2022 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension

5.1.1. Clinical presentation

Symptoms of PH are mainly linked to right ventricle (RV) dysfunction, and typically associated with exercise in the earlier course of the disease [25, 26]. The cardinal symptom is dyspnoea on progressively minor exertion. Other common symptoms are related to the stages and severity of the disease, and are listed in Figure 2 [60–62]. Potential clinical signs and physical findings are summarized in Figure 3 [60, 61]. Importantly, the physical examination may also be the key to identifying the underlying cause of PH (see Figure 3).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2

Symptoms in patients with pulmonary hypertension.

WHO-FC, World Health Organization functional class. ^aThoracic compression syndromes are found in a minority of patients with PAH with pronounced dilation of the pulmonary artery, and may occur at any disease stage and even in patients with otherwise mild functional impairment.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3

Clinical signs in patients with pulmonary hypertension.

CHD, congenital heart disease; CTEPH, chronic thrombo-embolic pulmonary hypertension; DVT, deep venous thrombosis; GORD, gastro-oesophageal reflux disease; HHT, hereditary haemorrhagic telangiectasia; PDA, patent ductus arteriosus; PH, pulmonary hypertension; PVOD, pulmonary veno-occlusive disease; RV, right ventricle; SSc, systemic sclerosis.

5.1.2. Electrocardiogram

Electrocardiogram (ECG) abnormalities (Table 8) may raise suspicion of PH, deliver prognostic information, and detect arrhythmias and signs of LHD. In adults with clinical suspicion of PH (e.g. unexplained dyspnoea on exertion), right axis deviation has a high predictive value for PH [63]. A normal ECG does not exclude the presence of PH, but a normal ECG in combination with normal biomarkers (BNP/NT-proBNP) is associated with a low likelihood of PH in patients referred for suspected PH or at risk of PH (i.e. after acute PE) [64, 65].

5.1.3. Chest radiography

Chest radiography presents abnormal findings in most patients with PH; however, a normal chest X-ray does not exclude PH [68]. Radiographic signs of PH include a characteristic configuration of the cardiac silhouette due to right heart (right atrium [RA]/RV) and PA enlargement, sometimes with pruning of the peripheral vessels. In addition, signs of the underlying cause of PH, such as LHD or lung disease, may be found (Table 9) [25, 26, 60, 69, 70].

5.1.4. Pulmonary function tests and arterial blood gases

Pulmonary function tests (PFTs) and analysis of arterial blood gas (ABG) or arterialized capillary blood are necessary to distinguish between PH groups, assess comorbidities and the need for supplementary oxygen, and determine disease severity. The initial work-up of patients with suspected PH should comprise forced spirometry, body plethysmography, lung diffusion capacity for carbon monoxide (DLCO), and ABG.

In patients with PAH, PFTs are usually normal or may show mild restrictive, obstructive, or combined abnormalities [71, 72]. More severe PFT abnormalities are occasionally found in patients with PAH associated with CHD [73], and those with group 3 PH. The DLCO may be normal in patients with PAH, although it is usually mildly reduced [71]. A severely reduced DLCO (<45% of the predicted value) in the presence of otherwise normal PFTs can be found in PAH associated with systemic sclerosis (SSc), PVOD, in PH group 3—associated with emphysema, interstitial lung disease (ILD), or combined pulmonary fibrosis and emphysema—and in some PAH phenotypes [74]. A low DLCO is associated with a poor prognosis in several forms of PH [75–78].

Patients with PAH usually have normal or slightly reduced partial pressure of arterial oxygen (PaO₂). Severe reduction of PaO₂ might raise suspicion for patent foramen ovale, hepatic disease, other abnormalities with right-to-left shunt (e.g. septal defect), or low-DLCO-associated conditions.

Partial pressure of arterial carbon dioxide (PaCO₂) is typically lower than normal due to alveolar hyperventilation [79]. Low PaCO₂ at diagnosis and follow-up is common in PAH and associated with unfavourable outcomes [80]. Elevated PaCO₂ is very unusual in PAH and reflects alveolar hypoventilation, which in itself may be a cause of PH. Overnight oximetry or polysomnography should be performed if there is suspicion of sleep-disordered breathing or hypoventilation [81].

5.1.5. Echocardiography

Independent of the underlying aetiology, PH leads to RV pressure overload and dysfunction, which can be detected by echocardiography [82–84]. When performed accurately, echocardiography provides comprehensive information on right and left heart morphology, RV and LV function, and valvular abnormalities, and gives estimates of haemodynamic parameters. Echocardiography is also a valuable tool with which to detect the cause of suspected or confirmed PH, particularly with respect to PH associated with LHD or CHD. Yet, echocardiography alone is insufficient to confirm a diagnosis of PH, which requires RHC.

Given the heterogeneous nature of PH and the peculiar geometry of the RV, there is no single echocardiographic parameter that reliably informs about PH status and underlying aetiology. Therefore, a comprehensive echocardiographic evaluation for suspected PH includes estimating the systolic pulmonary arterial pressure (sPAP) and detecting additional signs suggestive of PH, aiming at assigning an echocardiographic level of probability of PH. Echocardiographic findings of PH, including estimating pressure and signs of RV overload and/or dysfunction, are summarized in Figure 4.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4

Transthoracic echocardiographic parameters in the assessment of pulmonary hypertension.

Ao, aorta; IVC, inferior vena cava; LA, left atrium; LV, left ventricle; PH, pulmonary hypertension; RA, right atrium; RAP, right atrial pressure; RV, right ventricle; RVOT AT, right ventricular outflow tract acceleration time; sPAP, systolic pulmonary artery pressure; TAPSE, tricuspid annular plane systolic excursion; TR, tricuspid regurgitation; TRV, tricuspid regurgitation velocity. ^aRefers to collapse on inspiration.

Estimates of sPAP are based on the peak tricuspid regurgitation velocity (TRV) and the TRV-derived tricuspid regurgitation pressure gradient (TRPG)—after excluding pulmonary stenosis—taking into account non-invasive estimates of RA pressure (RAP). Considering the inaccuracies in estimating RAP and the amplification of measurement errors by using derived variables [85–87], these guidelines recommend using the peak TRV (and not the estimated sPAP) as the key variable for assigning the echocardiographic probability of PH. A peak TRV >2.8 m/s may suggest PH; however, the presence or absence of PH cannot be reliably determined by TRV alone [88]. Lowering the TRV threshold in view of the revised haemodynamic definition of PH is not supported by available data (key narrative question 2, Supplementary Data, Section 5.1) [89–92]. Tricuspid regurgitation (TR) velocity may underestimate (e.g. in patients with severe TR) [28] or overestimate (e.g. in patients with high CO in liver disease or sickle cell disease [SCD] [93, 94], misinterpretation of tricuspid valve closure artefact for the TR jet, or incorrect assignment of a peak TRV in the case of maximum velocity boundary artefacts) pressure gradients. Hence, additional variables related to RV morphology and function are used to define the echocardiographic probability of PH (Table 10) [82–84, 95], which may then be determined as low, intermediate, or high. When interpreted in a clinical context, this probability can be used to decide the need for further investigation, including cardiac catheterization in individual patients (Figure 5).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5

Echocardiographic probability of pulmonary hypertension and recommendations for further assessment.

CPET, cardiopulmonary exercise testing; CTEPH, chronic thrombo-embolic pulmonary hypertension; echo, echocardiography; LHD, left heart disease; N, no; PAH, pulmonary arterial hypertension; PH, pulmonary hypertension; RHC, right heart catheterization; TRV, tricuspid regurgitation velocity; Y, yes. ^aOr unmeasurable. The TRV threshold of 2.8 m/s was not changed according to the updated haemodynamic definition of PH. ^bSigns from at least two categories in Table 10 (A/B/C) must be present to alter the level of echocardiographic probability of PH. ^cFurther testing may be necessary (e.g. imaging, CPET). ^dRHC should be performed if useful information/a therapeutic consequence is anticipated (e.g. suspected PAH or CTEPH), and may not be indicated in patients without risk factors or associated conditions for PAH or CTEPH (e.g. when mild PH and predominant LHD or lung disease are present).

Echocardiographic measures of RV function include the tricuspid annular plane systolic excursion (TAPSE), RV fractional area change (RV-FAC), RV free-wall strain, and tricuspid annulus velocity (S′ wave) derived from tissue Doppler imaging, and potentially RV ejection fraction (RVEF) derived from 3D echocardiography. Furthermore, the TAPSE/sPAP ratio—representing a non-invasive measure of RV–PA coupling [96]—may aid in diagnosing PH [90, 97, 98]. The pattern of RV outflow tract (RVOT) blood flow (mid-systolic ‘notching’) may suggest pre-capillary PH [99, 100].

To separate between group 2 PH and other forms of PH, and to assess the likelihood of left ventricle (LV) diastolic dysfunction, LA size and signs of LV hypertrophy should always be measured, and Doppler echocardiographic signs (e.g. E/A ratio, E/E′) should be assessed even if the reliability of the latter is considered low [16]. To identify CHD, 2D Doppler and contrast examinations are helpful, but transoesophageal contrast echocardiography or other imaging techniques (e.g. computer tomography [CT] angiography, cMRI) are needed in some cases to detect or exclude sinus venosus atrial septal defects, patent ductus arteriosus, and/or anomalous pulmonary venous return [101]. The clinical value of exercise Doppler echocardiography in identifying exercise PH remains uncertain because of the lack of validated criteria and prospective confirmatory data. In most cases, increases in sPAP during exercise are caused by diastolic LV dysfunction [16].

5.1.6. Ventilation/perfusion lung scan

A ventilation/perfusion (V/Q) lung scan (planar or single-photon emission computed tomography [SPECT]) is recommended in the diagnostic work-up of patients with suspected or newly diagnosed PH, to rule out or detect signs of CTEPH [102, 103]. The V/Q SPECT is superior to planar imaging and is the methodology of choice; however, SPECT has been widely evaluated in assessing PE, but not to the same degree in CTEPH [68]. In the absence of parenchymal lung disease, a normal perfusion scan excludes CTEPH with a negative predicted value of 98% [104, 105]. In most patients with PAH, V/Q scintigraphy is normal or shows a speckled pattern but no typical perfusion defects characteristic of PE or CTEPH, whereas matched V/Q defects may be found in patients with lung disease (i.e. group 3 PH). Non-matched perfusion defects similar to those seen in CTEPH may be present in 7–10% of patients with PVOD/PCH or PAH [106, 107]. Deposition of the perfusion agent in extrapulmonary organs may hint to cardiac or pulmonary right-to-left shunting and has been reported in CHD, hepato-pulmonary syndrome, and pulmonary arteriovenous malformations (PAVMs) [68].

5.1.7. Non-contrast and contrast-enhanced chest computed tomography examinations, and digital subtraction angiography

Computed tomography (CT) imaging may provide important information for patients with unexplained dyspnoea or suspected/confirmed PH. The CT signs suggesting the presence of PH include an enlarged PA diameter, a PA-to-aorta ratio >0.9, and enlarged right heart chambers [68]. A combination of three parameters (PA diameter ≥30 mm, RVOT wall thickness ≥6 mm, and septal deviation ≥140° [or RV:LV ratio ≥1]) is highly predictive of PH [108]. Non-contrast chest CT can help determine the cause of PH when there are features of parenchymal lung disease, and may also point towards the presence of PVOD/PCH by showing centrilobular ground-glass opacities (which may also be found in PAH), septal lines, and lymphadenopathy [68].

Computed tomography pulmonary angiography (CTPA) is mainly used to detect direct or indirect signs of CTEPH, such as filling defects (including thrombus adhering to the vascular wall), webs or bands in the PAs, PA retraction/dilatation, mosaic perfusion, and enlarged bronchial arteries. Importantly, the diagnostic accuracy of CTPA for CTEPH is limited (at the patient level, sensitivity and specificity are 76% and 96%, respectively) [109], but was reported to be higher when modern, high-quality multi-detector CT scanners were used and when interpreted by experienced readers [109, 110]. Computed tomography pulmonary angiography may also be used to detect other cardiovascular abnormalities, including intracardiac shunts, abnormal pulmonary venous return, patent ductus arteriosus, and PAVMs.

