Clinical research in cystic fibrosis (CF) requires study endpoints that are sensitive to airways disease, repeatable and non-invasive. Despite significant advances in the treatment of CF, lung function assessments continue to rely on the forced expiratory volume in 1 second (FEV1). Although simple to perform, it lacks sensitivity, is difficult for younger subjects, and changes over time. An alternative method of assessing lung physiology is to derive measures of ventilation heterogeneity from inert gas washout tests. In early lung disease, measures of gas mixing appear to be more sensitive than spirometry. In addition, since only tidal breathing is required, they are more physiological and are more straightforward for younger subjects. Widespread use has been impaired by the lack of a robust and cost effective gas analyser technology. The work presented in this thesis concerns the adaptation, validation and then use of a novel gas analyser (Innocor) in a clinical system for the performance of multiple breath washouts. Lung clearance index (LCI), a simple measure of ventilation heterogeneity, has been calculated from washouts in 52 adults with CF and 50 healthy controls. LCI was more sensitive to disease than FEV1 in CF, being elevated in 11 of the 12 CF patients with normal spirometry. In healthy subjects, LCI has been shown to be repeatable and reproducible, with a narrow range of normal that is stable over a wide age range. In a separate study of 19 patients, LCI has also been shown to improve with treatment of an exacerbation in CF. Correlation with changes in other biochemical (serum CRP, peripheral blood white cell count, sputum IL-8, sputum neutrophil) clinical (symptom score) or structural (computed tomography) markers was poor. Short term change in LCI has also been demonstrated in CF patients in response to chest physiotherapy, although there was considerable heterogeneity of response in terms of both LCI and volume of lung ventilated by tidal breathing (as measured by washout functional residual capacity). In addition to LCI, multiple breath phase III slope analysis has been performed on washouts of CF patients and healthy controls, and this has been compared to other measures of lung physiology. Proposed measures of convective and diffusive gas mixing have been shown to be unreliable in CF. These studies have also been the first to demonstrate multi-centre use of washout tests as endpoints. The technology described here offers the possibility of a simple and reliable system for performing multiple breath washouts, though at present it is not available commercially. The studies have added to the understanding of the utility and reliability of washout tests, as well as some of their limitations. It is hoped that in future LCI will be an important clinical endpoint in therapeutic intervention studies in CF, and that it will also offer new ways to follow changes in lung physiology in other diseases.

Multiple breath washout (MBW) of inert gas for assessment of airway disease in children is an emerging technique. In many studies Lung Clearance Index (LCI), derived from multiple breath washout of SF6, is more able to detect early or mild lung disease than standard lung function measurements. It is also able to detect very early lung disease in progressive conditions such as Cystic Fibrosis (CF). Where infants born with this condition were thought to have minimal lung disease activity, LCI is higher in these children than healthy controls. Lack of available commercial devices has hampered expansion of this technique to centres other than specialist research teams. Innocor (Innovision, Dk), a photoacoustic mass spectrometer capable of performing multiple breath washout, was adapted within this research group for use in adults. This thesis describes the setup, adaptation and validation of Innocor for use in children. In 4 studies, healthy controls, children with asthma and children with CF were recruited to perform MBW. In one study, 29 healthy controls and 31 children with asthma were recruited. Healthy controls performed 1 set of washouts, establishing a normative range. Children with asthma performed measurements before and after bronchodilator. Results showed increased LCI in children with asthma even though they were clinically stable as defined by symptoms. LCI stayed high even following bronchodilator suggesting evidence of residual airway disease in well controlled asthmatics despite adequate symptom control. To investigate short term variability of MBW measurements, two other studies recruited 18 children with CF in each. They performed measurements before and after standard physiotherapy manoeuvres and during sitting and lying posture. LCI did not change significantly after airway clearance physiotherapy, compared with children who did no intervention. Variability was high in both groups however suggesting CF lung disease is a complex interaction of changing ventilation in adjacent lung units. Lying posture induced greater changes in lung function in children with CF than controls. LCI appears to be more sensitive to this change than standard lung function measurements (spirometry). In another study 32 children with CF were recruited to perform serial lung function measurements over 18 months. These were data collected as part of the UK Cystic Fibrosis Gene Therapy Consortium (CFGTC) clinical studies in preparation for planned gene therapy trials. LCI appears comparable to FEV1 and may be able to detect another aspect of airway disease. All initial studies were performed in older children (>5yrs). The basic Innocor device is unsuitable for testing of younger patients with low breath volume and high respiratory rate. In-house adaptations following detailed lung model experimentation led to a faster analyser response, potentially capable of MBW in younger children. The second part of this thesis concerns lab experiments and an in-vivo comparison with the current gold-standard MBW device, a respiratory mass spectrometer. 16 healthy volunteers and 9 children with CF were recruited. Ages ranged from 0.4 yrs to 49 yrs. Innocor values for lung volume estimation compared favourably with the mass spectrometer. No evidence of bias caused by Innocor error was seen, however intra-test variability was rather high, reducing the precision of the results. These studies indicate Innocor is a robust, simple to use device with potential as a commercial lung function system. Modifications were made to make it suitable for use in all ages. Further development will need to focus on the patient interface and software, which is the domain of the manufacturers. The experiments contained in this thesis are therefore of interest to the wider respiratory research community as well as manufacturers of MBW devices.

