Pulmonary arterial hypertension: pathogenesis and clinical managementBMJ 2018; 360 doi: https://doi.org/10.1136/bmj.j5492 (Published 14 March 2018) Cite this as: BMJ 2018;360:j5492
- Thenappan Thenappan, assistant professor1,
- Mark L Ormiston, assistant professor2,
- John J Ryan, assistant professor3,
- Stephen L Archer, professor2
- 1Department of Medicine, University of Minnesota, Minneapolis, MN, USA
- 2Department of Medicine, Queen's University, Kingston, ON, Canada
- 3Department of Medicine, University of Utah, Salt Lake City, UT, USA
- Correspondence to: S L Archer
Pulmonary hypertension is defined as a resting mean pulmonary artery pressure of 25 mm Hg or above. This review deals with pulmonary arterial hypertension (PAH), a type of pulmonary hypertension that primarily affects the pulmonary vasculature. In PAH, the pulmonary vasculature is dynamically obstructed by vasoconstriction, structurally obstructed by adverse vascular remodeling, and pathologically non-compliant as a result of vascular fibrosis and stiffening. Many cell types are abnormal in PAH, including vascular cells (endothelial cells, smooth muscle cells, and fibroblasts) and inflammatory cells. Progress has been made in identifying the causes of PAH and approving new drug therapies. A cancer-like increase in cell proliferation and resistance to apoptosis reflects acquired abnormalities of mitochondrial metabolism and dynamics. Mutations in the type II bone morphogenetic protein receptor (BMPR2) gene dramatically increase the risk of developing heritable PAH. Epigenetic dysregulation of DNA methylation, histone acetylation, and microRNAs also contributes to disease pathogenesis. Aberrant bone morphogenetic protein signaling and epigenetic dysregulation in PAH promote cell proliferation in part through induction of a Warburg mitochondrial-metabolic state of uncoupled glycolysis. Complex changes in cytokines (interleukins and tumor necrosis factor), cellular immunity (T lymphocytes, natural killer cells, macrophages), and autoantibodies suggest that PAH is, in part, an autoimmune, inflammatory disease. Obstructive pulmonary vascular remodeling in PAH increases right ventricular afterload causing right ventricular hypertrophy. In some patients, maladaptive changes in the right ventricle, including ischemia and fibrosis, reduce right ventricular function and cause right ventricular failure. Patients with PAH have dyspnea, reduced exercise capacity, exertional syncope, and premature death from right ventricular failure. PAH targeted therapies (prostaglandins, phosphodiesterase-5 inhibitors, endothelin receptor antagonists, and soluble guanylate cyclase stimulators), used alone or in combination, improve functional capacity and hemodynamics and reduce hospital admissions. However, these vasodilators do not target key features of PAH pathogenesis and have not been shown to reduce mortality, which remains about 50% at five years. This review summarizes the epidemiology, pathogenesis, diagnosis, and treatment of PAH.
Pulmonary hypertension is defined as a resting mean pulmonary artery pressure (mPAP) of 25 mm Hg or above. The classification system proposed by the Fifth World Symposium on Pulmonary Hypertension attempts to guide the clinical approach to pulmonary hypertension by dividing patients into five groups: group 1—pulmonary hypertension due to pulmonary vascular disease; group 2—pulmonary hypertension due to left heart disease; group 3—pulmonary hypertension due to lung disease or hypoxia; group 4—pulmonary hypertension due to chronic thromboembolic disease; and group 5—a miscellaneous collection of pulmonary hypertension syndromes caused by a variety of disorders, including hemolytic anemias and sarcoidosis (fig 1).1 In principle, patients in each of these groups share pathophysiology, prognosis, and therapeutic response; in reality, tremendous heterogeneity exists within each group.
This review focuses on group 1 pulmonary hypertension, also known as pulmonary arterial hypertension (PAH). PAH is a disease of the cardiopulmonary unit, affecting the pulmonary arterial and venous circulation and the right ventricle. Obstructive, hyperproliferative, vascular lesions, vasoconstriction of pre-capillary arterioles, and venous obstruction (in some forms of group 1 disease) increase pulmonary vascular resistance (PVR), increase right ventricular afterload, and promote right ventricular failure (RVF), which is the leading cause of death in PAH.2 Current treatments for PAH are primarily pulmonary vasodilators. They ameliorate symptoms and reduce hospital admissions, but they are expensive and not curative.
This review summarizes the epidemiology, diagnostic evaluation, and treatment of PAH. It also examines recent advances in basic science, noting potential therapeutic targets and future research questions.
Sources and selection criteria
We identified references for this review by doing a PubMed search for years 2007-17. We only included peer reviewed articles written in English. We used the following search terms in combination with the term “pulmonary hypertension”: “mechanisms”, “experimental models”, “diagnosis”, “epidemiology”, “survival”, “pulmonary arterial hypertension”, “guidelines”, “classification”, “imaging”, “hemodynamics”, and “therapy”. We included articles on the basis of the quality of study design and size, favoring randomized controlled trials, reports from large registries, and guidelines. For the pathogenesis section, we selected papers in reputable journals and highlighted evidence in which concordant data were available from more than one research group. We also included highly cited papers written before 2007. We excluded case reports and papers in non-peer reviewed journals. We screened approximately 1000 articles of evidence classes I-IV and included about 740 for detailed review.
Epidemiology and natural history of PAH
The incidence of PAH ranges from 2.0 to 7.6 cases per million adults per year, and its prevalence varies from 11 to 26 cases per million adults (table 1). The incidence of PAH is fourfold higher in women than in men, but survival is paradoxically worse in men with PAH.1314 Nearly half of the patients have idiopathic PAH (IPAH), heritable PAH, or anorexigen induced PAH, with connective tissue disease associated PAH (APAH) being the second most common subgroup. The National Institutes of Health (NIH) registry from the 1980s was the first major epidemiological study of PAH.3 Subsequently, 10 major registries have described the epidemiology of PAH (table 1).456789101112 These modern registries have provided two novel insights.
Firstly, in the current era, patients with IPAH, heritable PAH, or anorexigen associated PAH are older than those in the NIH registry (mean age at diagnosis 45-65 versus 36 years). Potential explanations include increased awareness of PAH due to the availability of PAH specific therapies and widespread use of Doppler echocardiography, leading to wider recognition of PAH in older patients. However, as the mean age of the group 1 cohort increases, the risk of misdiagnosis increases, as group 2 pulmonary hypertension in patients with heart failure and preserved ejection fraction (HFpEF) is much more prevalent and shares several echocardiographic abnormalities.
Secondly, long term survival in patients with PAH has significantly improved in the past two decades. The median survival is now six years, compared with 2.8 years in the 1980s.15 Similarly, one year survival rates for PAH patients range from 86% to 90%, improved from 65% in the 1990s.161718 Improved awareness of PAH, availability of PAH specific therapies, long term anticoagulation therapy, and better management of RVF are possible reasons for improved survival.
Nonetheless, PAH still causes a substantial clinical and economic burden. Although PAH related hospital admissions decreased between 2001 and 2012, the mean charge and length of stay per PAH related admission have increased with no significant decline in inpatient mortality.19
Diagnosis of PAH
PAH is defined as a resting mPAP of 25 mm Hg or above, pulmonary capillary wedge pressure (PCWP) below 15 mm Hg, and PVR above 3 Wood units20), in the absence of more prevalent causes of pulmonary hypertension such as left heart disease, chronic lung disease,21 or venous thromboembolism.
Most patients are referred for evaluation for pulmonary hypertension after detection of increased right ventricular systolic pressure (RVSP) by Doppler ultrasound. RVSP is calculated from Bernoulli’s principle, on the basis of the velocity of the tricuspid regurgitant jet (RVSP=4V2, where V is the maximum tricuspid regurgitant jet velocity) plus the estimated right atrial pressure (RAP). RAP is determined by observing the jugular venous pressure or, more commonly, by assessing the dimensions of the inferior vena cava, at baseline and in response to sniffing. On the basis of comparison with right heart catheterization (RHC), an inferior vena cava maximum diameter above 19 mm or a collapse of 30% or less on sniff testing implies an RAP above 10 mm Hg.22 The tricuspid regurgitant estimate of RVSP can underestimate or overestimate pulmonary artery systolic pressure and does not measure mPAP or indicate whether the pulmonary hypertension relates to intrinsic pulmonary vascular disease.23
Hence, echocardiographic findings suggestive of pulmonary hypertension should prompt referral to specialized centers for further evaluation to confirm the hemodynamic abnormality and classify the patient into the correct pulmonary hypertension group.24 This is important, as treatment and prognosis vary considerably by the cause of pulmonary hypertension. Figure 2 summarizes the diagnostic algorithm for PAH. Pulmonary function tests and overnight oximetry are needed to evaluate for chronic hypoxic lung disease. The ventilation/perfusion scan is the diagnostic test of choice for excluding chronic pulmonary thromboembolic pulmonary hypertension (CTEPH), as it is more sensitive than a computed tomography pulmonary angiogram in identifying distal CTEPH.25 An invasive pulmonary angiogram or computed tomography angiogram of the chest may be done for patients with a high probability ventilation/perfusion scan for confirmatory diagnosis. Exclusion of CTEPH is important, as it can be cured by surgical pulmonary thromboendarterectomy in suitable candidates. Serological testing helps to identify associated causes of PAH, such as connective tissue disease or cirrhosis. RHC is mandatory for confirming the diagnosis of PAH and is required before PAH targeted treatment is started. RHC also assesses the severity of pulmonary hypertension, evaluates potential left heart disease, and identifies the subset of patients who respond favorably to acute vasodilators. Figure 3 illustrates the test results in a typical patient with PAH.
