Hepatopulmonary Syndrome and Portopulmonary Hypertension: Current Status and Implications for Liver Transplantation

Abstract

Purpose of review:

Hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PoPH) are both pulmonary vascular complications of advanced liver disease; however, these syndromes have distinct pathophysiology, clinical implications, and management.

Recent findings:

While both conditions are associated with portal hypertension, HPS results from diffuse pulmonary capillary vasodilation and PoPH results from vasoconstriction and vascular remodeling of pulmonary arteries. In HPS, no medical therapies clearly improve outcomes; however, patients have excellent post-LT outcomes with near uniform reversal of hypoxemia. In PoPH, several medical therapies used in idiopathic pulmonary hypertension have been shown improve pulmonary hemodynamics, symptoms, and potentially LT outcomes; however, further study is needed to determine best treatment regimens, long-term outcomes on medical therapy, and role of LT.

Summary:

While HPS results in severe hypoxemia that is usually reversible by LT, PoPH patients develop progressive pulmonary hypertension that may improve with medical therapy.

Introduction

Over the last three decades, there has been increased recognition of two important pulmonary vascular complications of liver disease, hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PoPH)(1). While there is a growing literature regarding the impact of these complications on survival and post-liver transplant (LT) outcomes, the pathophysiology and optimal therapeutic strategies are still not well characterized. It is, however, clear that HPS and PoPH are distinct syndromes with differing pathophysiology and clinical implications. This review summarizes our current understanding of these conditions and highlights areas in need of further study.

Hepatopulmonary syndrome

Definition and pathophysiology

HPS is defined as an arterial oxygenation defect secondary to intrapulmonary vascular dilatations (IPVD) in the setting of advanced liver disease. It occurs most commonly in patients with cirrhosis and portal hypertension; however, this syndrome has also been observed in non-cirrhotic portal hypertension, acute liver failure, chronic hepatitis, and congenital portosystemic shunts(1–5). The salient feature on histology is gross dilation of the pulmonary precapillary and capillary vessels up to 100 microns in diameter and an increase in the absolute number of dilated vessels(6,7). Evaluation of explanted livers in HPS patients who underwent LT also revealed increased portal venule thrombosis and extensive vascular proliferation within the liver, suggesting that alterations in the vascular bed of the liver may be a contributing factor(8).

The mechanisms that result in pulmonary vasodilation and vascular remodeling are still under investigation with the current understanding highlighted in Figure 1A. Clear predisposing factors include the presence of hepatic dysfunction and decrease venous blood flow throughout the liver(1,6,8), and while either factor appears to be sufficient as HPS is found in patients with non-cirrhotic portal hypertension and in liver disease without portal hypertension(2–5), the disease course may be more severe in patients with both liver injury and portal hypertension, suggesting at least an additive and potentially synergistic effect(9). Imaging in patients with cirrhosis and HPS demonstrates increased obstruction of intrahepatic portal branches, slowed or hepatofugal portal blood flow, and increased large (> 10mm) abdominal portosystemic shunts compared to patients with cirrhosis and no HPS(8). Animal models and small human studies suggest that four main pulmonary changes are important in the pathogenesis of HPS: (i) decreased vascular tone through increased levels of nitrous oxide (NO) and endothelin-1-induced endothelial dysfunction; (ii) monocyte infiltration of pulmonary vasculature possibly due to increased bacterial translocation from the gut; (iii) intrapulmonary arteriovenous shunt formation through vascular remodeling and angiogenesis; and (iv) defective ventilation and diffusing capacity due to alveolar type II cell dysfunction(10).

Figure 1:

Pathophysiology of Hepatopulmonary Syndrome (A) and Portopulmonary Hypertension (B).

ET: endothelin 1; TNF-alpha: tumor necrosis factor- alpha; IL-6: interleukin-6; NO: nitrous oxide; VEGF: vascular endothelial growth factor; CO: carbon monoxide; AV: arteriovenous; V-Q mismatch: ventilation-perfusion mismatch; R -> L: right to left. Figure made using BioRender graphics.

