Paradoxical role of inhaled nitric oxide in advanced liver disease

From the Departments of Internal Medicine (Simpson), Anesthesiology and Pain Management (M. Ramsay, K. Ramsay, and Hein), Pulmonary Services (Lynch and Black), Gastroenterology (Crippin and Sarles), Cardiology (East and Escobar), and Baylor Institute of Transplantation Sciences (Goldstein, Husberg, Levy, and Klintmalm), Baylor University Medical Center, Dallas, Texas; and the Department of Anesthesiology and Pain Management, The University of Texas Southwestern Medical Center at Dallas (M. Ramsay).

Corresponding author: Michael A. E. Ramsay, MD, Department of Anesthesiology and Pain Management, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, Texas 75246 (e-mail: docram@baylordallas.edu).

he hyperdynamic circulation associated with severe liver cirrhosis has been attributed to the persistent induction of nitric oxide (NO) synthase (1, 2). The resultant increase in endogenous NO production causes the development of arteriolar vasodilation and vascular shunts, an increased cardiac index, and a low vascular resistance (3). This hypothesis has been supported by the increased levels of exhaled NO observed in patients with advanced liver disease (4, 5).

Three pathophysiological developments have been seen in the lungs of patients who have died of severe liver disease: vascular dilatations, arteriolar wall hyperplasia, and intravascular occlusion caused by thrombi (6). A preponderance of vascular dilatations is associated with hepatopulmonary syndrome (HPS), and an excess of vascular hyperplasia and/or occlusion is associated with increased pulmonary vascular resistance and the development of pulmonary hypertension.

The use of inhaled NO has not been shown to reverse portopulmonary hypertension except intraoperatively in response to an acute elevation of pulmonary artery (PA) pressures (7–9). Inhaled NO has been reported to reverse HPS in 2 patients despite the increased levels of endogenous NO, but it has not reversed HPS in other patients (10–12). We report on 3 patients with end-stage liver disease who, during evaluation for liver transplantation, were found to have either pulmonary hypertension or HPS. Upon exposure to inhaled NO, they demonstrated a significant improvement in either pulmonary hemodynamics or oxygenation.

Pulmonary hypertension was defined as a mean PA pressure of >25 mm Hg, normal PA occlusion pressure, and pulmonary vascular resistance of >120 dynes?sec^–1?cm^–5. HPS was defined as postural hypoxemia with a positive contrast echocardiogram demonstrating intrapulmonary shunting in a patient with severe liver disease.

All patients presenting for liver transplantation at Baylor University Medical Center with either pulmonary hypertension or HPS were invited to be included in the inhaled NO test study. After institutional review board approval and informed patient consent were obtained, patients were tested with inhaled NO delivered from an I-NOvent system (Ohmeda, Inc., Liberty Corner, NJ) via a face mask and a non-rebreathing circuit. Inhaled concentrations were increased in 10-ppm increments up to a maximum of 40 ppm. A positive response to inhaled NO was defined as a decline in mean PA pressure of at least 20% or an improvement in arterial partial pressure of oxygen of >=20%.

Of 434 consecutive patients who presented for liver transplantation evaluation, 20 patients were included in the study: 16 patients with moderate to severe pulmonary hypertension and 4 patients with HPS. Three patients from this group of 20 patients had a positive response to inhaled NO, 2 with pulmonary hypertension and 1 with HPS.

CASE REPORTS

Patient 1

A 62-year-old woman presented with cirrhosis of the liver secondary to chronic hepatitis C. She had a history of esophageal variceal bleeding, and a sclerosis procedure had been performed. On evaluation for liver transplantation, she had moderate pulmonary hypertension (defined as a mean PA pressure >35 mm Hg and normal PA occlusion pressure). Hemodynamic data included PA pressure of 63/18 mm Hg (mean, 38 mm Hg); pulmonary vascular resistance, 587 dynes?sec^–1?cm^–5; PA occlusion pressure, 5 mm Hg; and cardiac output, 4.5 L/min. Significant laboratory data included a prothrombin time of 13.2 seconds; total bilirubin, 1.6 mg/dL; and mildly elevated liver enzymes (serum aspartate aminotransferase, 71 U/L, and serum alanine aminotransferase, 55 U/L). An inhaled NO trial with concentrations of up to 40 ppm was performed using a tight-fitting face mask and a non-rebreathing circuit. At 20 ppm, a maximum and significant response in PA pressure and pulmonary vascular resistance was noted (Table 1). The PA pressure declined to 47/13 mm Hg (mean, 28 mm Hg), and the pulmonary vascular resistance declined to 383 dynes?sec^–1?cm^–5. This test was repeated the next day with similar results. The PA pressure returned to the elevated baseline levels within 30 minutes of cessation of the NO therapy.

