Cardiac ascites is frequently diagnosed, but there is a paucity of data regarding the predictors for its formation. In a group of patients with heart failure referred for orthotopic heart transplantation and ventricular assist device (VAD) placement, we attempted to identify patient characteristics and predictors associated with the development of ascites. Long-term outcomes of patients with and without ascites were examined. Patients were divided into 2 groups based on the presence or absence of significant ascites on imaging. Demographic information, laboratory values, and results of transthoracic echocardiograms and right-sided cardiac catheterizations were compared between the groups. Of the 196 patients, 29 patients (15%) had significant ascites. The group with significant ascites had higher mean creatinine (2.3 vs 1.6 mg/dl, p = 0.03). On transthoracic echocardiograms, the group with significant ascites had more severe right ventricular dilation (p = 0.03) and tricuspid valve regurgitation (p <0.01). On right-sided cardiac catheterizations, the group with significant ascites had higher mean right atrial (RA) pressure (17 vs 13 mm Hg, p = 0.01). There was no difference in pulmonary capillary wedge pressure between the groups (22 vs 23 mm Hg, p = 0.57). No threshold value of RA pressure was identified for the development of significant ascites. The presence of significant ascites was associated with decreased overall survival (p <0.01). In conclusion, impaired renal function and elevated right-sided cardiac pressures were more commonly seen in the group with significant ascites. No minimum RA pressure elevation was required for significant ascites formation. The presence of significant ascites was correlated with higher mortality.
Ascites is the most common complication of cirrhosis and develops in 58% of patients within 10 years of initial cirrhosis diagnosis. Low oncotic pressure from hypoalbuminemia and elevated portal pressures from cirrhosis contribute to the formation of ascites in the peritoneal space. It has been suggested that a portal pressure >10 mm Hg (normal <5 mm Hg) is needed to produce ascites from cirrhosis. Approximately 5% of portal hypertensive ascites is thought to be secondary to posthepatic pressure elevation from the heart. Heart failure (HF) can result in elevated pressures in the right atrium, inferior vena cava, and hepatic veins. This pressure is transmitted to the liver, causing portal hypertension, and theoretically, cardiac ascites. Although there is a high prevalence of heart disease and HF with resultant posthepatic portal hypertension, only a small subset of patients with HF and portal pressures >10 mm Hg present with significant ascites. The cause of heart disease, level of right atrial (RA) pressure, presence of tricuspid valve regurgitation (TVR), concurrent renal disease, electrolyte levels, and albumin levels may contribute to cardiac ascites formation. We sought to examine whether there are other risk factors for the development of ascites in addition to elevated pressures in the right heart.
Methods
The study was approved by the Mount Sinai Hospital institutional review board. The source population for this, retrospective cohort study, included all adult patients (aged 18 years or older), evaluated for an orthotopic heart transplantation (OHT) or ventricular assist device (VAD) placement at the Mount Sinai Hospital from January 1, 2010, to August 31, 2013. Patients with advanced HF were referred to our program for consideration of OHT or destination therapy VAD. The patients were identified from the cardiology clinic database. Patients without abdominal imaging interpretation available to determine the presence of ascites were excluded from the study.
Patients were divided into 2 groups based on the (i) presence or (ii) absence of significant ascites on abdominal imaging interpretation. Significant ascites was defined as having “moderate”-to-“large” ascites on abdominal imaging. Insignificant ascites was defined as “no,” “minimal,” “mild,” or “perihepatic” ascites. Abdominal imaging techniques included computerized tomography scan, magnetic resonance imaging, and ultrasound.
Demographic information, serum laboratory values, and results of transthoracic echocardiograms (TTEs) and right-sided cardiac catheterizations (RCC) were abstracted from the electronic medical records and compared between the groups. Glomerular filtration rate was calculated using the Modification of Diet Renal Disease equation.
Data were stored in a password-protected spreadsheet on the Mount Sinai server. No patient identifiers were included in the spreadsheet. A separate database with subject number and patient identifiers (name, date of birth) was maintained to protect the privacy of our patients.
Categorical variables were compared using the chi-square or Fisher’s test, where appropriate. Continuous variables were compared using the Student’s t test (if normality assumption was valid) or Wilcoxon rank-sum test (if normality assumption was invalid). Kaplan-Meier survival analysis was used to evaluate survival in the 2 groups. Statistical significance was defined as p value <0.05.
Results
Of 225 patients with HF evaluated for OHT or VAD placement, 29 patients were excluded because of lack of abdominal imaging. Of the 196 study patients, 29 patients (15%) had significant ascites (20 with “moderate” ascites and 9 with “large” ascites). Of these patients, all 29 patients had TTE pressure measurements available, and 26 patients had RCC pressure measurements available. Of the 167 patients without significant ascites (118 with “no” ascites and 49 with “minimal,” “mild,” or “perihepatic” ascites), 163 patients had TTE pressure measurements available and 150 patients had RCC pressure measurements available ( Figure 1 ).
