|Home | About | Journals | Submit | Contact Us | Français|
Imaging techniques of the lung continues to advance with improving ability to image the more distal airways. Two imaging techniques are reviewed, computerized tomography and magnetic resonance with hyperpolarized helium-3.
The underlying rationale for using regional ventilation as a surrogate for small airways disease in asthma is that the small airways are mainly responsible for the distribution of ventilation to the gas-exchanging surface of the distal lung; when the small airways are narrowed by disease, the distribution of ventilation becomes uneven and air that was inhaled becomes trapped behind prematurely (anatomically or functionally) closed small airways. This may not be reflected by changes in residual volume (measured by body plethysmography) since the latter measurement is a global measure of lung volume and is likely to be a less sensitive measure of regional ventilation that can be assessed by imaging techniques using quantitative image analysis (QIA) (1, 2). While it is also possible to use thoracic HRCT to directly assess changes in conducting airway structure (airway wall thickness, luminal area), with current CT technology such assessment can only reliably be carried out to the level of subsegmental bronchi (i.e., the 5th generation of the tracheobronchial tree), which are proximal to the portion of the respiratory tract that comprises the small airways (7th or 8th generation and distally). Thus, CT imaging currently provides only an indirect assessment of small airways disease.
Advances in CT technology over the last decade have led to the development of multi-detector CT scanners that allow shorter scanning times and provide higher spatial and temporal resolution. Therefore, whereas older scanners required 32 seconds of scanning time to acquire images throughout the lung with axial slice thickness of 5 mm, the latest 64-detector scanners can acquire whole-lung data in only 4 seconds with slice thickness down to 0.4 mm. Thus, earlier studies were greatly restricted in examination of the airways. More recent studies utilizing 64-detector scanners can acquire whole lung data that can process image data volumetrically with anatomic segmentation into the five lobes and 18 bronchopulmonary segments. Therefore, examining regional changes within all of the anatomically relevant regions of the lung and assessing the regional heterogeneity of small airways involvement between lobes and segments, as well as within these anatomic units can be performed. Of importance is the cooperation of the subject for the validity of the information. Imaging is routinely performed under controlled conditions with the patient supine at residual volume (RV) or functional residual capacity (FRC) during a suspended breathhold. Spirometry is used to assure that the subjects hold their breath at the stipulated volume. Assessment of regional air-trapping was based on analyses of frequency distribution curves of lung attenuation units (so-called “lung attenuation curves”, or LACs) using a validated software package developed at UCLA and thus not subject to inter- or intra-observer variation (1). Shifts in these curves toward lower or higher attenuation (indicating an increase or decrease in regional airtrapping, respectively) were also assessed using computer software (1,2). Studies conducted in healthy volunteers provided normal reproducibility data indicating that shifts in the LAC >7% represent an abnormal change (2).
An early interventional study tested the hypothesis that treatment for 1 month with an extra-fine steroid aerosol (beclomethasone [BDP]HFA) in 31 steroid-naïve asthmatic subjects would have a greater effect in reducing regional airtrapping than the same microgram dose of the same molecule delivered from a conventional suspension steroid aerosol (BDP CFC) using a randomized, double-blind parallel-group design (3). The results indicated a significant shift in the LAC in the direction of greater attenuation (less regional airtrapping) in the BDP-HFA subjects but no shift in either direction in the BDP-CFC-treated subjects, the difference in the mean shifts in the two groups being statistically significant (p<0.05). Moreover, while both groups of subjects showed similar shifts in the direction of lower attenuation (more airtrapping) with methacholine at baseline, this methacholine-induced shift increased in the CFC-treated group (a representative example is illustrated in Figure 1A) but diminished substantially in the HFA-treated group (a representative example is shown in Figure 1B), the difference between the two groups being statistically significant (p<0.04). The latter findings suggest that regional hyperresponsiveness of smaller airways is reduced by the extra-fine steroid aerosol but not by the larger-particle suspension steroid aerosol, implying that the former but not the latter penetrated the distal lung to an extent that suppressed inflammation (and thus hyperresponsiveness) in the small airways. The possible clinical significance of these findings is suggested by the observation that shortness of breath (assessed by daily diary using a Likert scale) improved significantly more in the BDP HFA than the BDP CFC group. Moreover, the same subjects showed no changes in the dimensions of larger airways assessed at functional residual capacity in response to either formulation of BDP, suggesting that changes in larger airways were not responsible for the reduction in lung attenuation or improvement in symptoms in the BDP HFA group.
