were placed in the supine position. The transducer was placed in the intercostal or subcostal space, and the patient's right arm was raised above the head to obtain a proper acoustic window. The tip of the probe transducer was placed perpendicular to the liver capsule to avoid refraction of the acoustic radiation force impulse. HV was identified in reference to continuity with the inferior vena cava, direction of blood flow, and other identifiers. The measurements were undertaken by approaching the HV 3 to 5 cm proximal to the inferior vena cava from the right intercostal space. We selected the right HV because it was detected easily. The angle of the ultrasound beam to right HV blood flow was less than 60°. A gray‐scale B‐mode evaluation of the liver was first performed, followed by color and spectral Doppler evaluation. As shown in Figure 2, HV waveforms were obtained using a pulsed‐wave Doppler device. The Doppler gate was placed in nonforced end inspiration. The downward Doppler signal indicates the venous return flow (antegrade flow), the upward Doppler signal indicates the reversed blood flow to the liver from the heart (retrograde flow), and no signal indicates stasis flow. We defined the total of reversed blood flow time to the liver from the heart and stasis flow time as HV stasis flow time. In reference to calculation of RVSI, ²¹ , ⁴² we calculated HVSI, which indicates the proportion of the HV statis flow time to the HV waveform cycle time, as follows: HV stasis flow time/HV waveform cycle time. In patients with sinus rhythm, we selected the most stable value among 5 HV waveform cycles. In patients with AF, an index beat (the beat following 2 preceding cardiac cycles of equal duration) was used for each measurement. When the venous return flow shows continuous flow, the cardiac cycle time was used as the HV waveform cycle time. Next, the maximal flow velocity (V), ventricular systolic (S), ventricular diastolic (D), and atrial reversal (A) waves were measured. We classified HV waveforms into 3 groups based on previous studies by the shape and position of the V wave: those in whom the continuous flow pattern or V wave was ambiguous (C group), those in whom the V wave ran under the baseline (U group), and those who had a reversed V wave (R group). ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ Figure 2 shows representative HVSI along with HV waveform groups. Figure 2A, HVSI=0 with C group; Figure 2B, HVSI=0 with C group; Figure 2C, low HVSI with C group; Figure 2D, HVSI=0 with U group; Figure 2E, low HVSI with U group; Figure 2F, high HVSI with R group; and Figure 2G, high HVSI with R group. Although the R group absolutely shows HVSI>0, the C and U groups sometimes show HVSI=0, and sometimes HVSI>0. We measured aorta preceliac peak systolic velocity, aorta preceliac velocity time integral, SWE, SWD, RVSI, and IRVF patterns, based on previous studies. ¹⁷ , ¹⁹ , ²⁰ , ²¹ In addition, we divided patients into 3 groups based on the median value of SWE (1.36): the low SWE group (SWE ≤1.36, n=246) and the high SWE group (SWE >1.36, n=250). We also divided patients into 3 groups based on the median value of SWD (10.0): the low SWD group (SWD ≤10.0, n=189) and the high SWD group (SWD >10.0, n=177).

Figure 2. Abdominal ultrasonography.

All abdominal ultrasonography studies were performed in the right HV. HV waveforms were obtained using a pulsed‐wave Doppler device. The downward Doppler signal indicates the venous return flow (antegrade flow), the upward Doppler signal indicates the reversed blood flow to the liver from the heart (retrograde flow), and no signal indicates stasis flow. We defined the total of reversed blood flow time to the liver from the heart and stasis flow time as HV stasis flow time. We calculated HVSI, which indicates the proportion of the HV statis flow time to the HV waveform cycle time, as follows: HV stasis flow time/HV waveform cycle time. Patients with an HVSI of 0 were defined as the HVSI 0 group (HVSI=0). Patients with HVSI above 0 were divided into 2 groups based on the median value of HVSI (0.20): the low HVSI group (HVSI 0.01–0.20) and the high HVSI group (HVSI>0.20). Next, the maximal flow velocity (V), ventricular systolic (S), ventricular diastolic (D), and atrial reversal (A) waves were measured. We classified HV waveforms into 3 groups based on previous studies by the shape and position of the V wave: those in whom the continuous flow pattern or V wave was ambiguous (C group), those in whom the V wave ran under the baseline (U group), and those who had a reversed V wave (R group). HVSI increased with increasing severity of congestion. This figure shows representative HVSI and HV waveform groups in patients with HF. A, HVSI=0 with C group. B, HVSI=0 with C group. C, Low HVSI with C group. D, HVSI=0 with U group. E, Low HVSI with U group. F, High HVSI with R group. G, High HVSI with R group. HV indicates hepatic vein; and HVSI, hepatic venous stasis index.

