Guidelines for the diagnosis and treatment of pulmonary hypertension

7.1.1 Clinical presentation

The symptoms of PAH are nonspecific and include breathlessness, fatigue, weakness, angina, syncope and abdominal distension 44. Symptoms at rest are reported only in very advanced cases. The physical signs of PAH include left parasternal lift, an accentuated pulmonary component of second heart sound, a pansystolic murmur of tricuspid regurgitation, a diastolic murmur of pulmonary insufficiency and an RV third sound. Jugular vein distension, hepatomegaly, peripheral oedema, ascites and cool extremities characterise patients in a more advanced state. Lung sounds are usually normal. The examination may also provide clues as to the cause of PH. Telangiectasia, digital ulceration and sclerodactyly are seen in scleroderma, while inspiratory crackles may point towards interstitial lung disease 45. The stigmata of liver disease such as spider naevi, testicular atrophy and palmar erythema should be considered. If digital clubbing is encountered in “IPAH”, an alternative diagnosis such as CHD or PVOD should be sought.

7.1.2 Electrocardiogram

The ECG may provide suggestive or supportive evidence of PH by demonstrating RV hypertrophy and strain, and right atrial dilatation. RV hypertrophy on ECG is present in 87% and right axis deviation in 79% of patients with IPAH 44. The absence of these findings does not exclude the presence of PH nor does it exclude severe haemodynamic abnormalities. The ECG has insufficient sensitivity (55%) and specificity (70%) to be a screening tool for detecting significant PH. Ventricular arrhythmias are rare. Supraventricular arrhythmias may be present in advanced stages, in particular atrial flutter, but also atrial fibrillation, which almost invariably leads to further clinical deterioration 46.

7.1.3 Chest radiograph

In 90% of patients with IPAH the chest radiograph is abnormal at the time of diagnosis 44. Findings include central pulmonary arterial dilatation, which contrasts with “pruning” (loss) of the peripheral blood vessels. Right atrium and RV enlargement may be seen in more advanced cases. The chest radiograph allows associated moderate-to-severe lung diseases (group 3, table 4⇓) or pulmonary venous hypertension due to left heart disease (group 2, table 4⇓) to be reasonably excluded. Overall, the degree of PH in any given patient does not correlate with the extent of radiographic abnormalities.

7.1.4 Pulmonary function tests and arterial blood gases

Pulmonary function tests and arterial blood gases will identify the contribution of underlying airway or parenchymal lung disease. Patients with PAH usually have decreased lung diffusion capacity for carbon monoxide (typically in the range of 40–80% predicted) and mild to moderate reduction of lung volumes. Peripheral airway obstruction can also be detected. Arterial oxygen tension is normal or only slightly lower than normal at rest and arterial carbon dioxide tension is decreased because of alveolar hyperventilation. COPD as a cause of hypoxic PH is diagnosed on the evidence of irreversible airflow obstruction together with increased residual volumes and reduced diffusion capacity for carbon monoxide and normal or increased carbon dioxide tension. A decrease in lung volume together with a decrease in diffusion capacity for carbon monoxide may indicate a diagnosis of interstitial lung disease. The severity of emphysema and of interstitial lung disease can be diagnosed using high-resolution computed tomography (CT). If clinically suspected, screening overnight oximetry or polysomnography will exclude significant obstructive sleep apnoea/hypopnoea.

7.1.5 Echocardiography

Transthoracic echocardiography provides several variables which correlate with right heart haemodynamics including PAP, and should always be performed in the case of suspected PH.

The estimation of PAP is based on the peak velocity of the jet of tricuspid regurgitation. The simplified Bernoulli equation describes the relationship of tricuspid regurgitation velocity and the peak pressure gradient of tricuspid regurgitation = 4×(tricuspid regurgitation velocity)². This equation allows for estimation of PA systolic pressure taking into account right atrial pressure: PA systolic pressure  =  tricuspid regurgitation pressure gradient + estimated right atrial pressure. Right atrial pressure can be estimated based on the diameter and respiratory variation of the inferior vena cava although often a fixed value of 5 or 10 mmHg is assumed. When peak tricuspid regurgitation velocity is difficult to measure (trivial/mild tricuspid regurgitation), use of contrast echocardiography (e.g. agitated saline) significantly increases the Doppler signal, allowing proper measurement of peak tricuspid regurgitation velocity. Also, potential systolic gradients between the RV and PA should be considered. Theoretically, calculation of P̄_pa from PA systolic pressure is possible (P̄_pa  = 0.61×PA systolic pressure + 2 mmHg) 47. This could allow the use of Doppler measurements, applying an accepted definition of PH as P̄_pa ≥25 mmHg. Unfortunately, despite the strong correlation of the tricuspid regurgitation velocity and tricuspid regurgitation pressure gradient, Doppler-derived pressure estimation may be inaccurate in the individual patient. In patients with severe tricuspid regurgitation use of the simplified form of the Bernoulli equation may lead to underestimation of PA systolic pressure. Also overestimations by >10 mmHg for PA systolic pressure are common 47. Therefore, PH cannot be reliably defined by a cut-off value of Doppler-derived PA systolic pressure.

Consequently, estimation of PAP based on Doppler transthoracic echocardiography measurements is not suitable for screening for mild, asymptomatic PH.

An alternative approach to echocardiographic diagnosis of PH is based on comparison of tricuspid regurgitation velocity with values reported in a healthy population. Ideally, the influence of age, sex and body mass should be taken into consideration 48. This method avoids cumulative error but is less directly linked to the accepted haemodynamic definition of PH as a P̄_pa ≥25 mmHg.