Many patients with PAH have more than one condition contributing to their total pulmonary hypertension burden. Endothelial dysfunction and hypoxia, although not causal, often exacerbate PAH. The clinician is often challenged to determine whether a patient has PAH that is somewhat exacerbated by concomitant mild lung disease, left heart disease, or pulmonary embolism. Conversely, many group 2 patients, such as those with mitral stenosis, and group 3 patients with chronic obstructive pulmonary disease or interstitial lung disease have a disproportionate pulmonary hypertensive response, which reflects underlying pulmonary arterial vasoconstriction, adverse remodeling, or both. Such group 2 and 3 pulmonary hypertension patients have physiology that is reminiscent of PAH, notably elevated PVR2627
Differentiating PAH from pulmonary hypertension due to HFpEF
Among the most common challenges in clinical practice is distinguishing between PAH and group 2 pulmonary hypertension secondary to HFpEF.27 Both syndromes have normal left ventricular systolic function and abnormal left ventricular diastolic parameters. Correct case classification is crucial, as HFpEF accounts for a considerable proportion of referrals to pulmonary hypertension centers,28 and PAH specific therapies are not effective in group 2 pulmonary hypertension.2930 Two metrics that differentiate the HFpEF group are increased left atrial size on echocardiography and increased PCWP on RHC. In normal older patients the mean PCWP is 9 (SD 3) mm Hg in men and 11 (3) mm Hg in women.31 Including patients with PCWP values above 15 mm Hg in the group 1 category is thus inappropriate. Skilled measurement of PCWP is critical to distinguish between group 2 pulmonary hypertension and PAH (fig 4).32 End expiratory PCWP as opposed to mean PCWP correlates better with left ventricular end diastolic pressure (LVEDP).3233 If an accurate PCWP cannot be obtained, left heart catheterization should be done to measure LVEDP.
Clinical and echocardiographic models can differentiate between group 2 pulmonary hypertension and PAH (fig 5).2734 Older patients with more cardiovascular comorbidities are more likely to have pulmonary hypertension due to HFpEF than PAH. Whereas the clinical model has greater than 90% accuracy in differentiating pulmonary hypertension secondary to HFpEF and PAH, the echocardiographic model has only 59% specificity for identifying PAH. Exercise hemodynamic assessment and volume challenge during RHC have also been proposed to distinguish PAH from pulmonary hypertension due to HFpEF.3536 However, these techniques have not been standardized and should be used only in experienced centers.
Differentiating PAH from pulmonary hypertension due to chronic lung disease or hypoxia
Table 2 lists the characteristics that can help physicians in differentiating PAH from group 3 pulmonary hypertension due to chronic lung disease. This is important, as PAH specific therapies have not been shown to be effective in pulmonary hypertension due to lung disease.38
Assessing right ventricular function
Right ventricular function is the major determinant of long term outcomes in PAH.39 Several echocardiographic parameters that reflect increased right ventricular and right atrial size or decreased right ventricular contractility are better prognostic indicators in PAH than RVSP.39 Increased mortality risk is predicted by reduced right ventricular systolic and diastolic function, the presence of pericardial effusion, increased right ventricular or right atrial size, decreased tricuspid annular plane systolic excursion (TAPSE) (fig 6A), decreased Tei index (fig 6B),40 reduced peak systolic tricuspid lateral annular velocity (S’) (fig 6C),4142 decreased isovolumic contraction velocity (fig 6C),434445464748 and abnormal right ventricular free wall strain.
Unlike the left ventricle, the right ventricle shortens from apex to base with contraction. TAPSE assesses longitudinal right ventricular shortening by measuring annular excursion in the apical four chamber view, using M mode ultrasound (fig 6A). A TAPSE below 18 mm predicts increased mortality in PAH.44 Although convenient and reproducible, TAPSE measures only basal right ventricular systolic function and is not as accurate a measure of global right ventricular function as cardiac magnetic resonance imaging (MRI) or three dimensional echocardiography.49 TAPSE can be influenced by left ventricular function.5051 In addition, the timing of the deterioration in TAPSE relative to the onset of RVF is poorly defined.52
The geometry of the right ventricle and its heavy trabeculation make two dimensional transthoracic estimation of right ventricular function and volume difficult. Three dimensional echocardiography has improved the ability to accurately assess the right ventricle.53 Right ventricular volume acquisitions and right ventricular ejection fraction (RVEF) measurements using three dimensional echocardiography are accurate and reproducible with minimal inter-observer variability (approximately 4%).54 Changes in right ventricular volumes and function measured by three dimensional echocardiography correlate with clinical outcomes in PAH.55 However, because of its low measurement variability, cardiac MRI is more accurate and reproducible than either the two or three dimensional transthoracic technique. For example, to detect a change of 5% in RVEF in a clinical trial population, two and three dimensional echocardiography would require twice and 2.5 times the patient sample size needed for cardiac MRI.49
Current treatments for PAH
The current treatment strategy for PAH can be broadly divided into general measures, supportive therapies (box 1), and PAH specific therapies.
General supportive measures and background therapies in PAH management
General supportive measures
Restrict salt and fluid intake to reduce volume overload in light of their limited right ventricular reserve
Supplemental oxygen therapy should be used if needed to maintain systemic oxygen saturation above 90% as it improves exercise capacity56
Exercise training is advocated in PAH. Exercise training increases six minute walk distance (+52 m), peak oxygen uptake, and quality of life in PAH.57 Low level, symptom limited, aerobic exercises are preferred, and intense isometric exercises can lead to syncope
Vaccination against influenza and pneumococcal pneumonia is recommended
Women in the reproductive age group are strongly counseled to use contraceptive measures. Pregnancy is associated with substantial mortality in PAH.58 Hysteroscopic sterilization is the preferred method of contraception, but other options include progesterone-only intrauterine devices and pills and tubal ligation.59 Estrogen containing contraceptives and injectable progesterone are contraindicated because they increase the risk of thrombosis59
Background therapies for PAH
Diuretics—Diuretics are used to treat venous congestion resulting from RVF
Digoxin—Digoxin improves right ventricular contractility and cardiac output in acute hemodynamic studies60; however, no data support its long term use in PAH. The benefits and risks should be weighed
Anticoagulation—In situ thrombosis occurs in the small resistance pulmonary arteries of PAH patients.61 Retrospective and prospective studies have reported improved survival with long term warfarin therapy in PAH.616263 Registry data have questioned the benefits of anticoagulation, especially in patients with APAH.6465 In the COMPERA registry, the use of anticoagulation was associated with a 21% improvement in survival in patients with IPAH but no benefit in those with APAH.64 The REVEAL registry found no survival benefit with anticoagulation in IPAH and reduced survival in APAH.65 Long term warfarin therapy is recommended only in IPAH, heritable PAH, or anorexigen associated PAH patients with an international normalized ratio goal of 1.5-2.5.2466 The role of novel oral anticoagulants is unknown
PAH specific therapies
Fourteen PAH specific therapies are available for PAH. They target components of four PAH relevant molecular pathways: voltage gated, L type calcium channels, nitric oxide cyclic guanosine monophosphate (cGMP), endothelin, and prostacyclin.2 These drugs do not directly target the adverse vascular remodeling that obliterates, obstructs, and stiffens the pulmonary vasculature, and most do not improve the function of the right ventricle. The vasodilator therapies may indirectly inhibit cell proliferation and regress adverse vascular remodeling when given chronically, particularly if they reduce mPAP. For example, pulmonary artery banding, which reduces pulmonary arterial pressure distal to the band, reduces downstream adverse vascular remodeling in the Sugen 5416/hypoxia rat model of PAH.67 This suggests that pulmonary hypertension begets pulmonary hypertension and conversely implies that improving hemodynamics, however achieved, may regress vascular remodeling.
