Hepatopulmonary Syndrome (HPS) - Full 10-Mark MD Exam Answer

Definition

1. Hepatic disease (cirrhosis or portal hypertension)
2. Intrapulmonary vascular dilation (IPVD)
3. Abnormal arterial oxygenation - specifically a widened alveolar-arterial oxygen gradient (A-a gradient >15 mmHg, or >20 mmHg if age >64 years)

• Fishman's Pulmonary Diseases and Disorders, p. 1719
• Murray & Nadel's Textbook of Respiratory Medicine, Ch. 126

Epidemiology

• Prevalence: 4% to 32% in cohorts of cirrhotic patients undergoing liver transplant evaluation
• HPS is reported in both acute and chronic liver diseases; most commonly in cirrhosis, but also documented with non-cirrhotic portal hypertension and acute/chronic hepatitis
• The severity of HPS does not correlate with the severity of the underlying liver disease
• Patients with HPS have a doubled risk of death compared with patients without HPS who have similar liver disease severity
• Fishman's, p. 1719

Pathogenesis

Vasoactive Substance Imbalance

• Nitric oxide (NO), prostaglandins, vasoactive intestinal peptide, endothelin, calcitonin, glucagon, substance P, atrial natriuretic factor

Nitric Oxide (NO)

• Strongly implicated due to its known pulmonary vasodilatory effects
• Increased NO production causes pulmonary capillary and pre-capillary dilation

Endothelin-1 / ET-B Receptor Pathway (Key Mechanism)

• Endothelin-1 normally causes vasoconstriction via ET-A receptors
• However, when ET-1 binds ET-B receptors, it activates endothelial nitric oxide synthase, causing pulmonary vasodilation
• In experimental HPS models, the ET-B receptor is upregulated, and experimental HPS was reversed by ET-B receptor blockade
• This ET-A/ET-B receptor dichotomy also explains why endothelin antagonism (used in PH) is NOT the treatment for HPS

Angiogenesis

• Aberrant angiogenesis within the pulmonary vasculature contributes to vascular remodeling

Structural Consequence

• Normal pulmonary capillaries: 8-15 μm in diameter
• In HPS, capillaries dilate to 15-100 μm in diameter
• Inhaled oxygen cannot diffuse to the center of these dilated vessels → deoxygenated blood returns to the left heart
• In a rare subset: true arteriovenous malformations form with no alveolar communication
• Fishman's, p. 1719

Mechanisms of Hypoxemia

1. Ventilation-Perfusion (V/Q) Mismatch (Primary Mechanism)

• Low V/Q is the major mechanism in most patients
• Explains the frequent response to supplemental oxygen
• In patients with orthodeoxia, V/Q worsens in the upright position due to increased basilar perfusion
• MIGET studies in cirrhotic patients confirm increased perfusion of low V/Q units even when resting PaO2 is normal

2. Diffusion Limitation

• Vasodilation + high cardiac output (seen with cirrhosis) reduce erythrocyte transit time
• The distance from alveolar gas to the erythrocyte (which tends to localize in the center of the dilated capillary) is increased
• This diffusion-limitation component explains oxygen responsiveness in many patients
• MIGET studies have neither confirmed nor excluded this mechanism definitively

3. True Shunt

• Becomes dominant in severe hypoxemia (PaO2 35-67 mmHg upright in MIGET studies)
• Demonstrated by passage of bubbles or macroaggregated albumin through the lung
• Paradoxically, even in these patients, there may be a response to 100% O2
• This is because at high FiO2, the shunt behaves more like a low V/Q area or diffusion-limited area - the key concept of a "shunt-like" mechanism
• The role of vasoactivity is confirmed by reversibility of V/Q and shunt abnormalities after liver transplantation
• Murray & Nadel's, p. 961-962

Clinical Features

Symptoms

• Most patients present primarily with symptoms of chronic liver disease; dyspnea is a minority presentation
• Platypnea: dyspnea that worsens on standing (pathognomonic when present)
• Orthodeoxia: PaO2 decrease of ≥5% or ≥4 mmHg on standing (due to increased basilar intrapulmonary vasodilation in the upright position)
• Note: platypnea-orthodeoxia is well-described but neither common nor pathognomonic in HPS

Physical Examination

• Spider angiomata
• Digital clubbing
• Peripheral cyanosis
• Signs of chronic liver disease (jaundice, ascites, etc.)