The reliability of several tricuspid regurgitation velocity cut-off values, using RHC as reference, has been assessed in two large screening studies. A trial evaluating the reliability of prospective screening of patients with scleroderma based on tricuspid regurgitation velocity >2.5 m·s⁻¹ in symptomatic patients or >3.0 m·s⁻¹ irrespective of symptoms, found that 45% of cases of echocardiographic diagnoses of PH were falsely positive 49. In symptomatic (dyspnoea) patients with HIV infection a PH criterion based on tricuspid regurgitation velocity >2.5 and 2.8 m·s⁻¹ was found to be a false positive in 72% and 29%, respectively 49.

Another trial selected a tricuspid regurgitation pressure gradient >40 mmHg (tricuspid regurgitation velocity >3.2 m·s⁻¹) with an assumed right atrial pressure of 10 mmHg (thus corresponding to a systolic PAP of >50 mmHg) as the cut-off value for diagnosis of PH 50. Those criteria were recently prospectively applied in systemic sclerosis patients 51. The Doppler diagnosis was confirmed in all 32 patients who were submitted to RHC. Like previous trials, the number of false-negative cases could not be assessed.

Other echocardiographic variables that might raise or reinforce suspicion of PH independently of tricuspid regurgitation velocity should always be considered. They include an increased velocity of pulmonary valve regurgitation and a short acceleration time of RV ejection into the PA. Increased dimensions of right heart chambers, abnormal shape and function of the interventricular septum, increased RV wall thickness and dilated main PA are also suggestive of PH, but tend to occur later in the course of the disease. Their sensitivity is questionable.

In table 9⇓ this Task Force suggests arbitrary criteria for detecting the presence of PH based on tricuspid regurgitation peak velocity and Doppler-calculated PA systolic pressure at rest (assuming a normal right atrial pressure of 5 mmHg) and additional echocardiographic variables suggestive of PH.

Echocardiography can be helpful in detecting the cause of suspected or confirmed PH. Two-dimensional, Doppler and contrast examinations can be used to identify CHD. High pulmonary blood flow found at pulsed wave Doppler in the absence of detectable shunt, or significant dilatation of proximal PA despite only moderate PH, may warrant transoesophageal examination with contrast or cardiac magnetic resonance imaging to exclude sinus venosus-type atrial septal defect or anomalous pulmonary venous return. In cases of suspicion of LV diastolic dysfunction, typical Doppler-echocardiographic signs should be assessed even if their reliability is considered low and a RHC may be required in specific circumstances (see section 9.1).

The practical clinical usefulness of exercise Doppler-echocardiography in the identification of cases with PH only on exercise is uncertain because of the lack of prospective confirmatory data 52.

7.1.6 Ventilation/perfusion lung scan

The ventilation/perfusion lung scan should be performed in patients with PH to look for potentially treatable CTEPH. The ventilation/perfusion scan remains the screening method of choice for CTEPH because of its higher sensitivity than CT 53. A normal- or low-probability ventilation/ perfusion scan effectively excludes CTEPH with a sensitivity of 90–100% and a specificity of 94–100%. While in PAH the ventilation/perfusion lung scan may be normal, it may also show small peripheral unmatched and nonsegmental defects in perfusion. Contrast-enhanced CT may be used as a complementary investigation but does not replace the ventilation/perfusion scan or traditional pulmonary angiogram. A caveat is that unmatched perfusion defects are also seen in PVOD.

7.1.7 High-resolution computed tomography, contrast-enhanced computed tomography and pulmonary angiography

High-resolution CT provides detailed views of the lung parenchyma and facilitates the diagnosis of interstitial lung disease and emphysema. High-resolution CT may be very helpful where there is a clinical suspicion of PVOD. Characteristic changes of interstitial oedema with diffuse central ground-glass opacification and thickening of interlobular septa suggest PVOD; additional findings may include lymphadenopathy and pleural effusion 54. Pulmonary capillary haemangiomatosis is suggested by diffuse bilateral thickening of the interlobular septa and the presence of small, centrilobular, poorly circumscribed nodular opacities.

Contrast CT angiography of the PA is helpful in determining whether there is evidence of surgically accessible CTEPH. It can delineate the typical angiographic findings in CTEPH such as complete obstruction, bands and webs, and intimal irregularities as accurately and reliably as digital subtraction angiography 55, 56. With this technique, collaterals from bronchial arteries can be identified.

Traditional pulmonary angiography is still required in many centres for the work-up of CTEPH to identify patients who may benefit from PEA 22. Angiography can be performed safely by experienced staff in patients with severe PH using modern contrast media and selective injections. Angiography may also be useful in the evaluation of possible vasculitis or pulmonary arteriovenous malformations.

7.1.8 Cardiac magnetic resonance imaging

Cardiac magnetic resonance imaging provides a direct evaluation of RV size, morphology and function, and allows noninvasive assessment of blood flow including stroke volume, CO, distensibility of PA, and RV mass 57. Cardiac magnetic resonance data may be used to evaluate right heart haemodynamics particularly for follow-up purposes. A decreased stroke volume, an increased RV end-diastolic volume, and a decreased LV end-diastolic volume measured at baseline are associated with a poor prognosis. Among the triad of prognostic signs, increased RV end-diastolic volume may be the most appropriate marker of progressive RV failure in the follow-up 58.

7.1.9 Blood tests and immunology

Routine biochemistry, haematology and thyroid function tests are required in all patients, as well as a number of other essential blood tests. Serological testing is important to detect underlying CTD, HIV and hepatitis. Up to 40% of patients with IPAH have elevated anti-nuclear antibodies, usually in low titre (1:80) 59. Systemic sclerosis is the most important CTD to exclude because this condition has a high prevalence of PAH. Anti-centromere antibodies are typically positive in limited scleroderma as are other anti-nuclear antibodies including dsDNA, anti-Ro, U3-RNP, B23, Th/To and U1-RNP. In the diffuse variety of scleroderma, U3-RNP is typically positive. In individuals with systemic lupus erythematosus, anti-cardiolipin antibodies may be found. Thrombophilia screening including anti-phospholipid antibodies, lupus anticoagulant and anti-cardiolipin antibodies should be performed in CTEPH. HIV testing is mandatory. Up to 2% of individuals with liver disease will manifest PAH and therefore liver function tests and hepatitis serology should be examined if clinical abnormalities are noted. Thyroid disease is commonly seen in PAH and should always be considered, especially if abrupt changes in the clinical course occur 60.