Echocardiography

Echocardiography was performed by experienced echocardiographers based on current recommendations. ⁴⁷ The echocardiographic parameters included left ventricular ejection fraction (LVEF), left ventricular outflow tract velocity time integral, left atrial volume, early trans mitral flow velocity to mitral annular velocity ratio, right atrium area, right ventricular areas, right ventricular fractional area change, inferior vena cava diameter, severity of tricuspid regurgitation, tricuspid regurgitation pressure gradient, tricuspid annular plane systolic excursion, and tissue Doppler‐derived tricuspid lateral annular systolic velocity. ⁴⁸ , ⁴⁹ The LVEF was calculated using Simpson's method. The right ventricular fractional area change, defined as [right ventricular end diastolic area−right ventricular end systolic area]/right ventricular end diastolic area×100, is a measure of right ventricular systolic function. ⁵⁰ All measurements were performed using ultrasound systems (ACUSON Sequoia, Siemens Medical Solutions USA, Inc., Mountain View, CA).

Right Heart Catheterizations and Hemodynamics

Of all 513 patients, 310 underwent RHC based on remedial judgment of the attending physician. All RHCs were performed under fluoroscopic guidance with patients in the resting supine position breathing room air at rest. Mean right atrial pressure (RAP), mean pulmonary artery pressure, mean pulmonary artery wedge pressure, cardiac index, and pulmonary vascular resistance (PVR) were measured using a 6F Swan Ganz catheter (Edwards Lifesciences, Irvine, CA). CI was calculated based on the thermodilution method. ⁵¹

Measurement of Laboratory Data

We measured BNP (B‐type natriuretic peptide) levels using a specific immunoradiometric assay (Shionoria BNP kit, Shionogi, Osaka, Japan). This assay was performed by experienced laboratory technicians who were blinded to characteristics of patients.

Statistical Analysis

Normally distributed data were presented as mean±SD, and nonnormally distributed data were presented as median and interquartile range (ie, median [25th percentile–75th percentile]). Categorical variables were expressed as numbers and percentages, and the chi‐square test was used for their comparisons. Among the 3 groups, parametric variables were compared using 1‐way ANOVA, and nonparametric variables were compared using the Kruskal–Wallis test. Kaplan–Meier analysis with log‐rank test was used to evaluate the cardiac event rate. HVSI was assessed as a predictor of postdischarge cardiac events by the Cox proportional hazard analysis. In addition, the statistical tests based on the Schoenfeld residuals which can assess the correlation between scaled residuals and time, was also conducted for a proportional hazard assumption. In the Cox proportional hazard analysis, to prepare for potential confounding, we considered the following clinical factors: age, sex, body mass index and confounding factors that were generally recognized as poor prognostic factors (ie, diabetes, CKD, AF, anemia, LVEF, and BNP). Univariate factors with P<0.05 were entered into the multivariable Cox proportional hazard analyses. We fit a Cox regression model with the calculated AHEAD score as the only independent variable, and we calculated the c‐index, which is regarded as a measure for model discrimination equivalent to the area under the curve for binary dependent variables. Receiver operating characteristic curve was computed by using the AHEAD score as the reference variable and the cardiac events as the binary reference variable. The modified AHEAD score was created by adding HVSI as a covariate and incorporated along with the AHEAD score as independent variables and cardiac events as the dependent variable. Model improvement was assessed by the regression method. A value of P<0.05 was defined as statistically significant for all comparisons. All statistical analyses were performed using Statistical Package for Social Science version 27.0 (IBM, Armonk, NY) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Table 1 shows the clinical and demographic characteristics of the study population. Of the 513 patients included in the study, the median age was 73.0 (63.0–81.0) years, and 289 (56.3%) were male. The breakdown of angiotensin‐converting enzyme inhibitor/angiotensin receptor blocker/angiotensin receptor neprilysin inhibitor was 203 (39.6%) with angiotensin‐converting enzyme inhibitor, 138 (26.9%) with angiotensin receptor blocker, and 0 (0%) with angiotensin receptor neprilysin inhibitor. During the study period, angiotensin receptor neprilysin inhibitor had not been approved in Japan. The percentage of patients who show New York Heart Association class III or IV, the prevalence of AF, and usage of mineralocorticoid receptor antagonist and diuretic increased in proportion to HVSI. In contrast, age, sex, body mass index, pathogenesis, and the prevalence of hypertension, dyslipidemia, diabetes, CKD, anemia, and other medication did not differ among the groups.

Table 1.

