|Home | About | Journals | Submit | Contact Us | Français|
Portopulmonary hypertension is defined as the combination of pulmonary arterial hypertension with portal hypertension and presents management complications in patients awaiting liver transplantation. The combination of these vascular disorders has a marked impact on mortality. At present the recommendations for management are limited because of the paucity of definitive clinical trials. We have reviewed the available data on prevalence, diagnosis and treatment. It is clearly time to more formally approach the study of this patient population.
Portopulmonary hypertension (PoPH) defined as the combination of pulmonary hypertension with portal hypertension was initially described in 1951 by Mantz and Craige. The original description was largely of academic interest as there were limited management options for end-stage liver disease (ESLD). The advent of liver transplant (LT) significantly increased the recognition of PoPH as a complication of ESLD as well as the use of therapies to optimize its management to potentially allow for LT. In 2008, there were 6318 liver transplants performed in the US and there are currently approximately 15000 patients on the transplant waiting list. Given the recent increase in liver transplants, a more thorough understanding of the pathophysiology and management of PoPH has become essential
Pulmonary hypertension is defined as hypertension in the pulmonary arterial bed. The 2003 World Health Organization (WHO) classification separates the various forms of pulmonary hypertension [Simonneau et al. 2004]. Notably, PoPH is in WHO Group I, also defined as pulmonary arterial hypertension. PoPH requires both the presence of pulmonary hypertension as well as portal hypertension. Diagnosis requires right heart catheterization and demonstration of: elevated mean pulmonary artery pressure (mPAP) (>25 mmHg at rest, >30 mmHg with exercise), normal pulmonary artery occluding pressure (PAOP) < 15 mmHg or elevated transpulmonary gradient (mPAP — PAOP > 12 mmHg and increased pulmonary vascular resistance (PVR) >240 dynes.sec.cm-5 [Rodriguez-Roisin et al. 2004].
There are relatively few studies that have investigated the relationship between right heart catheterization and noninvasive modalities such as Doppler echocardiography [Cotton et al. 2002; Kim et al. 2000; Castro et al. 1996]. These results have demonstrated a poor correlation between invasive and noninvasive studies and as a result, right heart catheterization is necessary to make a diagnosis [Krowka et al. 2006]. PoPH must be separated from other pulmonary processes associated with liver disease which can cause similar symptoms including hepatopul-monary syndrome, hepatic hydrothorax and high cardiac output states associated with ESLD.
Many patients with portal hypertension manifest a high flow circulatory state as a result of splanchnic vasodilation resulting in an elevated cardiac output (CO) and mPAP; PVR remains normal secondary to pulmonary vasodilation. The PAOP can also be increased as a result of increased blood volume resulting in mildly increased mPAP. In these circumstances the transpulmonary gradient and PVR are normal and this reflects the high flow state of portal hypertension not true PoPH. Approximately 30–50% of patients with portal hypertension demonstrate this high flow state [Iwakiri and Groszmann, 2006; Simonneau et al. 2004] and its management is different than that of PoPH [Krowka et al. 2000]. Finally, the diagnosis of PoPH assumes the exclusion of other diseases within WHO Group I pulmonary hypertension such as collagen vascular disease, toxins and HIV as well as fully evaluating for other potential causes such as valvular heart disease and chronic thromboembolic disease.
The prevalence of PoPH is difficult to estimate as it is often screened for in patients undergoing a transplant evaluation which introduces selection bias. Kawut et al.  in a recent multivariate analysis observed that female sex and presence of autoimmune liver disease had a significant association with PoPH. The female association is also observed in pulmonary arterial hypertension. There was no association between age and PoPH in this study. PoPH has been estimated to occur with a prevalence of 16.1% in patients with cirrhosis and refractory ascites [Benjaminov et al. 2003] and between 0.25% and 4% in patients with cirrhosis but without refractory ascites [Hoeper et al. 2004; Castro et al. 1996]. However, there does not appear to be a correlation between the severity of the liver disease or of the portal hypertension with the degree of pulmonary hypertension [Hadengue et al. 1991].