Investigations

• Fishman's, p. 1719-1720

Diagnosis

1. Cirrhosis or portal hypertension (hepatic disease)
2. Widened A-a gradient >15 mmHg (or >20 mmHg if age >64) on room air ABG, patient seated at rest
3. Demonstration of intrapulmonary vasodilation by bubble contrast-enhanced transthoracic echocardiography (CE-TTE)

Bubble Contrast Echocardiography (Most Sensitive Test)

• Agitated saline creates microbubbles ≥15 μm in diameter
• Normally trapped and absorbed in the pulmonary capillary bed
• In HPS: bubbles traverse dilated capillaries and appear in the left atrium 3-6 cardiac cycles after peripheral IV injection
• Intracardiac shunt: bubbles appear within 3 cardiac cycles (earlier)
• CE-TTE is qualitative but most sensitive

Technetium-Labeled Macroaggregated Albumin (Tc-MAA) Lung Perfusion Scan

• Tc-MAA injected IV; detected in lungs AND brain if shunt present
• Fractional brain uptake >5% = abnormal (shunt present)
• Advantages over echo: (a) quantifies shunt magnitude, (b) specific for HPS even in presence of intrinsic lung disease, (c) helps distinguish HPS from parenchymal lung disease hypoxemia
• Disadvantage: less sensitive than CE-TTE; cannot differentiate intracardiac from intrapulmonary shunt
• Shunt magnitude on MAA correlates poorly with degree of hypoxemia

ABG Positioning Protocol

• ABGs should be obtained in the seated position, at rest, on room air
• Also obtain supine ABG to document orthodeoxia if suspected
• Fishman's, p. 1719-1720

Severity Classification (Murray & Nadel)

Clinical Course and Prognosis

• Presence of significant HPS decreases exercise capacity, impairs quality of life, and increases mortality vs. comparable liver disease without HPS
• Average rate of decline: 5 mmHg/year in resting PaO2
• Multicenter prospective studies: HPS doubles the risk of death, independent of age, sex, race, or listing/transplant rates
• Guidelines recommend screening for HPS at the time of liver transplant evaluation
• Fishman's, p. 1720

Management

Medical Therapy

• No medical therapy has been proven effective
• Agents tried and failed (or with mixed results): norfloxacin, beta-blockade, nitric oxide inhibitors, nitric oxide itself, glucocorticoids, COX inhibitors, indomethacin, somatostatin, sorafenib, cyclophosphamide, plasma exchange
• Supplemental oxygen may ameliorate symptoms of hypoxemia but does not treat the underlying cause

Liver Transplantation (Definitive Treatment)

• The only proven therapy to resolve HPS
• Long-term follow-up: significant improvement or complete resolution of HPS in 85% of patients after liver transplantation
• Time to oxygenation improvement is variable - may take up to 1 year or longer
• Pre-operative PaO2 ≤44 mmHg is associated with increased post-operative mortality (retrospective UNOS data)
• However, transplantation may be safely performed even in patients with severe hypoxemia per other reports

MELD Exception Policy (UNOS)

• Because HPS increases mortality disproportionate to MELD score, the United Network of Organ Sharing (UNOS) has a MELD exception for HPS:
  • Confirmed HPS by bubble echocardiography + PaO2 <60 mmHg → eligible for MELD exception
  • Exception: MELD score upgraded to 3 points less than the median MELD at transplant for that donor service area, regardless of underlying liver disease severity
  • This policy has favorably improved outcomes for HPS patients
• Fishman's, p. 1720

Distinction from Portopulmonary Hypertension (POPH)

Summary (One-Liner for Viva)

• Fishman's Pulmonary Diseases and Disorders, 2-Volume Set, pp. 1718-1720 (Chapter on Portopulmonary Hypertension and HPS)
• Murray & Nadel's Textbook of Respiratory Medicine, 2-Volume, Ch. 44 (Mechanisms of Hypoxemia) pp. 961-962; Ch. 126 (Pulmonary Complications of Hepatic Disease) pp. 2299-2310