From a physiologic perspective, abnormal oxygenation in HPS is the result of three distinct mechanisms(6). First and most fundamentally, pulmonary capillary dilation allows for rapid passage of mixed venous blood into the pulmonary veins, leading to a ventilation-perfusion mismatch. Second, pleural and pulmonary arteriovenous connections result in direct right-to-left shunting. Finally, the diffusion capacity of oxygen is decreased due to increased diffusion distance in dilated capillaries and/or thickening of capillary walls in prolonged disease(6,11). While portopulmonary venous anastomoses have also been reported, their effect on arterial oxygenation in HPS seems negligible(12).

Diagnosis

The most common presenting symptom in HPS is dyspnea(13), though a common finding in patients with cirrhosis (ie due to ascites, hepatic hydrothorax, PoPH, anemia, and decreased functional status). Similarly, spider nevi, digital clubbing, and cyanosis on physical examination may suggest the presence of HPS, but are non-specific findings and often only seen in patients with advanced disease(13,14). Platypnea (increased dyspnea when sitting up from supine) and orthodeoxia (decrease in oxygen saturation ≥ 5% from supine to upright positioning) are more sensitive and specific findings for HPS, though not pathognomonic(14,15). Further, oxygen saturation of < 96% by pulse oximetry has a sensitivity of only 28% in diagnosing HPS and 71% in detecting those with a partial pressure of arterial oxygen (PaO₂) < 60 mmHg(16,17). As such, HPS should be considered in all patients with chronic liver disease, in patients with unexplained dyspnea, and those undergoing LT evaluation(18,19) (Figure 2).

Figure 2:

Work-up for Hepatopulmonary Syndrome (HPS).

Portal HTN: portal hypertension; LT: liver transplant; TTE: transthoracic echocardiography; IPVD: intrapulmonary vascular dilation; ABG: arterial blood gas; A-a gradient: alveolar-arterial oxygen gradient; PaO2: partial pressure of arterial oxygen; MELD: model for end-stage liver disease.

The diagnostic criteria for HPS includes (i) chronic liver disease, non-cirrhotic portal hypertension, or other causes of cavopulmonary shunt; (ii) arterial hypoxemia on an upright, room air arterial blood gas defined as an alveolar-arterial oxygen (A-a) gradient ≥ 15 mmHg (≥ 20 mmHg if ≥ 65 years old); and (iii) IPVD as demonstrated by positive contrast-enhanced transthoracic echocardiogram (CE-TTE)(1,18) (Table 1). While there has been debate regarding the appropriate definition of arterial hypoxemia(20), the A-a gradient incorporates the atmospheric pressure, adjusting for changes with altitude, and the partial pressure of arterial carbon dioxide, frequently low in patients with cirrhosis due to hyperventilation(1,6). The severity of HPS is defined by room air PaO₂ with ≥ 80 mmHg being mild disease, ≥ 60 to < 80 mmHg moderate disease, ≥ 50 to <60 mmHg severe, and < 50 mmHg very severe(1). Arterial hypoxemia should trigger further diagnostic work-up, including chest x-ray (CXR), pulmonary function tests (PFTs), and chest computed tomography (CT) to evaluate for other causes of hypoxemia. CXR and chest CT may show bibasilar interstitial markings and dilated peripheral pulmonary vessels with increased pulmonary artery to bronchus ratio, respectively(1,21). PFTs may demonstrate decrease diffusion capacity, especially in advanced disease.

Table 1.