In patients presenting with a clinical picture of acute PE, chest CT may be helpful in detecting signs of hitherto undetected CTEPH, which may include the presence of the above CTEPH signs, and RV hypertrophy as a sign for chronicity [111, 112]. Detecting ‘acute on chronic’ PE is important, as it may impact the management of patients with presumed acute PE.

Dual-energy CT (DECT) angiography and iodine subtraction mapping may provide additional diagnostic information by creating iodine maps [113], which reflect lung perfusion, thereby possibly increasing the diagnostic accuracy for CTEPH [114]. Although increasingly used, the diagnostic value of DECT in the work-up of patients with PH has not been established.

Digital subtraction angiography (DSA) is mainly used to confirm the diagnosis of CTEPH and to assess treatment options (i.e. operability or accessibility for BPA). Most centres use conventional two- or three-planar DSA. However, C-arm CT imaging may provide a higher spatial resolution, potentially identifying more target vessels for BPA and providing procedural guidance [115, 116].

5.1.8. Cardiac magnetic resonance imaging

Cardiac magnetic resonance imaging accurately and reproducibly assesses atrial and ventricular size, morphology, and function. Additional information on RV/LV myocardial strain can be obtained by applying tagging or by post-processing feature tracking. In addition, cMRI can be used to measure blood flow in the PA, aorta, and vena cava, allowing for quantifying stroke volume (SV), intracardiac shunt, and retrograde flow. By combining contrast magnetic resonance (MR) angiography and pulmonary perfusion imaging with late gadolinium-enhancement imaging of the myocardium, a complete picture of the heart and pulmonary vasculature can be obtained (see Supplementary Data, Table S2 for cMRI indices and normal values). A limitation is that there is no established method with which to estimate PAP. Even though the cost and availability of the technique precludes its use in the early diagnosis of PAH, it is sensitive in detecting early signs of PH and diagnosing CHD [117].

5.1.9. Blood tests and immunology

The initial diagnostic assessment of patients with newly diagnosed PH/PAH aims to identify comorbidities and possible causes or complications of PH. Laboratory tests that should be obtained at the time of PH diagnosis include: blood counts (including haemoglobin [Hb]); serum electrolytes (sodium, potassium); kidney function (creatinine, calculation of estimated glomerular filtration rate, and urea); uric acid; liver parameters (alanine aminotransferase, aspartate aminotransferase, and alkaline phosphatase, γ-glutamyl transpeptidase, bilirubin); iron status (serum iron, transferrin saturation, and ferritin); and BNP or NT-proBNP. In addition, serological studies should include testing for hepatitis viruses and HIV. Basic immunology laboratory work-up is recommended, including screening tests for anti-nuclear antibodies, anti-centromere antibodies, and anti-Ro. Screening for biological markers of antiphospholipid syndrome is recommended in patients with CTEPH. Additional thrombophilia screening is not generally recommended, unless therapeutic consequences are to be expected [118]. Pulmonary arterial hypertension and other forms of severe PH can be associated with thyroid function disorders; hence, laboratory screening should include at least thyroid-stimulating hormone.

5.1.10. Abdominal ultrasound

An abdominal ultrasound examination should be part of the comprehensive diagnostic work-up of patients with newly diagnosed PH, particularly if liver disease is suspected. A major objective is to search for liver disease and/or portal hypertension, or portocaval shunt (Abernethy malformation). During the course of the disease, patients with PH may develop secondary organ dysfunction mainly affecting the liver and kidneys [119]. In these patients, abdominal ultrasound is needed for differential diagnostic reasons and to assess the extent of organ damage.

5.1.11. Cardiopulmonary exercise testing

Cardiopulmonary exercise testing (CPET) is a useful tool to assess the underlying pathophysiologic mechanisms leading to exercise intolerance. Patients with PAH show a typical pattern, with a low end-tidal partial pressure of carbon dioxide (P_ETCO₂), high ventilatory equivalent for carbon dioxide (VE/VCO₂), low oxygen pulse (VO₂/HR), and low peak oxygen uptake (VO₂) [120]. These findings should prompt consideration of PVD. In patients with LHD or COPD, such a pattern may indicate an additional pulmonary vascular limitation [121, 122]. In populations at risk of PAH, such as those with SSc, a normal peak VO₂ seems to exclude the diagnosis of PAH [123].

5.1.12. Right heart catheterization, vasoreactivity, exercise, and fluid challenge

5.1.13. Genetic counselling and testing

Mutations in PAH genes have been identified in familial PAH, IPAH, PVOD/PCH, and anorexigen-associated PAH (Table 13) [148]. The screening recommendations herein specifically relate to patients with an a priori diagnosis of PAH and not ‘at-risk’ populations being screened for PAH (see Section 5.3). All patients with these conditions should be informed about the possibility of a genetic condition and that family members could carry a mutation that increases the risk of PAH, allowing for screening and early diagnosis [33, 148]. Even if genetic testing is not performed, family members should be made aware of early signs and symptoms, to ensure that a timely and appropriate diagnosis is made [148].

Genetic counselling by appropriately trained PAH providers or geneticists should be performed prior to genetic testing, to address the complex questions related to penetrance, genetically at-risk family members, reproduction, genetic discrimination, and psychosocial issues. Careful genetic counselling with genetic counsellors or medical geneticists is critical prior to genetic testing for asymptomatic family members [148].

If the familial mutation is known and an unaffected family member tests negative for that mutation, the risk of PAH for that person is the same as for the general population [148].

Many of the less common mutations outlined have a potential additional set of syndromic features. These are summarized in Table 13 where specific clinical history, examination, and investigations are suggested. In particular, clinicians should undertake a thorough history and examination, as syndromic PAH diagnoses may be missed if not interrogated. For example, in one of the largest studies to date, TBX4, ALK1, and ENG mutations were represented in the top six most common genetic findings in adults with previously diagnosed IPAH [149]. These findings have been confirmed and extended in international genetics consortia in 4241 patients with PAH [150]. It is therefore apparent that there is either phenotypic heterogeneity of these syndromes or missed diagnostic features. As more genes associated with PAH are discovered, it will become increasingly difficult to individually test for each. Next-generation sequencing has enabled the development of gene panels to simultaneously interrogate several genes [151]. It is, however, important to check the genes included in the panel at the time of testing, since the composition changes as genetic discoveries advance.

5.2.1 Step 1 (suspicion)

Patients with PH are likely to be seen by first-line physicians, mainly general practitioners, for non-specific symptoms. Initial evaluation should include a comprehensive medical (including familial) history, thorough physical examination (including measurement of blood pressure, heart rate, and pulse oximetry), blood test to determine BNP/NT-proBNP, and resting ECG. This first step may raise a suspicion of a cardiac or respiratory disorder causing the symptoms.

5.2.2. Step 2 (detection)

The second step includes classical, non-invasive lung and cardiac testing. Among those tests, echocardiography is an important step in the diagnostic algorithm (Figure 6), as it assigns a level of probability of PH, irrespective of the cause. In addition, it is an important step in identifying other cardiac disorders. Based on this initial assessment, if causes other than PH are identified and/or in case of low probability of PH, patients should be managed accordingly.

5.2.3. Step 3 (confirmation)

Patients should be referred to a PH centre for further evaluation in the following situations: (1) when an intermediate/high probability of PH is established; (2) in the presence of risk factors for PAH, or a history of PE. A comprehensive work-up should be performed, with the goal of establishing the differential diagnoses and distinguishing between the various causes of PH according to the current clinical classification. The PH centre is responsible for performing an invasive assessment according to the clinical scenario.

At any time, warning signs must be recognized, as they are associated with worse outcomes and warrant immediate intervention. Such warning signs include: rapidly evolving or severe symptoms (WHO-FC III/IV), clinical signs of RV failure, syncope, signs of low CO state, poorly tolerated arrhythmias, and compromised or deteriorated haemodynamic status (hypotension, tachycardia). Such cases must be immediately managed as inpatients for initial work-up at a nearby hospital or PH centre. The presence of RV dysfunction by echocardiography, elevated levels of cardiac biomarkers, and/or haemodynamic instability must prompt referral to a PH centre for immediate assessment.

This diagnostic process emphasizes the importance of sufficient awareness and collaboration between first-line, specialized medicine and PH centres. Effective and rapid collaboration between each partner permits earlier diagnosis and management, and improves outcomes.

5.3.1. Systemic sclerosis

In SSc, the prevalence of PAH is 5–19% [174], with an annual incidence of developing PAH of 0.7–1.5% [183–185]. Evidence for the clinical value of detecting PAH early in SSc was provided by a screening programme [186], which showed less severe haemodynamic impairment and better survival in screened patients compared with a contemporaneous, non-screened cohort [187], providing a strong rationale for screening for PAH in patients with SSc.

The diagnostic accuracy of echocardiography or other tests alone in detecting PAH is suboptimal [173]. Several screening algorithms have been studied using a combination of clinical features, echocardiography, PFTs, and NT-proBNP to select patients with SSc for RHC (DETECT [173]; Australian Scleroderma Interest Group [ASIG] [188]). Such combined approaches have improved diagnostic accuracy compared with the use of echocardiography, NT-proBNP, or PFTs alone, and are able to prevent unnecessary RHC and identify patients with mPAP 21–24 mmHg [189]. Therefore, a multimodal approach is warranted when screening patients with SSc for PAH; the echocardiographic assessment should follow the strategy described in Section 5.1.5.

Beyond initial screening, the frequency with which screening should be undertaken in asymptomatic subjects with SSc is unclear. A study from the Australian Scleroderma Study Cohort, where annual screening was recommended (some patients were screened up to 10 times), noted that most patients were diagnosed with PAH at their first screening; however, those diagnosed on subsequent screening had a lower mPAP, PVR, and WHO-FC, and better survival than those diagnosed at first screening [190]. Based on current evidence, annual screening for PAH in patients with SSc is sufficient. Given the financial and emotional cost associated with regular screening, stratifying subjects with SSc into those at highest and lowest risk of PAH would be desirable. Risk factors for PAH include: (1) clinical and demographic factors (i.e. breathlessness, longer disease duration, sicca symptoms, digital ulceration, older age, and male sex); and (2) the results of investigations (e.g. positive anti-centromere antibody profile, mild ILD, low DLCO, elevated FVC/DLCO ratio, or elevated NT-proBNP) [174, 191]. A recent meta-analysis showed that reduced digital capillary density, as assessed by video-capillaroscopy, or progression to a severe active/late pattern of vascular involvement is also a risk factor for PAH [192]. In addition to identifying patients at increased risk of PAH, a simple prediction model integrating symptoms, DLCO, and NT-proBNP identified subjects at very low probability of PAH who could potentially avoid further specific testing for PH [183]. Furthermore, CPET may help to identify patients with SSc with a low risk of having PAH and thus to avoid unnecessary RHC [123].

The recommendations on screening for PAH in SSc have been established based on key narrative question 3 (Supplementary Data, Section 5.2).

5.3.2. BMPR2 mutation carriers

In the evolving list of genes known to be associated with PAH, experience is largely restricted to BMPR2 mutation carriers who carry a lifetime risk of developing PAH of ∼20%, with penetrance higher in female carriers (42%) compared with male carriers (14%) [33, 148, 193]. There is currently no accepted screening strategy for evaluating PAH in BMPR2 mutation carriers. At present, based on expert consensus, asymptomatic relatives who screen positive for PAH-causing mutations are often offered yearly screening echocardiography [25, 26] The DELPHI-2 study, which prospectively screened carriers and relatives, recently demonstrated a 9.1% pick up over 47±27 months of PAH, with 2/55 diagnosed at baseline and 3/55 at follow-up; this equates to an incidence of 2.3%/year [33]. The screening schedule included ECG, NT-proBNP, DLCO, echocardiography, CPET, and optional RHC; however, none of the cases would have been picked up by echocardiography alone. Screening programmes should adopt a multimodal approach, although the optimal strategy and screening period remains undefined and will require multinational, multicentre study.