7.1.10 Abdominal ultrasound scan

Liver cirrhosis and/or portal hypertension can be reliably excluded by the use of abdominal ultrasound. The use of contrast agents and the addition of a colour-Doppler examination may improve the accuracy of the diagnosis 61. Portal hypertension can be confirmed by the detection of an increased gradient between free and occluded (wedge) hepatic vein pressure at the time of RHC 62.

7.1.11 Right heart catheterisation and vasoreactivity

RHC is required to confirm the diagnosis of PAH, to assess the severity of the haemodynamic impairment and to test the vasoreactivity of the pulmonary circulation. When performed at experienced centres, RHC procedures have low morbidity (1.1%) and mortality (0.055%) rates 63. The following variables must be recorded during RHC: PAP (systolic, diastolic and mean), right atrial pressure, P_pcw and RV pressure. CO must be measured in triplicate preferably by thermodilution or by the Fick method, if oxygen consumption is assessed. The Fick method is mandatory in the presence of a systemic-to-pulmonary shunt. Superior vena cava, PA and systemic arterial blood oxygen saturations should also be determined. These measurements are needed for the calculation of PVR Adequate recording of P_pcw is required for the differential diagnosis of PH due to left heart disease. In rare cases, left heart catheterisation may be required for direct assessment of LV end-diastolic pressure. A P_pcw >15 mmHg excludes the diagnosis of pre-capillary PAH. One of the most challenging differential diagnoses of PAH is heart failure with normal LV ejection fraction and diastolic dysfunction (see also section 9.1) 64. In this population, P_pcw may be mildly elevated or at the higher end of the normal range at rest. Exercise haemodynamics or volume challenge can show a disproportionate increase in P_pcw, although the relevance of this finding remains to be established. Coronary angiography may be required in the case of the presence of risk factors for coronary artery diseases and angina or in case of listing for double lung transplantation or PEA in patients with CTEPH.

In PAH, vasoreactivity testing should be performed at the time of diagnostic RHC to identify patients who may benefit from long-term therapy with calcium channel blockers (CCBs) (see also section 7.3.3) 65, 66. Acute vasodilator challenge should only be performed with short-acting, safe and easy to administer drugs with no or limited systemic effects. Currently the agent most used in acute testing is NO (table 9⇓) 66; based on previous experience 65, 67, 68 i.v. epoprostenol or i.v. adenosine may also be used as an alternative (but with a risk of systemic vasodilator effects) (table 10⇓).

Inhaled iloprost and oral sildenafil may be associated with significant vasodilator effects. Their role in the prediction of the response to CCB therapy has not yet been demonstrated. Due to the risk of potentially life-threatening complications, the use of CCBs given orally or i.v. as an acute test is discouraged. A positive acute response (positive acute responder) is defined as a reduction of P̄_pa ≥10 mmHg to reach an absolute value of P̄_pa ≤40 mmHg with an increased or unchanged CO 66. Only ∼10% of patients with IPAH will meet these criteria. Positive acute responders are most likely to show a sustained response to long-term treatment with high doses of CCBs and they are the only patients that can safely be treated with this type of therapy. About half of IPAH-positive acute responders are also positive long-term responders to CCBs 66 and only in these cases is the continuation of a CCB as a single treatment warranted. The usefulness of acute vasoreactivity tests and long-term treatment with CCBs in patients with other PAH types, such as heritable PAH, CTD and HIV patients is less clear than in IPAH. Nevertheless, experts recommend performing acute vasoreactivity studies in these patients and to look for a long-term response to CCBs in those in which the test is positive. No data are available on the usefulness of long-term CCB therapy in patients with PH associated with CHD and therefore the value of performing a vasoreactivity test in this setting is controversial. Acute vasoreactivity studies to identify patients with a long-term favourable response to CCBs is not recommended in clinical groups 2, 3, 4 and 5 (table 4⇓).

Recommendations for RHC and vasoreactivity test are summarised in table 11⇓.

7.1.12 Diagnostic algorithm

The diagnostic algorithm is shown in figure 1⇓: the diagnostic process starts with the identification of the more common clinical groups of PH (group 2–left heart disease and group 3–lung diseases), then distinguishes group 4–CTEPH and finally makes the diagnosis and recognises the different types in group 1–PAH and the rarer conditions in group 5.

PAH should be considered in the differential diagnosis of exertional dyspnoea, syncope, angina and/or progressive limitation of exercise capacity, particularly in patients without apparent risk factors, symptoms or signs of common cardiovascular and respiratory disorders. Special awareness should be directed towards patients with associated conditions and/or risk factors for development of PAH, such as family history, CTD, CHD, HIV infection, portal hypertension, haemolytic anaemia, or a history of intake of drugs and toxins known to induce PAH (table 8⇓). In everyday clinical practice such awareness may be low. More often PH is found unexpectedly on transthoracic echocardiography requested for another indication.