Calcium channel blockers
Calcium channel blockers (CCBs) are effective in the 5-10% of PAH patients who respond to acute vasodilatory challenge during RHC with a drop in mPAP by between 10 and 40 mm Hg, with no drop in cardiac output.68 Inhaled nitric oxide, intravenous epoprostenol, or intravenous adenosine are used for acute vasodilator testing. Vasoreactivity may be due to genetic predisposition, resulting from enrichment in vascular smooth muscle cell contraction gene variants. Early studies suggest that vasoreactivity can potentially be identified on the basis of a simple peripheral blood mRNA expression profile (reduced levels of mRNA for DSG2, a desmosomal cadherin, and RHOQ, a cytoskeletal protein).6970
CCBs are indicated only in PAH patients with a documented, positive vasodilator test. These patients are uncommon (5-10% of all cases) and have a different natural history with a five year survival rate of 90% with CCB monotherapy.62 Eligible PAH patients generally need higher than usual doses of CCB: amlodipine 20 mg, nifedipine 120-240 mg, or diltiazem 240-720 mg daily. Verapamil is not used because of its negative inotropic effects. Although diltiazem can also have a negative inotropic effect, it is preferred over amlodipine or nifedipine in patients with sinus or atrial tachycardia. Patients treated with CCBs should be closely monitored for adequate response and transitioned to PAH specific therapies if symptoms progress. Adequate long term response to CCBs in patients with APAH is rare.71
Drugs targeting the nitric oxide pathway
Nitric oxide is a potent pulmonary vasodilator that activates soluble guanylate cyclase (sGC) to generate cGMP. cGMP causes pulmonary artery smooth muscle cell (PASMC) relaxation through cGMP dependent protein kinases, which activate downstream targets, including the large conductance, calcium sensitive potassium channel BKca.72 Patients with PAH have reduced lung expression of endothelial nitric oxide synthase (eNOS), which synthesizes nitric oxide, and increased expression of phosphodiesterase 5 (PDE5), which degrades cGMP to 5′-GMP. The resulting decrease in cGMP is implicated in adverse pulmonary vascular remodeling in PAH.72 PDE5 inhibitors and sGC stimulators augment the nitric oxide-cGMP pathway and are approved for treating PAH. Although inhaled nitric oxide is useful for acute treatment in the intensive care unit and for acute vasodilator testing,6673 it is not used for chronic ambulatory therapy owing to the logistical challenges of delivering this short lived gas. A phase III trial is evaluating the safety and efficacy of long term inhaled nitric oxide in PAH (NCT02725372).
PDE5 inhibitors—Sildenafil and tadalafil are approved for treatment of PAH.72 Sildenafil improves six minute walk distance (6MWD), World Health Organization functional class, and mPAP at all three doses studied (20 mg, 40 mg, and 80 mg three times a day) (table 3).77 It had no effect on clinical worsening. Sildenafil is approved at 20 mg three times a day for treatment of PAH. However, long term extension studies showed significant improvement in hemodynamics with higher doses.77 Tadalafil (40 mg daily) increases 6MWD, time to clinical worsening, and health related quality of life (table 3).79 The major side effects of PDE5 inhibitors include headache, flushing, dyspepsia, and epistaxis.7779
sGC stimulators—sGC is a heterodimeric enzyme that generates cGMP. In PAH, sGC is often dysfunctional because it is oxidized or has lost its heme group.87 Riociguat directly stimulates sGC independent of nitric oxide, resulting in increased cGMP and pulmonary vasodilation. Riociguat 2.5 mg three times a day improves 6MWD, PVR, serum N terminal pro B type natriuretic peptide (NT-proBNP) concentrations, time to clinical worsening, and WHO functional class (table 3).81 The improvement in 6MWD is observed both in treatment-naive patients and those taking other PAH specific therapies.81 sGC and PDE5 inhibitors should not be given concurrently owing to the risk of hypotension. However, transition to sGC from PDE5 inhibitors improves exercise capacity and hemodynamics in patients who have inadequate responses to PDE5 inhibitors.88 Headache, dizziness, hypotension, dyspepsia, and gastroesophageal reflux are the most common adverse effects of riociguat.
Drugs targeting the endothelin pathway: endothelin receptors antagonists
Endothelin is a potent vasoconstrictor and smooth muscle mitogen. It acts through endothelin A and endothelin B receptors. Endothelin is overexpressed in the lungs and plasma of patients with PAH. The endothelial receptor antagonists (ERAs) bosentan, ambrisentan, and macitentan are beneficial in PAH.
Bosentan and macitentan are dual ERAs, blocking both endothelin A and endothelin B receptors. Bosentan 125 mg twice a day improves 6MWD, WHO functional class, Borg dyspnea score, and time to clinical worsening (table 3).76 Bosentan also improves PVR and 6MWD in patients with mild symptoms (WHO functional class II).89 Compared with bosentan, macitentan has greater tissue penetration and produces more sustained receptor blockade. Macitentan 3 mg and 10 mg daily reduce a composite endpoint of long term morbidity and mortality by 30% and 45%, respectively (table 3).84 Importantly, this is driven exclusively by the reduction in worsening of PAH, and no reduction is seen in either all cause or PAH related mortality.84 Ambrisentan (10 mg daily), a selective ERA antagonist, improves 6MWD, time to clinical worsening, WHO functional class, quality of life, and NT-proBNP when given as monotherapy.78
The major adverse effects of ERAs include hepatotoxicity, peripheral edema, anemia, and nasal congestion.90 In a meta-analysis, hepatotoxicity was more commonly observed with bosentan, anemia with bosentan and macitentan, and peripheral edema with bosentan and ambrisentan.90 Monthly liver function testing is warranted in patients taking bosentan. Although not mandated, serial liver function testing is advisable in patients taking macitentan and ambrisentan. ERAs should be discontinued if liver aminotransferases are more than five times the upper limit of normal, if aminotransferase elevations are accompanied by an increase in serum bilirubin concentrations (more than twice the upper limit of normal), or if patients develop signs of hepatic failure. Bosentan has several relevant drug interactions, notably decreasing serum sildenafil concentrations, which can impair the clinical benefits of combining these drugs.9192 Bosentan decreases warfarin concentrations, necessitating close monitoring of the international normalized ratio.93
Drugs targeting the prostacyclin pathway
Prostacyclin and prostanoids bind IP receptors, which increases cyclic adenosine monophosphate concentrations causing non-selective pulmonary vasodilatation.94 They also have antiplatelet, antithrombotic, antiproliferative, and anti-inflammatory properties.94 Prostacyclin expression is reduced in the lungs of patients with PAH.2 Depending on the preparation and specific molecule, prostanoids can be given as a continuous infusion (intravenously (epoprostenol and treprostinil) or subcutaneously (treprostinil)), via inhalation (iloprost and treprostinil), or orally (treprostinil and beraprost).
Epoprostenol is a synthetic prostacyclin analog. It has a very short half life (less than five minutes) and is unstable at room temperature. Therefore, it must be refrigerated and administered by continuous intravenous infusion. Epoprostenol improves exercise capacity, mPAP, PVR, cardiac output, quality of life, and survival (table 3).7475 The administration of epoprostenol requires a complex delivery system including a continuous infusion pump, central venous catheter, and ice packs to maintain temperature. It must be administered through a central venous catheter, as it is an irritant to the peripheral veins. A newer formulation (Veletri) is stable at room temperature, simplifying administration. Epoprostenol is usually started at 2 ng/kg/min, and the dose is gradually increased. The average dose of epoprostenol used for PAH ranges from 25 to 40 ng/kg/min. The most frequent adverse effects include gastrointestinal symptoms (nausea, vomiting, and diarrhea), headache, flushing, and jaw pain.
Treprostinil is a prostacyclin analog that has several advantages, including a longer half life (three hours) and stability at room temperature. Various formulations of treprostinil are approved to treat PAH (subcutaneous, intravenous, inhalational, and oral). Subcutaneous treprostinil improves 6MWD, hemodynamics, and quality of life (table 3).95 The major limitation of subcutaneous administration is infusion site pain, which occurs in 85% of patients.95 Although intravenous treprostinil has not been studied in a randomized trial, the US Food and Drug Administration approved its use for PAH on the basis of bioequivalence. Open label, long term extension studies have shown improvement in exercise capacity and hemodynamics and delay in time to clinical worsening with intravenous treprostinil.96 Subcutaneous or intravenous treprostinil is started at 2 ng/kg/min and increased gradually to an average target dose of 60-80 ng/kg/min. The adverse effects of treprostinil and epoprostenol are similar.
Catheter related bloodstream infection in patients receiving intravenous prostanoids can be life threatening. Patients must apply a meticulously sterile technique to reduce infection when refilling the pump at home. A fully implantable pump decreases catheter related side effects and increases patients’ satisfaction97; however, this pump is not approved for clinical use.
Inhaled treprostinil improves 6MWD and quality of life as an add-on therapy for patients who have symptoms despite taking sildenafil or bosentan (table 3).80 Inhaled treprostinil is started at three inhalations four times a day and increased gradually to a maximum of nine to 12 inhalations four times a day. It has minimal systemic effects owing to reduced systemic absorption and also lacks catheter related side effects. The common adverse effects include dry cough and headache.80
Treprostinil diolamine is an oral, salt form of treprostinil. It improves 6MWD in treatment-naive PAH patients with no improvement in WHO functional class or time to clinical worsening.1482 However, no significant benefit is seen when oral treprostinil is used as an add-on therapy in patients who have symptoms when taking an ERA or a PDE5 inhibitor.83 Thus, oral treprostinil is approved only as a monotherapy for improving exercise capacity in treatment-naive PAH patients. Oral treprostinil is started at 0.125 mg three times a day and increased by 0.125 mg every three to four days. Common adverse effects include nausea, diarrhea, headache, and jaw pain.82
Selexipag is an orally available, non-prostanoid activator of IP receptors.98 Both the parent drug and its metabolite have a high affinity for IP receptors. This selectivity minimizes adverse effects and facilitates dose escalation. In a long term, event driven trial, selexipag reduced the composite endpoint of death, lung transplantation, atrial septostomy, hospital admission for worsening PAH, or worsening of PAH by 40% (table 3).85 The benefits were mainly driven by reduction in hospital admissions for PAH, with no improvement in mortality. The reduction in morbidity was not dose dependent. Selexipag is started at 200 μg twice daily and increased weekly to a maximum of 1600 μg twice daily. The common adverse effects include nausea, vomiting, diarrhea, headache, and jaw pain.