5.3.3. Portal hypertension

An estimated 1–2% of patients with liver disease and portal hypertension develop PoPH [176, 194], which is of particular relevance in patients considered for transjugular portosystemic shunting or liver transplantation. In such patients, echocardiography is recommended to screen for PAH, even in the absence of symptoms. By using echocardiography, sPAP can be measured in ∼80% of patients with portal hypertension, which aids decisions to perform RHC. In patients assessed for liver transplantation, one study showed that an sPAP of >50 mmHg had 97% sensitivity and 77% specificity for detecting moderate-to-severe PAH [195]. Other investigators have recommended RHC when sPAP is >38 mmHg [196]. When screening for PoPH, it is advised to assess the echocardiographic probability of PH (see Section 5.1.5). In agreement with the International Liver Transplant Society, for patients awaiting liver transplantation, it is recommended to reassess for PAH annually, although the optimal interval remains unclear [175].

5.3.4. Pulmonary embolism

Chronic thrombo-embolic pulmonary hypertension is an uncommon and under-diagnosed complication of acute PE [112]. The reported cumulative incidence of CTEPH after acute, symptomatic PE ranges 0.1–11.8%, depending on the collective investigated [112, 178, 197–199]. A systematic review and meta-analysis reported a CTEPH incidence of 0.6% in all patients with acute PE, 3.2% in survivors, and 2.8% in survivors without major comorbidities [178]. A multicentre, observational, screening study reported a CTEPH incidence of 3.7/1000 patient-years and a 2 year cumulative incidence of 0.79% following acute PE [200]. A recent prospective observational study (FOCUS, Follow-up After Acute Pulmonary Embolism) showed a cumulative 2 year incidence of 2.3% and 16.0% for CTEPH and post-PE impairment, respectively, which were both associated with a higher risk of rehospitalization and death [201]. Due to insufficient awareness, some patients may have a delayed diagnosis of CTEPH because they may initially be misclassified as acute PE [112]. In this context, the current guidelines do not recommend routine follow-up of patients with PE by imaging methods of the pulmonary vascular tree, but suggest evaluating the index imaging test used to diagnose acute PE for signs of CTEPH. Echocardiography is the preferred first-line diagnostic test in patients with suspected CTEPH [103].

Up to 50% of patients have persistent perfusion defects after an acute PE; however, the clinical relevance is unclear [202–204]. All patients in whom symptoms can be attributed to post-thrombotic deposits within PAs are considered to have CTEPD, with or without PH [54]. While persistent dyspnoea is common after acute PE [205], the prevalence of CTEPD without PH is unknown and requires further study (see Section 10.1). A study exploring screening for CTEPH following acute PE identified, using echocardiography, a low yield of additional CTEPH diagnoses in asymptomatic patients [206]. Current PE guidelines recommend that further diagnostic evaluation may be considered in asymptomatic patients with risk factors for CTEPH at 3–6 months’ follow-up [103, 207]. Approaches to early detection of CTEPH following acute PE are based on identifying patients at increased risk [208]. In patients with persistent or new-onset dyspnoea after PE, non-invasive approaches use echocardiography to assess for PH and cross-sectional imaging to assess for persistent perfusion defects. Limited data exist on strategies using DECT, CT lung subtraction iodine mapping, or 3D MR perfusion imaging. Scoring systems, including the Leiden CTEPH rule-out criteria [206, 209] can be used to inform diagnostic strategies. Cardiopulmonary exercise testing may identify characteristic features of exercise limitation due to PVD, or suggest an alternative diagnosis. The optimal timing for assessing symptoms to aid early detection of CTEPH may be 3–6 months after acute PE, coinciding with the routine evaluation of anticoagulant treatment, but earlier assessment may be necessary in highly symptomatic or deteriorating patients [54, 103].

6.2.1. Clinical parameters

Clinical assessment is a key part of evaluating patients with PAH, as it provides valuable information for determining disease severity, improvement, deterioration, or stability. At follow-up, changes in WHO-FC (Table 15), episodes of chest pain, arrhythmias, haemoptysis, syncope, and signs of right HF provide important information. Physical examination should assess heart rate, rhythm, blood pressure, cyanosis, enlarged jugular veins, oedema, ascites, and pleural effusions. The WHO-FC is one of the strongest predictors of survival, both at diagnosis and follow-up [210–212], and worsening WHO-FC is one of the most alarming indicators of disease progression, which should trigger further investigations to identify the cause(s) of clinical deterioration [210, 213, 214].

6.2.2. Imaging

Imaging of the heart plays an essential role in the follow-up of patients with PAH. Several echocardiographic and cMRI parameters have been proposed to monitor RV function during the course of PAH. Table S2 provides a list of imaging parameters and relative cut-off values associated with increased and decreased risk of adverse events.

6.2.3. Haemodynamics

Cardiopulmonary haemodynamics assessed by RHC provide important prognostic information, both at the time of diagnosis and at follow-up [129, 212, 213, 216, 238, 243–245, 247, 248]. Currently available risk-stratification tools include haemodynamic variables for prognostication: RAP and PVR in REVEAL risk scores [213, 249, 250], and RAP, CI, and SvO₂ in the ESC/ERS risk-stratification table [25, 26]. The mPAP provides little prognostic information, except in acute vasodilator responders [129]. A recent study from France, which combined clinical and haemodynamic parameters, found that WHO-FC, 6MWD, RAP, and SVI (but not SV and SvO₂) were independent predictors of outcome [238].

To refine the risk-stratification table (Table 16), SVI criteria are now added with the cut-off values of >38 mL/m² and <31 mL/m² to determine low-risk and high-risk status, respectively [238].

The optimal timing of follow-up RHC has not been determined. While some centres regularly perform invasive follow-up assessments, others perform them as clinically indicated, and there is no evidence that any of these strategies is associated with better outcomes (Table 17).

6.2.4. Exercise capacity

The 6-minute walking test (6MWT) is the most widely used measure of exercise capacity in PH centres. The 6MWT is easy to perform, inexpensive, and widely accepted by patients, health professionals, and medicines agencies as an important and validated variable in PH. As with all PH assessments, 6MWT results must always be interpreted in the clinical context. The 6MWD is influenced by factors such as sex, age, height, weight, comorbidities, need for oxygen, learning curve, and motivation. Test results are usually given in absolute distance (metres) rather than the percentage of predicted values. Change in 6MWD is one of the most commonly used parameters in PAH clinical trials as a primary endpoint, key secondary endpoint, or component of clinical worsening [251]. A recent investigation showed that the best absolute-threshold values for 1 year mortality and 1 year survival, respectively, were 165 m and 440 m, respectively [252]. Improvements in 6MWD have had less predictive value than deterioration on key clinical outcomes (mortality and survival) [250, 252, 253]. These results are consistent with observations from clinical trials and registries [254, 255]; however, there is no single threshold that would apply to all patients [256]. Some studies have also suggested that adding SaO₂ measured by pulse oximetry and heart rate responses may improve prognostic relevance [246, 257]. Hypoxaemia observed during the 6MWT is associated with worse survival, but these findings still await confirmation in large multicentre studies.

The incremental shuttle walking test (ISWT) is an alternative maximal test for assessing patients with PAH. The ISWT has a potential advantage over the 6MWT in that it does not have a ceiling effect; furthermore, it keeps the simplicity of a simple-to-perform field test, in contrast to CPET. However, the ISWT experience in PAH is currently limited [258].

Cardiopulmonary exercise testing is a non-invasive method for assessing functional capacity and exercise limitation. It is usually performed as a maximal exercise test, and is safe even in patients with severe exercise limitation [259, 260]. Most PH centres use an incremental ramp protocol, although the test has not yet been standardized for this patient population. Robust prognostic evidence for peak VO₂ and VE/VCO₂ has been found in three studies, all powered for multivariable analysis [261–263]. When associated with SVI, peak VO₂ provided useful information to further stratify patients with PAH at intermediate risk [264]. However, the added value of CPET on top of common clinical and haemodynamic variables remains largely unexplored.

6.2.5. Biochemical markers

Considerable efforts have been made to identify additional biomarkers of PVD, addressing prognosis [265–272], diagnosis, and differentiation of PH subtypes [270, 273–276], as well as PAH treatment response [266]. Emerging proteins related to PAH and vascular remodelling include bone morphogenetic proteins 9 and 10 and translationally controlled tumour protein [270, 277, 278]. Proteome-wide screening in IPAH and HPAH identified a multimarker panel with prognostic information in addition to the REVEAL risk score [271]. Another study found that early development of SSc-associated PAH (PAH-SSc) was predicted by high circulating levels of C-X-C motif chemokine 4 in patients with SSc [276]. However, none of these biomarkers have been introduced in clinical practice.

Thus, BNP and NT-proBNP remain the only biomarkers routinely used in clinical practice at PH centres, correlating with myocardial stress and providing prognostic information [279]. Brain natriuretic peptide and NT-proBNP are not specific for PH, as they can be elevated in other forms of heart disease, exhibiting great variability. The previously proposed cut-off levels of BNP (<50, 50–300, and >300 ng/L) and NT-proBNP (<300, 300–1400, and >1400 ng/L) for low, intermediate, and high risk, respectively, in the ESC/ERS risk-assessment model at baseline and during follow-up are prognostic for long-term outcomes and can be used to predict response to treatment [266]. Refined cut-off values for BNP (<50, 50–199, 200–800, and >800 ng/L) and NT-pro-BNP (<300, 300–649, 650–1100, and >1100 ng/L) for low, intermediate–low, intermediate–high, and high risk, respectively, have recently been introduced as part of a four-strata risk-assessment strategy (see Section 6.2.7) [280].

6.2.6. Patient-reported outcome measures

A patient-reported outcome measure (PROM) is a term for health outcomes that are ‘self-reported’ by the patient. It is the patient's experience of living with PH and its impact on them and their caregivers, including symptomatic, intellectual, psychosocial, spiritual, and goal-orientated dimensions of the disease and its treatment. Despite treatment advances improving survival, patients with PAH present with a range of non-specific yet debilitating symptoms, which affect health-related quality of life (HR-QoL) [281, 282].

Patient-reported outcome measures remain an underused outcome measure. Tools validated in patients with PAH should be used to assess HR-QoL [282, 283] in individual patients. There has been a reliance on generic PROMs, which have been studied in patients with PAH but may lack sensitivity to detect changes in PAH [284, 285]. To address this, a number of PH-specific HR-QoL instruments have been developed and validated (e.g. Cambridge Pulmonary Hypertension Outcome Review [CAMPHOR] [286], emPHasis-10 [282, 287], Living with Pulmonary Hypertension [288], and Pulmonary Arterial Hypertension-Symptoms and Impact [PAH-SYMPACT]) [289]. These disease-specific PROMs track functional status, clinical deterioration, and prognosis in PAH, and are more sensitive to the differences in the risk status than generic PROMs [290, 291]. In addition, HR-QoL scores provide independent prognostic information [287].

6.2.7. Comprehensive prognostic evaluation, risk assessment, and treatment goals

In the 2015 ESC/ERS Guidelines for the diagnosis and treatment of PH, risk assessment was based on a multiparametric approach using a three-strata model to classify patients at low, intermediate, or high risk of death. Originally, these strata were based on estimated 1 year mortality rates of <5%, 5–10%, and >10%, respectively [25, 26]. Since then, registry data have shown that observed 1 year mortality rates in the intermediate- and high-risk groups were sometimes higher than predicted (i.e. up to 20% in the intermediate-risk group and >20% in the high-risk group). These numbers have been updated accordingly in the revised three-strata risk model (Table 16) [292–294].