Considerations for Liver Transplant Evaluation

Hepatopulmonary syndrome	Portopulmonary hypertension
Screening: All patient should be screened with CE-TTE(16,19) If positive, confirm with arterial blood gas(1)	Screening: All patient should be screened With doppler TTE(18,19) If eRVSP≥ 35 mmHg or other evidence of PAH, confirm with RHC(1,18,19,59)
Diagnostic criteria: Chronic liver disease and/or portal hypertension A-a gradient ≥ 15 mmHg(≥ 20 mmHg if ≥ 65 years old)(1) IVPD by CE-TTE (preferred) or 99mTcMAA lung perfusion scan(22–25)	Diagnostic criteria: Chronic liver disease and/or portal hypertension PAH by RHC: mPAP > 25 mmHg, PVR > 3 woods units, PCWP < 15 mmHg(1)
Indication for liver transplant: MELD exception points if PaO2 <60 mmHg on room air due to HPS regardless of degree of hepatic dysfunction(19)	Not an indication for liver transplant: Eligible if mPAP < 35 mmHg. Considered if mPAP ≥ 35 mmHg with PVR < 240 dynes/s per cm−5 and normal RV function(1,18,19,59)

CE-TTE: contrast-enhanced transesophageal echocardiography; eRVSP: estimated right ventricle systolic pressure; RHC: right heart catherization; A-a gradient : arterial alveolar oxygen gradient; PAH: pulmonary arterial hypertension; PVR: pulmonary vascular resistance; PCWP: pulmonary capillary wedge pressure; IVPD: intrapulmonary vascular dilation; ^99mTcMAA: technetium-99m macroaggregated albumin; MELD: model for end stage liver disease; PaO2: partial pressure of arterial oxygen; RV: right ventricle

CE-TTE is the preferred method for detection of IPVD(1,18,22). Hand-agitated saline is administrated peripherally, creating microbubbles of 10–20 microns, which in patients with diffuse capillary dilations or discrete arteriovenous connections will pass through the pulmonary circulation. CE-TTE is considered positive if microbubbles appear in the left atrium within 3 to 6 cardiac cycles after being visualized in the right heart, and, if within 3 cardiac cycles, suggestive of a right-to-left intracardiac shunt(22,23). An alternative method to detect right-to-left shunting is technetium-99m macroaggregated albumin (^99mTcMAA) lung perfusion scan, which follows the same principle as CE-TTE given aggregates of 20–90 microns(22,24). The study is considered positive if there is >6% uptake in the brain. Data suggests that CE-TTE is more sensitive for the detection of IPVD in adults, though the opposite may be true in children(22,25). Lung perfusion scans are unable to distinguish between intracardiac and intrapulmonary shunting; however, they are useful in quantifying the degree of hypoxemia caused by HPS, especially in patients with multiple causes of hypoxemia(26,27). Pulmonary angiography is not useful in the routine diagnosis of IPVD(1).

Prevalence and natural history

Epidemiologic data on HPS comes largely from transplant centers with reported prevalence from 4 to 47% in patients with chronic liver disease, averaging approximately 30%(13,14,20,27–31). The large variability in prevalence results from differences in the definition of hypoxemia, diagnostic method employed, and population studied. While more common in adult patients, HPS can also occur in children with chronic liver disease with an estimated prevalence of 3–20%(32). The majority of studies in adult patients report no difference in HPS prevalence based on age, gender, sex, or etiology of liver disease(27–30); however, the largest study of 973 patients with HPS listed for LT reported a higher percentage of woman, non-Hispanic white ethnicity, and hepatitis C and non-alcoholic steatohepatitis cirrhosis compared to patients listed without HPS(33). There is also conflicting data on the association between severity of liver disease or portal hypertension and the presence or severity of HPS(14,20,27–29,31,34,35). It is, however, clear that patients with chronic liver disease and HPS have decreased quality of life and greater than two-fold increase in mortality compared to patients without HPS(13,29,30). Without intervention, HPS patients have a 5-year survival of 20% with cause of death generally due to progressive hepatic dysfunction or other comorbidities rather than secondary to progressive hypoxemia; however, the severity of HPS does appear to be predictive of overall survival(29,30).