Survival in untreated PoPH does depend on the severity of the underlying pulmonary hypertension and portal hypertension. Patients not undergoing LT had a 1-year survival rate of 85% and 3-year survival rate of 38% [Kawut et al. 2005]. Survival data before therapy with vasodilators was available demonstrated a mean 15-month survival after diagnosis [Robalino and Moodie, 1991]. A recent longitudinal analysis of 154 patients with PoPH by Le Pavec et al.  utilizing multivariate analysis identified severity of cirrhosis and impairment in cardiac index as independent risk factors for mortality. Interestingly, the New York Heart Association (NYHA) functional class of heart failure was not related to survival in this study [Le Pavec et al. 2008].
Unfortunately, there is a significant mortality associated with LT in PoPH. Mortality appears to increase with worsening pulmonary hypertension. Patients with mPAP < 35 mmHg do not have increased risk, whereas mortality increased to approximately 50% for mPAP 35–49 mmHg and is greater then 70% for mPAP > 50 mmHg [Krowka et al. 2000]. The cause of death in these cases is attributed primarily to right ventricular dysfunction and the mortality persisted to 9-months post-transplant.
Pulmonary arterial hypertension results from several factors working simultaneously and synergistically including an asymmetry in regulators of vascular tone favoring vasoconstriction, endothelial remodeling as a result of increased pulmonary blood flow, proliferation of arterial smooth muscle, and microvascular thrombosis. This pathophysiology is also involved in PoPH. Notably, the histopathologic findings in the pulmonary vascular bed are identical in pulmonary arterial hypertension and PoPH [Schraufnagel and Kay, 1996].
The specific pathogenesis of PoPH is not well characterized but a general theory is that a blood-borne substance which is normally metabolized by the liver is able to reach the pulmonary circulation secondary to liver dysfunction and portosystemic collaterals causing PoPH [Lebrec et al. 1979]. There are numerous potential candidates for this substance including endothelin 1 (ET-1), vasoactive intestinal peptide, serotonin, thromboxane A2, interleukin 1, glucagon, and secretin [Egermayer et al. 1999; Mandell and Groves, 1996; Panos and Baker, 1996]. There has been particular interest in ET-1 as levels of this vasoconstrictor have been demonstrated to be elevated in cirrhotics with PoPH as opposed to cirrhotics without PoPH raising suspicion that ET-1 may play a causal role in PoPH [Benjaminov et al. 2003]. There has also been evidence of increased production of ET-1 by the liver in cirrhotic patients [Alam et al. 2000]. Finally, decreased levels of pulmonary prosta-glandin I2 synthase (PGI2) have been demonstrated in patients with PH including patients with cirrhosis as the underlying etiology [Tuder et al. 1999].
Until recently there has been relatively little data to guide the clinical therapy in PoPH as most randomized trials in the field of pulmonary arterial hypertension have specifically excluded this group of patients [Badesch et al. 2004]. However, therapy of PoPH has been based on the principles that guide pulmonary arterial hypertension management. In PoPH the use of vasodilator therapy has been assessed in case reports and case series; there are no prospective randomized controlled trials. Calcium channel blockers are contraindicated both during diagnostic right heart catheterization and in the management of PoPh as they can potentially cause systemic vasodilatation and precipitous drop in cardiac output. This sequence of events can instigate acute shock which is difficult to resuscitate and markedly increases the risk of the procedure.
Endothelin receptor antagonists (ETRA) block (either selectively or nonselectively) the ability of endothelin 1 (ET-1) to bind with the endothelin receptor A (ETA) or B (ET-B) on endothelial and arterial smooth muscle cells, inhibiting vasoconstriction. ETA is located on pulmonary vascular smooth muscle cells and induces a powerful vasoconstrictive response [Masaki, 1998]. In normal pulmonary vasculature ET-B is located on endothelial cells and mediates vasodilatation via production of prostacyclin and nitric oxide as well as clearance of ET-1. In PAH, ET-B is thought to mediate vasoconstriction through a separate population of receptors that are upregulated on vascular smooth muscle cells [Hirata et al. 1993].
There are currently two endothelin receptor antagonists; bosentan (Tracleer®) which is a sulfonamide-based nonselective agent and ambrisentan (Letairis®) which selectively inhibits endothelin A. Currently, only the use of bosentan has been reported in the management of PoPH.
