The distribution of lung damage in children with cystic fibrosis and its relationship to colonisation with Pseudomonas aeruginosa

Deposition of coarse particles in cystic fibrosis: model predictions versus experimental results

In patients with cystic fibrosis (CF), the lung regions most affected by infection are presumed to be poorly ventilated and, hence, difficult to treat with inhaled therapeutic aerosols. Current dosimetric models do not adequately describe regional particle deposition in CF. We have developed a multiple-path particle deposition model and compared model predictions with the observed pattern of coarse particle (5 microm, mass median aerodynamic diameter) deposition in ten CF patients and eight healthy volunteers. Our model divides the lung into quadrants, separated at lobar bronchi, representing apical and basal lung regions. The volume and ventilation of quadrants were experimentally determined from a xenon equilibrium and multi-breath washout, respectively. Regional ventilation in the healthy lung was assumed to be determined largely by regional compliance. In CF patients, the deviations in regional ventilation from that observed in the healthy subjects were assumed to be due to regional resistance. A "custom" lung morphology was calculated for each subject based on their lung volume (functional residual capacity plus one-half tidal volume) and ventilation to each quadrant. Input parameters for particle deposition calculations were "custom" lung morphology, breathing pattern, and inhaled particle size. Relative to healthy subjects, the CF patients had reduced ventilation to the apices and increased ventilation to the bases of the lung. In healthy subjects, the general pattern of particle deposition followed ventilation. However, in the CF patients, the model predicted increased particle deposition in the large airways of the apices (an obstructed and poorly ventilated region) and to a lesser extent in the basal lung (relatively healthier and better-ventilated region), whereas particle deposition in the parenchyma was only increased in the basal lung and was decreased or absent in the apical lung. Our modeling strategy improves estimates of regional aerosol deposition and may be useful for predicting breathing conditions and particle size for optimal drug delivery in a given CF patient.

Targeting delivery of aerosols to different lung regions

With the increasing use of aerosolized drugs, there is a need to understand the means by which these drugs can most effectively be targeted to desired regions of the lung. Several attempts have been made at targeting aerosols in the lung by changing particle sizes and breathing patterns with varying degrees of success. Recent use of such techniques as shallow, aerosol bolus delivery and extremely slow inhalations of aerosols in diagnostic lung tests may also prove beneficial for targeting drug delivery to the conducting airways. This review discusses the potential for utilizing aerosol delivery techniques for selectively targeting aerosol deposition along both serial and parallel pathways in the lung. Based on a review of previous studies concerning factors that determine aerosol and gas distribution in the lung, the potential for utilizing various breathing techniques in concert with variations in particle sizes are considered. Further research on the factors that determine distribution of aerosol in the diseased lung may help in designing successful targeting strategies for the future.

Regional deposition of coarse particles and ventilation distribution in healthy subjects and patients with cystic fibrosis

The efficacy of inhaled pharmaceuticals depends, in part, on their site of respiratory deposition. Markedly nonuniform ventilation distribution may occur in persons with obstructive airways diseases and may affect particle deposition. We studied the relationship between regional deposition (RDep) and regional ventilation (RVent) in a group of 12 cystic fibrosis (CF) patients with mild to moderate airway obstruction (63 +/- 8% predicted FEV1) and 11 healthy nonsmoking volunteers (104 +/- 13% predicted FEV1) using planar scintigraphic methods. RDep was assessed from initial deposition and 24-h retention images for monodisperse technetium-99m-labeled iron oxide particles (5-microm MMAD). Regional volumes and RVent were assessed from xenon-133 equilibrium and washout, respectively. Six regions of interest per lung were established by dividing each lung into thirds by height and approximately half by width. The two lower regions of the left lung were not analyzed due to activity in the stomach. Remaining regions were categorized as central (two interior-most regions) and peripheral (eight exterior regions). RDep and RVent were computed for the eight peripheral regions. Tracheobronchial (TB) deposition was estimated for each of the peripheral regions as the difference between initial activity and decay-corrected 24-h retention or parenchymal deposition. RDep was computed as the fraction of material within a region normalized to regional volume. RVent for each region was determined by normalizing the xenon washout rate for that region by the total washout rate for the eight peripheral regions. Significant linear associations were found between RDep and RVent in both the healthy subjects and CF patients. In healthy subjects, RDep in the TB airways was positively associated with RVent (p = 0.03). In CF patients, RDep in the TB airways was negatively associated with RVent (p = 0.04) and RDep in the parenchyma was positively associated with RVent (p < 0.001). The initial pattern of RDep in the lung was not significantly associated with RVent in either group. These data suggest that significant coarse particle deposition may occur in the TB airways of poorly ventilated lung regions in CF patients, whereas, particle deposition in the TB airways of the healthy subjects follows ventilation.

Distribution of disease in cystic fibrosis: correlation with pulmonary function

STUDY OBJECTIVE

We hypothesized that the radiographic distribution of disease is important in determining pulmonary function, with the lower lung zones of more importance than the upper lung zones. To test this hypothesis, we retrospectively studied patients with cystic fibrosis, a disease with a known proclivity for the upper lung zones.

SUBJECTS AND METHODS

Two hundred seventy-six chest radiographs obtained at 2- to 3-year intervals from 51 patients were scored by two radiologists using a 100-point visual severity scale. The distribution of disease was assessed by dividing the lungs into four equal horizontal zones and scoring each zone for the severity of disease. There were 146 concurrent chest radiographs and pulmonary function tests from which multiple linear regression was used to correlate these zonal scores with FEV1/FVC percent predicted.

RESULTS

There was excellent interobserver agreement, intraclass correlation coefficients >0.7. The distribution of disease became predominant in the upper lung zones as the patients aged. Although the median score in the upper lung zones was nearly twice that of the lower lung zones, the lower lung zones were nearly three times (partial F ratio; 6.9 lower zone score vs 2.4 upper lung zone score) as important in determining pulmonary function.

CONCLUSION

The regional distribution of disease is important in determining pulmonary function. Sparing of the lower lung zones is important in preserving pulmonary function.

Radioaerosol assessment of lung improvement in cystic fibrosis patients treated for acute pulmonary exacerbations

We compared bronchopulmonary distribution homogeneity of a radioaerosol before and after hospitalization in 20 patients with cystic fibrosis (CF) with pulmonary exacerbations in order to assess lung improvement. Deposition homogeneity was quantified in terms of skew (an index of distribution symmetry), derived from frequency distribution histograms generated from gamma camera images of the lungs following radioaerosol inhalation. Lower skew values indicate enhanced distribution homogeneity. Right lung skew (RLS) was significantly reduced following therapy (1.00 +/- 0.49 to 0.84 +/- 0.47), whereas skew in the left lung was unchanged (0.95 +/- 0.38 to 0.87 +/- 0.40). The reduction in RLS was significant in patients with Shwachman-Kulczycki (SK) clinical scores less than 50 (1.27 +/- 0.53 to 0.90 +/- 0.42), but not in patients with scores greater than 50 (0.81 +/- 0.38 to 0.80 +/- 0.52). These results indicate that treatment affected the right lung more than the left lung, particularly in patients with SK scores less than 50, and suggests that radioaerosol lung imaging may be valuable in identifying sites of impairment to be targeted during treatment. Statistically, skew was less sensitive an indicator of acute change than several other clinical indices that improved following hospital treatment.