Clinical and Demographic Characteristics of the Study Population

	Total (n=513)	HVSI=0 (HVSI=0, n=253)	Low HVSI (HVSI 0.01–0.20, n=132)	High HVSI (HVSI>0.20, n=128)	P value*
Demographics
Age, y	73.0 (63.0–81.0)	72.0 (62.0–80.0)	76.0 (68.0–82.0)	71.0 (59.0–79.0)	0.007
Male sex, n (%)	289 (56.3)	148 (58.5)	71 (53.8)	70 (54.7)	0.616
Body mass index, kg/m2	22.6 (20.3–25.4)	22.8 (20.6–25.6)	22.2 (19.9–24.8)	22.7 (20.2–25.1)	0.265
Systolic blood pressure, mm Hg	118.0 (106.0–131.0)	120.0 (107.0–131.0)	122.0 (109.0–137.0)	110.0 (100.3–125.8)	<0.001
Diastolic blood pressure, mm Hg	66.0 (59.0–75.0)	66.0 (59.5–76.0)	65.0 (58.0–75.0)	67.0 (58.0–74.0)	0.570
Heart rate, bpm	69.0 (61.0–81.0)	70.0 (61.0–82.0)	68.0 (60.0–77.0)	70.0 (62.0–84.8)	0.122
New York Heart Association class III or IV, n (%)	167 (33.9)	68 (28.6)	41 (31.8)	58 (46.0)	0.003
Etiology, n (%)
ischemic/myopathy/valvular/arrhythmia/pulmonary/congenital/others	101 (19.7)/126 (24.6)/178 (34.7)/48 (9.4)/41 (8.0)/8 (1.6)/11 (2.1)	56 (22.1)/57 (22.5)/90 (35.6)/21 (8.3)/21 (8.3)/4 (1.6)/4 (1.6)	28 (21.2)/35 (26.5)/48/(36.4)/8 (6.1)/7 (5.3)/2 (1.5)/4 (3.0)	17 (13.3)/34 (26.6)/40 (31.3)/19 (14.8)/13 (10.2)/2 (1.6)/3 (2.3)	0.311
Comorbidities
Atrial fibrillation, n (%)	181 (35.3)	75 (29.6)	44 (33.3)	62 (48.4)	0.001
Hypertension, n (%)	333 (64.9)	169 (66.8)	88 (66.7)	76 (59.4)	0.317
Dyslipidemia, n (%)	333 (64.9)	173 (68.4)	88 (66.7)	72 (56.3)	0.057
Diabetes, n (%)	186 (36.3)	96 (37.9)	57 (43.2)	33 (25.8)	0.057
Chronic kidney disease, n (%)	338 (65.9)	162 (64.0)	86 (65.2)	90 (70.3)	0.464
Anemia, n (%)	236 (46.0)	108 (42.7)	63 (47.7)	65 (50.8)	0.293
Medications
β‐blocker, n (%)	351 (68.4)	178 (70.4)	85 (64.4)	88 (68.8)	0.488
ACE‐I/ARB/ARNI, n (%)	320 (62.4)	166 (65.6)	74 (56.1)	80 (62.5)	0.185
ACE‐I, n (%)	203 (39.6)	103 (40.7)	48 (36.4)	52 (40.6)	0.682
ARB, n (%)	138 (26.9)	75 (29.6)	28 (21.2)	35 (27.3)	0.207
ARNI, n (%)	0 (0)	0 (0)	0 (0)	0 (0)	‐
Mineralocorticoid receptor antagonist, n (%)	181 (35.3)	70 (27.7)	45 (34.1)	66 (51.6)	<0.001
Sodium glucose cotransporter 2 inhibitor, n (%)	35 (6.8)	17 (6.7)	11 (8.3)	7 (5.5)	0.655
Diuretic, n (%)	333 (64.9)	146 (57.7)	84 (63.6)	103 (80.5)	0.001

Data are presented as mean±SD, median (interquartile range), or number (percentage). ACE‐I indicates angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor; and HVSI, hepatic venous stasis index.

^*A P value indicates statistically significance in comparison across all groups.

Regarding laboratory data (Table 2), levels of total bilirubin, lactate dehydrogenase, alkaline phosphatase, 7S domain of collagen type IV, and BNP became higher, and levels of albumin, cholinesterase, and white blood cell count became lower with increasing HVSI. However, aspartate aminotransferase, alanine aminotransferase, C‐reactive protein, blood urea nitrogen, creatinine, sodium, and hemoglobin did not differ among the groups. Regarding the parameters of echocardiography, there was a significant increase in left atrial volume, inferior vena cava diameter, tricuspid regurgitation pressure gradient, and severity of tricuspid regurgitation, and a significant decrease in tricuspid annular plane systolic excursion with increasing HVSI. However, LVEF did not differ among the groups. Cardiopulmonary hemodynamics evaluated by RHC worsened with increasing HVSI. There was a gradual increase in mean RAP, mean pulmonary artery pressure, pulmonary artery wedge pressure, and PVR with increasing HVSI, whereas there was no significant change in cardiac index. Regarding abdominal ultrasonography, severity of HV waveform groups, RVSI, IRVF patterns, SWE, and SWD became higher with increasing HVSI. However, aorta preceliac peak systolic velocity and aorta preceliac velocity time integral did not differ among the groups. These results suggest that elevated HVSI indicates mainly hepatic congestion and right‐sided overload rather than liver hypoperfusion.