Management

Currently, the only non-surgical treatment for HPS is supplemental oxygen. There is no proven survival benefit, but, based on experience with other pulmonary conditions and improvement in symptoms, supplemental oxygen is used to maintain oxygen saturation > 88% during rest, exercise, and sleep(18). LT is the only intervention that has been shown to reverse the underlying pathology of HPS and improve survival(27–30,33,36,37). In recent studies, the 5-year post-LT survival rate is 76% for HPS patients, comparable to survival in patients without HPS(33,36,37). Nearly all HPS patients who undergo LT have resolution of hypoxemia and pulmonary vasodilatation, though it may take several months, especially in patients with severe hypoxemia pre-LT(27–30). Recurrence of HPS is extremely rare and the reported cases are in patients who have developed recurrent cirrhosis and portal hypertension(38). POPH developing months to years after the resolution of HPS post-LT has also been reported(39).

There are several important considerations when evaluating patients with HPS for LT, including (i) which HPS patients should be considered; (ii) should these patients receive MELD exception points; and (iii) how can peri-operative risk be reduced in this population. Given the unclear correlation between severity of liver disease and severity of HPS, along with poor outcomes in HPS patients without LT, most major transplant centers and liver societies agree that severe hypoxemia (i.e. room air PaO2 <60 mmHg) from HPS is an indication for LT regardless of degree of hepatic dysfunction(1,18,19); however, there is debate regarding LT for HPS patients with very severe hypoxemia (i.e. room air PaO2 < 50 mmHg). Several multicenter studies, including an analysis of the United Network for Organ Sharing (UNOS) database, have reported increased post-LT morbidity and mortality in HPS patients if PaO2 ≤ 50 mmHg (or ≤ 44 mmHg in one study), raising the question of whether there is a degree of hypoxemia that should be considered too high risk for transplant(27,30,33). On the other hand, 5-year post-LT survival for these patients are still at or above survival for other indications for LT, and these patients have a dismal prognosis without LT. Finally, two more recent single-center studies reported no difference in post-LT mortality based on arterial hypoxemia, suggesting possible center-specific expertise and improved peri-operative care for these patients may play a role(36,37).

Currently, all patients with HPS who have PaO2 < 60 mmHg on room air and no clinically significant pulmonary comorbidities are eligible to receive standard MELD exception points (Table 1). Two analyses of UNOS data under the prior MELD exception system (i.e. 22 points at listing, increased by 3 points every 3 months) showed patients with HPS had lower waitlist mortality than patients without HPS but excellent post-LT outcomes, suggesting an advantage in mild HPS patients while disadvantaging those with severe HPS(33,36). In our current system for standard MELD exception points (listed with median MELD at transplant for center minus 3 points), there remains no difference in MELD exception based on degree of hypoxemia < 60 mmHg.

Regardless, with access to MELD exception points and increased experience caring for these patients, the mortality for HPS patients undergoing LT has decreased over the last few decades. There is, however, still substantial peri-operative morbidity and mortality. Current guidelines suggest the following recommendations to optimize peri-transplant care: (i) continuous intraoperative monitoring of mixed venous oxygen saturation and consideration of extracorporeal membrane oxygenation (ECMO) if < 65%; (ii) early extubation to decrease incidence of ventilator-associated pneumonia; (iii) delivery of 100% inspired oxygen via non-invasive methods to maintain oxygen saturation ≥ 85%; and (iv) goal-directed fluid therapy to avoid fluid overload(18). Risk factors for refractory hypoxemia post-operatively are pre-LT PaO2 ≤ 50 mmHg and intrapulmonary shunt ≥ 20%(40), though as mentioned previously the data are controversial regarding the impact of these risk factors on overall survival. Nayyar et al proposed an algorithm for the management of these patients with refractory hypoxemia, including Trendelenberg positioning, inhaled pulmonary vasodilators, intravenous methylene blue, and, ultimately, consideration of ECMO(41).

