Hepatic hydrothorax: Pathophysiology diagnosis and management
Dr Anastasios Roussos, Gastroenterology Clinic, Evangelismos Hospital, Athens, Greece. Email: firstname.lastname@example.org
Hepatic hydrothorax is defined as a significant pleural effusion (usually greater than 500 ml) in a cirrhotic patient, without an underlying pulmonary or cardiac disease. The diagnosis of hepatic hydrothorax should be suspected in a patient with established cirrhosis and portal hypertension, presenting with a unilateral pleural effusion, most commonly right-sided. In the vast majority of cases, patients with hepatic hydrothorax have end-stage liver disease. Therefore, they should be considered potential candidates for orthotopic liver transplantation. Until the performance of transplantation, other therapeutic modalities should be applied in order to relieve symptoms and prevent pulmonary complications.
Hepatic hydrothorax is defined as a significant pleural effusion, usually greater than 500 mL, in a cirrhotic patient, without an underlying pulmonary or cardiac disease.1,2 It seems to be a relatively, uncommon complication of portal hypertension with an estimated prevalence of 5–12% in patients with cirrhosis of the liver.3–5 Hepatic hydrothorax is usually right sided (65–87% of reported cases). However, it may be left-sided or bilateral. In the vast majority of cases, ascites are also present.6–8
It is well known that in cirrhotic patients, a large volume of ascetic fluid is generally well tolerated due to the capacitance of the peritoneal cavity. On the other hand, even modest volumes of pleural fluid can cause significant respiratory symptoms, including dyspnea and chest pain. Therefore, although infrequent, hepatic hydrothorax may represent a major clinical problem in the management of patients with portal hypertension.
The aim of the present report is to provide a review of the literature with regards to the pathogenesis, clinical manifestations, diagnosis, and treatment of hepatic hydrothorax.
Several mechanisms have been proposed in order to explain the development of hepatic hydrothorax in patients with cirrhosis and portal hypertension (Table 1).
Table 1. Proposed mechanisms in order to explain the development of hepatic hydrothorax
|Hypoalbuminemia—decreased colloid osmotic pressure|
|Azygos veins hypertension|
|Transdiaphragmatic migration of fluid via lymphatic channels|
|Leakage of ascitic fluid via diaphragmatic defects|
Hypoalbuminemia and subsequent decreased colloid osmotic pressure were originally thought to cause leakage of fluid into the pleural cavity.1,9 Both are common findings in patients with portal hypertension. However, hepatic hydrothorax exists only in a minority of these patients. Moreover, this theory can not explain why the pleural effusion in cirrhotics is predominantly right sided.
It has been suggested that an increased venous pressure in azygos veins might lead to leakage of plasma into the pleural space and the subsequent development of hepatic hydrothorax.1,10 However, it is well known from patients with congestive heart failure, that elevated systemic venous pressures without pulmonary venous hypertension do not give rise to pleural effusions.11 In addition, the leakage of pleural fluid due to increased blood flow in azygos veins fails to explain the right-side predomination of hepatic hydrothorax.
In order to explain the presence of pleural effusion in patients with ovarian tumors and malignant ascites, Meigs et al. suggested the transdiaphragmatic migration of fluid from the peritoneal cavity into the pleural space via the lymphatic channels.12 It is well known that radiolabeled substances injected in the peritoneal cavity are absorbed by the diaphragmatic lymphatic channels and flow into the large lymphatic vessels of the mediastinum. However, these vessels empty into systemic veins and not into the pleural cavity.13 Moreover, hepatic hydrothorax is a rare phenomenon in patients with portal hypertension, although most of them have increased flow states in their intra-abdominal lymphatic system.2 Finally, the theory of fluid leakage via the lymphatic channels fails to explain why hepatic hydrothorax is predominantly right sided.