Table 2.

Laboratory Data of the Study Population

	Total (n=513)	HVSI=0 (HVSI=0, n=253)	Low HVSI (HVSI 0.01–0.20, n=32)	High HVSI (HVSI>0.20, n=128)	P value*
Laboratory data
Total protein, g/dL	6.8±0.7	6.9±0.8	6.8±0.6	6.7±0.7	0.377
Albumin, g/dL	3.8 (3.4–4.2)	3.9 (3.5–4.3)	3.8 (3.5–4.1)	3.7 (3.3–4.1)	0.003
Total bilirubin, mg/dL	0.8 (0.6–1.0)	0.7 (0.6–1.0)	0.7 (0.5–1.0)	0.9 (0.7–1.2)	<0.001
Direct bilirubin, mg/dL	0.1 (0.1–0.1)	0.1 (0.1–0.1)	0.1 (0.1–0.1)	0.1 (0.1–0.2)	<0.001
Aspartate aminotransferase, U/L	23.0 (18.0–29.0)	23.0 (18.0–28.0)	22.0 (19.0–29.0)	23.5 (19.0–30.0)	0.321
Alanine aminotransferase, U/L	18.0 (12.0–27.0)	19.0 (13.0–29.8)	16.0 (12.0–24.0)	17.0 (13.3–26.0)	0.094
Lactate dehydrogenase, U/L	217.0 (188.0–260.0)	213.5 (182.0–250.8)	222.0 (186.0–266.0)	224.0 (200.0–267.8)	0.039
Alkaline phosphatase, U/L	239.0 (188.0–295.0)	231.0 (183.5–276.5)	239.0 (185.5–303.0)	262.0 (199.0–341.0)	0.025
Gamma‐glutamyl transferase, U/L	35.0 (20.0–70.0)	32.0 (19.0–62.0)	30.5 (18.0–59.8)	50.0 (27.0–92.0)	<0.001
Cholinesterase, U/L	250.0 (197.5–306.0)	266.0 (214.0–324.0)	240.5 (188.0–289.0)	234.0 (173.0–282.0)	<0.001
7S domain of collagen type IV, ng/mL	5.5 (4.5–6.9)	5.2 (4.3–6.5)	5.2 (4.5–6.4)	6.7 (5.5–8.5)	<0.001
C‐reactive protein, mg/dL	0.24 (0.09–0.98)	0.22 (0.08–0.86)	0.22 (0.08–1.19)	0.33 (0.10–1.22)	0.373
B‐type natriuretic peptide, pg/mL	183.4 (78.7–374.7)	129.1 (61.3–280.9)	217.8 (111.8–420.8)	223.1 (117.9–494.8)	<0.001
Blood urea nitrogen, mg/dL	19.0 (15.0–25.8)	19.0 (15.0–25.0)	20.0 (16.0–26.0)	20.0 (15.0–25.8)	0.740
Creatinine, mg/dL	1.00 (0.82–1.21)	0.97 (0.82–1.19)	0.99 (0.79–1.24)	1.04 (0.83–1.21)	0.684
Estimated glomerular filtration rate, mL/min per 1.73 cm2	52.0 (40.3–64.0)	53.0 (41.0–65.0)	52.0 (37.0–63.0)	50.0 (39.3–62.0)	0.392
Sodium, mEq/L	140.0 (138.0–141.0)	140.0 (138.0–141.0)	140.0 (139.0–141.0)	140.0 (138.0–142.0)	0.231
White blood cell, 103/μL	5.5 (4.6–6.7)	5.7 (4.8–6.9)	5.4 (4.4–6.4)	5.3 (4.5–6.6)	0.047
Hemoglobin, g/dL	12.8 (11.2–14.4)	12.9 (11.4–14.6)	12.5 (10.7–14.1)	12.6 (11.1–14.5)	0.051
Platelet, 103/μL	193.0 (153.0–248.0)	188.5 (152.0–243.0)	191.0 (154.0–245.0)	207.5 (154.3–257.3)	0.566
Echocardiography
Left ventricle ejection fraction, %	53.0 (34.7–62.0)	55.0 (36.0–63.1)	52.0 (35.0–62.0)	50.0 (30.0–61.0)	0.145
Left ventricular outflow tract velocity‐time integral, cm	16.0 (12.6–20.0)	16.5 (13.2–19.9)	16.3 (13.0–20.5)	14.5 (11.0–19.4)	0.015
Left atrial volume, mL	88.0 (66.0–119.7)	82.0 (58.5–105.5)	86.0 (67.0–117.2)	109.5 (79.3–148.0)	<0.001
Early trans mitral flow velocity to mitral annular velocity ratio	13.8 (9.7–19.1)	12.6 (9.1–16.6)	15.4 (10.9–20.1)	14.4 (10.8–21.6)	0.001
