Ascites with elevated protein content as the presenting sign of constrictive pericardial disease

Abstract

Two men, one 63 and one 52 years old, presented with ascites. Analysis of the ascitic fluid in both patients revealed a high protein content and an elevated serum-ascites gradient. Various studies showed the cause of the ascites to be constrictive pericardial disease. Total excision of their parietal pericardia relieved their symptoms, decreased their cardiac filling pressures, and increased their cardiac indices. These cases highlight the importance of suspecting pericardial constriction as an etiology for high-protein-count ascites.

We describe two patients who presented with high-protein-count ascites as the initial sign of constrictive pericardial disease. These cases highlight the importance of early recognition of a cardiac etiology for ascites, as prompt treatment with a total pericardiectomy significantly improved the patients’ quality of life.

DESCRIPTION OF CASES

Pertinent clinical features in each of the two patients prior to surgery are summarized in Table 1. The two patients had evidence of ascites for 6 and 60 months, respectively. Both received multiple therapeutic paracenteses and escalating doses of diuretics for refractory ascites. The first patient had no attributable hepatic cause for his ascites, but the second had a history of heavy alcohol use prior to its cessation at age 47. Despite receiving a Denver shunt, the second patient continued to have intractable ascites so was referred for possible transjugular intrahepatic portosystemic shunt and liver transplantation. Both patients had no previous cardiothoracic surgery, pericarditis, radiation treatment, or collagen vascular diseases. Neither patient had cardiopulmonary symptoms. Physical examination in these patients revealed normal blood pressure and heart rate, distended jugular veins, clear lungs, and no precordial murmurs or abnormal heart sounds. Their abdomens had clear evidence of ascites. Laboratory tests in the first patient were normal. The second patient had renal insufficiency and mildly elevated bilirubin. Analysis of their ascitic fluid revealed an elevated total protein content and a high serum-ascites albumin gradient.

Table 1.

Pertinent features of the two men prior to pericardiectomy

Variable	Case 1	Case 2
Age (years)	63	52
Age of onset of ascites (years)	62	47
Previous hepatitis or alcohol use	0	+
Serum creatinine (mg/dL)	1.3	1.8
International normalized ratio	1.0	1.1
Total bilirubin (mg/dL)	0.4	1.8
Aspartate aminotransferase (U/L)	17	19
Alanine aminotransferase (U/L)	15	8
Alkaline phosphatase (U/L)	100	169
Ascitic fluid total protein (g/dL)	4.2	4.6
Serum-ascites albumin gradient (g/dL)	1.6	1.7

The patients were then referred to cardiology for evaluation of a cardiac cause of the ascites. Transthoracic echocardiograms in both patients revealed a bright, thickened parietal pericardium, “septal bounce,” exaggerated respiratory variation across the tricuspid and mitral valves, a dilated inferior vena cava, expiratory hepatic vein diastolic flow reversal, normal tissue Doppler velocities, and no pericardial effusion (Figure 1). Cardiac catheterization pressure waveforms in each showed equalization of the elevated diastolic pressures and discordance of right and left ventricular systolic pressures (Figure 2). Cardiac magnetic resonance imaging scans disclosed marked thickening of the parietal pericardium (Figure 3).

Figure 1.

Case 1. Transthoracic echocardiogram. (a) Apical view of bright, thickened parietal pericardium (arrows). (b) Doppler analysis showing exaggerated variation of tricuspid valve blood flow with respiration. (c) Hepatic vein diastolic flow reversal during expiration (arrows). (d) Normal Doppler tissue velocities at the lateral mitral annulus.

Figure 2.

Case 1. Simultaneous right ventricular (RV) and left ventricular (LV) cardiac catheterization pressure waveform tracings demonstrating equalization of diastolic pressures (arrow) and discordance of systolic pressures. The arrow also depicts the square root sign or dip and plateau sign—a finding due to diastolic pressures rising abruptly to a level that is sustained until systole because of the restricted filling from the pericardium.