Given the morbidity associated with LT and the limited donor organ supply, there is great interest in developing novel interventions and medical therapies for HPS (Table 2). Procedural interventions have possible benefit in select cases, including TIPS(42–44), pulmonary angiography for coil embolization(45,46), and correction of congenital portosystemic shunts in children(5). Pentoxifylline has been shown to reduce hypoxemia in animal models, but human studies have been mixed and no randomized controlled studies exist(47–49). Interestingly, garlic extract has been reported to improve oxygenation and potentially even reverse HPS presumably through altering NO production; however, reports of significant drug-induced liver injury have limited its therapeutic potential(50,51). Further investigation into alternative angiogenesis inhibitors, modulators of endothelin and NO, agents that reduce systemic inflammation and pulmonary monocyte recruitment, and protectors of AT2 cell function are ongoing(10).

Table 2.

Non-surgical Therapies for Hepatopulmonary Syndrome and Portopulmonary Hypertension

Hepatopulmonary syndrome		Portopulmonary hypertension
Supplemental oxygen	Maintain SpO2 >88%, extrapolated from studies of hypoxemia from other etiologies(18)	Supplemental oxygen	Maintain SpO2 >88%, extrapolated from studies of hypoxemia from other etiologies(18)
TIPS	Transient improvement in hypoxemia for 1–3 months, unclear long-term benefit. Possible increased benefit using left branch of PV(42–44)	TIPS	Contraindicated in moderate to severe disease(1,83,84)
Coil embolization	Most useful in cases of severe hypoxemia and discrete AV connections, possible case of PAH following treatment(45,46,105)	Anti-coagulation	No clear benefit in PAH patients(78,79)
Congenital portosystemic shunts	Correction leads to improvement in oxygenation(5)	Calcium channel blockers	No clear benefit in PAH patients, possible harm(80,81)
Inhaled L-NAME	Reduced hypoxemia in experimental models, no clear benefit in human studies(106,107)	Beta-blockers	Consider stopping in moderate to severe disease, banding for variceal bleeding prevention(82)
Inhaled iloprost	No clear benefit in small study(108)	Prostacyclin analogues	IV Epoprostenol, treprostinil with improved hemodynamics in small studies. Ongoing studies of inhaled iloprost and oral berprost, treprostinil, and selexipag(87–90)
Garlic extract	Improved oxygenation in 3 small studies, though reports of DILI(50,51)	Endothelin receptor antagonists	Ambrisentan and bosentan with improved symptoms and hemodynamics in small studies. Currently enrolling RCT of macitentan vs placebo(92–94)
Norfloxacin	Reduced hypoxemia in experimental models, no clear benefit in human studies(109)	Phospho- diesterase-5 inhibitors	Sildenafil with increased functional class, exercise tolerance, and hemodynamics in small studies. Ongoing investigation of oral tadalafil(95–97)
Pentoxyfylline	Reduced hypoxemia in experimental models, no clear benefit in human studies(47–49)	Riociguat	Improved exercise capacity in RCT of PAH patients, including 13 patients with PoPH(85)
Sorafenib	No clear benefit(110)	Combination therapy	Sildenafil or bosentan with prostacyclin showed improved hemodynamics in small studies(91,98)

SpO2: peripheral oxygen saturation; TIPS: transjugular intrahepatic portosystemic shunt; PV: portal vein; AV: arteriovenous; PAH: pulmonary arterial hypertension; L-NAME: L-N-nitro arginine methyl ester; DILI: drug-induced liver injury; RCT: randomized controlled trial; PoPH portopulmonary hypertension

Portopulmonary hypertension

Definition and pathophysiology

PoPH is defined as pulmonary arterial hypertension (PAH) in the setting of portal hypertension with or without intrinsic liver disease. From a pathologic perspective, the pulmonary vascular remodeling of PoPH is identical to idiopathic PAH on autopsy and lung explant studies(52). As detailed in Figure 1B, findings include pulmonary artery vasoconstriction, medial hypertrophy due to smooth muscle proliferation, and, over time, intimal fibrosis and ultimately plexogenic arteriopathy from transition to fibroblasts. These vascular changes are associated with platelet aggregates and in situ thromboses, leading to pulmonary artery obstruction. The exact mechanisms of this progression are not fully understood, but hepatic injury, splanchnic vasodilation, hyperdynamic state, and portosystemic shunting have all been proposed as mechanisms that increase production and/or decreased clearance of circulating factors that facilitate remodeling(53). Patients with cirrhosis and PoPH have increased levels of endothelin-1 and interleukin-6(54,55) and decreased levels of prostacyclin compare to patients with cirrhosis without PoPH. Several other circulating factors, including thromboxane A2, serotonin, vasoactive intestinal peptide, estrogen, and other interleukins, have also been suggested to play a role, though further study is needed(53).