Right atrial end systolic area, cm2	18.6 (14.0–24.0)	18.0 (14.0–23.0)	15.0 (12.7–23.0)	21.0 (17.1–27.5)	<0.001
Right ventricle area diastole, cm2	19.4 (14.8–25.2)	19.4 (15.4–26.0)	15.6 (12.0–26.2)	20.6 (16.8–24.7)	0.073
Right ventricle area systole, cm2	11.2 (9.0–16.8)	11.9 (9.5–17.6)	9.2 (7.3–13.8)	13.5 (10.5–17.9)	0.023
Right ventricular fractional area change, %	38.0 (32.0–44.0)	38.0 (32.0–44.0)	38.5 (31.6–46.0)	32.0 (23.5–39.0)	<0.001
Inferior vena cava diameter, mm	15.0 (13.0–18.1)	14.0 (12.0–17.0)	15.0 (12.3–18.0)	18.0 (14.7–21.0)	<0.001
Tricuspid regurgitation pressure gradient, mm Hg	24.0 (19.0–32.0)	23.0 (18.8–31.0)	24.0 (19.0–31.0)	28.0 (20.5–36.5)	0.005
Tricuspid annular plane systolic excursion, mm	17.2 (14.4–20.0)	17.6 (14.8–20.3)	17.2 (14.7–20.2)	16.1 (12.1–19.5)	0.009
Tissue Doppler‐derived tricuspid lateral annular systolic velocity (tricuspid valve S′), cm/s	9.8 (7.9–11.3)	9.9 (8.0–11.3)	10.2 (8.7–11.5)	9.0 (7.5–10.5)	0.081
Tricuspid insufficiency, n (%)
None‐trivial/mild/moderate/severe	345 (67.3)/105 (20.5)/49 (9.6)/14 (2.7)	176 (69.6)/52 (20.6)/24 (9.5)/1 (0.4)	95 (72.0)/25 (18.9)/11 (8.3)/1 (0.8)	74 (57.8)/28 (21.9)/14 (10.9)/12 (9.4)	<0.001
Right heart catheterization (n=310)
Mean right atrium pressure, mm Hg	6.0 (4.0–9.3)	6.0 (3.0–8.0)	6.0 (4.0–9.0)	8.0 (5.0–11.0)	<0.001
Mean pulmonary artery pressure, mm Hg	22.0 (17.0–31.0)	20.0 (15.0–27.0)	21.0 (17.0–30.0)	25.0 (19.8–35.3)	<0.001
Mean pulmonary artery wedge pressure, mm Hg	13.0 (9.0–19.0)	12.0 (7.0–17.0)	14.0 (9.0–21.0)	16.0 (12.0–21.0)	<0.001
Cardiac index, L/min per m2	2.4 (2.0–2.8)	2.4 (2.1–2.9)	2.4 (2.1–2.8)	2.2 (2.0–2.6)	0.112
Pulmonary vascular resistance (WoodU)	2.0 (1.3–3.0)	1.9 (1.3–2.8)	1.9 (1.3–2.8)	2.5 (1.8–4.6)	0.005
Abdominal ultrasonography
Aorta preceliac peak systolic velocity, cm/s	60.4 (48.3–73.8)	59.7 (48.9–74.5)	60.7 (46.5–73.7)	60.9 (48.7–73.7)	0.804
Aorta preceliac velocity time integral, cm	19.2 (15.5–23.4)	19.2 (16.0–23.6)	19.8 (15.6–23.4)	18.3 (14.5–23.3)	0.354
Hepatic vein waveform groups
C/U/R group, n (%)	256 (49.9)/91 (17.7)/166 (32.4)	239 (94.5)/14 (5.5)/0 (0)	7 (5.3)/64 (48.5)/61 (46.2)	10 (7.8)/13 (10.2)/105 (82.0)	<0.001
Renal venous stasis index	0.09 (0.00–0.23)	0.00 (0.00–0.00)	0.00 (0.08–0.13)	0.24 (0.11–0.33)	<0.001
Intrarenal venous flow patterns, n (%)
Continuous/biphasic/monophasic pattern/others	339 (67.9)/97 (19.4)/61 (12.2)/2 (0.4)	210 (86.1)/20 (8.2)/14 (5.7)/0 (0.0)	85 (65.4)/29 (22.3)/16 (12.3)/0 (0.0)	44 (35.2)/48 (38.4)/31 (24.8)/2 (1.6)	<0.001
Shear wave elastography, m/s	1.36 (1.24–1.50)	1.33 (1.22–1.46)	1.33 (1.24–1.44)	1.46 (1.31–1.66)	<0.001
Shear wave dispersion, [m/s]/kHz	10.0 (8.90–12.0)	9.8 (8.8–11.0)	9.8 (8.5–11.9)	11.3 (9.5–14.3)	<0.001