The Multiple Breath Washout (MBW) test is increasingly being recognised as a sensitive method of detecting early small airways lung disease. Indices of MBW include lung clearance index (LCI), Scond and Sacin. Factors that affect MBW variability have not been fully established. This thesis presents five studies which examine MBW repeatability in children with and without cystic fibrosis (CF) or asthma. MBW was performed using 0.2% sulphur hexafluoride and the modified Innocor (Innovision). Testing was performed at the Clinical Research Facility of the Royal Hospital for Sick Children in Edinburgh. (1) MBW and spirometry were performed in children with and without CF (n=20 in each group), initially while sitting and then 30 minutes after assuming a supine posture. LCI was found to significantly rise on lying supine in healthy children (p < 0.01) and children with CF (p=0.03). (2) Thirty two children with CF performed MBW and spirometry on four study visits, results were correlated with findings from high resolution chest computed tomography scans taken on the first visit. LCI showed the strongest correlation with extent and severity of bronchiectasis (r=0.66, p < 0.01 and r=0.69, p < 0.01 respectively). Variability of LCI was similar to FEV1 over the 4 visits. (3) MBW and spirometry of 66 healthy children were compared to 63 children with stable asthma; lung function of asthmatic children was related to symptoms and medication use. LCI was higher in the asthmatic group (6.7 vs 6.3, p < 0.01); within the asthmatic group LCI was significantly higher if asthma was less well controlled (p=0.02). (4) Children with and without asthma (n=21 in each group) performed MBW and spirometry before and after exercise and again after salbutamol, symptom data was collected from asthmatic children. Baseline LCI was abnormal in the asthmatic group who had severe exercise induced bronchospasm during testing. (5) Asthmatic children admitted to hospital due to exacerbation performed MBW and spirometry. Mean (SD) LCI was abnormally high at 8.5 (1.7) in the nine patients recruited and returned to normal 6.7 (0.6) in three patients who attended follow up. I have presented evidence that LCI is repeatable and sensitive to early disease in CF and asthma. I have described for the first time the effects of exercise and exacerbation on MBW indices in asthmatic children. MBW is potentially a very useful tool in paediatrics; standardisation of testing and equipment may enable clinical use.