If noninvasive assessment is compatible with PH, clinical history, symptoms, signs, ECG, chest radiograph, transthoracic echocardiogram, pulmonary function tests (including nocturnal oximetry, if required) and high-resolution CT of the chest are requested to identify the presence of group 2–left heart disease or group 3–lung diseases. If these are not found or if PH seems “out of proportion” to their severity, less common causes of PH should be looked for. Ventilation/perfusion lung scan should be considered. If a ventilation/perfusion scan shows multiple segmental perfusion defects, a diagnosis of group 4–CTEPH should be suspected. The final diagnosis of CTEPH (and the assessment of suitability for PEA) will require CT pulmonary angiography, RHC and selective pulmonary angiography. The CT scan may also show signs suggestive of group 1'– PVOD. If a ventilation/perfusion scan is normal or shows only subsegmental “patchy” perfusion defects, a tentative diagnosis of group 1–PAH or the rarer conditions of group 5 is made. In table 12⇓ the further management according to the likelihood of PAH is given including indications for RHC. Additional specific diagnostic tests including haematology, biochemistry, immunology, serology and ultrasonography will allow the final diagnosis to be refined. Open or thoracoscopic lung biopsy entails substantial risk of morbidity and mortality. Because of the low likelihood of altering the diagnosis and treatment, routine biopsy is discouraged in PAH patients.

Recommendations for diagnostic strategy are summarised in table 13⇓.

7.2.1 Clinical, echocardiographic and haemodynamic parameters

Both clinical and haemodynamic assessments yield important prognostic information which may guide clinical management. These data have been derived from cohorts of patients and may not accurately reflect the prognosis of individuals. Prognosis is significantly affected by the aetiology of PAH 69.

Despite large interobserver variation in the measurement, World Health Organization functional class (WHO-FC) (table 14⇓) remains a powerful predictor of survival. In untreated patients with IPAH or heritable PAH, historical data showed a median survival of 6 months for WHO-FC IV, 2.5 yrs for WHO-FC III, and 6 yrs for WHO-FC I and II 8. Extremes of age (<14 yrs or >65 yrs), falling exercise capacity, syncope, haemoptysis and signs of RV failure also carry a poor prognosis in IPAH.

Echocardiography generates many indices, and those with the best prognostic value identified by multivariate analysis are pericardial effusion 70, 71, indexed right atrium area 71, LV eccentricity index 71 and the RV Doppler index 72, 73. Estimated systolic PAP derived from tricuspid regurgitant jet velocity is not prognostic 71. The tricuspid annular plane systolic excursion (TAPSE) has been reported to be of prognostic value 74.

Resting haemodynamics measured at RHC predict prognosis 8. These include PA oxygen saturation, right atrial pressure, CO, PVR and a marked vasoreactivity response. PAP is also prognostic but less reliable as it may fall towards the end stage of the disease as the RV fails. Some studies suggest that reduced arterial O₂ saturation, low systolic blood pressure and increased heart rate carry a worse prognosis 75.

Right atrial pressure, CI and P̄_pa have been incorporated in a formula to predict prognosis 8. It is unclear whether this formula is applicable to current clinical practice.

7.2.2 Exercise capacity

For objective assessment of exercise capacity, the 6-minute walking test (6-MWT) and cardiopulmonary exercise testing are commonly used in patients with PAH.

The 6-MWT is technically simple, inexpensive, reproducible and well standardised 77. In addition to distance walked, dyspnoea on exertion (Borg scale) and finger O₂ saturation are recorded. Walking distances <332 m 78 or <250 m 79 and O₂ desaturation >10% 80 indicate impaired prognosis in PAH. With respect to treatment effects, absolute values >380 m following 3 months of i.v. epoprostenol correlated with improved survival in IPAH patients while the increase from baseline did not 79. The increase in 6-MWT distance has remained the primary end-point in most pivotal PAH RCTs. The test is not sufficiently validated in PAH subgroups and is influenced by (but not corrected for) body weight, sex, height, age and patient motivation 77.

With cardiopulmonary exercise testing gas exchange and ventilation are continuously recorded throughout incremental exercise. In PAH, O₂ uptake at the anaerobic threshold and peak exercise are reduced in relation to disease severity, as are peak work rate, peak heart rate, O₂ pulse and ventilatory efficiency 81. Following multivariate analysis of clinical, haemodynamic and exercise parameters peak O₂ uptake (<10.4 mL O₂·kg⁻¹·min⁻¹) and peak systolic arterial pressure during exercise (<120 mmHg) independently predicted a worse prognosis in IPAH patients 75.

While the results of both methods do correlate in PAH, cardiopulmonary exercise testing failed to confirm improvements observed with 6-MWT in RCTs 82, 83. Although lack of standardisation and insufficient expertise in performing cardiopulmonary exercise testing were identified as the main reasons explaining this discrepancy 81, the 6-MWT remains until now the only Food and Drug Administration- and European Agency for the Evaluation of Medicinal Products-accepted exercise end-point for studies evaluating treatment effects in PAH. Despite detailed recommendations 84, 85, a generally accepted standardisation of cardiopulmonary exercise testing with respect to data acquisition and analysis in PAH is lacking.

7.2.3 Biochemical markers

Biochemical markers emerged within the last decade as an attractive noninvasive tool for assessment and monitoring of RV dysfunction in patients with PH.

Serum uric acid is a marker of impaired oxidative metabolism of ischaemic peripheral tissue. High uric acid levels were found to relate to poor survival in patients with IPAH 86. However, allopurinol is often prescribed to patients with PAH, and hyperuricaemia and diuretics influence its plasma levels, impairing the value of clinical monitoring based on uric acid levels.

Atrial natriuretic peptide and brain natriuretic peptide (BNP) share similar physiological properties. Both induce vasodilatation and natriuresis and are released from myocardium in response to wall stress. Interest in the clinical application of natriuretic peptides in monitoring RV failure due to chronic PH has focused on BNP.

