Article Text

Download PDFPDF

Non invasive assessment of pulmonary hypertension severity and disease progression
  1. Philip J Kilner
  1. Correspondence to Philip J Kilner, Royal Brompton Hospital and Imperial College, London, Sydney Street, London SW36NP, UK; p.kilner{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Pulmonary arterial hypertension (PAH) causes morbidity and mortality in both paediatric and adult age ranges. It tends to be progressive but its symptoms can respond, in many cases, to drug treatments that have gained clinical acceptance in the past decade. A CT approach to the quantification of peripheral pulmonary arterial pruning in PAH has been reported by Moledina and colleagues1 in this journal (see page 1245). This is considered here in relation to more established methods for monitoring the severity and prognostic implications of PAH.

Exercise capacity

Although PAH may not initially be associated with noticeable symptoms, it tends to result in shortness of breath and limitation of everyday activities that involve moderate exertion.2 For assessing and monitoring the severity of the condition, the 6 min walk test is widely used.3 4 Limitation of the distance that a patient is able to walk on the flat in 6 min reflects a combination of pathophysiological factors: increased pulmonary vascular resistance and its effect on pulmonary perfusion, elevation of pulmonary arterial and right ventricular pressures and any resultant deterioration in right ventricular contractile function. The diastolic filling and the subsequent contractile function of the left ventricle may then be compromised by both the presence of a dilated, high pressure right ventricle adjacent to it in the pericardial space and by limitation of pulmonary venous return. There is a consequent reduction in cardiac output and its responsiveness to exercise. The 6 min walk test has the advantages of simplicity, inexpensiveness and sensitivity to change, particularly when supervised serially by the same experienced staff. Treadmill cardiopulmonary exercise testing, if necessary modified to be sensitive to changes in relatively modest levels of exercise, is also a safe, objective measure of exercise capacity,5 6 with oxygen uptake and ventilatory efficiency providing indirect information on pulmonary blood flow. However, such tests of performance can be affected by a number of factors, not only PAH. Individual motivation, exercise preconditioning and a concomitant illness could each modify the results, although this is less likely to be an issue in the longitudinal follow-up of a single patient than in comparisons between different patients.

Non-invasive imaging in relation to pathophysiology

Doppler echocardiography is used to measure the velocities of tricuspid regurgitant jet flow in screening for suspected PAH.7 8 It allows estimation, by the modified Bernoulli equation, of the systolic difference of pressures between the right ventricle and the right atrium. Any left to right shunting or obstruction of the right ventricular outflow tract or proximal pulmonary arteries need to be excluded before an elevation of right ventricular pressure can be attributed to elevation of the peripheral pulmonary resistance. However, cardiac catheterisation remains the gold standard for establishing the diagnosis of PAH and allows an assessment of potential pulmonary vasoreactivity, which carries important prognostic information.2 9

In PAH, the rate of flow from the right ventricle to the pulmonary trunk peaks and declines relatively early in systole as the arteries reach their limit of expansion, with peripheral run-off being limited by the high resistance. There then tends to be a plateau of relatively slow and slightly prolonged forward flow as the hypertrophied right ventricle continues to drive blood out through the high resistance. Detailed analyses of the time curves of local blood velocity by Doppler ultrasound or of the flow rate by cardiovascular magnetic resonance (CMR) are potentially valuable. So are assessments of right ventricular size, function and hypertrophy by either modality, particularly by CMR, although the analyses of these are by no means straightforward given the highly trabeculated boundaries of the right ventricle. Furthermore, both modalities can visualise and potentially measure the abnormal curvature and time course of displacement of the interventricular septum in PAH. Systolic septal flattening and further transient leftward displacement when systole of the right ventricle continues beyond that of the left ventricle reflect the changing relative pressures and time courses of right and left ventricular contraction. CMR can visualise and measure the diameters and systolic expansion of the pulmonary arteries. The diameters are typically abnormally large in PAH, but with limited expansion because the arterial walls remain near the limit of their elastic range. CMR, using cine imaging by steady state free precession or contrast-enhanced angiography, is also capable of identifying pulmonary arterial thrombus, which may have arrived by embolisation as a potential cause of PAH, or have been formed in situ secondary to the sluggish flow and altered haematological properties in pre-existing PAH. Contrast-enhanced CT, with its superior spatial resolution, is usually better than CMR for identifying thrombus and any resultant occlusion of pulmonary arterial branches, and can be acquired and analysed to give right ventricular volumetric data.10 All three modalities can also provide information on the underlying cardiovascular anatomy, particularly any atrial or ventricular septal defect or patent arterial duct as a potential cause of secondary PAH. In this context, CMR has the advantageous combination of unrestricted access, freedom from ionising radiation and the ready provision of functional information on ventricular performance and the volumes of flow through the aorta and pulmonary trunk.11 CT, on the other hand, has the advantages of high spatial resolution, speed of acquisition and ability to provide information on parenchymal lung disease The non-invasive modalities of echocardiography, CMR and CT can thus, between them, assess several distinct aspects of the pathophysiology in PAH.

