Unilateral pleural effusion—but where from?

Abstract

In December, 2007, a 63-year-old woman presented with a week’s history of shortness of breath, 4 months of ankle swelling, and weight gain. She had no chest pain, palpitations, cough, or fever. She had been diagnosed with non-A, non-B hepatitis in 1980. There was a large right-sided pleural eff usion on chest radiograph, shown to be loculated by CT. Thoracentesis yielded 1·5 L of clear, yellow transudate. CT of the abdomen showed a shrunken liver consistent with cirrhosis without ascites. No abnormalities were found in full blood count, renal function, liver function, concentrations of thyroid hormones, pancreatic enzymes, and brain natriuretic peptide, electrocardiogram, tuberculin test, pelvic ultrasonography, echocardiogram, or urine analysis. Hepatitis B and C, Wilson’s disease, haemochromatosis, α-1 antitrypsin defi ciency, and autoimmune hepatitis were excluded. A liver biopsy sample showed regenerative nodule formation consistent with cirrhosis of unknown cause. Doppler ultrasonography of the portal vein was normal, whereas oesophagogastroduodenoscopy showed mosaic patterning of the stomach consistent with portal hypertensive gastropathy. Repeat thoracentesis on day 4 drained 1·5 L of fl uid. Treatment with fl uid restriction, low salt diet, furosemide, and spironolactone was started, but fl uid reaccumulated within 3 days. On day 8, talc pleurodesis was done and she was discharged. 2 weeks later she was readmitted with recurrent pleural eff usion. Hepatic hydrothorax was considered a possible diagnosis. Hepatic vein catheterisation (fi gure) revealed a hepatic vein pressure gradient (HVPG) of 21 mm Hg (normal value: 1-5 mm Hg). Transjugular intrahepatic portosystemic shunt (TIPS) was undertaken, after which the HVPG was 4 mm Hg and no reaccumulation of pleural fl uid occurred. 4 days later, the patient became encephalopathic with a rise in concentrations of bilirubin (68 μmol/L) and aminotransferases (aspartate amino transferase 750 U/L, alanine aminotransferase 291 U/L). Intracranial processes were ruled out as a potential cause of altered mental status. Lactulose treatment was initiated and a reduction in TIPS size was considered but she developed aspiration pneumonia necessitating intubation and died within 4 days. The family declined autopsy. Hepatic hydrothorax is a signifi cant pleural eff usion in a cirrhotic patient in the absence of a primary cardiopulmonary cause; it is a complication of portal hypertension, with a prevalence of 5–12%. Although usually right-sided, it can be left-sided or bilateral. The cause is thought to be leakage of ascitic fl uid via a diaphragmatic defect. Diagnosis of hepatic hydrothorax without ascites and obvious signs of portal hypertension is challenging. Doppler ultrasonography of the portal vein can show portal hypertension, but HVPG is the gold standard. Scintigraphic demonstration of a diaphragmatic defect by use of radiolabelled human serum albumin is helpful, and in retrospect might have been useful in our patient. Treatment of hepatic hydrothorax is challenging, with liver transplant the ultimate goal. Diuretics, fl uid restriction, and repetitive thoracenteses are used to prevent pulmonary complications. Diaphragmatic repair, pleurodesis, and TIPS are reserved for refractory cases. Rapid migration of fl uid from peritoneum to the pleural space means pleurodesis is often unsuccessful. 30-day mortality with TIPS ranges from 19% to 68%; complications include hepatic encephalopathy, stent occlusion, infection, bleeding, acute renal failure, and acute respiratory distress syndrome. Rapid onset of encephalopathy after TIPS can be due to infection or to insuffi cient arterial liver perfusion as a result of arterial-stent fi stula, arterial-biliary fi stula, or stent-related stenosis of the artery causing liver infarction. In patients with unilateral pleural eff usion, after exclusion of cardiac and pulmonary causes, a hepatic cause should be considered. Defects should be demonstrated by scintigraphy if the diagnosis is in doubt before extensive work-up. TIPS is an eff ective treatment, but serious side-eff ects can occur even in patients with well-compensated liver disease.