Several abbreviated approaches of the 2015 ESC/ERS risk-stratification tool have been introduced and independently validated using the Swedish Pulmonary Arterial Hypertension Registry (SPAHR) [292], the Comparative, Prospective Registry of Newly Initiated Therapies for PH (COMPERA) [293], and the French PH Registry (FPHR) [295]. Other risk-stratification tools have been developed from the US REVEAL, including the REVEAL 2.0 risk score calculator, and an abridged version (REVEAL Lite 2) [249, 296]. In all these studies, WHO-FC, 6MWD, and BNP/NT-proBNP emerged as the variables with the highest predictive value.

The main limitation of the 2015 ESC/ERS three-strata, risk-assessment tool is that 60–70% of the patients are classified as intermediate risk [292–295, 297–303]. An initial attempt to substratify the intermediate-risk group has been proposed, using a modified mean score in the SPAHR equation (with low–intermediate, 1.5–1.99 and high–intermediate, 2.0–2.49 as cut-offs), where the high–intermediate group was associated with worse survival [302]. There have also been attempts to further improve risk stratification by exploring the additional value of new biomarkers [304], or by measuring RV structure and function by echocardiography and cMRI [231, 305, 306]. Other strategies have included incorporating renal function [307] or combining 6MWD with TAPSE/sPAP ratio [96, 97]; however, all of these strategies have to be further validated.

Two recent registry studies have evaluated a four-strata, risk-assessment tool based on refined cut-off levels for WHO-FC, 6MWD, and NT-proBNP (Table 18) [280, 308]. Patients were categorized as low, intermediate–low, intermediate–high, or high risk. Together, these studies included >4000 patients with PAH and showed that the four-strata model performed at least as well as the three-strata model in predicting mortality. The four-strata model predicted survival in patients with IPAH, HPAH, DPAH, and PAH associated with CTD (including the SSc subgroup), and in patients with PoPH. The observed 1-year mortality rates in the four risk strata were 0–3%, 2–7%, 9–19%, and >20%, respectively. Compared with the three-strata model, the four-strata model was more sensitive to changes in risk from baseline to follow-up, and these changes were associated with changes in the long-term mortality risk. The main advantage of the four-strata model over the three-strata model is better discrimination within the intermediate-risk group, which helps guide therapeutic decision-making (see Section 6.3.4). For these reasons, the four-strata model is included in the updated treatment algorithm (see Figure 9). However, the three-strata model is maintained for initial assessment, which should be comprehensive and include echocardiographic and haemodynamic variables, for which cut-off values for the four-strata model have yet to be established.

Several studies have identified WHO-FC, 6MWD, and BNP/NT-proBNP as the strongest prognostic predictors [293, 295, 296]. With the abbreviated risk-assessment tools, missing values become an important limitation. REVEAL Lite 2 provides accurate prediction when one key variable (WHO-FC, 6MWD, or BNP/NT-proBNP) is unavailable, but is no longer accurate when two of these variables are missing [293, 296]. The original three-strata SPAHR/COMPERA risk tool was developed with at least two variables available, while the four-strata model was developed and validated in patients for whom all three variables were available. It is therefore recommended to use at least these three variables for risk stratification. However, two components may be used when variables are missing, especially when a functional criterion (WHO-FC or 6MWD) is combined with BNP or NT-proBNP [296].

Collectively, the available studies support a risk-based, goal-orientated treatment approach in patients with PAH, where achieving and/or maintaining a low-risk status is favourable and recommended (key narrative question 4, Supplementary Data, Section 6.1) [298, 300, 303, 309, 310]. For risk stratification at diagnosis, use of the three-strata model is recommended taking into account as many factors as possible (Table 16), with a strong emphasis on disease type, WHO-FC, 6MWD, BNP/NT-proBNP, and haemodynamics. At follow-up, the four-strata model (Table 18) is recommended as a basic risk-stratification tool, but additional variables should be considered as needed, especially right heart imaging and haemodynamics. At any stage, individual factors such as age, sex, disease type, comorbidities, and kidney function should also be considered.

6.3.1. General measures

Managing patients with PAH requires a comprehensive treatment strategy and multidisciplinary care. In addition to applying PAH drugs, general measures and care in special situations represent integral components of optimized patient care. In this context, the systemic consequences of PH and right-sided HF, often contributing to disease burden, should be appropriately managed [119].

6.3.2. Special circumstances

6.3.3. Pulmonary arterial hypertension therapies

6.3.3.1. Calcium channel blockers

Patients with PAH who respond favourably to acute vasoreactivity testing (Figure 8) may respond favourably to treatment with CCBs [129, 146]. Less than 10% of patients with IPAH, HPAH, or DPAH are responders, while an acute vasodilator response does not predict a favourable long-term response to CCBs in patients with other forms of PAH [129, 146, 378]. The CCBs that have predominantly been used in PAH are nifedipine, diltiazem, and amlodipine [129, 146]. Amlodipine and felodipine are increasingly being used in clinical practice due to their long half-life and good tolerability. The daily doses that have shown efficacy in PAH are relatively high and they must be reached progressively (Table 19). The most common adverse events are systemic hypotension and peripheral oedema.

View this table:

• View inline
• View popup

TABLE 19

Dosing of pulmonary arterial hypertension medications in adults

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8

Vasoreactivity testing algorithm of patients with presumed diagnosis of idiopathic, heritable, or drug-associated pulmonary arterial hypertension.

BNP, brain natriuretic peptide; I/H/D-PAH, idiopathic, heritable, drug-associated pulmonary arterial hypertension; mPAP, mean pulmonary arterial pressure; N, no; NT-proBNP, N-terminal pro-brain natriuretic peptide; PH, pulmonary hypertension; PVR, pulmonary vascular resistance; WHO-FC, World Health Organization functional class; WU, Wood units; Y, yes. ^aInhaled nitric oxide and inhaled iloprost are recommended; intravenous epoprostenol can be used if inhaled nitric oxide or inhaled iloprost are unavailable. ^bSee text for details. ^cmPAP ≤30 mmHg and PVR ≤4 WU.

Patients who meet the criteria for a positive acute vasodilator response and treated with CCBs should be closely followed for safety and efficacy, with a complete reassessment after 3–6 months of therapy, including RHC. Additional acute vasoreactivity testing should be performed at re-evaluation to detect persistent vasodilator response, supporting possible increases in CCB dosage. Patients with a satisfactory chronic response present with WHO-FC I/II and marked haemodynamic improvement (ideally, mPAP <30 mmHg and PVR <4 WU) while on CCB therapy. In the absence of a satisfactory response, additional PAH therapy should be instituted. In some cases, a combination of CCBs with approved PAH drugs is required because of clinical deterioration with CCB withdrawal attempts. Patients who have not undergone a vasoreactivity study or those with a negative test should not be started on CCBs because of potentially severe side effects (e.g. severe hypotension, syncope, and RV failure), unless prescribed at standard doses for other indications [379].

Recommendations	Classa	Levelb
High doses of CCBs are recommended in patients with IPAH, HPAH, or DPAH who are responders to acute vasoreactivity testing	I	C
Close follow-up with complete reassessment after 3–4 months of therapy (including RHC) is recommended in patients with IPAH, HPAH, or DPAH treated with high doses of CCBs	I	C
Continuing high doses of CCBs is recommended in patients with IPAH, HPAH, or DPAH in WHO-FC I or II with marked haemodynamic improvement (mPAP <30 mmHg and PVR <4 WU)	I	C
Initiating PAH therapy is recommended in patients who remain in WHO-FC III or IV or those without marked haemodynamic improvement after high doses of CCBs	I	C
In patients with a positive vasoreactivity test but insufficient long-term response to CCBs who require additional PAH therapy, continuation of CCB therapy should be considered	IIa	C
CCBs are not recommended in patients without a vasoreactivity study or non-responders, unless prescribed for other indications (e.g. Raynaud's phenomenon)	III	C

CCB, calcium channel blocker; DPAH, drug-associated pulmonary arterial hypertension; HPAH, heritable pulmonary arterial hypertension; IPAH, idiopathic pulmonary arterial hypertension; mPAP, mean pulmonary arterial pressure; PAH, pulmonary arterial hypertension; PVR, pulmonary vascular resistance; RHC, right heart catheterization; WHO-FC, World Health Organization functional class; WU, Wood units. ^aClass of recommendation. ^bLevel of evidence.

RECOMMENDATION TABLE 7

Recommendations for the treatment of vasoreactive patients with idiopathic, heritable, or drug-associated pulmonary arterial hypertension

6.3.4. Treatment strategies for patients with idiopathic, heritable, drug-associated, or connective tissue disease-associated pulmonary arterial hypertension

Pulmonary arterial hypertension is a rare and life-threatening disease and should be managed, where possible, at PH centres in close collaboration with the patient's local physicians.

This section describes drug treatment and is focused on non-vasoreactive patients with IPAH/HPAH/DPAH and on patients with PAH associated with connective tissue disease (PAH-CTD). Information on the dosing of PAH medication is summarized in Table 19. For other forms of PAH, treatment strategies have to be modified (see Section 7). The approach to vasoreactive patients with IPAH/HPAH/DPAH is described in Section 6.3.3.1.

In addition to targeted drug treatment, the comprehensive management of patients with PAH includes general measures that may include supplementary oxygen, diuretics to optimize volume status, psychosocial support, and standardized exercise training (Section 6.3.1) [315]. Prior to the treatment decisions, patients and their next of kin should be provided with appropriate and timely information about the risks and benefits of the treatment options so they can make the final, informed, and joint decision about the treatment with the medical team. Treatment decisions in patients with IPAH/HPAH/DPAH or PAH-CTD should be stratified according to the presence or absence of cardiopulmonary comorbidities (Section 6.3.4.3) and according to disease severity assessed by risk stratification (Section 6.2.7).

6.3.4.2. Treatment decisions during follow-up in patients without cardiopulmonary comorbidities

Patients with PAH require regular follow-up, including risk stratification and an assessment of patient concordance with therapy. Patients who achieve a low-risk status have a much superior long-term survival compared with patients with intermediate- or high-risk status [292, 295, 296]. Achieving and maintaining a low-risk profile is therefore a key objective in managing patients with PAH.

Several clinical trials have assessed the safety and efficacy of sequential combination therapy in patients with PAH. SERAPHIN enrolled 742 patients with PAH, mostly with IPAH/HPAH/DPAH and PAH-CTD, of whom 63.7% were receiving other PAH medication at the time of enrolment, mostly sildenafil [167]. In the subgroup of patients with background PAH therapy, macitentan at a daily dose of 10 mg reduced the risk of clinical worsening events compared with placebo (HR 0.62; 95% CI, 0.43–0.89) [167].

GRIPHON assessed the safety and efficacy of selexipag [419]. This study enrolled 1156 patients with PAH, also mostly with IPAH/HPAH/DPAH or PAH-CTD, who were treatment naïve or receiving background therapy with an ERA, PDE5i, or a combination of both. Selexipag at a dose of up to 1600 µg b.i.d. was associated with a reduced risk of clinical worsening events independent of the background medication. In patients receiving ERA/PDE5i combination therapy (n=376), the risk of clinical worsening events was lower with selexipag than with placebo (HR 0.63; 95% CI, 0.44–0.90) [431].

The effects of combination therapy on long-term survival in patients with PAH remain unclear. A 2016 meta-analysis demonstrated that combination therapy (initial and sequential) was associated with a significant risk reduction for clinical worsening (relative risk [RR] 0.65; 95% CI, 0.58–0.72; P<0.0001) [432]; however, all-cause mortality was not improved (RR 0.86; 95% CI, 0.72–1.03; P=0.09) and a substantial proportion of patients had clinical worsening events or died despite receiving combination therapy. In addition, registry data showed that the use of combination therapy increased since 2015 but there was no clear improvement in overall survival rates [428, 433, 434]. These data were corroborated by a study showing that less than half of patients receiving initial combination therapy with an ERA and a PDE5i achieved and maintained a low-risk profile [422].