Hoeper et al.  performed a retrospective case series including 31 consecutive patients with Child class A or B cirrhosis and severe PoPH (mPAP > 50mmHg). Patients received either bosentan (18) or iloprost (13) in a nonran-domized fashion and were followed over a 3-year period to assess effects on survival, exercise capacity and hemodynamic measurements. In the bosentan group the 1-, 3- and 5-year survival rates were 94%, 89% and 89%, respectively, versus 77%, 62% and 46% for iloprost over the same time period. Event-free survival (i.e. survival in the absence of transplantation, right heart failure or clinical worsening requiring additional treatment for PoPH) was also significantly better in the bosentan group (p = 0.017). Bosentan demonstrated a significant improvement in 6-minute walk test compared to baseline at 1 year (p< 0.001) although there was no difference when compared to iloprost. Finally, bosentan demonstrated improvement in PVR with a net decrease of 345 ±361 dynes.sec.cm-5 [Hoeper et al. 2007].
Therapy was well tolerated in both groups with only one patient in the bosentan group experiencing an increase in liver function tests. The therapeutic effects of bosentan seem to support the importance of ET-1 in the pathogenesis of PoPH. Despite the absence of notable liver toxicity in this particular study, bosentan has been associated with liver injury in PAH. A dose-dependent rise in liver function tests was observed in patients receiving bosentan in the BREATHE-1 trial [Rubin et al. 2002]. Ambrisentan which has not yet been utilized in patient with PoPH has recently been reported to have minimal effect on liver function tests in patients with PAH who had discontinued the use of other ETRAs because of elevations in serum aminotransferases [McGoon et al. 2009].
Epoprostenol (Flolan®) was one of the first agents used in the management of IPAH and subsequently PoPH. It is a potent pulmonary and systemic vasodilator as well as inhibitor of platelet aggregation delivered by continuous intravenous infusion [Kuo et al. 1997; Barst et al. 1996, 1994]. Krowka et al.  was one of the first to look at the utility of intravenous epoprostenol in a series of 15 patients with PoPH (mPAP > 35mmHg). Statistically significant improvement in PVR (34%) and CO (±2L) was noted during administration of epoprostenol during right heart catheterization. Long-term use of epoprostenol with repeat right heart catheterization in six patients demonstrated further significant improvement in PVR (–47% from baseline and –31% from the acute change) [Krowka et al. 1999]. Fixet al.  retrospectively analyzed pulmonary hemodynamics and liver chemistries in patients with moderate-to-severe PoPH in 19 patients treated with epoprostenol and 17 patients not treated with epoprostenol (although 10 of this latter group were treated with other pulmonary vasodilators). They found epoprostenol therapy led to significant improvement in mPAP (48.1–36.1 mmHg, p< 0.001), PVR (632 to 282dynes.sec.cm-5, p< 0.001) and cardiac out put (5.5–7.7l/min, p = 0.0009) after a median of 15.4 months of therapy. Notably, there was no follow-up catheterization data in the group not treated with epo-prostenol and no difference in survival between the two groups (hazard ratio 0.85, p = 0.77). Finally, there were no significant changes to liver function tests [Fix et al. 2007]. Earlier studies have also indicated continuous epoprostenol therapy results in improvement in NHYA functional class after 1 year of therapy [McLaughlin et al. 1999].
The side-effect profile of epoprostenol is increased by the mechanism required for its continuous delivery and includes infection, thrombosis, and concern for rebound pulmonary hypertension in the setting of disruption of drug delivery. There have also been reports of hypers-plenism in patients with PoPH treated with epoprostenol [Findlayet al. 1999]. It is not clear if the risk profile for epoprostenol is greater in patients with PoPH as opposed to other forms of PAH.
Iloprost (Ventavis®) is a prostacyclin analog that has been used in the management of PoPH as both a nebulized and intravenous formulation. A comparative study with bosentan [Hoeper et al. 2007] demonstrated that iloprost is better than no therapy but inferior to bosentan. A case report using inhaled iloprost in PoPH demonstrated improvement in mPAP, PVR and 6 minute walk test [Halank et al. 2004]. A subsequent case report utilizing intravenous iloprost as rescue therapy was effective and allowed for LT by reducing mPAP from 54 mmHg to 38 mmHg after 3 months of continuous infusion [Minder et al. 2004].