Approximately 25% of Fel D 1 antigen shed by cats is in the submicronic size range capable of reaching the small airways and inducing an inflammatory response in this region of the lung. We therefore examined whether naturalistic exposure to cat allergen in a 50 m3 room that contained two neutered cats could induce late allergic responses in the small airways detectable by HRCT using techniques of QIA. Ten subjects with cat-induced asthma were examined both physiologically and using HRCT before and 6 and 22 hours after a cat-room challenge (4). While there was no significant decline in FEV1 at 6 or 22 hours, HRCT QIA revealed increased regional air trapping at both 6 and 22 hours after challenge; a representative example is shown in Figure 2A. FEF25–75% did decline significantly at 6 hrs but not at 22 hours and closing volume, but not plethysmographically determined RV, increased significantly at both times periods after challenge. Interestingly, while the PC20 measured using FEV1 did not change from baseline after 22 hours, the change in CT attenuation in a selected region of the lung increased in response to methacholine at 22 hours after challenge occurred, indicating an increase in regional hyperresponsiveness (example is illustrated in Figure 2B). These findings indicate that a real-world exposure to cat allergen can cause both narrowing of small airways evident at 6 hrs (late asthmatic reaction) that persists for as long as 22 hours and is associated with regional hyperresponsiveness detectable by HRCT 22 hours after challenge that is not detectable spirometrically. Moreover, changes in two of the sensitve physiologic measures (closing volume and FEF25–75%) paralleled the HRCT findings at one or both time points, supporting the validity of the latter findings. It is noteworthy, however, that subjects had no symptoms of asthma 22 hrs after the cat room challenge despite the significant HRCT changes. While the clinical significance of these findings is therefore uncertain, it is possible that these changes represent clinically and physiologically silent distal lung inflammation that could serve as a substratum on top of which clinically significant asthma events could be superimposed.
Thus, quantitative image analysis of HRCT scans is a sensitive noninvasive method of assessing involvement of small airways in asthmatic patients and is capable of detecting narrowing of the distal airways (regional air-trapping) and increased responsiveness of these airways to methacholine (regional increases in air-trapping after methacholine) in response to various interventions when changes in traditional physiologic tests are largely undetectable. At present QIA of HRCT remains a research tool, the clinical utility of which remains to be determined. Limitations of this technique include radiation exposure, particularly with repeated imaging, and expense.
Magnetic resonance (MR) imaging with inhaled “hyperpolarized” helium-3 (3He) gas is an investigational technique with which high resolution images of the lung airspaces can be acquired (5, 6). Before imaging the 3He gas needs to be hyperpolarized in a special device (polarizer) and it typically takes several hours to produce liter quantities of the gas. When 3He gas is inhaled and imaging performed, the gas polarization quickly decreases (loses signal) by the radiofrequency pulses of the MR scanner and as a result, the number of images that can be obtained from one inhalation of gas is limited. This is in contrast to conventional hydrogen-based (1H) MR imaging, a technique with which unlimited numbers of images can be acquired, because the strong magnetic field of the MR scanner causes the polarization of the nuclei to quickly recover after each pulse. However, 1H MR imaging is poorly suited for airway imaging because the concentration of 1H nuclei (i.e. water) in air is low.
When the 3He gas is inhaled, the gas distributes evenly in the healthy lung and the airspaces are shown as homogeneous bright (white) structures (7–9). When there is airflow obstruction, for instance from airway closure as in asthma, the airspaces distal to the obstruction do not fill and are shown as black because of the lack of signal (figure 3). These foci of decreased or absent signal are also referred to as “ventilation defects”, and quantification of the defects provides a measure of the regional airflow. Since airflow obstruction forms the mainstay of the disease, ventilation defects are commonly found in patients with asthma, and we demonstrated in a recent study involving 58 asthmatics with differing severity and 18 age-matched healthy subjects that the number and size of the defects correlate with clinical measures of disease severity, and with spirometry (10)(figures 4–5). It was also shown that the mean number of ventilation-defects-per-image slice (VDS) for the asthmatics with moderate and severe persistent disease (11) was significantly greater than that for those with mild intermittent and mild persistent disease (1.37±0.24 versus 0.53±0.12; p<0.001). Another interesting finding was that there were more asthmatics with abnormal VDS than with abnormal FEV1, FVC, FEV/FVC or PEF. The number of asthmatics with abnormal VDS and abnormal FEF25%–75% was about the same. This could imply that VDS most closely correlates with FEF 25%–75%. It has been suggested that FEF25%–5% is sensitive to narrowing of the small airways and thus the abnormal VDS may be related to changes involving those airways (12, 13). However, other studies have had difficulty identifying objective changes in the small airways (14, 15) and it remains therefore uncertain which airways are involved when abnormalities in VDS are found in patients with asthma.