Switching from PDE5is to riociguat has also been investigated as a treatment-escalation strategy [429, 435]. REPLACE was a randomized, controlled, open-label study that enrolled patients on a PDE5i-based therapy who were in WHO-FC III and had a 6MWD of 165–440 m [429]. The study predominantly included patients with IPAH/HPAH/DPAH or PAH-CTD who were randomized to continue their PDE5i or to switch from a PDE5i to riociguat up to 2.5 mg t.i.d. The study met its primary endpoint, termed ‘clinical improvement’, which was a composite of pre-specified improvements in 6MWD, WHO-FC, and NT-proBNP at week 24. Clinical improvement at week 24 was demonstrated in 41% of the patients who switched to riociguat and in 20% of the patients who maintained their PDE5i (odds ratio [OR] 2.78; 95% CI, 1.53–5.06; P=0.0007). In addition, fewer patients in the riociguat group experienced a clinical worsening event (OR 0.10; 95% CI, 0.01–0.73; P=0.0047).

Based on the evidence summarized above, the following recommendations for treatment decisions during follow-up are:

• i) In patients who achieve a low-risk status with their initial PAH therapy, continuation of treatment is recommended.

• ii) In patients who are at intermediate–low risk despite receiving ERA/PDE5i therapy, adding selexipag should be considered to reduce the risk of clinical worsening. In these patients, switching from PDE5i to riociguat may also be considered.

• iii) In patients who are at intermediate–high or high risk while receiving oral therapies, the addition of i.v. epoprostenol or i.v./s.c. treprostinil and referral for LTx evaluation should be considered [309, 436]. If adding i.v./s.c. prostacyclin analogues is unfeasible, adding selexipag or switching from PDE5i to riociguat may be considered.

Recommendations	Classa	Levelb
General recommendation for sequential combination therapy
It is recommended to base treatment escalations on risk assessment and general treatment strategies (see Figure 9)	I	C
Evidence from studies with a composite morbidity/mortality endpoint as primary outcome measure
Addition of macitentan to PDE5is or oral/inhaled prostacyclin analogues is recommended to reduce the risk of morbidity/mortality events [167, 168, 437]	I	B
Addition of selexipag to ERAsc and/or PDE5is is recommended to reduce the risk of morbidity/mortality events [418, 419]	I	B
Addition of oral treprostinil to ERA or PDE5i/riociguat monotherapy is recommended to reduce the risk of morbidity/mortality events [412, 413, 415]	I	B
Addition of bosentan to sildenafil is not recommended to reduce the risk of morbidity/mortality events [419]	III	B
Evidence from studies with change in 6MWD as primary outcome measure
Addition of sildenafil to epoprostenol is recommended to improve exercise capacity [392, 438]	I	B
Addition of inhaled treprostinil to sildenafil or bosentan monotherapy should be considered to improve exercise capacity [411, 439]	IIa	B
Addition of riociguat to bosentan should be considered to improve exercise capacity [395, 440]	IIa	B
Addition of tadalafil to bosentan may be considered to improve exercise capacity [393]	IIb	C
Addition of inhaled iloprost to bosentan may be considered to improve exercise capacity [441, 442]	IIb	B
Addition of ambrisentan to sildenafil may be considered to improve exercise capacity [443]	IIb	C
Addition of bosentan to sildenafil may be considered to improve exercise capacity [419, 444]	IIb	C
Addition of sildenafil to bosentan may be considered to improve exercise capacity [444–446]	IIb	C
Other sequential double- or triple-combination therapies may be considered to improve exercise capacity and/or alleviate PH symptoms	IIb	C
Evidence from studies with safety of combination therapy as primary outcome measure
Combining riociguat and PDE5is is not recommendedd [389]	III	B

6MWD, 6-minute walking distance; ERA, endothelin receptor antagonist; PDE5i, phosphodiesterase 5 inhibitor; PH, pulmonary hypertension. ^aClass of recommendation. ^bLevel of evidence. ^cERAs used in the GRIPHON study were bosentan and ambrisentan. ^dThe PATENT plus study investigated the combination of sildenafil and riociguat; however, combining riociguat with any PDE5i is contraindicated.

RECOMMENDATION TABLE 10

Recommendations for sequential drug combination therapy for patients with idiopathic, heritable, or drug-associated pulmonary arterial hypertension

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9

Evidence-based pulmonary arterial hypertension treatment algorithm for patients with idiopathic, heritable, drug-associated, and connective tissue disease-associated pulmonary arterial hypertension.

DLCO, lung diffusion capacity for carbon monoxide; ERA, endothelin receptor antagonist; I/H/D-PAH, idiopathic, heritable, or drug-associated pulmonary arterial hypertension; i.v., intravenous; PAH-CTD, PAH associated with connective tissue disease; PCA, prostacyclin analogue; PDE5i, phosphodiesterase 5 inhibitor; PH, pulmonary hypertension; PRA, prostacyclin receptor agonist; ReCo, recommendation; s.c., subcutaneous; sGCs, soluble guanylate cyclase stimulator. ^aCardiopulmonary comorbidities are conditions associated with an increased risk of left ventricular diastolic dysfunction, and include obesity, hypertension, diabetes mellitus, and coronary heart disease; pulmonary comorbidities may include signs of mild parenchymal lung disease and are often associated with a low DLCO (<45% of the predicted value). ^bIntravenous epoprostenol or i.v./s.c. treprostinil.

6.3.5. Drug interactions

Among PAH drugs, clinically relevant pharmacokinetic interactions are observed between bosentan and sildenafil (reduced sildenafil plasma concentration [385]), bosentan and hormonal contraceptives (reduced contraception efficacy [361]), and bosentan and vitamin K antagonists (VKAs) (potential need for VKA dose adjustment [386]). Additional pharmacokinetic interactions of potential clinical relevance are listed in Supplementary Data, Table S4.

6.3.6. Interventional therapy

6.3.7. Advanced right ventricular failure

6.3.8. Lung and heart–lung transplantation

Lung transplantation remains an important treatment option for patients with PAH refractory to optimized medical therapy. In patients with PAH, referral to an LTx centre should be considered early (Table 20): 1) when they present with an inadequate response to treatment despite optimized combination therapy; 2) when they present with an intermediate–high or high risk of death (i.e. 1-year mortality >10% when estimated with established risk-stratification tools) [471] (see Section 6.2.7), which exceeds the current mortality rate after LTx [472]; 3) when patients have a disease variant that poorly responds to medical therapy, such as PVOD or PCH.

Both heart–lung and bilateral LTx have been performed for PAH. Currently, most patients receive bilateral LTx, while combined heart–lung transplantation is reserved for patients who have additional non-correctable cardiac conditions [473]. With the introduction of the lung allocation score (LAS), waiting list mortality has decreased and the odds of receiving a donor organ have increased [474]. In some countries, an ‘exceptional LAS’ can be obtained for patients with severe PH. Some other countries not using the LAS have successfully implemented high-priority programmes for these patients [475]. The patient and their next of kin should be fully engaged in the transplant assessment process and informed of the risks and benefits, and the final decision should be jointly made between the patient and medical team (see Section 6.3.1.8). For patients with PAH who survive the early post-transplant period, long-term outcomes are good. A study found that for primary transplant patients with IPAH who survived to 1 year, conditional median survival was 10.0 years [476].

6.3.9. Evidence-based treatment algorithm

A treatment algorithm for patients with IPAH/HPAH/DPAH or PAH-CTD is shown in Figure 9. The evidence supporting this algorithm has mainly been generated in patients with IPAH/HPAH/DPAH or PAH-CTD who present without cardiopulmonary comorbidities. Patients with HIV-associated PAH, PoPH, and PAH associated with congenital heart disease were not enrolled or under-represented in most PAH therapy trials. Treatment recommendations for these patients are provided in Section 7.

6.3.10. Diagnosis and treatment of pulmonary arterial hypertension complications

6.3.11. End-of-life care and ethical issues

The clinical course of PAH may be characterized by progressive deterioration and occasional episodes of acute decompensation. Life expectancy is difficult to predict, as patients may either die slowly because of progressive right HF or experience sudden death.

Patient-orientated care is essential in managing PAH. Information about disease severity and possible prognosis should be provided at initial diagnosis but empathic and hopeful communication, as well as yielding hope, is essential, in line with Section 6.3.1.8. At the right time, open and sensitive communication will enable advanced planning and discussion of a patient's fears, concerns, and wishes, and will ultimately contribute to making the final, well-informed, and joint decision about treatment with the medical team.

Patients approaching end of life require frequent assessment of their full needs by a multidisciplinary team. In advanced stages, recognizing that cardiopulmonary resuscitation in severe PAH has a poor outcome may enable a do not resuscitate order; this may facilitate patients being in their preferred place of care at end of life. Attention should be given to controlling distressing symptoms and prescribing appropriate drugs while withdrawing medication that is no longer needed, which may include PAH medication. Well-informed psychological, social, and spiritual support is also vital. Specialist palliative care should be consulted for patients whose needs are beyond the expertise of the PH team [346].

6.3.12. New drugs in advanced clinical development (phase 3 studies)

Pulmonary arterial hypertension remains an incurable condition with a high mortality rate, despite use of PAH drugs mainly targeting imbalance of vasoactive factors. Novel agents, which are currently in phase 3 development, are ralinepag and sotatercept. Ralinepag is an orally available prostacyclin receptor agonist, which, in a phase 2 RCT that included 61 patients with PAH, improved PVR compared with placebo after 22 weeks of therapy [495]. Sotatercept—a fusion protein comprising the extracellular domain of the human activin receptor type IIA linked to the Fc domain of human immunoglobulin G1—acts as a ligand trap for members of the transforming growth factor (TGF)-β superfamily, thus restoring balance between growth-promoting and growth-inhibiting pathways [496]. In a phase 2 RCT that included 106 patients with PAH treated over 24 weeks, s.c. sotatercept reduced PVR in patients receiving background PAH therapy [496]; improvements were also observed in 6MWD and NT-proBNP [496].

7.2.1. Epidemiology and diagnosis

There is a strong female predominance in PAH-CTD (female/male ratio 4:1), and mean age at diagnosis is commonly >50 years, especially in SSc [173, 502–511, 513, 515, 516]. In the setting of a CTD, patients may present with concomitant disorders such as ILD, and have shorter survival compared with patients with IPAH [503]. The unadjusted risk of death for PAH-SSc compared with IPAH is 2.9, and the predictors of outcome are broadly similar to those for IPAH [516, 517]. Symptoms and clinical presentation are also similar to IPAH, and some patients thought to have IPAH can be identified as having an associated CTD by careful clinical examination and immunological screening tests. Chest CT is recommended for evaluating the presence of associated ILD or PVOD/PCH [504, 513, 515]. An isolated reduction of DLCO is common in PAH-CTD [173, 502–504].

Resting echocardiography combined with other tests is recommended as a screening test in asymptomatic patients with SSc, followed by annual assessments. Screening/early detection is discussed in Section 5.3.1. In other CTDs, PH screening in the absence of suggestive symptoms is not recommended, while echocardiography should be performed in the presence of symptoms. As in other forms of PAH, RHC is recommended in all cases of suspected PAH-CTD to confirm diagnosis, determine severity, and rule out LHD [504].

7.2.2. Therapy

Drugs for PAH should be prescribed in PAH-CTD according to the same treatment algorithm as in IPAH (Figure 9). Patients with PAH-CTD have been included in most of the major RCTs for regulatory approval of PAH therapy [518]. Some aspects of PAH-CTD treatment differ according to the associated CTD [506]. Immunosuppressive therapy combining glucocorticosteroids and cyclophosphamide may result in clinical improvement in patients with SLE- or mixed CTD-associated PAH [506], while it is not recommended in PAH-SSc [519]. Patients with SSc and other CTDs may have ILD and/or HFpEF, which needs to be considered when initiating PAH therapy [504, 515]. In SSc, the long-term risk/benefit ratio of oral anticoagulation is unfavourable because of an increased risk of bleeding, while VKAs are recommended in PAH-CTD with a thrombophilic predisposition (e.g. antiphospholipid syndrome) [319].