Data are presented as mean±SD, or median (interquartile range), or number (percentage). HVSI indictes hepatic venous statis index.

^*A P value indicates statistically significance in comparison across all groups.

During the follow‐up period (median 517 days; range 1 to 1594 days, interquartile range 140.5 to 931.5 days), death from any cause occurred in 51 patients (30 cardiac deaths and 21 noncardiac deaths). We observed 87 cardiac events (29 cardiac deaths and 58 worsening HF). The 29 cardiac deaths included 26 deaths from HF and 3 deaths from ventricular fibrillation. There were 15 patients who experienced cardiac death and 4 patients who experienced noncardiac death after the first event of worsening HF. In the Kaplan–Meier analysis for cardiac event rate stratified by HVSI, the cumulative cardiac event rate significantly increased with increasing HVSI (Figure 3; log‐rank, P=0.002). In the Cox proportional hazard analysis (Table 3), due to the limited number of cardiac events (87 events) and to avoid overfitting, we selected the following factors: age, sex, body mass index, and confounding factors that were generally recognized as poor prognostic factors (ie, diabetes, CKD, AF, anemia, LVEF, and BNP). In the univariate Cox proportional hazard analysis, high HVSI was associated with high cardiac event rate (Table 3; low HVSI group versus HVSI 0 group: hazard ratio [HR], 1.739 [95% CI, 1.014–2.984], P=0.045, high HVSI group versus HVSI 0 group: HR, 2.335 [95% CI, 1.429–3.814], P<0.001). In the multivariable Cox proportional hazard analyses, we selected univariate factors with P<0.05 (ie, age, body mass index, diabetes, CKD, anemia, LVEF, and BNP). After adjusting for these confounding factors, high HVSI was an independent prognostic factor (Table 3; high HVSI group versus HVSI 0 group: HR, 2.037 [95% CI, 1.209–3.434], P=0.008). In the subgroup analysis regarding LVEF, there was no significant interaction between prognostic impact of HVSI and LVEF (P=0.340). In addition, in the Kaplan–Meier analysis for cardiac event rate stratified by HV waveform groups, the cumulative cardiac event rate was lowest in the C group (Figure 4; log‐rank, P=0.012). Furthermore, in the Kaplan–Meier analysis for cardiac event rate stratified by SWD, the cumulative cardiac event rate was significantly higher in the high SWD group than in the low SWD group (log‐rank, P=0.002). In addition, in the Kaplan–Meier analysis for cardiac event rate stratified by SWE, the cumulative cardiac event rate tended to be higher in the high SWE group than in the low SWE group (log‐rank, P=0.060).

Figure 3. Kaplan–Meier analysis for cardiac event rate stratified by HVSI.

The cumulative cardiac event rate significantly increased with worsening HVSI (log‐rank, P=0.002). HVSI indicates hepatic venous stasis index.

Table 3.

Cox Proportional Hazard Model for Predicting Cardiac Events

Cardiac events	Univariate			Multivariable*
	Hazard ratio	95% CI	P value	Hazard ratio	95% CI	P value
Age	1.024	1.006–1.042	0.008	1.018	0.996–1.040	0.109
Sex (male=1)	1.153	0.749–1.773	0.519
Body mass index	0.917	0.874–0.962	<0.001	0.945	0.896–0.996	0.034
Diabetes	2.045	1.343–3.115	<0.001	2.004	1.290–3.113	0.002
Chronic kidney disease	2.476	1.473–4.164	<0.001	1.515	0.873–2.631	0.140
Atrial fibrillation	1.259	0.820–1.933	0.293
Anemia	1.780	1.164–2.721	0.008	1.368	0.865–2.165	0.180
Left ventricle ejection fraction	0.980	0.968–0.992	<0.001	0.984	0.971–0.998	0.025
B‐type natriuretic peptide	1.001	1.001–1.002	<0.001	1.001	1.001–1.001	<0.001
HVSI
Low HVSI vs HVSI 0	1.739	1.014–2.984	0.045	1.375	0.794–2.381	0.255
High HVSI vs HVSI 0	2.335	1.429–3.814	<0.001	2.037	1.209–3.434	0.008

HVSI indicates hepatic venous stasis index.

