AbstractIntroduction and Objectives
Collateral ventilation is important in pathophysiology of Chronic Obstructive Pulmonary Disease (COPD), complicated pneumothorax, and bronchoscopic lung volume reduction surgery but limited observations of it in vivo have been attained. Current techniques capable of imaging collateral ventilation require monitoring over multiple breathing cycles and use ionising radiation. Here we present the first single-breath demonstration of collateral ventilation, by using hyperpolarised gas MRI.Methods
Ten patients with moderate to severe COPD as defined by NICE guidelines were scanned using a 1.5T whole body MRI system. A mix of 200 ml hyperpolarised 3He and 800 ml N2 was inhaled, and 3He MR images were acquired. 3D images with full lung coverage were acquired at six time-points during a single breathhold.Results
Abstract S118 figure 1 shows example sequential images in one patient, with insets highlighting an area with late ventilation. Hyperpolarised 3He MR signal is non-renewable and diminishes over time. In normally ventilated regions of the lung this expected signal decay is observed. However, in the regions indicated with arrows the signal increases over time due to collateral ventilation, with a progressive influx of polarised gas from the edge of the defects towards the centre. The long time constants for ventilation of some areas are not compatible with their ventilation via their feeding bronchi. Instances of collateral ventilation were observed in eight of the ten patients scanned. These examples varied in defect size, number and fill-rate from the strongest case (Abstract S118 figure 1) to much more subtle effects.Conclusions
A method for direct visualisation of collateral ventilation within a single breath-hold has been demonstrated in COPD patients for the first time. The technique gives 3D full lung coverage, and is non-invasive and non-ionising. The ability to image collateral ventilation directly may help to understand pathophysiology in COPD and aid assessment for therapy.