Subgroup analyses of patients with PAH-SSc enrolled in RCTs performed with monotherapy or combination therapy of ERAs, PDE5is, sGC stimulators, prostacyclin receptor agonists, epoprostenol, and prostacyclin analogues have shown positive effects versus placebo [301, 401, 519, 520]. In some of these trials, the magnitude of the response in the PAH-CTD subgroup was lower than in the IPAH subgroup [519, 520]. Continuous i.v. epoprostenol therapy improved exercise capacity, symptoms, and haemodynamics in a 3-month RCT in PAH-SSc [401]. However, a retrospective analysis showed a better effect of i.v. epoprostenol on survival in IPAH compared with PAH-SSc [521]. The choice of PAH therapy in the context of SSc and its systemic manifestations may consider other vascular damage such as digital ulcers [522].

Connective tissue disease should not be considered as an a priori contraindication for LTx [523]. This has been extensively studied in SSc, where a multidisciplinary approach optimizing SSc management before, during, and after surgery is recommended [523]. Indications and contraindications for transplantation have to be adapted to the specificities of CTD, with a special focus on digestive (gastro-oesophageal reflux disease and intestinal disease), cardiac, renal, and cutaneous involvement [523].

7.3. Pulmonary arterial hypertension associated with human immunodeficiency virus infection

The use of highly active antiretroviral therapy (HAART), and advances in managing opportunistic infections have contributed to increased life expectancy in patients with HIV [525, 526]. Consequently, the spectrum of complications has shifted towards other long-term conditions, including PAH. Clinical and histopathological findings in PAH associated with HIV infection (PAH-HIV) share many similarities with IPAH [1, 527]. With the availability of HAART given in combination with PAH therapies, the prognosis of PAH-HIV has markedly improved in recent years [526, 528]. In addition, the incidence of PAH-HIV has declined in parallel with the increasing availability of HAART [528]. Taken together, these effects on survival and incidence have resulted in a stable PAH prevalence in patients with HIV over recent decades. A French population study indicated that the prevalence of PAH in individuals with HIV infection was 0.46%, which is very similar to the prevalence before the HAART era [177].

The pathogenesis of PAH-HIV remains unclear. There is no evidence of a direct role of HIV in the pathogenesis of PAH and, although present in inflammatory cells in the lungs, the virus itself has never been found in pulmonary vascular lesions of patients with PAH-HIV [529]. This suggests that an indirect action of viral infection on inflammation and growth factors may act as a trigger in a predisposed patient.

7.3.1. Diagnosis

Pulmonary arterial hypertension associated with HIV shares a clinical presentation with IPAH. Before the availability of HAART most patients were in WHO-FC III or IV at diagnosis. Nowadays, patients are diagnosed with much less severe symptoms and haemodynamics. Patients may present with other risk factors for PAH such as liver disease (chronic viral hepatitis B or C) or exposure to drugs or toxins. Patients with PAH-HIV are more likely to be male and i.v. drug abusers [403, 526]. There is no correlation between the severity of PAH and the stage of HIV infection or the degree of immunodeficiency [403, 530]. Because of its low prevalence, asymptomatic patients with HIV should not be screened for PAH. However, echocardiography should be performed in patients with unexplained dyspnoea to detect HIV-related cardiovascular complications such as myocarditis, cardiomyopathy, or PAH. Right heart catheterization is mandatory to confirm the diagnosis of PAH-HIV and to rule out LHD [527].

Pulmonary arterial hypertension is an independent risk factor for death in patients with HIV. In the 1990s, before the availability of HAART, patients with PAH-HIV had poor outcomes, with a 3 year survival of <50% [403]. The overall survival has now improved and patients with PAH-HIV have a better prognosis than most patients with other forms of PAH [526].

7.3.2. Therapy

Current recommendations for the treatment of PAH-HIV are largely based on data from IPAH [25, 26].

Treatment of PAH-HIV with HAART has improved functional status and survival in some retrospective studies [525, 526, 531]. The use of HAART in PAH-HIV is therefore recommended, irrespective of viral load and CD4+ cell count.

Anticoagulation is not recommended because of an increased risk of bleeding and drug interactions [319, 527]. Patients with PAH-HIV are usually non-responders to acute vasoreactivity testing and therefore should not receive CCBs [378].

The prospective, open-label, BREATHE-4 study showed that bosentan markedly improved WHO-FC, exercise capacity, quality of life, and haemodynamics after 16 weeks in patients with PAH-HIV [532]. In a long-term, retrospective series, bosentan therapy was associated with haemodynamic normalization in 10/59 patients [533]. Bosentan potentially interacts with antiretroviral drugs, and close monitoring is required when combined with HAART. Very few patients with PAH-HIV have been included in RCTs with ambrisentan and macitentan, and no definite conclusion can be drawn from those studies.

Positive effects of sildenafil and tadalafil in PAH-HIV have been established in case studies [534, 535]. Interactions have been reported between PDE5is and protease inhibitors, resulting in major increases in PDE5i concentrations; these drugs should be introduced at low dosages with careful monitoring of potential side effects, including hypotension [536, 537]. There are no data on the use of the sGC stimulator riociguat in PAH-HIV.

Treatment with i.v. epoprostenol resulted in significant improvement in WHO-FC, exercise capacity, haemodynamics, and survival in selected patients with PAH-HIV [403, 538]. There are very few data on the use of i.v. or s.c. treprostinil or inhaled iloprost in PAH-HIV [539, 540].

There are no clinical trial data on the use of combination therapy for PAH-HIV. Given the lack of supporting evidence and potential safety concerns when PAH drugs are co-administered with antiretroviral drugs, initial monotherapy with PAH medication is recommended, followed by an individualized use of combination therapy in patients who do not reach a low-risk profile.

7.4.1. Diagnosis

The diagnosis of PoPH is based on the presence of otherwise unexplained pre-capillary PH in patients with portal hypertension or a portosystemic shunt. The diagnostic approach is the same as in other patients with suspected or newly detected PH. Transthoracic echocardiography is usually the first non-invasive assessment in patients with suspected PH, and echocardiography is also recommended as a screening tool in patients evaluated for liver transplantation. As patients with liver disease often have an elevated CO, TRV tends to overestimate PAP in these patients. Hence, RHC with comprehensive haemodynamic assessment is essential to confirm the diagnosis of PH and to distinguish PAH (with elevated PVR) from unclassified PH (with a normal PVR).

7.4.2. Therapy

Patients with unclassified PH (i.e. mPAP >20 mmHg, elevated CO, and PVR ≤2.0 WU) should be regularly followed-up but should not be treated with drugs approved for PAH.

In patients with an established diagnosis of PoPH, treatment should follow the same general principles as in other patients with PAH, taking into account the severity of underlying liver disease, the indication for liver transplantation, and the potential effects of PAH medication on gas exchange, which may deteriorate with vasodilators in patients with PoPH [548, 549]. All drugs approved for PAH can principally be used to treat patients with PoPH, bearing in mind that these patients are usually excluded from registration studies. Nevertheless, various case series support the use of approved PAH medication in patients with PoPH. The largest series published so far reported on 574 patients with PoPH treated with various PAH drugs, mostly PDE5is or ERAs, alone and in combination [545]. Most patients (56.8%) were in Child–Pugh class A at the time of PAH diagnosis. At the first follow-up, which took place 4.5 months after starting treatment, improvements were seen in haemodynamics, WHO-FC, 6MWD, and BNP/NT-proBNP; survival at 5 years was 51%. In patients presenting with mild liver disease, the main causes of death were PAH progression and malignancy, whereas complications of liver disease were the most common causes of death in patients with advanced liver disease. The 5 year survival of patients who underwent liver transplantation (n=63) was 81%.

The only RCT dedicated to the treatment of PoPH was PORTICO, a 12 week study that randomized 85 patients to macitentan (n=43) or placebo (n=42) [168]. PORTICO met its primary endpoint, demonstrating a significant reduction in PVR from baseline (ratio of geometric mean 0.65; 95% CI, 0.59–0.72; P<0.0001). There were, however, no differences between the two treatment groups in secondary outcome measures, including WHO-FC, 6MWD, and NT-proBNP.

7.5.1. Diagnosis and risk assessment

The diagnostic work-up of PAH associated with adult CHD should be based on the presence of symptoms and includes medical history, physical examination, PFTs, ABG, imaging (especially echocardiography), and exercise and laboratory testing. Of note, standard echocardiographic criteria for detecting PH may not be applicable in complex adult CHD [559]. Right heart catheterization with compartmental oximetry for calculating pulmonary blood flow/systemic blood flow (Qp/Qs) is required to confirm PAH diagnosis and guide therapeutic interventions. Thermodilution should be avoided in the presence of intracardiac shunts, and direct Fick is the most accurate method. Pulmonary vascular resistance may be overestimated due to erythrocytosis [560]. Interpreting invasive haemodynamics (see Section 5.1.12) should be made in the context of multiparametric assessment of exercise capacity, laboratory testing, and imaging.

Predictors of worse outcomes in adult CHD-associated PAH are WHO-FC III–IV, exercise intolerance assessed by 6MWD or peak VO₂, history of hospitalization for right HF, biomarkers (NT-proBNP >500 pg/mL, C-reactive protein >10 mg/mL, high serum creatinine, and low albumin levels), iron deficiency, and echocardiographic indices of RV dysfunction [559, 561]. When compared with patients with IPAH, patients with Eisenmenger syndrome may have a relatively stable long-term clinical course. The right ventricle is unloaded by the right-to-left shunt, sustaining CO at the expense of hypoxaemia and cyanosis. However, due to immortal time bias, prognosis of Eisenmenger syndrome is not as favourable as previously thought [562].

As in other forms of PAH, risk assessment is important to guide therapy, and specific risk factors have been described in Eisenmenger syndrome. A large multicentre study showed that mortality in adults with Eisenmenger syndrome was predicted by the presence of pre-tricuspid shunt, advancing age, low rest oxygen saturation, absence of sinus rhythm, and presence of pericardial effusion [563].

7.5.2. Therapy

Outcomes in adult CHD-associated PAH have improved with the availability of new PAH therapies, advances in surgical and peri-operative management, and a team-based, multidisciplinary approach in PH centres. These patients should be managed by specialized health professionals. Patient education, behavioural modifications, and social and psychological support are all important aspects of management.

Shunt closure (surgical or interventional) may only be considered in patients with prevalent systemic-to-pulmonary shunting without significantly increased PVR. Criteria for defect closure based on Qp/Qs ratio and (baseline and/or after targeted PAH treatment) PVR have been proposed by the 2020 ESC Guidelines for the management of adult congenital heart disease [101]. Decisions on shunt closure should not be made on haemodynamic numbers alone, and a multiparametric strategy should be followed. For instance, shunt closure is not indicated in the case of desaturation during exercise in the 6MWT or CPET, or when there is secondary erythrocytosis suggesting dynamic reversal of shunt. There is no evidence for a long-term benefit of a treat-and-repair approach in patients with adult CHD-associated PAH with prevalent systemic-to-pulmonary shunts; therefore, there is a need for future prospective studies [564]. Defect closure is contraindicated in all patients with Eisenmenger syndrome, and may also adversely affect patients with small/coincidental defects that behave similarly to IPAH [565]. There are no prospective data available on the usefulness of vasoreactivity testing, balloon closure testing, or lung biopsy for assessing operability and normalization of PVR after closure [566].

Patients with adult CHD-associated PAH may present with clinical deterioration in different circumstances, such as arrhythmia, during non-cardiac surgery requiring general anaesthesia, dehydration or bleeding, thrombo-embolism, and lung infections. Surgeries should be limited to those deemed essential, and performed in specialized centres with anaesthetists experienced in adult CHD and PAH. Endocarditis should be suspected in patients with sepsis, whereas a cerebral abscess should be excluded in those with neurological symptoms or new headache, especially in those with low oxygen saturations and complex anatomies. It is recommended to avoid strenuous exercise, but mild and moderate activities seem to be beneficial [567]. Patients should receive all recommended vaccinations and endocarditis prophylaxis in the presence of cyanosis. Although pregnant patients with left-to-right shunts and stable, well-controlled PAH have tolerated pregnancy well under specialized care, pregnancy is still associated with both high maternal mortality and foetal complications in Eisenmenger syndrome and should be discouraged in this setting [568, 569]; hence, effective contraception is highly recommended. Levonorgestrel-based, long-acting, reversible contraception implants or intrauterine devices have been recommended for these patients [570].