Pulmonary arterial pruning

A widely known radiographic sign of PAH is the peripheral ‘pruning’ of pulmonary arteries visible on chest radiography. Moledina et al1 describe the use of contrast-enhanced CT pulmonary angiography as a basis for what is effectively quantification of the pruning of peripheral pulmonary arteries, expressed in terms of the fractal dimension of the arterial branch patterns. This method presumably reflects luminal changes in relatively small arterial branches that result from the crucial pathological processes involving stenosis and occlusion of peripheral pulmonary arteries and arterioles.


Fractal self-similarity of spatial patterning across size scales can be recognised throughout the natural world, notably in the branch patterns of trees, blood vessels and pulmonary airways; also in Brownian motion, in the complex eddies and counter-eddies of turbulently flowing water or air and in the disintegration of mountains to boulders and fragments; and also in the fluctuations of commodity markets and share prices. Progressively small-scale self-similarities of geometric forms were considered by Gottfried Leibnitz in the 17th century. In progress since then, the mathematicians Lewis Fry Richardson (1881–1953) and Benoit Mandelbrot (1924–October 2010) stand out. Both were interested in the complexities of social and economic as well as biological and geographical phenomena. Both contributed to the recognition that measuring the length of a geographical boundary or coastline is not a straightforward problem, and that the answer arrived at depends on the measuring scale used.12 Mandelbrot13 is also famous for identifying a simple mathematical formula that could generate phenomenally complex and beautiful patterns, and for coining the word ‘fractal’. The way that fractal curves, branches or fragments occupy space challenges familiar concepts of spatial dimensionality. Fractal lines or branches can be considered to have spatial dimensions of ‘one point something’, the number after the point depending on the intricacy of form. Practically speaking, for natural examples such as the coastline or the branching of arteries, the apparent dimension also depends on the measuring scale or image resolution used.

Returning to the paper by Moledina and colleagues,1 CT is the only modality that can provide data with three-dimensional spatial resolution approaching that needed to distinguish the more peripheral pulmonary arterial branches from one another, and from veins or other structures. As the authors discuss, the timing of acquisition relative to opacification by the contrast agent is critical. The issues of spatial resolution and the adjustments of the threshold of attenuation selected to delineate boundaries must also be crucial. The effective resolution can be compromised by any respiratory or bodily movement during acquisition. In the young age group studied, cooperation during imaging and the spatial resolution relative to body size may not have been as suitable as that potentially achievable in adults. Nevertheless, the correlations reported between the fractal analysis and patient outcomes seem very promising. Correlation of CT-based fractal analysis with any available histological findings, and the study of any reversibility with treatment, would also be of interest.

Is there a ‘best test’?

There is not yet a single, optimal test for PAH severity and disease progression. While the 6 min walk remains a valuable, inexpensive but non-specific test of patient performance, each of the three main non-invasive imaging modalities offer measurements and insights relative to specific aspects of PAH pathophysiology, which may vary in relative importance between individual patients.


The author is grateful to Professor Michael Gatzoulis, Dr Gerhard Diller and Dr Sonya Babu-Narayan for their helpful suggestions.


View Abstract


  • Linked articles 214130.

  • Funding PK is supported by the British Heart Foundation.

  • Competing interests None.

  • Provenance and peer review Commissioned; not externally peer reviewed.

Linked Articles