In our own experience we have treated five patients with PoPH (mPAP > 40 mmHg) with the combination of a prostanoid with sildenafil (one with epoprostenol; four with intravenous treprostinil). In all cases, patients had significant improvement in their pulmonary vascular hemo-dynamics (mPAP decreased to less than 35 mmHg) after 3 months of therapy. To date, three of the patients have been successfully transplanted. The other patients are cleared for transplant but have been stable and have not been transplanted because of improved MELD scores. There were no adverse changes to liver functions tests.
Sildenafil (Revatio®) inhibits cyclic guanosine monophosphate (cGMP) phosphodiesterase type 5 enzyme which normally functions to metabolize cGMP. cGMP induces vasodilation through relaxation of arterial smooth muscle cells. Sildenafil is an oral agent with selectivity to the pulmonary arterial bed as a result of the high concentration of cGMP present there. Makisalo et al.  reported the first use of sildenafil as monotherpay for a patient with PoPH in preparation for liver transplant. Mean pulmonary pressures were decreased from 56 mmHg to 28–31 mmHg and PVR from 398 to 179 dynes.sec.cm-5 after 4 weeks of therapy with no notable liver function abnormalities [Makisalo et al. 2004]. The patient successfully underwent LT however did experience postoperative bleeding which was thought to be in part secondary to the inhibition of collagen induced platelet aggregation caused by sildenafil [Berkels et al. 2001].
Reichenberger et al.  treated 14 patients with severe PoPH (mPAP 55 mmHg and PVR 1070 dynes.sec.cm-5) of which 8 were treated with sildenafil alone and an additional 6 patients in which it was used as adjunctive therapy with prostanoids (either iloprost or treprostinil). Three-month interval right heart catheterization demonstrated significant improvement in mPAP (46 mmHg p = 0.01) as well as a significant decrease in PVR (698 dynes.sec.cm-5 p=<0.05). However, these improvements were not observed at the 12-month interval catheterization [Reichenberger et al. 2006]. Reported improvement in pro-brain natriuretic peptide in this study is interesting but of unclear clinical significance. Two recent small, open-labeled and uncontrolled trials have also suggested a benefit to the use of sildenafil for reduction in mPAP, PVR and as a bridge to LT [Gough and White 2009; Hemnes and Robbins, 2009].
There is limited data reported on the use of combination pulmonary vasodilator therapy. Austin et al.  treated a single patient with severe PoPH (mPAP = 70 mmHg) with iloprost, bosentan and sildenafil. The therapies were added sequentially beginning with inhaled iloprost which did not result in significant improvement of mPAP followed by addition of sildenafil which resulted in reduction of mPAP to 45 mmHg (after 14 days of combined therapy). The addition of bosentan for an additional month resulted in a further reduction of mPAP to 32 mmHg. There was no hepatotoxicity reported [Austin et al. 2008].
The absence of randomized controlled data makes the current management of PoPH challenging. At present there are three unique classes of agents which have each demonstrated the ability to improve the mPAP and PVR in PoPH. It is difficult to assess the superiority of one agent over another as the data reports on small numbers of patients in a retrospective and uncontrolled fashion.
The utilization of combination therapy, although promising, requires further evaluation before its approach can be recommended. We would encourage a referral to practitioners experienced in the care of pulmonary hypertension and the use of these complex medications. Prospective randomized controlled trials between agents are required to further delineate the efficacy of these classes as well as enhance our understanding of their side-effect profile in addition to determining which, if any, of the classes are superior in the management of PoPH.
Dr. Waxman receives research funding from Gilead Sciences and is a clinical trials investigator for Gilead Sciences, United Therapeutics, Pfizer, and Actelion. Dr. Troy has no disclosures.
Patrick J. Troy, Pulmonary and Critical Care Unit, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, 02114, USA.
Aaron B. Waxman, Pulmonary and Critical Care Unit, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, 02114, USA ; Email: gro.srentrap@namxawba.