The final step of BNP synthesis consists of a high molecular weight precursor, proBNP cleaved into biologically inactive N-terminal segment (NT-proBNP) and the proper low molecular weight BNP. NT-proBNP has a longer half-life and a better stability both in circulating blood and after sampling. RV failure is the main cause of death in PAH, and BNP/NT-proBNP levels reflect the severity of RV dysfunction. Nagaya et al. 87. showed that the baseline median value of BNP (150 pg·mL⁻¹) distinguished patients with a good or bad prognosis. In 49 out of 60 patients, BNP measurement was repeated after 3 months of targeted therapy and again the supramedian level (>180 pg·mL⁻¹) was related to worse long-term outcome. Plasma BNP significantly decreased in survivors but increased in nonsurvivors despite treatment. In a trial involving 68 patients with PAH associated with scleroderma, NT-proBNP below a median of 553 pg·mL⁻¹ was related to better 6-month and 1-yr survival 88. Using receiver operating characteristic (ROC) analysis, an NT-proBNP cut-off point at 1400 pg·mL⁻¹ was predictive of a 3-yr outcome in 55 patients with severe pre-capillary PH 89. Serum NT-proBNP below 1,400 pg·mL⁻¹ seemed particularly useful for identification of patients with good prognosis, who would not need escalation of treatment in the immediate future, and this has been independently confirmed 90. Larger outcome trials are still required to verify the suggested cut-off levels for NT-proBNP.

Increases in NT-proBNP plasma levels on follow-up have been associated with worse prognosis 88. Several recent trials assessing new drugs in PAH or CTEPH reported a significant decrease in NT-proBNP in the actively treated versus placebo patients.

Elevated plasma levels of cardiac troponin T and troponin I are established specific markers of myocardial damage and are prognostic indicators in acute coronary syndromes and acute pulmonary embolism. Elevated cardiac troponin T was an independent predictor of fatal outcome during 2-yr follow-up in a single trial on 51 patients with PAH and five with CTEPH 91. In some patients cardiac troponin T disappeared from plasma either temporarily or permanently after introduction of treatment. The value of monitoring of the cardiac troponin T level in patients with PH still requires confirmation in future studies. Other biomarkers are currently under investigation 92, 93.

In conclusion, several circulating biomarkers convey prognostic information in patients with PAH, but their value in everyday clinical practice is still not established.

BNP/NT-proBNP plasma levels should be recommended for initial risk stratification and may be considered for monitoring the effects of treatment, in view of their prognostic implications. Low and stable or decreasing BNP/NT-proBNP may be a useful marker of successful disease control in PAH.

7.2.4 Comprehensive prognostic evaluation

Regular evaluation of patients with PAH should focus on variables with established prognostic importance as outlined above. Treatment decisions should be based on parameters that reflect symptoms and exercise capacity and that are relevant in terms of predicting outcome. Not all parameters obtained repeatedly in PAH patients are equally well suited to assess disease severity. For example, PAP is measured on a regular basis, either by RHC or by echocardiography. The magnitude of the PAP correlates poorly with symptoms and outcome as it is determined not only by the degree of PVR increase but also by the performance of the RV. Thus, the PAP alone should not be used for therapeutic decision making. Table 15⇓ lists several parameters of known prognostic importance that are widely used as follow-up tools. Not all parameters need to be assessed at every visit (table 16⇓), but in order to obtain a clear picture it is important to look at a panel of data derived from clinical evaluation, exercise tests, biochemical markers, and echocardiographic and haemodynamic assessments. It is crucial not to rely just on a single parameter as several assessments may provide divergent results. In addition, no clear-cut thresholds for any single parameters can be identified to separate patients with good prognosis from those with a poor one. In table 15⇓, patients with better or worse prognosis are separated by an intermediate group for which prognostication is more difficult. In these cases, additional factors not included in table 15⇓ should be considered such as age, aetiology and comorbidities.

7.2.5 Definition of patient status

Based on the clinical, noninvasive and invasive findings the clinical condition of a patient can be defined as stable and satisfactory, stable but not satisfactory, unstable and deteriorating:

7.2.6 Treatment goals and follow-up strategy (see also section 7.3.7 and table 22⇓)

Treatment goals for PAH patients which may be considered are those listed in the “stable and satisfactory definition” and in the “better prognosis” column of table 15⇓.

Target values and treatment goals should be adjusted to the individual patient. For example, a 6-MWT >400 m is usually considered acceptable in PAH patients. Younger patients are often capable of walking 500 m or more despite the presence of severe PH and RV dysfunction. In these patients, additional exercise testing with cardiopulmonary exercise test and/or RHC is particularly useful in order to obtain a more reliable assessment of RV function. Peak V′_O2, O₂ pulse, peak systolic blood pressure during exercise and the minute ventilation/carbon dioxide production slope (ventilator efficacy) provide important information about RV function during exercise 75, 96. Biomarkers, echocardiography and RHC are useful additional tools to decide whether or not the patient can be considered stable. Suggested follow-up strategies for patients with PAH are reported in table 16⇓.

There is no universally accepted consensus about when and how often to perform follow-up RHC. Some, but not all, expert centres perform RHC regularly, for example, once a year. Some centres use RHC whenever a change in treatment is considered, while others regularly perform RHC 3–6 months after new treatments have been instituted to ensure that haemodynamics are in the desired range. In terms of prognostic importance, the most relevant haemodynamic variables are cardiac output, right atrial pressure (RAP) and mixed-venous oxygen saturation, i.e. those variables that reflect RV function. Recommendations for the use of RHC in PAH patients are provided in table 11⇓.

Recommendations for evaluation of severity and follow-up are summarised in table 17⇓.

7.3.1 General measures

Patients with PAH require sensible advice about general activities of daily living and need to adapt to the uncertainty associated with a serious chronic life-threatening disease. The diagnosis usually confers a degree of social isolation 100. Encouraging patients and their family members to join patient support groups can have positive effects on coping, confidence and outlook.