In 1955, Emerson described for the first time a diaphragmatic fenestration in a patient with cirrhosis and pleural effusion.14 More recent studies have demonstrated the existence of such diaphragmatic defects in a large number of patients with hepatic hydrothorax.15,16 Microscopic examination of these defects revealed discontinuities in the collagen bundles that make up the tendinous portion of the diaphragm.15 Increases in the intra-abdominal pressure, for example, as a result of ascites, coughing, or straining, might lead to small hernations of the peritoneum through these gaps into the pleural cavity. These hernations, also called pleuroperitoneal blebs, may rupture. The ensuing defects, which are typically less than 1 cm in diameter, allow free communication between the peritoneal and pleural space.3,7,17
The leakage of ascitic fluid via diaphragmatic defects seems to explain the pathogenesis of hepatic hydrothorax. The negative intrapleural pressure compared to that of the peritoneal cavity facilitates the one-way transfer of fluid and its subsequent trapping into the pleural space. Hepatic hydrothorax occurs when the accumulation of fluid surpasses the absorptive capacity of the pleura. This theory is further supported by the fact that air, dyes, or radiolabeled substances intra-abdominally injected in patients with hepatic hydrothorax move rapidly into the pleural cavity.13–15,18 Moreover, the direct passage of peritoneal fluid into the pleural space via diaphragmatic defects explains the right-side predomination of hepatic hydrothorax. Autopsy studies suggest that pleuroperitoneal blebs occur less frequently in the left hemidiaphragm, as it seems to be thicker and more muscular.19
It is well known that hepatic hydrothorax may occasionally occur in the absence of ascites.20–23 Recent studies using intra-abdominally-injected, radiolabeled colloids and/or radioactive isotope scans confirmed the communication between the peritoneal cavity and the pleural space in these cases.21,24 Moreover, ultrasonography showed a large defect in the right hemidiaphragm of a patient with hepatic hydrothorax and no evidence of ascites.25 Therefore in patients with hepatic hydrothorax, ascites seem to occur only if the rate of the collection of fluid in the abdomen exceeds the capacity of the pleural space to maintain a pressure gradient and drain the peritoneal cavity.
Clinical presentation and diagnosis
The diagnosis of hepatic hydrothorax should be suspected in a patient with established cirrhosis and portal hypertension, presenting with a unilateral pleural effusion, most commonly right sided.4–6 In most cases, the dominating clinical manifestations are those of cirrhosis and ascites. However, a variety of respiratory symptoms, including dyspnea, non-productive cough, pleural chest pain, and fatigue due to hypoxemia may also occur. The severity of these symptoms depends on the volume of pleural fluid. Patients rarely present with symptoms and signs of respiratory failure due to an acute tension hydrothorax.26,27 In a minority of cases, pleural effusion may be noted incidentally on a chest radiography performed for other reasons.28
Hepatic hydrothorax may occur in the absence of ascites. Several authors have reported cases of unilateral pleural effusions in cirrhotic patients without clinical or even ultrasonic evidence of ascites.20–23 In these cases, respiratory symptoms may confuse the clinician leading to an extensive diagnostic evaluation, as liver cirrhosis is not commonly considered a distinct cause of pleural effusion.
A diagnostic thoracentesis should be performed in all patients with suspected hepatic hydrothorax.29 Pleural effusions associated with cirrhosis and portal hypertension are transudative (Table 2).2,7,8,30 However, total protein, albumin, and total lipid levels may be slightly higher in pleural fluid than in the peritoneal fluid derived from the same patient. This difference might be attributed to the greater efficacy of water absorption by the pleural surface.31
Table 2. Composition of fluid in patients with uncomplicated hepatic hydrothorax
|Cell count <1000/mm3|
|Polymorphonuclear leukocyte count <250/mm3|
|Protein concentration <2.5 g/dL|
|Pleural fluid/serum total protein ratio <0.5|
|Pleural fluid/serum lactate dehydrogenase ratio >0.6|
|Pleural fluid/serum albumin gradient >1.1|
|Pleural fluid glucose concentration similar to that of serum|
In patients with cirrhosis and pleural effusion, a thorough investigation for primary cardiopulmonary disorders should be performed. In a recent study a diagnosis other than hepatic hydrothorax was established in 18% of cirrhotics with pleural effusions.29 Therefore, in all cases, a computed tomographic scan of the chest should be performed in order to exclude pulmonary, mediastinal, or pleural disorders. If malignancy is strongly suspected and a cytological examination of pleural fluid is non-diagnostic, thoracoscopy and pleural biopsy may be required. Echocardiography is indicated if there is suspicion of pericarditis or right heart failure.