Patients with positive acute vasodilator testing should be treated initially with CCBs and monitored closely. Non-responders to vasodilator are stratified on the basis of clinical, echocardiographic, and hemodynamic evaluations that assess right ventricular function as high or low risk (table 4). Patients with a high risk profile have worse survival.99100101 Thus, patients at high risk should be considered for initial parenteral prostanoid therapy. Epoprostenol is preferred in these patients given its survival benefit. In a retrospective review of 19 patients at high risk presenting with cardiac index less than 2 L/min/m2 and PVR greater than 20 WU, starting triple combination therapy with epoprostenol, a PDE5 inhibitor, and an ERA at diagnosis was associated with improvement in exercise capacity and hemodynamics and 100% survival at three years.102 The efficacy, safety, and cost effectiveness of this approach needs further assessment in a randomized controlled trial.
Patients with a low risk profile are treated by either sequential combination therapy or initial combination therapy. In the first approach, patients are started on oral monotherapy. A second drug, targeting a different pathway, is added if the response to monotherapy is inadequate. In the initial combination therapy approach, patients are started on an oral combination of PDE5 inhibitor and ERA simultaneously at diagnosis.66 Compared with initial monotherapy, initial combination therapy with ambrisentan and tadalafil improved a composite endpoint of death, lung transplantation, hospital admission for worsening PAH, worsening PAH, and persistent worsening after six months by 50% (table 3)86; however, no mortality advantage was seen.86 On the basis of these results, initial combination therapy has been preferred over sequential combination therapy in most patients. Sequential combination therapy is considered primarily for patients with mild disease and preserved right ventricular function.
Patients should be routinely monitored to ensure an adequate response; in those with inadequate responses to combination therapy, a third agent should be added.66 Patients who have high risk features despite receiving combination therapy should receive parenteral prostacyclin therapy. For patients who are not at treatment goal but who lack high risk features, addition of inhaled treprostinil or selexipag or starting a transition from PDE5 inhibitor to sGC stimulators can improve outcomes. The role of oral treprostinil therapy in patients who are already receiving combination therapy is limited, as it is known to improve exercise capacity only in treatment-naive patients.
Patients who continue to have WHO functional class III or IV symptoms on maximal medical therapy should be referred as early as possible for lung transplantation.24 Atrial septostomy is considered as a bridge to transplant in patients who are unstable while awaiting transplant or as a palliative procedure in patients who are not candidates for transplant.24 Veno-arterial extracorporeal membrane oxygenators and pumpless oxygenators (Novo lung) have also been used successfully as a bridge to transplant in patients with PAH and RVF.103
Pathogenesis of PAH
Despite the benefits of current therapies to the quality of life and time to clinical worsening, these treatments do not decrease mortality, apart from epoprostenol which improved survival in PAH patients with WHO functional class IV.74 The average improvement in functional capacity (6MWD) and hemodynamics (mPAP) for each of the four classes of PAH specific therapies is modest. In randomized clinical trials, the average fall in mPAP with an optimal dose of sGC, PDE5 inhibitor, ERA, or prostanoid is less than 6 mm Hg (table 5).87 The primary criticism of current classes of PAH specific therapies, beyond their limited hemodynamic efficacy, is that they primarily target excessive vasoconstriction, which is a dominant pathophysiologic feature in less than 10% of patients with PAH.4 Better understanding of disease pathogenesis is thus needed to identify new targets for therapy. Ongoing research has provided new understanding of the cellular, genetic, and epigenetic changes that drive pathological vascular remodeling in the lungs of patients with PAH (fig 7). Box 2 describes the five emerging concepts in the pathogenesis of PAH.
Concepts in PAH
Characterization of a cancer-like phenotype in PAH that accounts for the obstructive pulmonary vasculopathy. This is defined by altered mitochondrial metabolic function (Warburg metabolism) and mitochondrial dynamic function (fragmentation). These changes drive cell proliferation and impair apoptosis. They are downstream from genetic mechanisms, such as BMPR2 mutation, and epigenetic mechanisms
PAH is promoted by genetic and epigenetic factors that contribute to disease pathogenesis. The most common genetic mechanism is mutation of BMPR2
PAH is in part a disease of altered immunity and increased inflammation. Improved understanding of the role of chronic inflammation, fibrosis, and immune mediated mechanisms offers potential therapies to reverse adverse vascular remodeling
PAH is predominantly a disease of women, although afflicted men have worse prognosis. There is an emerging appreciation for the importance of sex differences in the incidence, therapeutic responsiveness, and outcomes of PAH
Patients with PAH die from right ventricular failure. There is an increasing recognition of the unique embryologic origins and response to increased afterload of the right ventricle, and the crucial role that right ventricular adaptation plays, both in determining prognosis and as a target for therapy. It seems that the ischemic, metabolically remodeled right ventricle in PAH is an example of hibernating myocardium
Cancer-like metabolic and mitochondrial cellular phenotype of PAH
In 1891 Romberg described medial thickening and occlusive vascular lesions in the lungs of patients with PAH.104 Controversies remain regarding the cellular and molecular origins of this pathological remodeling, including some uncertainly as to which cells proliferate (PASMC, clonal endothelial cells, fibroblasts, inflammatory cells, or some combination of these). There is also debate about the relative importance of an early excess of endothelial apoptosis (at the time of an initial vascular insult) versus a later resistance to apoptosis that promotes vascular obstruction. However, strong evidence from multiple laboratories, using both patient derived primary cell lines and animal models, shows that a proliferative, apoptosis resistant, cancer-like phenotype occurs in PASMCs, endothelial cells, and fibroblasts in established PAH.105106107
Cytosolic calcium and ion channels
Increased cytosolic calcium ([Ca2+]cyt) contributes to the contractile, hyperproliferative, and anti-apoptotic phenotype of PAH PASMCs. [Ca2+]cyt is regulated by several ion channels that control calcium influx, as well as Ca2+ sequestration within the sarcoplasmic reticulum and mitochondria. Elevated [Ca2+]cyt in PASMCs from patients with PAH has been linked to the activation of store operated Ca2+ channels, including the transient receptor potential channel TrpC6,108 and downregulation of voltage gated potassium channels, such as Kv1.5.109 Decreased expression of Kv1.5 channels, which normally maintain PASMC membrane potential,110 leads to membrane depolarization and influx of Ca2+ through voltage dependent calcium channels (CaL). In rodent models of pulmonary hypertension, restoring Kv1.5 expression through adenoviral gene transfer improves hemodynamics.111 In addition to promoting smooth muscle cell contractility, prolonged increases in [Ca2+]cyt can increase proliferation by driving cells into the cell cycle.109 Activation of the Ca2+/calcineurin sensitive transcription factor nuclear factor of activated T cells (NFAT) in PAH PASMCs perpetuates the elevation in [Ca2+]cyt by suppressing Kv1.5 expression.112 NFAT activation also promotes apoptosis resistance by increasing expression of bcl-2.112 Failure of mitochondrial calcium uptake, caused by downregulation of the mitochondrial calcium uniporter (MCU), further contributes to the cytosolic calcium overload in PAH.113 Rho kinase activation in PAH causes calcium sensitization, leading to greater vasoconstriction for a given level of [Ca2+]cyt. Moreover, the vasoconstriction caused by rho kinase activation is not reversed by conventional vasodilators and likely contributes to vascular stiffening.114 Rho kinase inhibitors, such as fasudil, have been studied in limited human PAH cohorts.115 They seem to be safe and effective and may reduce mortality in PAH patients with right ventricular failure.116
Although there is little debate that calcium concentrations are elevated in PAH PASMCs, the relative importance of excessive calcium entry via store operated channels and/or CaL channels (which are therapeutically targeted by CCBs) versus impaired organelle uptake of calcium (in mitochondria and endoplasmic reticulum) versus calcium sensitization (by rho kinase) remains uncertain.
Mitochondrial metabolic dysfunction
Mitochondrial dysfunction in PAH PASMCs includes a metabolic shift from glucose oxidation toward uncoupled aerobic glycolysis, a metabolic pattern first described by Otto Warburg in cancer cells. In the presence of oxygen, rates of glycolysis in normal vascular cells are closely coupled to rates of glucose oxidation. In Warburg metabolism, uncoupled glycolysis is increased because, whereas mitochondrial respiration (glucose oxidation) is actively suppressed, glycolysis is disproportionately elevated providing the abnormal cell with sufficient energy to thrive.