7.3.2 Supportive therapy

7.3.3 Specific drug therapy

7.3.4 Treatment of arrhythmias

Arrhythmias are an increasing clinical problem in PAH patients. In contrast to patients with left heart disease, malignant ventricular arrhythmias such as ventricular tachycardia, ventricular flutter and ventricular fibrillation are rare in PAH patients. In a series of 132 witnessed cardiac arrests in patients with PAH, ventricular fibrillation was observed in only 8% of the cases 168. Another series of 231 patients with PAH or CTEPH followed over a 6-yr period did not report any case of malignant ventricular arrhythmia 46. In that series, supraventricular tachyarrhythmias occurred with an annual incidence of 2.8%. Atrial flutter and atrial fibrillation were equally common and both arrhythmias invariably led to clinical deterioration with signs of right heart failure. Treatment of atrial flutter proved to be more successful than treatment of atrial fibrillation. Restoration of a stable sinus rhythm was associated with a favourable long-term survival while persistent atrial fibrillation was associated with a 2-yr mortality of >80% 46. Although prospective and controlled data are lacking, these findings suggest that maintenance of a stable sinus rhythm should be considered an important treatment goal in patients with PAH. In order to achieve a stable sinus rhythm, prophylaxis with antiarrhythmic drugs without negative inotropic effects such as amiodarone (see interactions in table 20⇓) should also be considered even if specific data regarding its efficacy are lacking.

7.3.5 Balloon atrial septostomy

Patients with Eisenmenger's syndrome and patients with IPAH with a patent foramen ovale have a survival advantage over those without a patent foramen ovale 169, supporting the concept of atrial septostomy as a treatment for IPAH. The creation of an inter-atrial right-to-left shunt can decompress the right heart chambers, and increase LV preload and CO 170, 171. In addition, this improves systemic O₂ transport despite arterial O₂ desaturation 170 and decreases sympathetic hyperactivity. The recommended technique is graded balloon dilation atrial septostomy, which produces equivalent improvements in haemodynamics and symptoms but reduced risk compared with the original blade technique. Other techniques are considered to be experimental 172.

A careful pre-procedure risk assessment ensures reduced mortality. Balloon atrial septostomy (BAS) should be avoided in end-stage patients presenting with a baseline mean RAP of >20 mmHg and O₂ saturation at rest of <80% on room air. Patients should be on optimal medical therapy, which may include pre-conditioning with i.v. inotropic drugs, prior to considering BAS. Evidence suggests a benefit in patients who are in WHO-FC IV with right heart failure refractory to medical therapy or with severe syncopal symptoms 170, 171. It may also be considered in patients awaiting transplantation or when medical therapy is not available. Severe IPAH has been the main indication for BAS in adults, although other indications include PAH associated with surgically corrected CHD, CTD, distal CTEPH, PVOD and pulmonary capillary haemangiomatosis.

Evidence shows improvements in CI, decreases in right atrial pressure with improvement in 6-MWT 170, 171.

The impact of BAS on long-term survival has not been established in RCTs 170, 171. BAS should be regarded as a palliative or bridging procedure to be performed only by centres with experience in the method 42.

7.3.6 Transplantation

The advent of disease-specific therapy for severe PAH has reduced patient referral for lung transplant programmes 42. The long-term outcomes of medically treated patients remain uncertain and transplantation should remain an important option for those who fail on such therapy. Studies indicate that up to 25% of patients with IPAH may fail to improve on disease-specific therapy and the prognosis of patients who remain in WHO-FC III or IV is poor 79, 97. International guidelines to aid referral and listing have been published by the International Society for Heart and Lung Transplantation 173.

The prognosis of PAH varies according to the aetiology, and PAH associated with CTD has a worse prognosis than IPAH even when treated with prostanoids, while patients with PAH associated with CHD have a better survival. The worst prognosis is seen in patients with PVOD and pulmonary capillary haemangiomatosis because of the lack of effective medical treatments, and those patients should be listed for transplantation at diagnosis.

In any case, patients with features identifying a worse prognosis profile (table 15⇓) despite maximal medical therapy should be referred for transplant listing.

Both heart–lung and double-lung transplantation have been performed for PAH, although the threshold of unrecoverable RV systolic dysfunction and/or LV diastolic dysfunction is unknown. As a consequence, each centre has developed its own strategy for the choice of the type of transplantation in the individual patient. However, due to the shortage of donor organs, most patients are considered for double-lung transplantation. While RV afterload is immediately reduced after double-lung transplantation, RV systolic and LV diastolic functions do not improve immediately and haemodynamic instability is a common problem in the early post-operative period. Both single and bilateral procedures have been performed, apparently with similar survival. However, any complication occurring in the allograft following single lung transplantation is associated with severe hypoxaemia. Currently the vast majority of patients worldwide receive bilateral lungs, as evidenced by the International Society for Heart and Lung Transplantation Registry figures 174.

Although patients with Eisenmenger's syndrome due to simple shunts have been treated by isolated lung transplantation and repair of the cardiac defect 174, patients with ventricular septal defects may have a better outcome with heart–lung transplantation 175.

The overall 5-yr survival following transplantation for PAH is 45–50%, with evidence of continued good quality of life 174.

7.3.7 Treatment algorithm

1. A treatment algorithm for PAH patients 176 is shown in figure 2⇓. The grades of recommendation and levels of evidence for the PAH treatments 177 are shown also in table 21⇓. The definition of clinical response to treatments is reported in table 22⇓. Country-specific regulatory approval and labelling for PAH medical treatments are shown in table 23⇓. Potential drug interactions are shown in table 20⇓.

2. Drug classes are listed in alphabetical order (ERAs, phosphodiesterase type-5 inhibitors, prostanoids) and single compounds are listed in alphabetical order within each class in figure 2⇓ and tables 21⇓ and 23⇓.

