Article Text


Pulmonary arterial hypertension
P38 Which patients with pulmonary arteriovenous malformations are dyspnoeic? Retrospective analysis of a single centre 2005–2010 cohort
  1. V Santhirapala1,
  2. JT Springett2,
  3. H Wolfenden2,
  4. HC Tighe2,
  5. CL Shovlin1
  1. 1Imperial College London, London, UK
  2. 2Imperial College Healthcare NHS Trust, London, UK


Introduction/objectives Pulmonary arteriovenous malformations (PAVMs) are aberrant connections between pulmonary arteries and veins, creating a right-to-left shunt. Hypoxaemia is common,1 but dyspnoea is usually not the presenting complaint.2 We hypothesised this may be relevant to dyspnoea mechanisms in the general population.

Methods With ethical approval, new patients presenting with CT-proven PAVMs between June 2005 and July 2010 were studied retrospectively. Based on self-reported exercise tolerance at presentation, and blinded to physiological measurements, two investigators assigned patients to the MRC dyspnoea scale, grading dyspnoea on strenuous exertion only (normal, Grade 1); on incline/stairs (2); and on the level (3–5 according to severity). Individuals undertaking regular/intense sport were reclassified as Grade ‘0’. Four separate SpO2 measurements, after standing for 7, 8, 9 and 10 min, were used for statistical calculations using GraphPad software. Interim analyses were performed on the first 88 patients.

Results In the absence of severe pulmonary hypertension or emphysema, only 18% of patients were dyspnoeic (Abstract P38 Table 1). To determine which factors might influence dyspnoea, single variables were examined. There was considerable overlap in the SpO2 values between the three groups of Grade ‘0’, 1 and 2, and no relationship between dyspnoea grade and SpO2 demonstrated by Spearman's rank correlation coefficient (r=0.09; p=0.39). The patients however, ranged in age from 17 to 87 years, and older patients were more dyspnoeic (Spearman r=0.33; p=0.0016). Multiple regression analyses were therefore performed to determine whether there was a relationship between SpO2 and dyspnoea that was masked by differences in age. These suggested SpO2 may make an independent contribution to dyspnoea (p=0.064), although age was still more strongly associated (p=0.0044). However, these factors alone did not account for most of the population variation in dyspnoea grade (overall model: r2=11.37%, p=0.0073).

Abstract P38 Table 1

Population stratification by dyspnoea grade: *Quartile distribution (Q1, Q2, Q3) where Q2 represents the median value

Summary These data imply that it is unusual for PAVMs alone to account for moderate to severe dyspnoea, and that there is less of an influence of SpO2 than expected. Further study is warranted.

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.