Although this disease signature in PAH was initially identified in PASMCs,117118 similar metabolic and mitochondrial changes occur in pulmonary artery endothelial cells (PAECs)104119120 and adventitial fibroblasts of PAH patients.121 More recently, examination of the diseased right ventricle in PAH patients and animal models has identified a similar Warburg phenotype in cardiomyocytes, where it reduces contractility.122
In healthy cells, the pyruvate produced by glycolysis enters the mitochondria via the mitochondrial pyruvate transporter. There it is converted by pyruvate dehydrogenase (PDH) into acetyl coenzyme A, which fuels the Krebs cycle. In PAH, oxidative phosphorylation is actively suppressed by upregulated expression of pyruvate dehydrogenase kinase (PDK). PDK phosphorylates and inhibits PDH.123124 This shifts the cell to rely on glycolysis, which is energetically inefficient. However, in PAH, glycolysis is upregulated, both by an increase in glucose influx mediated by increased glucose transporter (glut) expression and by alterations in splice variant expression of the terminal glycolytic enzyme pyruvate kinase.106107 This metabolic shift supports rapid proliferation while avoiding mitochondrial apoptosis.106107124
Warburg metabolism accounts for the increase in fluorodeoxyglucose uptake in the lungs and right ventricle of PAH patients and preclinical PAH models observed using positron emission tomography.125 In patients with PAH, a lactate gradient exists from superior vena cava to pulmonary artery.126 Whether this pre-pulmonary lactate gradient reflects right ventricular ischemia or Warburg metabolism is unknown.
Emerging metabolic therapies that exploit this pathway include the small molecule PDK inhibitor dichloroacetate. Dichloroacetate reverses the Warburg phenotype in PASMCs and right ventricular cardiomyocytes and regresses PAH in preclinical models.124127 Dichloroacetate has been safely used to chronically treat children with mitochondrial disease and lactic acidosis.128 A four month, open label study of dichloroacetate (3 to 6.25 mg/kg twice daily) in patients with IPAH taking approved PAH therapies showed that dichloroacetate reduced mPAP and PVR while improving 6MWTD; however, some patients did not respond to dichloroacetate, and these patients had functional variants of SIRT3 and UCP2.129
Vascular cells isolated from patients with PAH also show additional signs of mitochondrial dysfunction, including fragmentation and membrane hyperpolarization. Hyperpolarization of mitochondrial membranes contributes to apoptosis resistance by blocking the release of pro-apoptotic mediators, such as cytochrome C(103). The mitochondria in vascular cells exist in dynamic networks that are continuously dividing (fission) and joining together (fusion).130 For nuclear division to occur, mitosis must be coordinated with mitochondrial division. Increased rates of mitotic fission in PAH cause PAH PASMCs to show fragmented mitochondria (fig 8), and this can be therapeutically targeted.117 A fission/fusion imbalance in PAH results from increased activation of the mitochondrial fission mediator dynamin related protein 1 (Drp1) and reduced expression of the fusion mediator, mitofusin 2.117130 Drp1 activation and mitofusin 2 downregulation are secondary to other changes in the PAH milieu, including elevated [Ca2+]cyt, increased mitogen concentrations (platelet derived growth factor, endothelin 1), and impaired expression of the mitochondrial biogenesis promoter peroxisome proliferator activated receptor γ coactivator 1-α.117131
Reduced expression of the MCU in PASMCs from PAH patients and rodent models provides a potential unifying mechanism linking dysregulated mitochondrial metabolism and dynamics and partially explains the observed increase in PASMC proliferation and apoptosis resistance in PAH.113 MCU is the major functional subunit of the MCU complex that allows for the influx of Ca2+ into the mitochondrial matrix. Reduced expression of MCU (and increased expression of MICU1, a negative regulator of MCU) increases [Ca2+]cyt in PAH while lowering intramitochondrial calcium. Increased [Ca2+]cyt promotes vasoconstriction and enhances mitochondrial fission and proliferation. Low intramitochondrial calcium inhibits calcium sensitive dehydrogenases in the mitochondrial matrix (including PDH) and thereby inhibits glucose oxidation.113 MCU downregulation in PAH is epigenetically mediated by increases in the micro-RNAs miR25 and miR138. Anti-miRs or MCU gene transfer reverse the mitochondrial phenotype in PAH PASMCs and regress PAH in the monocrotaline model.113
Genetic contributions to PAH
Heritable forms of PAH account for 6-10% of all PAH.24 Heterozygous, germline mutations in BMPR2, the gene encoding the type II bone morphogenetic protein (BMP) receptor (BMPR-II), account for 70-80% of heritable PAH cases,132133134135 as well as 15-25% of IPAH cases.132134135136137Box 3 describes the discovery of BMPR2 gene mutations in PAH. At risk siblings from families with PAH carrying a BMPR2 mutation have a disease penetrance in mutation carriers of only 27%. Disease penetrance is greater in females (42%) than in males (14%).138 Thus, a BMPR2 mutation increases an individual’s chance of developing PAH 25 000-fold, from roughly 1 in 100 000 to 1 in 4.8138 A meta-analysis of PAH patients also showed that, compared with patients without BMPR2 mutations, mutation carriers are younger at diagnosis, develop more hemodynamically severe disease, are less likely to respond to vasodilators, and are at an increased risk of death.139 However, BMPR2 mutations are considered to be permissive of disease, requiring additional genetic, epigenetic, or environmental factors for the development of PAH in people with mutations.
Discovery of BMPR2 mutations in PAH
1951—First description of “primary pulmonary hypertension” by Dresdale includes recognition of heritable PAHRETURN TO TEXT
BMPR-II protein concentrations are reduced by about 75% in lung tissue and endothelial cells from patients with PAH.140141 This reduction is greater than expected from haploinsufficiency alone and occurs in patients lacking BMPR2 mutations, as well as in non-genetic rodent models.140141 This suggests that factors associated with disease, independent of mutation status, suppress BMPR-II expression and provides biologic plausibility that targeting BMPR-II deficiency might be beneficial even in PAH patients lacking BMPR2 mutations.
Impaired BMPR-II signaling has been shown to promote accelerated cell proliferation,142 while potentially contributing to disease initiation by enhancing the susceptibility of PAECs to apoptosis.143 Loss of BMPR2 in the endothelium also causes mitochondrial dysfunction and inflammation,119 providing a potential mechanism linking BMPR2 mutations to mitochondrial dysfunction in PAH. Emerging BMPR-II related therapies for PAH include strategies to rescue the functionality of mutated BMPR2 alleles or enhance BMPR-II signaling through the functional receptors that are produced by the non-mutated allele. Rescue strategies include the use of read through compounds such as Ataluren (PTC-124), which promote the transcriptional read through of premature termination codons and the production of functional, full length BMPR-II protein from mutated alleles.144 Missense mutations can also be rescued in preclinical studies by using chemical chaperones, such as 4-phenylbutyrate, which increase trafficking of misfolded, but otherwise functional, BMPR-II protein from the endoplasmic reticulum to the cell surface.145 Preclinical trials of lung targeted BMPR-II gene therapy have had mixed results.146147 Examples of enhancing signaling via non-mutated BMPR-II protein include the delivery of recombinant BMP ligands, such as BMP9,143 or small molecule agonists of canonical BMP signaling, such as FK506.148 These approaches enhance BMPR-II mediated signaling in the endothelium and can reverse established disease in the Sugen 5416/hypoxia rat model.143148 Inhibition of BMPR-II degradation, through lysosomal inhibitors such as hydroxychloroquine, also enhances BMPR-II mediated signaling in human PAECs and prevents disease in the monocrotaline model.149
Mutations in other components of the BMP signaling pathway including ACVRL1, which encodes the type I BMP receptor ALK1, ENG, the gene encoding the accessory receptor endoglin, and SMAD9, which encodes the BMP transcriptional mediator, Smad 8, contribute to a small percentage of PAH cases.150151152153 Whole exome sequencing studies of families with PAH have also identified mutations in genes unrelated to canonical BMP signaling pathways in 1-3% of all cases of PAH. These include mutations in KCNK3, which encodes the pH sensitive potassium channel TASK-1,154 and CAV1, which encodes caveolin 1, a membrane protein that is essential for the formation of lipid rafts, known as caveolae.155 Whole exome sequencing approaches have also been used in the examination of pulmonary veno-occlusive disease (PVOD) and pulmonary capillary hemangiomatosis (PCH), which are classified as group 1 pulmonary hypertension. Independent studies examining families with autosomal recessive forms of either PVOD or PCH identified bi-allelic mutations of EIF2AK4 (or GCN2), the gene encoding eukaryotic translation initiation factor 2 α kinase, as the mutation underlying the hereditary forms of these conditions.156157 Thus, genotyping shows that PVOD and PCH are phenotypic variants of the same disease. Bi-allelic EIF2AK4 mutations were also identified in 20-25% of sporadic cases of both conditions.156158
Preliminary data from the PAH Biobank sponsored by the National Heart, Lung, and Blood Institute at Cincinnati Children’s Hospital Medical Center have been provided by study principal investigator William Nichols (CCHMC/University of Cincinnati). The incidence of pathogenic/suspected pathogenic gene variants identified, using panel sequencing of 12 genes in the 2251 PAH patients, is 10.8%. Pathogenic/suspected pathogenic variants are less frequent in APAH (5.8%) and more common in IPAH (13.2%) (personal communication, W Nichols). Nicholas Morrell (University of Cambridge) is conducting a study of IPAH and heritable PAH patients (n=1250), using whole genome sequencing as part of the National Institutes for Health Research Bioresource for Rare Diseases Study. These two studies will ultimately define the genetic contribution to various PAH patient populations and help to determine the value of routine genetic testing in patients with PAH.
