A 34-year-old Malaysian man presented with esophageal variceal bleeding, which was successfully treated with sclerotherapy. Investigations at that time showed evidence of portal hypertension with prominent portosystemic collateral vessels and splenomegaly. Hepatic synthetic function was normal. A liver biopsy revealed minimal fibrosis of nonspecific etiology and a diagnosis of noncirrhotic portal hypertension (NCPH) was made.
Five years later he presented with increasing exertional breathlessness. On examination he had developed finger clubbing and was cyanotic. O2 saturation and PaO2 on room air were reduced at 91% and 60 mmHg, respectively. The PO2 rose to 576 mmHg on 100% O2. Pulmonary angiography showed no evidence of discrete AV malformations; however, hepatic vein wedge pressure was measured at the time and was elevated at 30 mmHg, in keeping with marked portal hypertension. Doppler studies of the hepatic vasculature showed patent vasculature. A repeat liver biopsy revealed hepatoportal sclerosis, in keeping with the original diagnosis of NCPH. Subsequently, a contrast-enhanced echocardiogram was performed which demonstrated delayed appearance of echo-contrast in the left atrium indicating an intrapulmonary shunt confirming the diagnosis of hepatopulmonary syndrome (HPS).
The patient's respiratory symptoms insidiously worsened. By 4 years later his PaO2 on room air had fallen to 49 mmHg. A technetium-labeled macroaggregated albumin (99mTcMAA) lung perfusion scan revealed markedly abnormal brain uptake of radioactivity tracer and a shunt fraction of 15.25% (normal <6%) was calculated using standard methods. The patient was listed for orthotopic liver transplantation (OLT) and the surgery was successfully performed 6 months later. Examination of the explanted liver was again consistent with hepatoportal sclerosis with no evidence of cirrhosis or active liver disease. Following OLT his breathlessness improved steadily over several months. By 3 months post-OLT his O2 saturation on room air was 96% and PO2 was 90 mmHg.
Three years post-OLT he presented with right upper quadrant discomfort and worsening dyspnea associated with the redevelopment of hypoxia. At rest, in the supine position his O2 saturation on room air was 93% with a PaO2 of 64 mmHg. A liver biopsy revealed no evidence of significant graft pathology and a computed tomography (CT) chest was unremarkable. A 99mTcMAA lung perfusion scan was repeated and indicated 8.9% brain uptake of tracer in keeping with recurrence of HPS (Fig. 1A). Given his abdominal discomfort and an abnormal hepatic Doppler study, he proceeded to an abdominal angiogram, which revealed a severe stenosis between the donor and recipient cava with minimal hepatic venous (HV) outflow (Fig. 2). The intercaval anastomosis was crossed with a guide wire and a balloon dilatation to 8 mm was performed resulting in markedly improved hepatic venous outflow (Fig. 3). Two months later the patient's dyspnea had completely resolved and PaO2 on room air had risen to 89 mmHg (Fig. 4). A repeat 99mTcMAA lung perfusion scan (Fig. 1B) revealed no significant brain uptake and thus confirmed resolution of HPS.
To our knowledge, the recurrence of HPS in an adult NCPH patient post-OLT has not been reported. HPS has been reported to recur posttransplantation in the context of serious graft disease or cirrhosis. In our case it redeveloped in the absence of any evidence of graft damage, suggesting that the hemodynamic alterations associated with impairment of hepatic outflow were solely responsible. While HV outlet obstruction is an unusual cause of HPS, the syndrome has been described in individuals with IVC obstruction with amelioration of hypoxia on restoration of flow.
Why the disease occurs in only a minority of cirrhosis patients and only occasionally in patients with NCPH is uncertain. However, this presumably reflects an underlying “susceptibility” in those who develop the syndrome which is absent in most patients. What factors govern this susceptibility is as yet unknown. The redevelopment of clinically significant HPS in our patient many years posttransplant after apparent complete resolution of the syndrome shows that this susceptibility persists for life and that it cannot be related to liver disease per se or some host susceptibility factor within the liver.