Diagnosis

The most common presenting symptom in PoPH is exertional dyspnea(1). Other non-specific symptoms can include dyspnea at rest, chest pain, palpitations, pre-syncope, and syncope. Physical examination signs include those of pulmonary hypertension (accentuated and split S2, right ventricular heave) and right heart failure (right-sided S3 gallop, jugular venous distension, ascites, and lower extremity edema); however, in mild POPH, patients may be asymptomatic with a normal exam(57). Mild hypoxemia is a common finding, but it is rare to have clubbing, cyanosis, and severe hypoxemia as seen in HPS. CXR may show cardiomegaly and enlargement of central pulmonary arteries with rightward axis and right bundle branch block on electrocardiogram but these are late findings. As such, all patients with unexplained dyspnea and/or hypoxemia and all patients being evaluated for LT should undergo TTE for further evaluation(18,58,59) (Figure 3). Elevated estimated right ventricular systolic pressure (RVSP) requires further hemodynamic assessment by right heart catheterization (RHC) to confirm mean pulmonary arterial pressure (mPAP) and rule out other causes of pulmonary hypertension, There is debate over the appropriate cutoff for RSVP requiring further evaluation, and some have argued that RSVP is too inaccurate be used for screening, especially in this population, as 20% of patients with cirrhosis and portal hypertension have a mildly elevated mPAP due to their high-flow state(58–60). Other TTE parameters have been proposed to increased specificity for PAH(61,62). Currently, the AASLD recommends RHC in patients with RVSP ≥ 45 mmHg and/or other evidence of elevated mPAP on TTE, while the ERS Task Force and Mayo Clinic use RVSP ≥ 50 mmHg(1,18,19,59).

Figure 3:

Work-up for Portopulmonary Hypertension (PoPH).

Portal HTN: portal hypertension; LT: liver transplant; TTE: transthoracic echocardiography; PAH: pulmonary arterial hypertension; mPAP: mean pulmonary artery pressure; PCWP: pulmonary capillary wedge pressure; PVR: pulmonary vascular resistance; CI: cardiac index; WHO II-IV: World Health Organization functional class II-IV; RVSP: right ventricular systolic pressure; RV: right ventricle; ePVR: estimated pulmonary vascular resistance; ERS: European Respiratory Society

Diagnostic criteria for PoPH include a clinical diagnosis of portal hypertension and evidence of PAH as defined by (i) mPAP > 25 mmHg on RHC; (ii) peripheral vascular resistance (PVR) > 3 Woods units (240 dynes/s per cm⁻⁵); and (iii) pulmonary capillary wedge pressure (PCWP) < 15 mm Hg (Table 1). The severity of PoPH is based on RHC and graded as mild (25 ≤ mPAP < 35 mmHg), moderate (35 ≤ mPAP < 45 mmHg), and severe (mPAP ≥ 45 mmHg(1). Notably, not all patients with an elevated mPAP have PAH, as the high-flow state observed in many cirrhotic patients can lead to high mPAP but normal PVR. Left-sided volume overload can also lead to an elevated mPAP, but normal PVR and increased PCWP. Similarly, patients with PoPH can also have left-sided dysfunction, which, especially in the setting of concurrent renal dysfunction, can worsen overall volume overload and lead to an elevated PCWP. In these cases, a transpulmonary gradient (i.e. mPAP – PCWP) greater than 10 mmHg is suggestive of increased pulmonary resistance and should be re-evaluated once volume status is optimized(1). Work-up also requires excluding other causes of PAH with PFTs, sleep study, ventilation-perfusion scan, antinuclear antibody, screening for human immunodeficiency virus, and any other testing indicated by the clinical presentation(57). Compared with idiopathic PAH, patients with PoPH generally have higher cardiac output (CO) and less severe PAH by mPAP and PVR(63).