Involvement of epigenetic factors in PAH
The expression of genes in PAH is also influenced through epigenetic processes, defined as mechanisms that alter gene expression without changing the sequence of genomic DNA. Epigenetic mechanisms in PAH include DNA methylation, histone modification, and RNA interference via micro-RNAs (box 4).
Epigenetic mechanisms in PAH
DNA methylation—DNA methylation involves the covalent attachment of a methyl group to cytosine residues in CpG dinucleotide sequences. Methylation occurs in CpG rich regions of the genome known as CpG islands. These CpG islands are usually near the gene promoters, and their hypermethylation interferes with gene transcription
Histone acetylation—The modification of histones influences the transcriptional activity of genes through the regulation of their accessibility to transcription factors. Histone acetylation, which is associated with increased transcriptional activity, is carried out by histone acetyltransferases and reversed by histone deacetylases
Production of micro-RNA—miRs are encoded by intronic DNA and regulate gene expression through RNA interference. Short (20-22 nucleotide) segments of RNA bind to complementary sequences in the 3′-untranslated region of mRNA, leading to mRNA degradation or the repression of translation. As a single miR holds the potential to bind and regulate multiple gene targets, these molecules can serve as “master regulators” for programs of targeted gene expressionRETURN TO TEXT
Epigenetic regulation of a pulmonary hypertension phenotype was first identified in the assessment of spontaneous PAH in the fawn hooded rat model.159 Hypermethylation and silencing of the gene encoding superoxide dismutase 2 in this model caused a 50% reduction in pulmonary artery superoxide dismutase 2 protein expression (also seen in PAH patients). Superoxide dismutase 2 is a mitochondrial enzyme that converts superoxide to hydrogen peroxide. Hydrogen peroxide serves as a diffusible redox signaling molecule. Reduced hydrogen peroxide in the fawn hooded rat model causes normoxic activation of hypoxia inducible factor (HIF-1α), which triggers the Warburg effect by upregulating PDK and glut expression and increased PASMC proliferation,160 and also inhibits Kv1.5 channel function/expression, causing vasoconstriction.118 This epigenetic mechanism results from lung specific increases in the expression of DNA methyltransferases 1 and 3B (DNMT). A similar mechanism of reduced superoxide dismutase 2 and elevated HIF-1α occurs in endothelial cells from patients with IPAH.120 Increased DNA methylation is implicated in the exaggerated pulmonary hypertensive response of mice born to mothers subjected to a restrictive diet during pregnancy.161 Future therapies may involve inhibiting DNA methyltransferases,162 as is currently done in some hematologic malignancies (using decitabine),163 modulating the enzymes that demethylate DNA (ten eleven dioxygenases),164 and/or manipulating methyl binding proteins, which alter the transcription of methylated genes.
An examination of lung tissue from patients with IPAH, as well as lungs and right ventricles from rats with chronic hypoxic pulmonary hypertension, identified increased expression of histone deacetylases (HDAC1 and HDAC5) in remodeled pulmonary vessels.165 Additional work in PAECs from PAH patients identified increased nuclear accumulation of HDAC4 and HDAC5, leading to downregulation of the transcription factor myocyte enhancer factor 2 and a reduction of several genes involved in pulmonary vascular integrity and homeostasis, including Krüppel-like factors 2 and 4 and connexins 37 and 40.166 Interestingly some metabolic enzymes, such as PDH, act within the nucleus to generate the acetyl coenzyme A required for histone regulation.167 Increased histone acetylation has also been reported in a rat model of persistent pulmonary hypertension in the newborn,168 where a compensatory sixfold increase in eNOS expression was attributed to enhanced acetylation of the eNOS promoter and a modest decrease in gene methylation.
The HDAC inhibitors valproic acid and suberoylanilide hydroxamic acid reduced established hypoxia induced pulmonary hypertension in rodent models.165 Similar results were also obtained with the HDAC inhibitor MGCD0103, which prevented hypoxic pulmonary vascular remodeling in rats.169 Although valproic acid improves RVH in both the monocrotaline rat model and the pulmonary artery banding model of RVF,170 another HDAC inhibitor, trichostatin A, worsened RVH and fibrosis in the rat pulmonary artery banding model.171 These differences may reflect the need to target specific subtypes of HDAC in PAH and/or differences in the role of HDACs in various forms of RVH.
Micro-RNA dysregulation occurs in whole lung tissue in PAH, as well as in cultured PAECs, PASMCs, and fibroblasts. Table 6 summarizes the miRs identified in PAH and their effect on in vitro or in vivo processes.172173174175176177178180181183184185186 Many of these dysregulated miRs influence pathways that are critical to creating the cancer-like, mitochondrial-metabolic phenotype of PAH.
Therapeutic modulation of disease associated miRs, using miR-mimics or antago-miRs, has shown potential to prevent or reverse PAH in rodent models.182 Although these findings are promising, the ability of miRs to regulate a broad array of genes, which is the very basis of their value, also raises the potential for undesirable off-target effects that could limit their translation into the treatment of human disease. miRs, particularly those present at altered concentrations in the circulation, represent useful biomarkers that can predict survival.179184
Two recent studies provide an example of the intersection of genetics, epigenetics, and metabolism in the pathogenesis of PAH. Caruso et al and Zhang et al showed that altered activity of the enzyme pyruvate kinase contributes to the Warburg phenomenon in endothelial cells and fibroblasts in PAH, respectively.106107 Pyruvate kinase is the final step in glycolysis, catalyzing phosphate transfer from phosphoenolpyruvate to ADP, producing pyruvate and ATP. Both groups showed that this pro-glycolytic mechanism is stimulated by upregulation of a heterogeneous nuclear ribonucleoprotein, polypyrimidine tract binding protein (PTBP1), in response to an epigenetic mechanism—namely, downregulation of miR124. The increase in PTBP1 changes the splice variant expression of pyruvate kinase, favoring the proglycolytic variant pyruvate kinase M2. The mitochondrial supply of pyruvate is further compromised by downregulation of the mitochondrial pyruvate transporter. Remarkably, BMPR2 mutation/downregulation leads to the same Warburg phenotype and causes proliferation of endothelial cells. This work shows coherence between apparently disparate theories, illustrating how seemingly unrelated gene mutation or epigenetic changes can promote the Warburg mitochondrial metabolic phenotype, a final common pathway of the proliferative PAH phenotype in many vascular cells.
Inflammation and fibrosis in PAH
Chronic inflammation and maladaptive fibrosis play an essential role in PAH. Immune dysfunction is a central feature of APAH, especially in patients with connective tissue diseases (for example, systemic lupus erythematosus and scleroderma) and infections (for example, HIV and schistosomiasis).187 The prevalence of schistosomiasis in the developing world makes it the single largest cause of PAH, with an estimated 2 000 000 cases worldwide.188
Chronic inflammation is also observed in IPAH. Histologic examinations of lung tissue from IPAH patients have identified the presence of immune cell infiltrates, composed of lymphocytes, macrophages, dendritic cells, and mast cells in pulmonary vascular lesions.189190191 Circulating concentrations of inflammatory cytokines including interleukins 1β, 2, 4, 6, 8, 10, and 12p70,192193 tumor necrosis factor α,192193 and the chemokines CXC3L1 (or fractalkine),194 CCL2 (or monocyte chemotactic protein 1),195 and CCL5 (or RANTES)196 are elevated in IPAH. Elevated concentrations of interleukins 6, 8, 10, and 12 predicted survival in a cohort of 21 PAH patients better than traditional prognostic markers, such as 6MWD or cardiopulmonary hemodynamics.193
Fibroblasts isolated from the adventitia of pulmonary arteries from both PAH patients and chronic hypoxia models of pulmonary hypertension show a hyperproliferative, apoptosis resistant, pro-inflammatory phenotype, marked by the production of inflammatory molecules and expression of myofibroblast markers.181197198 These changes are associated with epigenetic modifications including increased HDAC activity and decreased miR124 expression.181 Interestingly, the depletion of circulating monocytic cell populations blocks adventitial remodeling in a rat model of hypoxic pulmonary hypertension, suggesting that fibrotic remodeling is at least partially dependent on the recruitment of circulating monocytic mesenchymal precursors.199 The interplay between adventitial fibroblasts and monocytic cells is further supported by the demonstration that human and experimental PAH fibroblasts activate a STAT3 mediated, pro-inflammatory phenotype in macrophages by means of increased interleukin 6 secretion.197
Recent mechanistic studies examining the contribution of cellular immunity to the pathogenesis of pulmonary hypertension have also focused on the role of specific lymphocyte populations, including T cell, B cell, and natural killer cell subsets. The role of T lymphocytes in disease pathogenesis is controversial. Athymic nude rats lack mature T cells and develop more severe pulmonary hypertension in response to either monocrotaline or the vascular endothelial growth factor receptor blocker SU-5416,200201 suggesting a protective role for regulatory T cells.202 In Sugen rats, PAH is attenuated by blockade of leukotriene B4 activity with leukotriene A4 hydrolase inhibitors. More recently, blockade of interleukin 6, using the anti-interleukin 6 receptor antibody MR16-1, was shown to prevent chronic hypoxia induced pulmonary hypertension in mice by blocking the accumulation of Th17 cell derived interleukin 21 and M2 macrophages in the lungs of these mice.203 Thus the role of T cell subsets in disease is context specific and likely varies among species.