The ventilation defects can be remarkably persistent over time. Investigation recently showed that patients who were imaged twice within 42–82 minutes had 75% of defects remaining in the same location, and that most (71%) of these defects did not change size (16). When patients were imaged at longer (1–476 days) intervals there was relatively little change in total number of defects for each asthmatic, although the number of defects that remained in the same location decreased over time. However, a considerable number of defects remaining in the same location did not change size, with 67% of these still unchanged in size after 31 days (median interval), 43% after 41 days, and 38% after 85 days (figure 6). Thus, in certain regions of the lungs the airways appeared to remain persistently narrowed with time. This trend was regardless of disease severity or whether the patients used asthma medications.
It has also been shown that ventilation defects increase in number and extent when bronchoconstrictor (methacholine) is given, as can be expected from the increase in airflow obstruction caused by the medication. This can also be demonstrated with exercise in patients who are sensitive to this (17). However, when methacholine is given repeatedly, that is, the same dose on two different days, there appears to be a marked tendency for certain airways to respond to the same degree with each challenge (18). More specifically, we found that 69% of the ventilation defects that appeared at the first challenge also occurred at the second, and of those 43% had the same size (figure 7). These findings seem to suggest that certain airways are more sensitive to the methacholine than others, and some not at all. We have also shown that when β-agonist is given the defects decrease in number and FEV1 increases, as would be expected from the improvement of airflow caused by the bronchodilator (19). When the effects of the medication wear off the airways close down again and the defects recur (20). In few patients one can see an increase in defects after a bronchodilator is given suggesting paradoxic bronchoconstriction (21, 22).
Using 3He MR one can also obtain spatial information about the lung microstructure by measuring the diffusivity of the gas in the lung, that is the degree to which 3He atoms move within the airspaces. Studies have demonstrated that the 3He diffusivity is increased in patients whose airspaces are known to be enlarged, such as in chronic obstructive pulmonary disease (COPD) (23). It has also been found in an animal study that the increase in the apparent diffusion coefficient (ADC) of 3He correlates directly with size of the alveolar spaces (24). Initial investigation using this technique have shown variable results in patients with asthma (25). However, using a more sensitive method for measuring 3He diffusivity in 14 patients with severe, difficult-to-treat asthma, we found a slight but significant increase in ADC compared to healthy subjects although the values were substantially lower than those seen in COPD patients (26). An interesting finding was also that the ADC values remained elevated when patients were treated with oral steroids while there was improvement of the regional airflow evidenced by a decrease in the number of ventilation defects (27). In many of these patients no abnormalities were shown with computed tomography (CT). Therefore, the cause of the ADC elevation in the patients with severe asthma is uncertain but may reflect permanent alterations of the airway structure or tissue remodeling.
In summary, with hyperpolarized 3He MR imaging the changes of airflow in the distal airways can be assessed in great detail and are depicted as ventilation defects. These defects can be markedly persistent in some areas of the lung suggesting a relative regional distribution of the disease. A focal distribution of disease is also suggested from the relative persistent, regional response that can occur following repeat methacholine challenge. While medications such as short acting bronchodilators and oral steroids lead to improvement of the regional airflow, evidenced by the decrease in number and size of the ventilation defects, persistent alterations in 3He diffusivity can be demonstrated in the lungs of severe asthmatics possibly reflecting permanent structural changes of the airways or tissue remodeling. Hyperpolarized helium-3 MR imaging may become an important tool for assessing the distal airways in patients with asthma.
Symposium funded by an unrestricted grant from Teva Pharmaceutical Industries.
The authors thank Drs. Michelle Zeidler, Eric Kleerup, Jonathan Goldin, Talissa Altes, Chengbo Wang and John Mugler for their contributions to the research presented in this article. The 3He MR imaging studies were funded by the National Institutes of Health/National Heart, Lung, and Blood Institute (grant RO1 HL66479) and by contributions from the Commonwealth of Virginia Technology Research Fund (grant IN2002–01) and Siemens Medical Solutions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.