Secondary erythrocytosis is beneficial for adequate oxygen transport and delivery, and routine phlebotomy should be avoided whenever possible. Symptoms of hyperviscosity in the presence of haematocrit >65% should be approached with appropriate hydration. Iron deficiency should be corrected. When i.v. iron supplementation is administered, special care should be taken to avoid air emboli during administration [571]. Supplemental oxygen therapy has not been shown to impact survival.

Oral anticoagulant treatment with VKAs may be considered in patients with large PA aneurysms with thrombus, atrial arrhythmias, and previous thrombo-embolic events, but with low bleeding risk. In patients with very high Hb levels (>20 mg/dL), standard international normalized ratio measures are less accurate, and citrate-adjusted blood bottles must be used. Regarding using novel oral anticoagulants (NOACs), a large, nationwide, German, adult CHD database (including 106 NOAC-treated patients with Eisenmenger syndrome) showed that NOAC users had higher long-term risk of bleeding, major adverse cardiovascular events, and mortality compared with those on VKAs, suggesting that initiating NOACs should be reserved for experienced adult CHD centres, carefully weighing potential benefits and risks [572, 573].

Compared with other group 1 subgroups, limited data exist on the use of drugs approved for PAH in patients with adult CHD-associated PAH. Bosentan improved 6MWD and decreased PVR in patients with Eisenmenger syndrome in WHO-FC III [574]. Patients with more complex lesions were less likely to respond to PAH therapies compared with patients with simple lesions. An RCT investigating the efficacy of macitentan found no effect on 6MWD in a mixed cohort of patients with Eisenmenger syndrome (6MWD improved in both treatment and placebo arms), although decreases in NT-proBNP and PVR were noted in the macitentan arm [575].

Experiences with other ERAs and PDE5is have shown favourable functional and haemodynamic results in Eisenmenger syndrome [576]. In a small, single-centre, pilot study, adding nebulized iloprost to a background of oral PAH therapy failed to improve 6MWD in Eisenmenger syndrome [577]. In case symptoms persist or in clinical deterioration, a sequential and symptom-orientated treatment strategy is recommended in Eisenmenger syndrome, starting with an oral ERA (or PDE5i) and escalating therapy. Should symptoms not adequately improve with oral therapies, i.v./s.c. options should be proactively considered [578]. There is a theoretical risk of paradoxical embolism in right-to-left shunt lesions with the presence of a central venous catheter for i.v. therapy; therefore, s.c. prostacyclin analogue infusion may be considered.

The effect of PAH therapies in patients with prevalent systemic-to-pulmonary shunts is less well established. Patients with small/coincidental defects should be treated with PAH medication [557]. This is also the case for patients with PAH after defect correction who have increased mortality compared with those with Eisenmenger syndrome [579]. These patients were included in major RCTs with PAH therapies and should be evaluated based on comprehensive risk assessment (Table 16) [580]. The effect of PAH therapies in patients with segmental PH remains a matter of debate [101, 581]. While some series have reported promising results, there have been cases where therapies were not tolerated [581]. Similarly, using PAH therapies in Fontan circulation has yielded conflicting results, and results of further studies are awaited [582–584].

Heart–lung transplantation or LTx with heart surgery is an option in highly selected cases not responsive to medical treatment; however, it is limited by organ availability and lesion complexity. Mortality is high during the first year after surgery, especially after heart–lung transplantation, but remains relatively low thereafter [585].

7.7.1. Diagnosis

Most patients complain of non-specific dyspnoea on exertion and fatigue [590]. Physical examination may reveal digital clubbing and bibasal crackles on lung auscultation [590]. Pulmonary arterial hypertension and PVOD/PCH share the same haemodynamic profile as pre-capillary PH [590, 591]. The PAWP is not elevated because the pulmonary vascular changes occur in small venulae and capillaries, while the LA filling pressure remains normal [590, 591]. A diagnosis of PVOD/PCH is based on the results of tests suggesting venous post-capillary involvement, chronic interstitial pulmonary oedema, and capillary proliferation [1, 590, 591]. These tests include PFTs (decreased DLCO, frequently <50% theoretical values), ABG (hypoxaemia), and non-contrast chest CT (subpleural thickened septal lines, centrilobular ground-glass opacities, and mediastinal lymphadenopathy) [1, 425, 589, 591, 592]. Importantly, these patients are at risk of drug-induced pulmonary oedema with PAH therapy, a finding suggestive of PVOD/PCH [425, 591]. Detecting biallelic EIF2AK4 mutations is sufficient to confirm a diagnosis of heritable PVOD/PCH [158, 591, 592]. Lung biopsy is hazardous in PH and is not recommended for diagnosing PVOD/PCH [1, 425].

7.7.2. Therapy

There is no established medical therapy for PVOD/PCH [425]. Compared with IPAH, PVOD/PCH has a poor prognosis and limited response to PAH therapy, with a risk of pulmonary oedema due to pulmonary venous obstruction [425, 591]. However, there are reports of incomplete and transient clinical improvement in individual patients with PVOD/PCH treated with PAH therapy, which should be used with great caution in this setting [425, 591]. Diuretics, oxygen therapy, and slow titration of PAH therapy can be used on an individual basis [425]. Therefore, therapy for PVOD/PCH should be undertaken at centres with extensive experience in managing PH, and patients should be fully informed about the risks [425]. Anecdotal reports suggest a potential benefit of immunomodulatory treatments, but this approach requires further study [593]. The only curative therapy for PVOD/PCH is LTx, and eligible patients should be referred to a transplant centre for evaluation upon diagnosis [425, 591]. Pathological examination of the explanted lungs will confirm the diagnosis [590].

7.8.1. Epidemiology and classification

The reported annual incident rate for paediatric PH is 64/million children [594]. The distribution of the various aetiologies of PH in childhood differs from PH in adulthood [594–596]. Pulmonary arterial hypertension is the most frequent type of PH in children, with the vast majority (82%) of cases being infants with transient PAH (i.e. PPHN or repairable cardiac shunt defects). Of the remaining children with PAH, most have either IPAH, HPAH, or irreversible CHD-associated PAH. The reported incidences of IPAH/HPAH and (non-transient) CHD-associated PAH are 0.7 and 2.2/million children, respectively, with a prevalence of 4.4 and 15.6/million children, respectively [594]. Other conditions associated with PAH (Table 6) do occur in children but are rare.

Another significant proportion (34–49%) of children with non-transient PH are neonates and infants with PH associated with respiratory disease, especially developmental lung diseases, including bronchopulmonary dysplasia (BPD), congenital diaphragmatic hernia (CDH), and congenital pulmonary vascular abnormalities [594–598]. These children form a prominent and distinctive group in paediatric PH and are currently classified as PH group 3 associated with developmental lung disease (Table 6; Table S7). A significant and growing proportion of children with PH associated with respiratory disease is made up of pre-term infants with BPD. Also, newly recognized genetic developmental lung disorders—including alveolar capillary dysplasia, TBX4-mutation-related lung disorders, and surfactant abnormalities—are currently classified in this category (Figure 10) [599].

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 10

Neonatal and paediatric versus adult pulmonary hypertension.

ASD, atrial septal defect; AVT, acute vasoreactivity testing; CHD, congenital heart disease; COPD, chronic obstructive pulmonary disease; CTEPH, chronic thrombo-embolic pulmonary hypertension; GORD, gastro-oesophageal reflux disease; HPAH, heritable pulmonary arterial hypertension; ILD, interstitial lung disease; IPAH, idiopathic pulmonary arterial hypertension; MS, mitral stenosis; NO, nitric oxide; PAH, pulmonary arterial hypertension; PAH-CHD, PAH associated with congenital heart disease; PAH-CTD, PAH associated with connective tissue disease; PAH-HIV, PAH associated with HIV infection; PH, pulmonary hypertension; PoPH, porto-pulmonary hypertension; PPHN, persistent pulmonary hypertension of the newborn; PS, pulmonary arterial stenosis; PVS, pulmonary vein stenosis; RCT, randomized controlled trial; RHC, right heart catheterization; VSD, ventricular septal defect. ^aIn patients with idiopathic, heritable or drug-associated PAH. Pulmonary hypertension in neonates and infants significantly differs in aetiology, pathophysiology, risk assessment, and treatment from older children and adults, while PH in older children has more similarities with PH in adults.

Another distinctive feature of PH in children is the high burden of genetic disorders. Childhood PH is often associated with chromosomal, genetic, and syndromic anomalies (11–52%). Like in adults, gene mutations implicated in the pathogenesis of HPAH are found in 20–30% of sporadic cases, where paediatric HPAH seems to be characterized by an enrichment in TBX4 and ACVRL1 variations [600, 601]. Additionally, 17% of children with PAH have other disorders known to be associated with PAH, including trisomy 21. Finally, 23% of children with PAH have copy number variations not previously associated with PH [600, 602, 603].

Given the frequent association of paediatric PAH with chromosomal, genetic, and syndromic anomalies (for which the mechanistic basis for PAH is generally uncertain), genetic testing may be considered for defining aetiology and comorbidities, stratifying risk, and identifying family members at risk; however, this should be after appropriate expert genetic counselling for the child and family (see Section 5.1.13).

The clinical PH classification (Table 6) is also followed for paediatric PH. To improve applicability of this classification in infants and children with PH, it has been adapted to give room to PH associated with various congenital cardiovascular and pulmonary diseases or specific paediatric conditions (Tables S5–S8) [599].

7.8.2. Diagnosis and risk assessment

Historically, the definition of PH in children aged >3 months has been the same as in adults. The definition for PH has now been redefined to mPAP >20 mmHg in adults as well as in children. The impact of an mPAP 21–24 mmHg on outcomes in children is unknown. However, in the interest of consistency and to facilitate transition from paediatric to adult PH care, it is recommended that the updated definition for PH also be followed in children. No treatment recommendations currently exist for this group of children (mPAP 21–24 mmHg).

Regarding the newly introduced criterion to include PVR >2 WU to identify pre-capillary PH in adults, PVR had previously been included in the definition for PAH in children. In children, blood flows are traditionally indexed assuming that systemic and pulmonary blood flows change proportionally with body size, while the transpulmonary pressure gradient does not. Since blood flow is the denominator in the equation for calculating PVR, the need for indexing of PVR in children is emphasized, and the criterion of pulmonary vascular resistance index (PVRI) ≥3 WU·m² in the definition for PAH in children remains unchanged [599].

Since the aetiology of paediatric PH is very diverse, a methodical and comprehensive diagnostic approach is crucial to reach an accurate diagnosis and treatment plan. As in adults, IPAH is a diagnosis ‘per exclusion’. A diagnostic work-up, similar to that in adults but customized for paediatric PH, is recommended [599]. Pre-term infants with BPD should be screened for PH, since PH is prevalent in this population and seriously affects outcome [604].

Also in children, RHC is the gold standard for definitively diagnosing and establishing the nature of PH, and provides important data for stratifying risk [604a, 605]. To identify those suitable for high-dose CCB treatment, acute vasoreactivity testing is recommended in children with IPAH/HPAH. The criteria used in adults for a positive acute response have identified children who will show sustained benefit from CCB therapy; however, these criteria do not define reversibility of PAH or operability in children with CHD. Since RHC in children with PH may be associated with major complications (in 1–3% of cases, especially in young infants and those in worse clinical condition), risks and benefits have to be balanced in the individual child [605]. Heart catheterization in children with PAH should be exclusively performed in experienced paediatric PH centres. Indications for repeated RHC in children with PH are currently not well defined.