The contribution of humoral immunity and excessive B cell activation to the pathogenesis of PAH is supported by multiple studies identifying circulating autoantibodies and the ectopic expansion of pulmonary lymphoid tissues in patients with PAH.204 Increased bronchus associated lymphoid tissues and autoantibody production has also been reported in the monocrotaline rat model, with passive autoantibody transfer being sufficient to transmit disease between animals.205
Impairment of innate lymphocytes, such as natural killer cells, is reported in human and experimental PAH.206 Natural killer cells can adopt an angiogenic phenotype and contribute to vascular remodeling in pregnancy and cancer through the production of angiogenic cytokines and matrix degrading enzymes.207208 In PAH, natural killer cells have an impaired phenotype, marked by increased transforming growth factor β and matrix metalloproteinase 9 production.206 Deficiencies in natural killer cells and CD8+ killer T cells were linked to an increased risk of death in a small cohort of IPAH and APAH patients.209
Right ventricular dysfunction in PAH
The right ventricle’s ability to adapt to pressure overload determines functional status and prognosis in PAH. A dilated right ventricle or a small left ventricle independently predicts poor survival in PAH.210 Adult PAH patients with RVF admitted for inotropic therapy have a 41% acute mortality rate,211 far exceeding the mortality of patients admitted with left ventricular failure.
Chronic pressure overload in PAH stimulates RVH in a compensatory attempt to maintain cardiac output while minimizing wall stress (fig 9A, right panel). However, in some patients, the right ventricle dilates owing to maladaptive remodeling and eventually fails (fig 9A, left panel, and fig 9B). In general, PAH patients with congenital heart disease (Eisenmenger’s syndrome) show adaptive, concentric RVH and are free from RVF for decades215; conversely, APAH patients, particularly those with scleroderma, have maladaptive remodeling with worse right ventricular function and a higher incidence and severity of RVF, despite lower mPAP.216 Adaptive RVH has been associated with a lower risk of clinical worsening compared with maladaptive RVH in patients with IPAH.38217
Abnormalities of right ventricular perfusion, angiogenesis, autonomic signaling, metabolism, and fibrosis contribute to the development of maladaptive RVH and RVF. Many of the changes in the right ventricle during RVH, such as activation of the autonomic nervous system and uncoupled glycolysis,218219 are ultimately maladaptive. Whether these chamber specific, cardiac abnormalities reflect primary insults to the right ventricle or are simply the maladaptive responses to increased right ventricular afterload remains uncertain. Nonetheless, they represent important pathophysiologic abnormalities and offer many therapeutic targets.
Right ventricular ischemia
Patients with PAH often present with evidence of myocardial ischemia, including angina-like chest pain and elevated concentrations of serum troponin.220 Right ventricular ischemia may reflect reduced right coronary artery perfusion pressure. However, right ventricular function is usually preserved as long as coronary perfusion pressure, the difference in pressure between the aorta and the right ventricle in systole or diastole, exceeds 50 mm Hg.221 In PAH, systemic (aortic) hypotension and a rise in right ventricular pressure (in both systole and diastole) combine to reduce the pressure gradient that drives perfusion of the right coronary artery, contributing to right ventricular ischemia. Ischemia in RVH may also reflect loss of small vessels (microvascular rarefaction).222 The observation that RVF is more prevalent in patients with scleroderma, who have circulating autoantibodies toxic to the endothelium, and in PAH models that are characterized by endothelial injury (induced by SU5416) suggests a potential role for primary endothelial damage in the right ventricular microcirculation.223 Capillary rarefaction in experimental PAH is reversible with β-adrenergic blockers.222
Autonomic activation and downregulation of β1 receptors in RVH
The right ventricle is less responsive to inotropes than the left ventricle.224 Circulating catecholamines are elevated in PAH patients with RVF.225 The autonomic activation exceeds that seen in left ventricular failure.226 Like PAH patients, rats with PAH induced maladaptive RVH show selective downregulation and uncoupling of β1-adrenoreceptors, α1-adrenoreceptors, and dopaminergic receptors in right ventricular myocytes. This is due to G protein coupled receptor kinase 2 (GRK2) mediated uncoupling of receptor signaling and impairs responses to inotropes. GRK2 inhibition seems to be therapeutically beneficial in PAH.227228 In rodent models of PAH and RVH, dobutamine is the optimal right ventricular inotrope; however, this has not been established in PAH patients.227 The observed downregulation of the cardiomyocyte β1-receptors suggests biologic plausibility for using β-adrenergic blockers for RVF in PAH patients. However, in a study of 19 patients, bisoprolol reduced cardiac output and exercise capacity with no improvement in RVEF.229 In contrast, carvedilol reduced glycolytic rate and increased β-adrenergic receptor expression in the right ventricle with improvement in right ventricular fractional area change and no reduction in exercise capacity and cardiac output.230 More definitive studies are needed.
Right ventricular metabolism in RVH
The fetal right ventricle relies on glycolysis and glucose oxidation as the major sources of ATP production.231 In the adult right ventricle, there is a transition to reliance on fatty acid oxidation (FAO), which accounts for 60-90% of ATP production. FAO needs about 12% more oxygen per mole of ATP than does glucose oxidation. The hypertrophied right ventricle depends on glucose metabolism.232 The metabolic fate of glucose in RVH is altered because processes, including PDK mediated inhibition of PDH, suppress mitochondrial metabolism. In this scenario, lactate (the end product of glycolysis) accumulates, causing acidosis and exacerbating right ventricular hypokinesis.228 Activating glucose oxidation by using the PDK inhibitor dichloroacetate improves right ventricular function while regressing RVH. Thus, the right ventricle in PAH can be resuscitated metabolically. Consistent with this, increased right ventricular uptake of fluorodeoxyglucose was observed in a small series of PAH patients studied with positron emission tomography, a consequence of reliance on uncoupled glycolysis, and was reduced by effective PAH targeted therapy (fig 9C).214
In PAH patients, there is evidence of cardiac steatosis and lipotoxicity, which murine data suggest relates to reduced FAO.93 Whether differences in FAO occur in adaptive versus maladaptive RVH remains to be determined. In light of this report of impaired FAO in the right ventricles of patients with PAH, it is uncertain whether FAO inhibitors, such as trimetazidine and ranolazine, would be helpful in PAH, as they are in preclinical models of RVH induced by pulmonary artery banding.223
Right ventricular hibernation
Right ventricular ischemia leads to right ventricular hibernation, defined as a reduction in right ventricular function caused by a chronic decrease in perfusion with persistent myocardial viability.228 Thus, future therapeutic opportunities might include anti-ischemic or angiogenic therapy. Alternatively, it might be empirically easier to treat RVF and ischemia by pharmacologically changing metabolism or adrenergic signaling to “do more with less” (that is, use of drugs that alter cardiac metabolism or anti-adrenergic drugs).228229 However, the clinical trial data that would be needed to alter practice are lacking.
Right ventricular fibrosis
The role of right ventricular fibrosis in RVF in PAH is unknown. However, late gadolinium enhancement on MRI at right ventricular insertion points may indicate localized fibrosis (fig 9D)233 and is an independent risk factor for poor prognosis.234
Chamber specific responses to RVH
Quantitative differences in the expression of many transporters and pumps in the right versus left ventricle, seen in fetal hearts, persist into adulthood.235 This raises the possibility that chamber selective therapeutic targets may exist in RVH. For example, PDE5 expression, normally absent in the right ventricle, is induced during RVH. Sildenafil has a direct right ventricular inotropic effect that occurs only in RVH.72236
Sex differences in PAH
Although woman are more likely to acquire PAH, men (particularly those aged over 60 years) have a worse prognosis. This may relate to more maladaptive response of the male right ventricle to PAH. Male patients are less likely to improve their RVEF after starting PAH targeted therapy. Male and female patients showed a similar reduction in PVR one year after starting PAH targeted therapy, but RVEF improved in female patients, whereas it deteriorated in male patients.13 This difference accounted for much of the male survival disadvantage. Even healthy people have differences in right ventricular function, such that men have lower RVEF than women (by about 4%), despite greater right ventricular mass (approximately 8%) and volumes.237 Beyond the adaptive response of the right ventricle, other important sex differences in the pathogenesis of PAH exist in both the development and progression of vascular obstruction and the responsiveness to PAH targeted therapy.238 These are beyond the scope of our review, and the interested reader is referred to the literature.99100101239
Referral to PAH centers
All three documents:
Referral to centers expert in the management of PAH
CHEST guideline and expert panel report 2007 and ACCF/AHA 2009 expert consensus document:
Warfarin anticoagulation in IPAH with a target international normalized ratio 1.5-2.5
2015 ESC/ERS guidelines:
May be considered in patients with IPAH, heritable PAH, and anorexigen associated PAH (level IIb recommendation)
All three documents:
Perform vasoreactivity testing before starting PAH targeted therapy
Use of high dose CCBs in patients with positive vasoreactivity testing
All three documents:
Indicated in patients with IPAH, heritable PAH, and anorexigen related PAH
2015 ESC/ERS guidelines:
Vasodilator responsiveness does not predict favorable long term response to CCB therapy in APAH
Combination PAH targeted therapy
CHEST guideline and expert panel report 2007:
Start monotherapy in patients who are functional class II and III. Add second agent for patients “who remain symptomatic on stable and appropriate doses” of ERAs or PDE5 inhibitors
ACCF/AHA 2009 expert consensus document:
The safety and efficacy of combination therapy in PAH is a subject of active investigation
2015 ESC/ERS guidelines (reflecting results of AMBITION trial published in 2015 86):
Level I recommendation for the initial use of ambrisentan and tadalafil as therapy in patients who are functional class II and III. Other combination therapies are given level IIa and IIIb recommendations
Table 7 summarizes emerging therapies for PAH that address novel targets.