Prevalence and Natural History

PoPH, more rare than HPS, is present in approximately 0.7% of patients with cirrhosis, 2% of patients with portal hypertension, and 5–9% of patients undergoing LT evaluation(59,64–67), and accounts for 5–15% of PAH cases(68,69). While uncommon, it can also occur in children(70). Female sex, autoimmune etiologies of liver disease, elevated estradiol levels, large spontaneous or surgical portosystemic shunts, portocaval shunts, hepatofugal blood flow, and splenectomy have been reported as possible risk factors for PoPH(71–74); however, the severity of liver disease determined by MELD or CTP scores and the degree of portal hypertension assessed by hepatic venous pressure gradient (HVPG) do not correlate with the presence or severity of PoPH(65,71). Without treatment, patients have a 1- and 5-year survival between 35–46% and 4–14%, respectively(75,76), and cause of death is split between right heart failure and complications of liver disease. As demonstrated in the REVEAL registry, PoPH patients have worse survival than idiopathic PAH despite higher CO and lower PVR (40% vs 60% survival at 5 years, respectively)(63). Notably, given the baseline high output state in patients with cirrhosis, the finding of “normal CO” in a patient with PoPH is actually a sign of significant RV dysfunction, which may partially explain this discrepancy. Additionally, the investigators noted a delay in initiating PAH-specific therapy in the PoPH group, likely due to the exclusion of PoPH patients from randomized controlled trials of PAH treatments.

Management

Several studies suggest improved survival and quality of life after initiation of medical therapy in PoPH patients regardless of LT; however, these were retrospective cohort studies, often with inclusion of historical controls, and small uncontrolled prospective studies(77). The true impact of medical therapy on survival and optimal PAH-directed regimens are still unknown and are extrapolated from data in other etiologies of PAH (Table 2). As there is uncertain benefit from anticoagulation in PAH and increased bleeding risk in patients with portal hypertension, anticoagulation is not recommended(78,79). Calcium channel blockers should be avoided due to lack of established benefit and potential to worsen portal hypertension(80,81). In moderate-to-severe PoPH, withdrawal of beta-blockade should be considered along with an alternative method for prevention of variceal bleeding (i.e. band ligation)(82). Severe PoPH and heart failure are absolute contraindications to TIPS, and moderate PoPH is a relative contraindication and TIPS should be avoided in these patients if possible(83,84).

In addition to diuretic therapy, PAH-directed therapy should be considered for PoPH patients with mPAP ≥ 35 mmHg (Table 2)(1,59). The only randomized controlled trial that included PoPH patients evaluated riociguat, a soluble guanylate cyclase stimulator, and included 13 PoPH patients of which 11 received riociguat. As such, no conclusions could be made about efficacy in PoPH, though overall patients with PAH had improved exercise capacity(85). Continuous infusion of prostacyclin analogues (esoprostenol, treprostinil) have been used to improve hemodynamics as a bridge to LT with success, though there are technical challenges to continuous infusion in patients prone to hepatic encephalopathy(86–89). Investigation of inhaled iloprost, oral beraprost (not available in the US), and several other oral prostacyclin anaglogues (oral treprostinil, selexipag) are ongoing(90,91). Oral endothelin receptor antagonists (ambrisentan, bosentan) have shown promising results in small studies of PoPH patients, including improved hemodynamics, decreased symptoms, and, for bosentan, improved survival compared to those treated with inhaled iloprost(92–94). Notably, approximately 10% of PAH patients have had liver toxicity with bosentan, though this has not been reported with ambrisentan to date. Similarly, phosphodiesterase-5 inhibitors (sildenafil, possibly tadalafil) may increase functional class, exercise tolerance, and improve hemodynamics in PoPH patients(95–97). Sildenafil has also been used in combination with prostacylins and bosentan with improved hemodynamics noted in 11 and 7 patients, respectively(91,98).