3. The treatment algorithm does not apply to patients in other clinical groups, and in particular not to patients with PH associated with group 2–left heart disease or with group 3–lung diseases. In addition, the different treatments have been evaluated by RCTs mainly in IPAH, heritable PAH, PAH due to anorexigen drugs, and in PAH associated with CTD or with CHD (surgically corrected or not). The grades of recommendation and levels of evidence for the other PAH subgroups are lower (see the section on specific PAH subsets).

4. The suggested initial approach after the diagnosis of PAH, is the adoption of the general measures, the initiation of the supportive therapy and referral to an expert centre.

5. Acute vasoreactivity testing should be performed in all patients with group 1–PAH, although patients with IPAH, heritable PAH and PAH associated with anorexigen use are the most likely to exhibit an acute positive response and to benefit from high-dose CCB therapy. Vasoreactive patients, as defined above, should be treated with optimally tolerated doses of CCBs (see section 7.3.3); adequate response should be confirmed after 3–4 months of treatment.

6. Nonresponders to acute vasoreactivity testing who are in WHO-FC II should be treated with an ERA or a phosphodiesterase type-5 inhibitor.

7. Nonresponders to acute vasoreactivity testing, or responders who remain in (or progress to) WHO-FC III should be considered candidates for treatment with either an ERA or a phosphodiesterase type-5 inhibitor, or a prostanoid.

8. As head-to-head comparisons among different compounds are not available, no evidence-based first-line treatment can be proposed. In this case the choice of the drug is dependent on a variety of factors including the approval status, the route of administration, the side-effect profile, patients' preferences and physicians' experience. Some experts still use first-line i.v. epoprostenol in WHO-FC III patients because of its survival benefits.

9. 1. Continuous i.v. epoprostenol is recommended as first-line therapy for WHO-FC IV PAH patients because of the survival benefit in this subset. S.c. and i.v. treprostinil have been also approved for the treatment of WHO-FC IV patients in the USA. Although no RCTs have been performed with the i.v. delivery of iloprost, this prostacyclin analogue has been approved in New Zealand.

   2. Although ambrisentan, bosentan and sildenafil are approved in WHO-FC IV patients in the USA, only a small number of these patients were included in the RCTs of these agents. Accordingly, most experts consider these treatments as a second line in severely ill patients.

   3. In WHO-FC IV patients, initial combination therapy should also be considered.

   4. In the case of inadequate clinical response (table 22⇓, fig. 2⇓), sequential combination therapy should be considered. Combination therapy can include either an ERA plus a phosphodiesterase type-5 inhibitor, or a prostanoid plus an ERA, or a prostanoid plus a phosphodiesterase type-5 inhibitor. Appropriate protocols for timing and dosing to limit possible side-effects of the combination have still to be defined. In expert centres also triple combination therapy is considered.

   5. BAS and/or lung transplantation are indicated for PAH with inadequate clinical response (table 22⇓) despite optimal medical therapy or where medical treatments are unavailable. These procedures should be performed only in experienced centres.

7.3.8 End of life care and ethical issues

The clinical course of PH is one of progressive deterioration interspersed with episodes of acute decompensation. It is difficult to predict when patients will die because death may come either suddenly or slowly because of progressive heart failure. It has been shown that physicians caring for patients tend to be overoptimistic in their prognostication and frequently misunderstand their patients' wishes.

Open and sensitive communication with patients allows advanced planning and discussion about their fears, concerns and wishes, which is essential to good care. Opportunities to discuss prognosis should be created at the time of initial diagnosis. Recognition that cardiopulmonary resuscitation in severe PH has poor outcomes may enable a “do not resuscitate” order. This may increase the chance of patients being in their preferred place of care at the end of life.

Patients approaching the end of life require frequent assessment of their full needs by a multidisciplinary team. Attention should be given to controlling the distressing symptoms and prescribing appropriate drugs while withdrawing unnecessary medication. Well-informed psychological, social and spiritual support is also vital. Specialist palliative care should be consulted about patients whose needs are beyond the expertise of the PH team.

7.4.1 Paediatric pulmonary arterial hypertension

Paediatric PH is similar to adult disease even if the lungs are still developing in a growing child. The worse prognosis in children with a median survival estimated at 10 months compared with 2.8 yrs in the adult 8 has not been confirmed. The exact incidence and prevalence of PH in children is not known. All forms of PH included in the clinical classification (table 4⇓) have been described in children, but the majority of patients present with PH associated with CHD or idiopathic/heritable forms. In contrast, the prevalence of PH associated with CTD, portal hypertension, HIV infection and drugs and toxins is lower. Patients with chronic lung disease of prematurity are a growing population. Persistent PH of the neonate is also classified under PAH. Its natural history, treatment and outcome are sufficiently different to justify its exclusion from this discussion 178.

No clear differences have been identified among the mechanisms involved in the development of PAH in children and adults.

7.4.2 Pulmonary arterial hypertension associated with congenital cardiac shunts

PAH associated with CHD is included in group 1 of the PH clinical classification (table 4⇓). A specific clinical classification (table 6⇓) and an anatomical–pathophysiological classification (table 7⇓) are useful to better define each individual patient with PAH associated with CHD 1, 16.

Epidemiological data remain scarce as no studies were designed to assess the prevalence of PAH in adult CHD, although 5–10% was suggested in a recent study 189. The persistent exposure of the pulmonary vasculature to increased blood flow due to systemic-to-pulmonary shunts as well as increased pressure may result in a typical pulmonary obstructive arteriopathy (identical to other PAH forms) that leads to the increase of PVR. If PVR approaches or exceeds systemic vascular resistance, the shunt is reversed (Eisenmenger's syndrome) 190.

7.4.3 Pulmonary arterial hypertension associated with connective tissue disease

PAH is a well-known complication of CTDs such as systemic sclerosis,114 systemic lupus erythematosus, mixed CTD and, to a lesser extent, rheumatoid arthritis, dermatomyositis and Sjögren's syndrome. PAH associated with CTD is the second most prevalent type of PAH after IPAH in registries 3, 4.