Figure 3.

Case 2. A magnetic resonance imaging scan showing marked asymmetric pericardial thickening preferentially over the right ventricle (arrows) and gynecomastia in this patient with cirrhosis (ovals).

Both patients underwent total excision of their parietal pericardia with resolution of their ascites and elevated cardiac filling pressures. Hemodynamics before and after pericardiectomy are shown in Table 2. The parietal pericardium in each was severely thickened by dense fibrous tissue (Figure 4).

Table 2.

Hemodynamics before and after pericardiectomy

	Case 1		Case 2
Measurement	Preop	Postop	Preop	Postop
Mean right atrial pressure (mm Hg)	19	–	19	–
Right ventricular pressure (mm Hg)	35/19	–	28/15	–
Pulmonary artery pressure (mm Hg)	38/18	–	41/19	–
Pulmonary capillary wedge pressure (mm Hg)	19	–	20	–
Left ventricular pressure (mm Hg)	111/22	–	95/16	–
Central venous pressure (mm Hg)	19	3∗	27	13∗
Cardiac index (L/min/m2)	2.1	3.9∗	1.2	2.8∗

^∗Intraoperatively.

Figure 4.

Necropsy specimens contrasting (a) abnormal thickened parietal pericardium in Case 2 constrictive disease with (b) normal, thin parietal pericardium.

DISCUSSION

Constrictive pericardial disease is a rare cause of recurrent ascites. Both of our patients had high-protein-count ascites with an elevated serum-ascites albumin gradient, a finding well described in previous reports of constrictive pericarditis (1–3). A serum-ascites albumin gradient ≥1.1 g/dL and an ascites fluid total protein >2.5 g/dL is typical of constrictive pericardial disease and other postsinusoidal causes of ascites. Sinusoidal diseases, such as liver cirrhosis, exhibit a serum-ascites albumin gradient >1.1 g/dL but an ascites fluid total protein <2.5 g/dL (1).

A common finding in our cases was jugular venous distention, a sign not usually seen in patients with hepatic cirrhosis. Previous case series indicate that about 80% of patients with constrictive pericardial disease present with elevated jugular venous pressures. Neither of our patients had dyspnea or orthopnea. As many as half of the patients who undergo a pericardiectomy lack cardiopulmonary symptoms (4). Therefore, a high index of suspicion is required to diagnose this entity, especially in patients with elevated-protein-count ascites, jugular venous distention, and no cardiopulmonary symptoms.

The echocardiograms in our two patients were quite specific for constrictive pericardial disease. However, a transthoracic echocardiogram typically has low sensitivity in detection of this entity, and a constrictive pericardium may easily be overlooked. Thus, a physician should not be reassured with a negative echocardiogram if there is a high concern for this disease. Constrictive disease differs from restrictive disease by having normal tissue Doppler velocities on echocardiogram and discordance of right and left ventricular systolic pressures on cardiac catheterization.

Constrictive pericardial disease can lead to significant morbidity. Both patients suffered from the sequelae of ascites with repeated therapeutic paracenteses and escalating doses of diuretics before the proper diagnosis was made.

Pericardiectomy is the treatment of choice for patients with symptomatic chronic constrictive pericardial disease. The early hospital mortality is about 7% (5). The most common cause of death in the perioperative period is low-output heart failure (6). The long-term survival curves after pericardiectomy differ according to the etiology of the constrictive pericarditis and the type of surgery. Idiopathic/viral and postsurgical constrictive pericardial disease have the best 10-year survival rates after pericardiectomy of about 67% and 56%, respectively, while postradiation pericarditis has the worst at 11% (5). Also, total pericardiectomy has a better survival rate than partial pericardiectomy (7).

Acknowledgments

We gratefully acknowledge William Clifford Roberts, MD, for editorial assistance with the report and Jong Mi Ko, BA, for the photography of the surgical specimens.