Demographic information of the groups with and without significant ascites is listed in Table 1 . There were no significant differences in age, gender, ethnicity, cause of heart disease, or image presence of splenomegaly 2 groups. The mean body mass index was noted to be higher in the group without significant ascites.
| Variable | Significant Ascites | p -value | |
|---|---|---|---|
| No (N=167) | Yes (N=29) | ||
| Mean age at imaging (years ± SD) | 56 ± 12 | 55 ± 15 | 0.73 |
| Male | 124 (74%) | 24 (83%) | 0.33 |
| Mean body mass index (kg/m 2 ± SD) | 28 ± 6 | 25 ± 6 | 0.01 |
| Ethnicity | 0.14 | ||
| Asian | 7 (4%) | 2 (7%) | |
| Black | 36 (22%) | 6 (20%) | |
| Caucasian | 49 (29%) | 10 (35%) | |
| Hispanic | 32 (19%) | 1 (3%) | |
| Others or unknown | 43 (26%) | 10 (35%) | |
| Cause of heart disease | 0.76 | ||
| Valvular | 19 (11%) | 7 (24%) | |
| Ischemic | 67 (40%) | 11 (38%) | |
| Idiopathic | 49 (29%) | 5 (17%) | |
| Others | 32 (19%) | 6 (21%) | |
| Presence of splenomegaly | 7 (4%) | 4 (14%) | 0.06 |
Available serum tests were compared between the 2 groups ( Table 2 ). The group with significant ascites had higher mean creatinine, higher mean blood urea nitrogen, higher mean brain natriuretic peptide, lower mean sodium, and lower mean albumin. Not enough patients had available ascitic fluid analysis to perform a statistical analysis on ascitic fluid results.
| Variable | Significant Ascites | p -value | |
|---|---|---|---|
| No (N=167) | Yes (N=29) | ||
| Mean sodium (mEq/L ± SD) | 136 ± 4 | 135 ± 6 | 0.05 |
| Mean blood urea nitrogen (mg/dL ± SD) | 33 ± 20 | 50 ± 24 | <0.01 |
| Mean creatinine (mg/dL ± SD) | 1.7 ± 1.4 | 2.3 ± 1.2 | 0.03 |
| Glomerular filtration rate (mL/min) | 0.004 | ||
| <30 | 21 (13%) | 9 (31%) | |
| 30-60 | 76 (46%) | 16(55%) | |
| >60 | 70 (42%) | 4 (14%) | |
| Mean protein (g/dL ± SD) | 6.6 ± 1.0 | 6.3 ± 0.9 | 0.11 |
| Mean albumin (g/dL ± SD) | 3.6 ± 0.7 | 3.3 ± 0.6 | 0.03 |
| Median alkaline phosphatase (U/L [IQR]) | 91 (67-126) | 124 (91-190) | <0.01 |
| Median total bilirubin (mg/dL [IQR]) | 0.9 (0.5-1.7) | 1.2 (0.9-1.9) | 0.02 |
| Median direct bilirubin (mg/dL [IQR]) | 0.4 (0.2-0.8) | 0.5 (0.4-0.9) | 0.10 |
| Median aspartate transaminase (U/L [IQR]) | 29 (22-42) | 34 (24-39) | 0.88 |
| Median alanine transaminase (U/L [IQR]) | 25 (17-39) | 17 (13-34) | 0.88 |
| Mean white blood count (x10 3 cells/μL ± SD) | 8.5 ± 4.4 | 8.6 ± 4.7 | 0.92 |
| Mean hemoglobin (g/dL ± SD) | 11.6 ± 2.0 | 10.3 ± 2.1 | <0.01 |
| Mean platelet (x10 3 cells/μL ± SD) | 202 ± 86 | 161 ± 75 | 0.01 |
| Mean brain natriuretic peptide (pg/mL ± SD) | 1105 ± 984 | 1612 ± 1529 | 0.04 |
TTE measurements were available for 163 of 167 patients without significant ascites and 29 of 29 patients with significant ascites. The median time interval between abdominal imaging and TTE was 6 days (interquartile range [IQR] 2 to 24 days). TTE results are summarized in Table 3 . The group with significant ascites had more severe right ventricular (RV) dilation and TVR. However, the group with significant ascites had a higher left ventricular ejection fraction.
| Variable | Significant Ascites | p -value | |
|---|---|---|---|
| No (N=163) | Yes (N=29) | ||
| Mean left ventricular ejection fraction (% ± SD) | 20 ± 9 | 33 ± 18 | <0.01 |
| Mean right atrial size (cm 2 ± SD) | 26 ± 10 | 30 ± 9 | 0.07 |
| Right ventricular dilatation | 0.03 | ||
| Mild | 96 (59%) | 11 (39%) | |
| Moderate | 55 (34%) | 11 (39%) | |
| Severe | 12 (7%) | 6 (21%) | |
| Not available | 0 (0%) | 1 (3%) | |
| Mean right ventricular systolic pressure (mmHg ± SD) | 44 ± 15 | 41 ± 19 | 0.43 |
| Tricuspid valve regurgitation | <0.01 | ||
| None or minimal | 47 (29%) | 4 (14%) | |
| Mild | 86 (53%) | 8 (28%) | |
| Moderate | 26 (16%) | 12 (41%) | |
| Severe | 4 (3%) | 5 (17%) | |
RCC measurements were available for 150 of 167 patients without significant ascites and 26 of 29 patients with significant ascites. The median time interval between abdominal imaging and RCC was 7 days (IQR 2 to 52 days). The group with significant ascites had a higher mean RA pressure and RV end-diastolic pressure. There was no difference in pulmonary capillary wedge pressure between the groups. The results are summarized in Table 4 . No clear threshold value of RA pressure was identified for the development of significant cardiac ascites ( Figure 2 ).