Most published studies on LT outcomes in PoPH patients are small, retrospective, and uncontrolled, using variable hemodynamic cutoffs and medical therapy. Overall, post-LT outcomes are highly variable, though there is increasing literature to support the use of medical therapy as a bridge to LT and normalization of pulmonary hemodynamics following LT in some patients.(28,67,76,86,98,99). Studies prior to availability of medical therapy for PoPH reported no increase in perioperative mortality if mPAP was < 35 mmHg, but significantly increased mortality if mPAP ≥ 35 mmHg (100% mortality if mPAP ≥ 50 mmHg in one study)(28,100,101). As such, PoPH patients are generally considered for LT if their baseline mPAP is < 35 mmHg, though the presence of mild PoPH itself is not an indication for transplant (Table 1)(1,18,59). Patients with PoPH and mPAP ≥ 35 mmHg are recommended for medical optimization first with consideration for listing and allocation of MELD exception points if on treatment mPAP < 35 mmHg and PVR < 5 Woods units (400 dynes/s per cm⁻⁵⁾. Patients can also be considered for LT if mPAP ≥ 35 on treatment but PVR < 240 dynes/s per cm⁻⁵ with normal RV function. Per the ERS Task force, mPAP ≥ 45 mmHg is considered an absolute contraindication to transplant, though some transplant centers use mPAP ≥ 50 mmHg(18,59). With these general guidelines, 5-year post-LT survival for PoPH patients is 63–67% based on available data(76,86,102,103). Analysis of the UNOS database from 2006 to 2012 reported increased waitlist and post-LT mortality in patients who met PoPH criteria and received MELD exception points compared to patients without PoPH (5-year survival 64% vs 70%, respectively)(103). Notably, in this study and others, mortality is high within the first year- some studies reporting over 50% of deaths within 30-days of LT and up to 36% in-hospital mortality(28,103).

There are several important considerations to optimization of post-LT outcomes for PoPH patients. All patients should have a pulmonary artery catheter placed before the abdominal incision is made(1,18), and, if mPAP ≥ 35 mmHg, a full hemodynamic assessment with CO, PCWP, and PVR repeated, with case cancellation if mPAP ≥ 45–50 mmHg or acute therapy is unable to lower mPAP to < 40 mmHg with normal PVR. PAH-targeted therapy should be continued during the transplant surgery with transesophageal echocardiography (TEE) available if needed for RV monitoring. Reperfusion of the transplanted liver is an especially critical time during the procedure, as the acute rise in PA pressures may precipitate RV failure(104). Various interventions have been utilized including inhaled NO, intravenous prostacyclin, milrinone, and ECMO. RV assist devices should not be used due to potential for PA rupture. Immediately following transplant, major changes in PAH-directed therapy should be avoided and patients should be monitored with serial TTE every 4 to 6 months and consideration for weaning of medical therapy in patients with normalized hemodynamics. There is no clear method for the execution of this process and studies report variable outcomes with 33–82% of patients weaned off therapy over months to years post-LT(67,76,86,98,99).

Conclusions

While often discussed together, HPS and PoPH are distinct pulmonary vascular sequelae of liver disease with unique pathophysiology and clinical implications. HPS is characterized by arterial hypoxemia secondary to IPVD, and, while there are no effective medical therapies, LT usually facilitate complete resolution of hypoxemia and outcomes are excellent. On the other hand, patients with PoPH rarely have significant hypoxemia and instead develop progressive right heart failure from PAH. PoPH patients have highly variable outcomes with LT and high peri-operative mortality; however, several promising medical therapies are available to help minimize these risks in well selected LT candidates. Further study is needed in both HPS and PoPH to determine the best allocation strategy to minimize waitlist mortality and maximize post-LT survival.