Patient 2

A 46-year-old woman presented with end-stage liver disease as a result of primary biliary cirrhosis. She complained of progressive fatigue, had a history of an esophageal variceal bleed, and had undergone banding of the esophageal varices. The pretransplant evaluation demonstrated mild encephalopathy and severe pulmonary hypertension (defined as a mean PA pressure >45 mm Hg with a normal PA occlusion pressure). PA pressures of near systemic levels were recorded: PA pressure, 90/31 mm Hg (mean, 52 mm Hg); PA occlusion pressure, 10 mm Hg; cardiac output, 7.1 L/min; and pulmonary vascular resistance, 473 dynes?sec^–1?cm^–5 (Table 2). An echocardiogram demonstrated normal left ventricular morphology and function (ejection fraction, 0.55) and increased right heart pressure as indicated by paradoxical septal wall motion, a dilated right ventricle, and severe tricuspid regurgitation. Laboratory data included serum aspartate aminotransferase, 57 U/L; serum alanine aminotransferase, 46 U/L; total bilirubin, 1.1 mg/dL; and prothrombin time, 12.2 seconds. Inhaled NO was delivered via a face mask and non-rebreathing circuit with concentrations up to 40 ppm, and a maximum positive response was noted at 20 ppm with the mean PA pressure declining from 52 to 36 mm Hg. Baseline parameters returned within 5 minutes of cessation of inhaled NO.

Patient 3

A 49-year-old woman presented with severe cirrhosis as a result of hepatitis C. On evaluation for transplantation, she was found to have HPS. She was dyspneic and cyanotic, with a history of orthodeoxia and platypnea. Multiple dermal spider angiomata and marked digital clubbing were present. The room air arterial blood gas analysis in an upright position revealed a PaO₂ of 48 mm Hg (82% saturation). Prothrombin time was 11.7 seconds; serum aspartate aminotransferase, 81 U/L; serum alanine aminotransferase, 38 U/L; and total bilirubin, 2.6 mg/dL. HPS was confirmed by a positive contrast echocardiogram demonstrating significant intrapulmonary shunting. This patient did not respond to inhaling 100% oxygen. The PaO₂ remained low at 57 mm Hg (89% saturation), therefore categorizing her with type 2 HPS (13). The calculated shunt fraction was 27.5%. Upon inhaling 20 ppm of NO with a well-fitted face mask and non-rebreathing circuit, the patient's arterial PaO₂ increased to 205 mm Hg (98% saturation) (Table 3); a further increase to 40 ppm of NO had no additional effect. Baseline parameters returned 5 minutes after the cessation of inhaled NO, the oxygen saturation declining to 84%. This was a repeatable effect of inhaled NO on this patient.

DISCUSSION

All 3 patients who responded to NO underwent successful orthotopic liver transplantation. At 3-month follow-up, patient 1 had evidence of mild pulmonary hypertension on echocardiogram; patient 2 had moderate pulmonary hypertension; and patient 3 had improved oxygen saturation (91% on breathing room air). Both pulmonary hypertensive patients were also on continuous infusions of epoprostenol. These infusions were started after the NO study.

Why these 3 patients should demonstrate a positive response to inhaled NO is intriguing. It would have been informative if endogenous NO levels could have been measured in these patients' exhaled breath to determine if they were producing excessive endogenous NO or normal NO. It would appear likely that patient 1, who did not present with a hyperdynamic circulation, did not have an elevated level of endogenous NO. Patient 2 had an elevated cardiac output but certainly not the typically very high cardiac output levels frequently seen with end-stage liver disease patients. This possibly indicated lower levels of endogenous NO. Although both of these patients had evidence of portal hypertension, as demonstrated by esophageal varices, it is possible that the pulmonary hypertension was unrelated. Coincidental primary pulmonary hypertension may have developed concomitantly with severe liver disease. Alternatively, perhaps patients in the early development of portopulmonary hypertension may still possess some vasoreactivity in response to inhaled NO. Pulmonary hypertension may be the result of several disease processes with different phenotypic expressions and, therefore, different responses to therapy.

Patient 3 presented with the more uncommon type 2 HPS, which does not respond to inhaled 100% oxygen (13). This is thought to be the result of a few, discrete, significant pulmonary vascular shunts formed in otherwise normal lung anatomy. In this situation, it could be hypothesized that inhaled NO caused dilatation of the normal pulmonary arterioles, perfusing normal alveoli, resulting in a reduction in the shunt fraction, and, therefore, an improvement in oxygenation.

The role of endogenous and exogenous NO still needs to be clearly defined in end-stage liver disease.