In cases where the diagnosis of hepatic hydrothorax is in doubt, in particular when pleural effusion is left sided and/or ascites are absent an intraperitoneal administration of 99mTc-human serum albumin or 99mTc-sulfur colloid should be performed. The migration of the radioisotope into the pleural cavity within hours confirms the presence of a communication between the pleural and peritoneal spaces.32–34 According to a recent study, this technique has sensitivity and specificity rates of 71% and 100%, respectively.35 Moreover, it can be used even in the absence of ascites, through the ultrasound-guided administration of the radiolabeled substance in a 500-mL volume of normal saline.36 Finally, other diagnostic modalities, including magnetic resonance imaging and thoracoscopy could be performed in order to detect the underlying diaphragmatic defects.37,38
Spontaneous bacterial empyema (SBEM) represents a distinct complication of hepatic hydrothorax. It is defined as the spontaneous infection of pleural fluid.39,40 The pathogenesis of SBEM remains unclear. It might occur as a result of a direct bacterial spread from the peritoneal cavity. However, in a recent study, nearly 45% of episodes were not associated with spontaneous bacterial peritonitis.39 Moreover, SBEM may occur even in the absence of ascites. In those cases, a transient bacteremia that infects the pleural space could be the underlying pathogenetic mechanism.41 A high Child–Pugh score, a decreased pleural fluid total protein, and low levels of C3 component in pleural fluid are risk factors for the development of SBEM. The responsible bacteria are usually Escheria coli, Streptococcus species, Enterococcus, and Klebsiella.42
Any patient with hepatic hydrothorax who presents with fever, pleuritic pain, and/or encephalopathy should be evaluated for SBEM.43 The thoracentesis with the subsequent culture of pleural fluid is the diagnostic procedure of choice. Table 3 shows the diagnostic criteria of SBEM. A recent study showed that an analysis of pleural fluid with a reagent strip for leukocyte esterase might represent a rapid, easy-to-use, and inexpensive tool for the diagnosis of SBEM in cirrhotic patients. However, more studies are needed to confirm these results.44
Table 3. Diagnostic criteria of spontaneous bacterial empyema
|Serum/pleural fluid albumin gradient >1.1 g/dL|
|Polymorphonuclear leukocyte count >500/mm3 or positive fluid culture|
|Absence of pneumonia or a contiguous infection process on chest radiography|
SBEM is a serious complication of hepatic hydrothorax.39,40 A recent study in 15 patients with SBEM showed a mortality rate of 20%.40 More studies are needed to clarify the underlying pathogenetic mechanism and the natural course of SBEM.
The vast majority of cases patients with hepatic hydrothorax have end-stage liver disease. Therefore, they should be considered as potential candidates for orthotopic liver transplantation. Until the performance of transplantation, other therapeutic modalities should be applied in order to relieve symptoms and prevent pulmonary complications. Table 4 shows the existing therapeutic modalities in patients with hepatic hydrothorax.
Table 4. Therapeutic modalities in patients with hepatic hydrothorax
| Sodium-restricted diet|
|Transjugular intrahepatic portosystemic shunts|
| Tube thoracostomy—injection of a sclerosing agent|
| Surgical repair of diaphragmatic defects|
| Peritoneovenous shunting|
Hepatic hydrothorax is ascetic fluid that has entered the pleural cavity. Therefore, the principles of medical management are identical to those of ascites in cirrhotic patients.45 The aim of the therapy should be to create and subsequently maintain a negative sodium balance. Therefore, a sodium-restricted diet (less than 90 mEq/day) is recommended. However, usually diet therapy is not sufficient to achieve a negative sodium balance,46 therefore, diuretics are required in the vast majority of cases. A combination of a distal acting agent (e.g. spironolactone 100 mg/day) plus a loop diuretic (e.g. furosemide 40 mg/day) is indicated. If there is no response, and incompliance with treatment is excluded, the doses may be increased in a stepwise fashion; spironolactone up to 400 mg/day, and furosemide up to 160 mg/day.47,48 A recent report of a case showed that terlipressin, a drug known to be beneficial in hepatorenal syndrome, might also be effective in the treatment of hepatic hydrothorax.49 The successful use of octreotides50,51 and nasal continuous positive airway pressure52 have also been reported. However, studies in series of patients are needed to confirm these findings.
More invasive therapeutic modalities should be considered only in patients who do not respond to traditional medical management.