^*Adjusted for age, body mass index, diabetes, chronic kidney disease, anemia, left ventricle ejection fraction, and B‐type natriuretic peptide.

Figure 4. Kaplan–Meier analysis for cardiac event rate stratified by HV waveform groups.

The cumulative cardiac event rate was lowest in the C group among the groups (log‐rank, P=0.012). HV indicates hepatic vein.

Subsequently, we used AHEAD score to test predictive ability of HVSI. In this study, the original AHEAD score demonstrated modest discrimination (c‐index=0.63 [95% CI, 0.57–0.70]). When HVSI was added to AHEAD score (Figure 5), the model discrimination was significantly improved (c‐index=0.67 [95% CI, 0.60–0.73], P=0.034).

Figure 5. Comparison of the ROC curves between the original AHEAD score and the modified AHEAD score (added HVSI).

The AUCs for the original AHEAD score and the modified AHEAD score were 0.63 (95% CI, 0.57–0.70) and 0.67 (95% CI, 0.60–0.73), respectively. The AHEAD score is based on the following comorbidities: A–atrial fibrillation, H–hemoglobin levels of <12.0 g/dL in women and <13.0 g/dL in men (anemia), E–elderly (age >70 years), A–abnormal renal parameters (creatinine >130 μmol/L), D, diabetes. AUC indicates area under the curve; HVSI, hepatic venous stasis index; and ROC, receiver operating characteristic.

DISCUSSION

In the present study, we investigated the associations of HVSI assessed by abdominal ultrasonography with laboratory tests, echocardiography, and RHC, as well as its prognostic impact in patients with HF. The present study is, to the best of our knowledge, the first to report that HVSI is a quantitative marker of hepatic congestion and associated with liver dysfunction, right‐sided overload, and higher cardiac events in patients with HF.

Cardio‐Hepatic Interaction

As it receives up to 25% of total cardiac output, the liver is very sensitive to hemodynamic modulation and heart dysfunction. ⁷ Cardio‐hepatic interaction such as congestive hepatopathy (passive congestion secondary to increased CVP) and hypoxic hepatitis (reduced arterial perfusion) in patients with HF has been reported. ¹ , ⁷ , ⁸ , ¹⁰ , ¹¹ , ¹² , ¹³ Passive congestion prevails in congestive hepatopathy secondary to chronic HF, whereas arterial hypoperfusion predominates in acute HF leading to hypoxic hepatitis. However, these forward and backward factors often coexist and potentiate each other. ¹² The starting point of congestive hepatopathy is an increase in the CVP secondary to right HF with a high level of filling pressure. Due to the absence of valves in HV, increased CVP affects the sinusoidal bed causing centrilobular congestion, sinusoidal dilation, and perivenular fibrosis. ¹ , ⁷ , ¹⁰ , ¹³ , ⁵² , ⁵³ , ⁵⁴ Sinusoidal dilation is positively correlated with the degree of elevation of RAP and explains the hepatomegaly frequently observed in patients with congestive hepatopathies. ⁵⁵ The major damage occurs in the Rappaport acinus, which surrounds the venae centralis hepatis. ¹² , ¹³ Centrilobular necrosis can extend to peripheral areas if HF persists and worsens, and causes deposition and spread of connective tissue bridging one venae centralis hepatis to the other, ultimately leading to cirrhosis. ⁷ , ⁵⁶ Hypoxic hepatitis is a clinical and histological syndrome leading to the reduction of hepatic arterial perfusion due to acute fall in cardiac output. ⁵² , ⁵⁴ Reduced hepatic arterial perfusion can cause hypoxic hepatopathy, and hypoxia can cause centrilobular necrosis and sinusoidal damage in the liver, leading to elevation of transaminase. ⁵⁴ Concerning heart disease that be caused by a liver disease, cirrhotic cardiomyopathy is present. ⁵⁷ It is a syndrome characterized by systolic dysfunction, impaired diastolic relaxation, prolonged isovolumetric relaxation time, and electrophysiological disturbances such as prolonged QTc interval. ⁵⁸