Treatment of children with PAH is based on risk stratification [599]. Predictors of worse outcome in paediatric PAH are similar to those in adults, and include clinical evidence of RV failure, progression of symptoms, WHO-FC III–IV, certain echocardiographic parameters (e.g. TAPSE), and elevated serum NT-proBNP. A 6MWD <350 m has also been suggested as a predictor of worse outcome in paediatric PH, but its value in young children is less established. Further prognosticators identified in paediatric PAH are failure to thrive and haemodynamic variables, such as RAP >10 mmHg, the ratio of mean pulmonary-to-systemic blood pressure >0.75, and PVRI >20 WU·m² [602, 606, 607]. Paediatric risk-assessment tools based on these parameters have been retrospectively validated in observational paediatric registries [599, 604a].

7.8.3. Therapy

The ultimate goal of treatment should be to improve survival and facilitate normal childhood activities without limitations. In the absence of RCTs in paediatric PAH, recommended treatment algorithms are extrapolated from those in adults and enhanced with data from observational studies in children with PAH [599].

Observational cohort studies support treatment algorithms designed for adults to be used for children (including the superiority of combination therapy over monotherapy) [608]. Drugs investigated in children, with or without formal approval by the European Medicines Agency (EMA) for treating children with PAH, are shown in Table 22.

A paediatric treatment algorithm, derived from that for adults, is based on risk stratification, recommending general measures, high-dose CCB therapy for responders to acute vasoreactivity testing (where close follow-up is mandatory, as some patients may fail long-term therapy), oral or inhaled combination therapy for children at low risk, and combination therapy with i.v./s.c. prostacyclin analogues for those at high risk [599].

In the case of insufficient response to recommended drug therapy, or when drugs are unavailable, a Potts shunt (a surgical or interventional connection between the left PA and the descending aorta), BAS, or LTx may be considered in children with severe PH (see Sections 6.3.6.1 and 6.3.8) [599]. Reported clinical experience with Potts shunts is limited to just over 100 patients, predominantly children, with a mortality of 12–25% and long-term clinical benefit in a subset of children with long-term follow-up [456–459].

Monitoring of treatment effect and disease course is pivotal in managing all patients with PAH (adults and in children). In children with PAH, clinical risk scores including WHO-FC, TAPSE, and serum NT-proBNP are potential treatment targets for goal-orientated treatment [604a, 609].

Contemporary treatment algorithms for infants with PPHN have been proposed but are outside the scope of these guidelines [610].

The recommendations discussed above apply to children with PAH, whereas the specific group of infants with neonatal PVD, mostly classified as PH associated with developmental lung disease and with heterogeneous aetiology, require a distinct and customized approach (Figure 10).

In pre-term infants with BPD and PH, the underlying lung disease should primarily be treated. Frequently, these infants are additionally treated with therapies for PAH, including sildenafil and bosentan; however, these are not approved by the EMA for use in infants with group 3 PH and developmental lung diseases (BPD, CDH). Their effects on outcomes in this population are unclear, and data enabling robust treatment recommendations are lacking. These children should be treated by multidisciplinary teams involving cardiologists, neonatologists, pulmonologists, and nutritionists. Pulmonary hypertension in these infants may disappear with lung healing, although long-term cardiovascular sequelae have been reported [611, 612].

8.2.1. Diagnosis and control of the underlying left heart disease

Patients with suspected PH-LHD will have an established diagnosis of LHD, such as HFrEF/HFmrEF, HFpEF, valvular heart disease, and/or CHD. The distinction between PH associated with HFpEF and other forms of PH (e.g. PAH, CTEPH) may be challenging, particularly given the increased burden of cardiovascular comorbidities in real-world PAH populations [142, 450, 661]. In this context, validated scores for diagnosing HFpEF (HFA-PEFF, H2FPEF) [16, 662, 663] may be helpful for detecting it as an underlying condition in PH, and the presence or absence of risk factors for PAH or CTEPH should be determined. Patients with signs of predominant RV strain and/or PH should be further evaluated. Patients should be assessed or reassessed when they are fully recompensated and in a clinically stable condition.

8.2.2. Evaluation of pulmonary hypertension and patient phenotyping

Patients with LHD and suspected PH should be evaluated following the diagnostic strategy for PH (see Section 5). This requires identifying clinical features and a multimodal approach using non-invasive diagnostic tests such as echocardiography, ECG, and BNP/NT-proBNP levels. In the presence of mild PH and predominant LHD, no further testing may be necessary. Otherwise, CTEPH and significant lung disease should be ruled out by V/Q scan and PFTs, and additional cardiac imaging including cMRI may be considered in selected cases. For phenotyping, a combination of variables may help to determine the likelihood of LHD, and HFpEF in particular, versus other causes of PH (Table 23). Pulmonary hypertension associated with left heart disease is likely in the presence of known cardiac disease, multiple cardiovascular comorbidities/risk factors, atrial fibrillation at diagnosis, and specific imaging findings (LV hypertrophy, increased LA size, and reduced LA strain). Although exercise echocardiography has been proposed to uncover HFpEF, it is unable to diagnose or classify PH in this context. A combination of clinical findings and phenotyping is required to decide about the need for further invasive assessment.

8.2.3. Invasive assessment of haemodynamics

The decision to perform cardiac catheterization and to invasively assess cardiopulmonary haemodynamics should depend on the presence of an intermediate to high echocardiographic probability of PH, and should be determined by the need to obtain relevant information for prognostication or management. In patients with a high likelihood of LHD as the main cause of PH, or with established underlying LHD and mild PH (Table 23), invasive assessment for PH is usually not indicated. Indications for RHC in LHD include: (1) suspected PAH or CTEPH; (2) suspected CpcPH with a severe pre-capillary component, where further information will aid phenotyping and treatment decisions (Figure S2); and (3) advanced HF and evaluation for heart transplantation. While several haemodynamic measures (mPAP, PVR, pulmonary arterial compliance [PAC], transpulmonary pressure gradient, and DPG) are associated with outcomes in PH-LHD [142, 632, 635], the most robust and consistent data are available for PVR. Invasive assessment should be conducted in experienced centres, when management of the underlying LHD has been optimized and patients are in a clinically stable condition. With respect to respiratory variations of intrathoracic pressures, all pressure readings should be taken at end-expiration.

Additional testing during RHC may be useful for distinguishing between PAH and HFpEF [18, 23, 664–669], and to uncover LHD in patients with a high likelihood of PH-LHD and normal resting PAWP [670–673]; both exercise testing and fluid challenge may be considered in special situations (see Section 5.1.12). Conditions associated with reduced LV diastolic compliance or valvular heart disease may be associated with a rapid increase in PAWP when challenged with increased systemic venous return [674]. While the upper limit of normal remains controversial [142, 143, 665, 667], a PAWP cut-off of >18 mmHg has been suggested to identify HFpEF as the underlying cause of PH, despite normal PAWP at baseline [143]. While this may help to classify PH, therapeutic consequences of such testing remain to be determined.

As differentiating between severe PH associated with HFpEF and IPAH with cardiac comorbidities is challenging, patients with an unclear diagnosis, particularly those with a predominant pre-capillary component (e.g. PVR >5 WU), should be referred to a PH centre for individualized management.

8.3.1. Pulmonary hypertension associated with left-sided heart failure

8.3.2. Pulmonary hypertension associated with valvular heart disease

Pulmonary hypertension frequently occurs as a consequence of valvular heart disease. While surgical or interventional approaches for valvular repair improve cardiopulmonary haemodynamics by reducing PAWP and PAP and improving forward SV [699], persistent PH after correcting valvular heart disease is frequent and associated with adverse outcomes [634, 700].

8.3.3. Recommendations on the use of drugs approved for PAH in PH-LHD

The recommendations on the use of drugs approved for PAH in patients with PH-LHD have been established based on key narrative question 5 (Supplementary Data, Section 8.3).

The recommendations on the use of PDE5is in patients with CpcPH associated with HFpEF are based on PICO question II (Supplementary Data, Section 8.4). Two RCTs that enrolled patients with HFpEF and PH were identified, but no study that specifically enrolled patients with HFpEF and CpcPH. Harmful effects cannot be excluded, even if the available data from clinical studies, case series, and registries suggest that PDE5is may be safely administered to patients with HFpEF-associated CpcPH. As a result, a general recommendation for or against the use of PDE5is in patients with HFpEF and CpcPH cannot be made. However, it is clinically relevant to make a recommendation against their use for patients with HFpEF and IpcPH.

9.2.1. Pulmonary hypertension associated with chronic obstructive pulmonary disease or emphysema

Studies using drugs approved for PAH in patients with PH associated with COPD or emphysema have yielded conflicting results and are mostly limited by small sample size, short duration, and insufficient haemodynamic characterization of PH [739, 741, 742]. In a 16 week RCT of 28 patients with COPD and severe PH confirmed by RHC, sildenafil therapy resulted in statistically significant improvements in PVR and quality of life [743]. Registry data identified that ∼30% of patients with COPD and severe PH, predominantly treated with PDE5is, had improved WHO-FC, 6MWD, and PVR versus baseline, and those with a treatment response had improved transplant-free survival [51, 718]. However, in the absence of large randomized trials, the evidence is insufficient to support the general use of medication approved for PAH in patients with COPD and PH. Patients with COPD and suspected or confirmed severe PH should be referred to PH centres for individual decision-making.

9.2.2. Pulmonary hypertension associated with interstitial lung disease

Numerous phase 2 and phase 3 studies have investigated the use of ERAs to treat ILD, all with negative results [740, 744, 745]. In addition, the PDE5i sildenafil has been investigated in phase 3 trials of patients with ILD, also with negative results [746, 747]. Few data from RCTs are available for patients with PH associated with ILD, and many of the studies performed for this indication [748, 749] suffered from the same limitations as the aforementioned studies in PH associated with COPD. In addition, there were several adverse safety signals: ambrisentan was associated with an increased risk of clinical worsening in patients with ILD with and without PH [740, 750], while riociguat was associated with an increased risk of clinical worsening events, including potential excess mortality, in patients with PH associated with idiopathic interstitial pneumonia [181].

In contrast, promising results have been obtained with the use of inhaled treprostinil. A phase 3 RCT (INCREASE) examined inhaled treprostinil at a target dose of 72 µg given four times daily in 326 patients with PH associated with ILD [734, 751]. The PH diagnosis was confirmed by RHC within 1 year prior to enrolment. At week 16, the placebo-corrected 6MWD improved by 31 m with inhaled treprostinil. There were also improvements in NT-proBNP and clinical worsening events, the latter driven by a lower proportion of patients whose 6MWD declined by >15% from baseline.

Given the significant impact of even non-severe PH in patients with lung disease, eligible patients should be referred for LTx evaluation. In patients with ILD and PH, inhaled treprostinil may be considered based on the findings from the INCREASE study, but further data are needed, especially on long-term outcomes. The routine use of other medication approved for PAH is not recommended in patients with ILD and non-severe PH. For patients with severe PH and/or severe RV dysfunction, or where there is uncertainty regarding the treatment of PH, referral to a PH centre is recommended for careful evaluation, to facilitate entry into RCTs, and consider PAH therapies on an individual basis (Figure S3). Registry data show that some patients with group 3 PH are being treated with PAH medication, predominantly PDE5is [718, 752, 753], but it is unclear if and to what extent these patients benefit from this treatment.

9.2.3. Recommendations on the use of drugs approved for PAH in PH associated with lung disease

The recommendations on the use of drugs approved for PAH in patients with PH associated with COPD and ILD have been established based on key narrative questions 6 and 7 (Supplementary Data, Sections 9.1 and 9.2, respectively).

The recommendations on the use of PDE5is in patients with severe PH associated with ILD are based on PICO question III (Supplementary Data, Section 9.3). There are no direct data from RCTs on the safety, tolerability, and efficacy of PDE5is in patients with PH associated with ILD. The indirect data included in the guidelines do not enable firm conclusions to be drawn. Given the lack of robust evidence, the Task Force members felt unable to provide a recommendation for or against the use of PDE5is in patients with ILD and severe PH, and recommend that these patients are referred to a PH centre for individualized decision-making.