Tremendous advances have been made in the diagnosis and management of PAH. As a result, patients have access to four classes of PAH specific agents. However, none of these agents is curative and all are expensive. A concerning trend in clinical trials is to use composite endpoints that include mortality and imply survival benefits when outcomes are driven by reductions in hospital admission or improved functional capacity. Arguably, the available PAH specific therapies do not target key features of this syndrome, including its neoplastic, mitochondrial-metabolic, inflammatory, and fibrotic phenotypes. Although basic science studies have clarified several key molecular pathways that underlie PAH, the translation of these studies into clinically available biomarkers and new therapeutic agents has been slow, and no clear path exists to move a drug or biomarker from the bench top to the bedside. Moreover, preclinical studies are siloed, and researchers (us included) often fail to consider or integrate opposing or complementary theories of disease. As a result, PAH is variably viewed as a genetic disease, an epigenetic syndrome, an autoimmune inflammatory syndrome, or a disease of disordered mitochondrial dynamics and metabolism. Like the parable of the blind man and the elephant, each reductionist perspective fails to integrate the totality of the data and thus fails to fully portray the broader picture that is the PAH elephant. It is increasingly clear that these mechanisms intersect, offering new therapeutic targets.
Questions for future research
How should right ventricular function best be measured in PAH patients?
MRI versus three dimensional echocardiography versus two dimensional echocardiography
What is the role of right ventricle relevant biomarkers?
Is right ventricle-pulmonary artery coupling an important metric in PAH?
Can right ventricle-pulmonary artery coupling be modified therapeutically using drugs or mechanical devices?
What is the role of assessing right ventricular contractile reserve in patients with PAH?
Should genetic testing be routinely done in patients with confirmed PAH and their family members?
Screening for associated PAH
How often should patients with risk factors for PAH (eg, scleroderma or HIV infection) have screening echocardiograms for diagnosing PAH?
Is the initial combination PAH targeted therapy strategy superior to a sequential combination therapy strategy?
What is the health economic implication of initial combination therapy?
The use of inotropes in PAH patients with RVF is associated with high mortality rates, and the inotrope of preference varies tremendously between and within centers.
What is the optimal right ventricular inotrope for use in PAH patients?
Are inotropes increasing mortality in PAH?
Is there a role for right ventricular assist devices or oxygenators in supporting PAH patients?
Therapeutic application of drugs approved for group 1 pulmonary hypertension in other groups (with the exception of riociguat for group 4 disease) is not supported by evidence
How can therapeutic creep best be avoided?
Are there subsets of group 2 and group 3 patients with elevated PVR that might benefit from a PAH targeted therapeutic agent?
How can we accelerate translation of investigator initiated research to advance therapy and diagnosis of PAH?
Can we create a forum to allow industry to examine potential therapeutic targets and biomarkers discovered by investigators to enhance knowledge translation?
Is there a need for more advanced and sensitive endpoints to evaluate progression and regression of PAH?
What is the role of routine cardiac MRI, MRI lung perfusion, exercise right heart catheterization, and measurements of right ventricle-pulmonary artery coupling?
Patients are not consulted routinely in the design of clinical trials of PAH therapies
What endpoints are most meaningful to PAH patients?
Can PAH patients advocate for the funding of investigator initiated therapies?
PAH is a very heterogeneous syndrome
How can we begin to apply a personalized medicine approach to these patients?
Is the best level to personalize PAH therapy identified at the genomic, epigenomic, or proteomic level?
Much of the epidemiology in PAH comes from registries and specialized clinics
What is the epidemiology of PAH in population based studies in adults and children?
Should we routinely compare the epidemiology of group 1 disease with the other pulmonary hypertension groups?
Does the changing epidemiology of group 1 pulmonary hypertension (with more older, obese patients) reflect misclassification of group 2 pulmonary hypertension patients with HFpEF?
What is the basis for the female preponderance in PAH?
How best can regional and national PAH registries be standardized and shared?
Basic and translational science
How does the BMPR2 pathway intersect with other important PAH pathways?
Why is BMPR2 downregulated in patients and animals lacking gene mutations?
Does the BMPR2 pathway offer viable therapeutic targets?
What classes of drugs should be the focus of phase I clinical trials in PAH?
Candidates include anti-inflammatory agents, immunomodulators, metabolic modulators that target Warburg metabolism, agents targeting fibrosis, agents modulating mitochondrial dynamics, inhibitors of rho kinase, and agents that enhance expression of or function of BMPR2
Although animal models of PAH are imperfect, it may be that the experimental design of preclinical studies is more to blame for false positive reports of therapeutic benefits that may not be reproducible in patients.
What is the role for randomization and blinding in preclinical pulmonary hypertension studies?
Should papers assessing preclinical therapies for pulmonary hypertension be required to include measures of pulmonary arterial pressure, cardiac output, and/or PVR?
Is there a standard duration of observation and ethically acceptable mortality surrogate that should be used in preclinical studies to avoid falsely optimistic assessment of drug benefit?
Can we create standardized protocols and models for preclinical assessment of PAH to make findings more robust and representative of the human disease?
How can we better share DNA, mRNA, proteins, cells, and tissues from patients with PAH to advance discovery?
Should resource sharing always result in authorship, or would a fee for service repository of biomaterials be beneficial?
Are the peripheral vasculature, right ventricular coronary vasculature, or skeletal muscles directly affected in PAH?
Are circulating inflammatory, autoimmune, or epigenetic factors directly targeting tissues beyond the pulmonary vasculature in PAH?
Glossary of abbreviations
6MWD—six minute walk distance
APAH—associated pulmonary arterial hypertension
BMP—bone morphogenetic protein
BMPR-II—type II bone morphogenetic protein receptor
CCB—calcium channels blocker
CTEPH—chronic pulmonary thromboembolic pulmonary hypertension
cGMP—cyclic guanosine monophosphate
eNOS—endothelial nitric oxide synthase
ERA—endothelin receptor antagonists
FAO—fatty acid oxidation (FAO)
GRK2—G protein-coupled receptor kinase-2
HFpEF—heart failure and preserved ejection fraction
IPAH—idiopathic pulmonary arterial hypertension
Kv—voltage-gated potassium channels
LVEDP—left ventricular end-diastolic pressure
MCU—mitochondrial calcium uniporter
mPAP—mean pulmonary artery pressure
MRI—magnetic resonance imaging
NFAT—nuclear factor of activated T cells
NIH—National Institutes of Health
NT-proBNP—N terminal pro B type natriuretic peptide
PAEC—pulmonary arterial endothelial cells
PAH—pulmonary arterial hypertension
PASMC—pulmonary arterial smooth muscle cells
PCH—pulmonary capillary hemangiomatosis
PCWP—pulmonary capillary wedge pressure
PDK—pyruvate dehydrogenase kinase
PTBP1—polypyrimidine tract-binding protein
PVOD—pulmonary veno-occlusive disease
PVR—pulmonary vascular resistance
RHC—right heart catheterization
RVEF—right ventricular ejection fraction
RVF—right ventricular failure
RVH—right ventricular hypertrophy
RVSP—right ventricular systolic pressure
sGC—soluble guanylate cyclase stimulators
TAPSE—tricuspid annular plane systolic excursion
Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors
Contributors: All authors defined intellectual content, conducted literature research, acquired data, and participated in manuscript preparation, editing, and critical review.
Funding: SLA is supported by Canada Foundation for Innovation (229252 and 33012), NIH RO1s HL113003 and HL071115, a Tier 1 Canada research chair in mitochondrial dynamics and translational medicine (950-229252), and a grant from the William J Henderson Foundation. TT is funded by AHA scientist development grant 15SDG25560048. MLO is supported by a CIHR project grant (PJT-152916) and a Tier 2 Canada research chair.
Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: TT received a modest honorarium from Gilead and Actelion for participating in an advisory board.
Provenance and peer review: Commissioned; externally peer reviewed.
Patient involvement: No patients were involved in the creation of this article.