Therapeutic thoracentesis should be performed in order to relieve symptoms of dyspnea in patients with large effusions. Moreover, it is indicated in cases with recurrent or refractory hydrothorax.3–5 Chest radiographs (posteroanterior and lateral decubitus) and computed tomography scan of the chest are required to define the size of the hydrothorax. No more than 2 L of fluid should be removed as otherwise there is a significant risk of hypotension or re-expansion of pulmonary edema.53 Thoracentesis is a safe procedure. Pain at the puncture site seems to be the more frequent complication. Other rare complications include pneumothorax, hemothorax, soft tissue infection, empyema, air embolism hemoptysis, and subcutaneous emphysema.54
Therapeutic thoracentesis is not a maintenance treatment of hepatic hydrothorax. Thoracentesis dependence may be associated with both the deterioration of clinical status and impaired quality of life. Therefore, when therapeutic thoracentesis is required more than once every 2 weeks in patients on strict sodium restrictions and optimal diuretics, alternative therapeutic modalities should be considered.7,8
Transjugular intrahepatic portosystemic shunts
The performance of transjugular intrahepatic portosystemic shunts (TIPS) is the procedure of choice in patients with hepatic hydrothorax who do not respond to diuretic treatment and/or repetitive therapeutic thoracentesis. The TIPS are created by the placement, under the fluoroscopic guidance of an expandable metal stent in the hepatic parenchyma between the portal and the hepatic venous system. Therefore, TIPS reduce portal pressure and address the elevated hepatic sinusoidal pressure that leads to ascites.55,56
Data in the literature concerning the effectiveness of TIPS in the management of refractory hepatic hydrothorax are encouraging. The TIPS may lead to a symptomatic improvement in 70–80% of patients with refractory hydrothorax.57–60 However, shunt occlusion and hepatic encephalopathy are common complications encountered after the TIPS placement (25% and 31% of patients, respectively, in a large series).60 The worsening of pre-existing pulmonary hypertension may also occur. Moreover, an increased mortality rate (20–25%), within the first 2 months after TIPS placement has been reported. In the vast majority of cases, death is attributed to the complications of end-stage liver disease.56 A recent study showed that the mortality rates were the same for patients with similar Child–Pugh scores, irrespective of the indication for the procedure.61
In conclusion, TIPS may lead to symptomatic relief in most patients with refractory hepatic hydrothorax. However, a variety of serious complications are related to TIPS placement. Moreover, TIPS do not improve the overall prognosis of patients with end-stage liver disease. Therefore, they must be considered only as a ‘bridge’ to liver transplantation and not as treatment of choice for patients with refractory ascites.
Three surgical approaches have been used in the management of hepatic hydrothorax: tube thoracostomy with chemical pleurodesis, repair of defects in the diaphragm, and peritoneovenous shunts.
Tube thoracostomy with subsequent injection of a sclerosing agent, for example, talc, may lead to ablation of the space between the parietal and visceral pleura.62–64 However, the rates of fluid re-accumulation are extremely high. Moreover, the procedure is associated with a variety of complications, including fever, empyema, chest pain, pneumonia, incomplete re-expansion, and wound infection.64 Therefore, it is not recommended for the treatment of hepatic hydrothorax in clinical practice.
A few case reports described the successful surgical repair of diaphragmatic defects responsible for fluid migration into the pleural cavity.65–67 In a study concerning eight patients with refractory hepatic hydrothorax, Moroux et al. used video-assisted thoracoscopy in order to repair diaphragmatic defects in addition to pleurodesis. There was a complete response in six cases and no recurrence was detected.68 Moreover, the use of pleura and mesh onlay reinforcement of the diaphragm seems to be an encouraging treatment for refractory hepatic hydrothorax.69 However, more studies in larger numbers of patients are needed to confirm these results.
Peritoneovenous shunting for the management of hepatic hydrothorax has been used in a limited number of patients;70,71 the results are conflicting. This procedure is associated with serious complications, including hepatic encephalopathy and shunt occlusion that cause significant morbidity and mortality.72 Therefore, peritoneovenous shunting for the treatment of hepatic hydrothorax was abandoned nearly a decade ago.
Recently, Xiol et al. studied the outcome of liver transplantation in patients with hepatic hydrothorax. They showed that there were no differences between the hydrothorax group and the control group in the days of mechanical ventilation after surgery, transfusion requirements, postoperative mortality, and long-term survival. Long-term evolution was similar between patients with refractory hepatic hydrothorax or spontaneous bacterial empyema and those with non-complicated hepatic hydrothorax. Therefore, they suggested that hepatic transplantation might be an excellent therapeutic option for patients with hepatic hydrothorax.73
Hepatic hydrothorax is an uncommon complication of portal hypertension in patients with liver cirrhosis. The leakage of ascitic fluid via diaphragmatic defects seems to explain the pathogenesis of hepatic hydrothorax. In most cases, dominating clinical manifestations are those of cirrhosis and ascites. However, a variety of respiratory symptoms, including dyspnea, non-productive cough, pleural chest pain, and fatigue due to hypoxemia may also occur. A diagnostic thoracentesis should be performed in any patient with suspected hepatic hydrothorax. In cases where the diagnosis of hepatic hydrothorax is in doubt, an intraperitoneal administration of a radiolabeled substance should be performed. The migration of the radioisotope into the pleural cavity within hours confirms the presence of a communication between the pleural and peritoneal spaces.
In the vast majority of cases, patients with hepatic hydrothorax have end-stage liver disease. Therefore, they should be considered potential candidates for orthotopic liver transplantation. Until the performance of transplantation, other therapeutic modalities, including a sodium-restricted diet, diuretics, therapeutic thoracentesis, and transjugular intrahepatic portosystemic shunts should be applied in order to relieve symptoms and prevent pulmonary complications.