Background: This work presents a generalized technique to estimate pulmonary ventilation-to-volume (v/V) distributions using the multiple-breath nitrogen washout, in which both tidal volume (VT) and the end-expiratory lung volume (EELV) are allowed to vary during the maneuver. In addition, the volume of the series dead space (vd), unlike the classical model, is considered a common series unit connected to a set of parallel alveolar units. Methods: The numerical solution for simulated data, either error-free or with the N2 measurement contaminated with the addition of Gaussian random noise of 3 or 5 % standard deviation was tested under several conditions in a computational model constituted by 50 alveolar units with unimodal and bimodal distributions of v/V. Non-negative least squares regression with Tikhonov regularization was employed for parameter retrieval. The solution was obtained with either unconstrained or constrained (VT, EELV and vd) conditions. The Tikhonov gain was fixed or estimated and a weighting matrix (WM) was considered. The quality of estimation was evaluated by the sum of the squared errors (SSE) (between reference and recovered distributions) and by the deviations of the first three moments calculated for both distributions. Additionally, a shape classification method was tested to identify the solution as unimodal or bimodal, by counting the number of shape agreements after 1000 repetitions. Results: The accuracy of the results showed a high dependence on the noise amplitude. The best algorithm for SSE and moments included the constrained and the WM solvers, whereas shape agreement improved without WM, resulting in 97.2 % for unimodal and 90.0 % for bimodal distributions in the highest noise condition. Conclusions: In conclusion this generalized method was able to identify v/V distributions from a lung model with a common series dead space even with variable VT. Although limitations remain in presence of experimental noise, appropriate combination of processing steps were also found to reduce estimation errors.

Background: This work presents a generalized technique to estimate pulmonary ventilation-to-volume (v/V) distributions using the multiple-breath nitrogen washout, in which both tidal volume (VT) and the end-expiratory lung volume (EELV) are allowed to vary during the maneuver. In addition, the volume of the series dead space (vd), unlike the classical model, is considered a common series unit connected to a set of parallel alveolar units. Methods: The numerical solution for simulated data, either error-free or with the N2 measurement contaminated with the addition of Gaussian random noise of 3 or 5 % standard deviation was tested under several conditions in a computational model constituted by 50 alveolar units with unimodal and bimodal distributions of v/V. Non-negative least squares regression with Tikhonov regularization was employed for parameter retrieval. The solution was obtained with either unconstrained or constrained (VT, EELV and vd) conditions. The Tikhonov gain was fixed or estimated and a weighting matrix (WM) was considered. The quality of estimation was evaluated by the sum of the squared errors (SSE) (between reference and recovered distributions) and by the deviations of the first three moments calculated for both distributions. Additionally, a shape classification method was tested to identify the solution as unimodal or bimodal, by counting the number of shape agreements after 1000 repetitions. Results: The accuracy of the results showed a high dependence on the noise amplitude. The best algorithm for SSE and moments included the constrained and the WM solvers, whereas shape agreement improved without WM, resulting in 97.2 % for unimodal and 90.0 % for bimodal distributions in the highest noise condition. Conclusions: In conclusion this generalized method was able to identify v/V distributions from a lung model with a common series dead space even with variable VT. Although limitations remain in presence of experimental noise, appropriate combination of processing steps were also found to reduce estimation errors.

Detection of early pathophysiological changes of breathing in children with chronic respiratory disease MD. Vaclav Koucky - Ph.D. thesis Abstract Introduction: Currently, there are different methods for infant pulmonary function testing (iPFT) and morphological assessment of microscopic changes in endobronchial biopsy samples (EBB). In research setting, they allow detection of early pathophysiological changes of breathing in small children with chronic respiratory disease, respectively in risk of its development. Their clinical significance, however, is not fully acknowledged. The aim of this thesis is to evaluate the safety, feasibility and clinical significance of iPFT and EBB in infants younger than 2 years of age. In addition, the relationship between functional and morphological changes of respiratory tract and the function of peripheral chemoreceptors was studied in selected patients' subgroups. Methods: Fifty-five infants with cystic fibrosis (CF), 35 physician-confirmed recurrent wheezers (AB), 9 infants with congenital diaphragmatic hernia, 7 with interstitial lung disease (chILD) and 3 with primary ciliary dyskinesia (PCD) were enrolled. All infants underwent iPFT and relevant clinical history data were recorded. Based on patients' age, CF group was divided into CFmalí (< 6 months) and CFvelcí (>...