Hepatic Vein Waveforms

HV waveforms are reported to be useful for the assessment of increased CVP and prognosis of cardiac event rate in patients with HF. ²² The downward Doppler signal indicates the venous return flow (antegrade flow), the upward Doppler signal indicates the reversed blood flow to the liver from the heart (retrograde flow), and no signal indicates stasis flow. HV waveform is primarily composed of 2 distinct forward systolic and diastolic waves (S wave and D wave) during ventricular systole and early diastole, and 2 small reverse flow during late ventricular systole and atrial systole (V wave and A wave). ⁴³ , ⁴⁴ , ⁴⁶ , ⁵⁹ , ⁶⁰ However, the V wave is transitional, which may be antegrade, retrograde, or neutral. ⁶¹ Along with the increasing of RAP, the flow reverses to the vena cava in the right atrial systole. In the HV waveform, it shows as a retrograde A wave (Figure 2E). After the end systole, as the ventricular contraction intensity decreases and the closed tricuspid valve begins to return to its original resting position, the right atrium fills, venous return velocity decreases, and temporary equilibrium is reached, making V waves. ⁴³ Therefore, the V wave corresponds to right atrial overfilling. In patients with high right ventricular end diastolic pressure, the right atrium must contract with greater force to drive blood into the right ventricle. This results in increased RAP, leading to an augmentation of the retrograde blood flow into the vena cava (Figure 2F). If the A, S, and V waves are all retrograde, they may fuse into a single retrograde wave and become biphasic waveforms, alternating with the D wave (Figure 2G). ⁴⁴

With regard to the HV waveforms, it has been suggested that the significance of HV waveforms differ depending on the disease. ⁴³ , ⁶² In patients with liver cirrhosis, the continuous HV waveform (namely Figure 2A with HVSI=0 in this study) is mainly caused by inflammatory or fibrotic changes, intrahepatic fat deposition, and changes in the compliance of the venous wall, suggesting the presence of severe liver fibrosis. ⁶³ , ⁶⁴ , ⁶⁵ Because we excluded the patients with distinct liver disease (including liver cirrhosis), we were able to measure the change of HV waveform due to not liver disease but HF, and observe associations between HVSI and RAP. Therefore, the results of this study may have been in contrast to those of previous studies on HV waveform in patients with liver cirrhosis.

The Utility of HVSI

We here report the utility of HVSI for estimating hepatic congestion, liver dysfunction, RAP, and prognosis. First, the classification of HV waveform groups has a weakness; it may be difficult to classify some patterns. HVSI can complement the weaknesses by quantifying HV waveforms. We consider that quantifying HVSI can detect hepatic congestion more sensitively than classification of HV waveform groups. Second, the measurement of SWE and SWD requires specific abdominal ultrasonographic equipment and is not easy for cardiologists. HVSI needs neither special equipment nor complex measurements, cardiologists can measure HVSI through training a few times. Third, according to the past study, RVSI and IRVF flow patterns are useful as a prognostic indicator for HF. ¹⁷ , ¹⁸ , ²¹ , ²³ , ²⁴ However, they differ in that HVSI reflects hepatic congestion and RVSI and IRVF flow patterns reflect renal congestion. In terms of assessing organ congestion, there is value in measuring both HVSI and RVSI or IRVF flow patterns. RVSI and IRVF flow patterns have to detect both intrarenal arterial flow and intrarenal venous flow at the same time, which is often difficult. HVSI has only to detect HV waveform. The size of hepatic vein is 6 to 7 mm and enough to be easily measured. In addition, HV is closer in anatomic proximity to the right heart than intrarenal vessels. In the present study, the measurement of HVSI seemed to be superior to those of RVSI and IRVF patterns, in terms of ease of measurement and accuracy. As shown in Figure 4, regarding comparison of HVSI with HV waveform groups for predicting prognosis, Kaplan–Meier analysis of HV waveform groups showed overlap and inversion between the U group and R group. Our data suggest that the HVSI can be superior to HV waveform groups. Patients with HF and high HVSI despite treatment can have remaining hepatic congestion and lead to adverse prognosis. Thus, it appears that cardioprotective agents or decreasing central venous pressure by diuretic agents may be useful to decrease HVSI and improve prognosis in these patients.

Study Limitations

There are some limitations in the present study. First, the study might be underpowered because it was a single center prospective cohort study with a relatively small sample size and a short follow‐up period. Second, although documented liver disease was excluded in this study, we could not fully exclude other liver diseases that may have affected the HVSI. Third, we used only data during hospitalization and did not consider changes in medical parameters or treatments after discharge. Future studies need to assess alterations of HVSI in response to hemodynamic changes because the hemodynamics of patients with HF change dynamically. Fourth, we could not perform abdominal ultrasonography in all hospitalized patients. In addition, there might be a potential selection bias since the attending physicians made decisions to perform RHC. Therefore, the present results need to be viewed as preliminary, and further studies with a larger number of patients are needed.

CONCLUSIONS

HVSI assessed by abdominal ultrasonography reflects hepatic congestion and right‐sided HF, and is associated with adverse prognosis in patients with HF.

Sources of Funding

This study was supported in part by a grant‐in‐aid for Scientific Research (20K07828 and 20K16529) from the Japan Society for the Promotion of Science.

Disclosures

None.