Systemic sclerosis, particularly in its limited variant (CREST syndrome), represents the main CTD associated with PAH. The prevalence of haemodynamically proven PAH in large cohorts of patients with systemic sclerosis is between 7 and 12% 2, 114. In these patients, PAH may occur in association with interstitial fibrosis or as a result of an isolated pulmonary arteriopathy. In addition, pulmonary venous hypertension from left heart disease may be present. It is imperative to determine which mechanism is operative since this dictates treatment.

Histopathological changes in PAH associated with CTD are generally indistinguishable from those of classical IPAH, although there is more frequent involvement of the pulmonary veins 199. The pathophysiological mechanisms leading to PAH in patients with CTD remain unknown. The presence of antinuclear antibody, rheumatoid factor, immunoglobulin G and complement fraction deposits in the pulmonary vessels wall suggest a role for an immunological mechanism.

7.4.4 Pulmonary arterial hypertension associated with portal hypertension

PAH is a well-recognised complication of chronic liver diseases 204, 205. Portal hypertension rather than the hepatic disorder itself seems to be the main determining risk factor for developing PAH 204.

The condition is not uncommon since PAH associated with portal hypertension (also named as portopulmonary hypertension) represents ∼10% of the PAH population 3. It is believed that 1–2% of patients with liver disease and portal hypertension develop PAH, but the prevalence of PAH may reach 5% among patients with advanced liver disease evaluated for liver transplantation 206. The pathogenesis is unclear and may be related to toxic substances derived from the gastrointestinal tract, which are not eliminated by the liver due to portosystemic shunts, thus causing damage to the lung endothelium. Another possibility is that the high CO state is inducing PAH.

7.4.5 Pulmonary arterial hypertension associated with HIV infection

The use of highly active antiretroviral therapy and aggressive management of opportunistic infections has contributed to increased life expectancy in HIV-infected patients. Consequently, the spectrum of complications has shifted towards other long-term conditions, including cardiovascular diseases such as dilated cardiomyopathy, pericardial disorders, noninfectious thrombotic endocarditis, accelerated atherosclerosis and PAH. The initial prevalence of PAH was found to be 0.1–0.5% with an estimated incidence of 0.1% per year 214. It was suggested that the use of highly active antiretroviral therapy could decrease the rate of PAH associated with HIV infection, but a population study recently conducted in France contradicts this hypothesis because the calculated minimal prevalence of HIV-related PAH was 0.46%, very similar to before the highly active antiretroviral therapy era 49.

The pathogenesis of HIV-related PAH remains unclear. The absence of viral particles in the complex plexiform lesions found from these patients suggests that an indirect action of viral infection on inflammation and growth factors may act as a trigger on a predisposed patient 49.

8.1.1 Diagnosis

The diagnosis of PVOD can be established with a high probability by the combination of clinical suspicion, physical examination, bronchoscopy and radiological findings 221. This noninvasive approach may avoid lung biopsy (the gold standard to confirm a diagnosis of PVOD) in most of the cases. Most patients complain of dyspnoea on exertion and fatigue, a clinical presentation that is indistinguishable from IPAH. Physical examination may reveal digital clubbing and bi-basal crackles on lung auscultation, these being unusual in other forms of PAH. Case series suggest that patients with PVOD are more severely hypoxaemic and have a much lower diffusion capacity of carbon monoxide than in other forms of PAH 54, 221, 222. This can be explained by the presence of chronic interstitial pulmonary oedema typical of PVOD and/or a low CO state and/or the presence of a patent foramen ovale.

Chest radiograph may reveal Kerley B lines and peripheral interstitial infiltrate in addition to other typical signs of PH.

High-resolution CT scanning is the investigation of choice. Typical findings suggestive of PVOD are the presence of subpleural thickened septal lines, centrilobular ground-glass opacities (contrasting with a panlobular distribution found in IPAH) and mediastinal lymphadenopathy 54. The association of these three findings was found to be 100% specific for PVOD in cases of PAH, with a 66% sensitivity 54. In addition, their presence appears to correlate closely with the risk of pulmonary oedema with epoprostenol therapy 223, 224.

Because PVOD may be associated with occult alveolar haemorrhage, bronchoscopy with bronchoalveolar lavage may be a useful tool in the diagnostic strategy. In a retrospective study, the results of bronchoalveolar lavage performed in 19 patients with suspicion of IPAH were analysed 222. Compared with IPAH, the eight cases of PVOD presented with a significantly elevated cell count, a higher percentage of haemosiderin-laden macrophages and a markedly elevated Golde score. The percentage of macrophages, lymphocytes and neutrophils was similar.

Haemodynamic presentation of PVOD is similar to IPAH. Importantly, P_pcw is almost invariably normal because the pathological changes occur in small venulae and do not affect the larger pulmonary veins. Vasoreactivity testing may be complicated by acute pulmonary oedema.

8.1.2 Therapy

There is no established medical therapy for PVOD. Most importantly, vasodilators and especially prostanoids must be used with great caution because of the high risk of pulmonary oedema 223, 224. However, there are reports of sustained clinical improvement in individual patients treated with these medications. There are no data regarding the use of newer medical therapies such as ERAs or phosphodiesterase type-5 inhibitors in the treatment of PVOD and pulmonary capillary haemangiomatosis. Therefore, therapy for PVOD should be undertaken only at centres with extensive experience in the management of PH, and patients should be fully informed about the risks. Atrial septostomy may be considered but is usually limited by hypoxaemia. The only curative therapy for PVOD and pulmonary capillary haemangiomatosis is lung transplantation, and similarly to IPAH there are no reports of recurrence of disease following transplantation. Patients with PVOD should be referred to a transplant centre for evaluation as soon as the diagnosis is established.

Recommendations for PVOD are summarised in table 29⇓.