A clinical study of the hepatopulmonary syndrome in Iraqi patients with chronic liver disease


Objective To evaluate the frequency, clinical, and laboratory features of HPS and to determine their usefulness in its diagnosis in Iraqi patients with liver disease. Patients and Methods Fifty-two patients with chronic liver disease were evaluated for the presence of HPS using pulse oximetry, arterial blood gas analysis (ABG), and transthoracic contrast-enhanced echocardiography (CEE). Patients with SaO2 of ≤92% and/or a decrease in SaO2 of ≥4% after change from supine to upright position; were further tested with ABG and CEE. Patients who had a PaO2 of ≤70 mmHg together with positive CEE were considered to have HPS. Results Of 52 patients studied, 13 (26.9%) were in Child-Pugh class A, 16 (30.7%) in class B, and 23 (42.3%) in class C. Four patients (7.6%) proved to have HPS; one was in Child-Pugh class B and the other three were in class C (p=0.3574). Among the clinical features assessed as predictors of HPS, dyspnea (p = 0.0027), cyanosis (p < 0.0001), and finger clubbing (p = 0.0514) reached statistical significance. Neither upper GI endoscopy nor biochemical liver tests were statistically different between patients with and without HPS. Conclusions HPS is not rare. Dyspnea is a useful marker for its presence in the appropriate clinical setting. Platypnea, cyanosis, and finger clubbing are much less sensitive but more specific features of the syndrome. No particular pattern of biochemical liver tests is useful in predicting the presence or absence of HPS. There was a trend for HPS to occur in patients with more advanced liver disease.