INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The noninvasive identification of pulmonary emphysema is an attractive proposition, but it is made virtually impossible by the pathologic basis of its definition. Pulmonary emphysema is characterized by abnormal, permanent enlargement of air spaces distal to the terminal bronchioles accompanied by the destruction of their walls (1). Thus, the clinical diagnosis of emphysema is difficult in vivo, because plain chest radiographs and conventional lung function tests have a low sensitivity and specificity (2), and lung tissue is usually not available. However, noninvasive identification of this disease not only allows assessment of the extent of damage but also paves the way for epidemiologic studies of emphysema as well as interventions such as antiprotease replacement and persuading a patient to give up smoking. At present, high-resolution computed tomography (HRCT) is the best noninvasive modality for detecting pulmonary emphysema (3, 4). However, HRCT is expensive, logistically difficult to use in field studies, and associated with a considerable radiation dose for the patients.

Recent studies using the new aerosol methods, aerosol- derived airway morphometry (ADAM) (5, 6) and aerosol bolus dispersion (ABD) (7, 8) showed that these methods are powerful noninvasive tools for the detection of emphysematous lung injury in patients with chronic obstructive pulmonary disease (COPD): ADAM detected significantly increased peripheral air-space dimensions in patients with emphysema as a marker for structural changes of the alveolar space (9, 10). In addition, aerosol bolus dispersion as a marker of ventilation inhomogeneities is enhanced clearly in patients with emphysema (10, 11). Recent investigations in patients with COPD (22) and in asymptomatic smokers (12, 13) showed that the aerosol methods have a higher sensitivity and specificity than conventional lung function tests in detecting emphysema and peripheral airway lesions.

However, since prior investigations analyzed the diagnostic power of the aerosol methods only in patients with COPD or smokers, no data are provided on the sensitivity and specificity of the aerosol methods in detecting pulmonary emphysema in unselected patients with different lung diseases in comparison with the noninvasive "gold-standard" HRCT. Also, the recommendation of the aerosol methods for field studies in COPD research (10, 13) warrants prospective evaluation in patients with different lung diseases. For this reason, we performed a prospective study to evaluate the diagnostic power of ADAM and ABD in an unselected group of 50 patients who underwent HRCT during a clinical work-up for various lung diseases and symptoms.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

This prospective study included 56 consecutive patients referred between April 1996 and May 1997 to the Department of Radiology, Center for Respiratory Medicine, Munich-Gauting, for HRCT during a clinical work-up for various pulmonary symptoms or lung diseases. Anamnestic data were collected using a questionnaire based on recommendations of the American Thoracic Society (14). The smoking history of the patients was quantified using the cumulative cigarette consumption expressed in pack-years (PY). Exclusions criteria were previous thoracic surgery, pleural effusions, obstructing bronchial carcinoma and pneumothorax. Four patients could not be evaluated because they refused to participate in the study, and two patients were not able to complete the lung function tests. A total of 50 patients (30 male, 20 female) 23 to 76 yr of age (mean age ± standard deviation, 56 ± 12 yr) were thus included in the present study. The distribution of HRCT biopsy-proven clinical diagnoses of the study group is shown in Table 1. Pulmonary emphysema was visually assessed by an expert chest radiologist unaware of the clinical and lung function data, as described by Goddard and colleagues (15).

Informed written consent was obtained from each subject. The protocol was approved by the Ethics Committee of the Medical School of the Ludwig-Maximilians-University (Munich, Germany).

Measurements

Pulmonary function testing. Body plethysmography and spirometry were performed using a Jäger-Masterlab (Erich Jäger GmbH, Wuerzburg, Germany). The following parameters were measured: TLC, VC, thoracic gas volume (Vtg), residual volume (RV), airway resistance (Raw), FEV₁, and maximal expiratory flows at 50% vital capacity (MEF₅₀). Relative values of these lung function parameters were calculated by normalization to the reference values proposed by the European Community for Coal and Steel (16). The transfer factor for carbon monoxide (DL_CO,sb) was calculated as proposed by Cotes and colleagues (17). Carbon monoxide labeled with the stable oxygen isotope (¹⁸O) was measured as a function of the respired volume by a respiratory mass spectrometer (DLT 100R, improved to fit scientific standard; Wagner, Worpswede, Germany).

Aerosol-derived airway morphometry (ADAM). ADAM and ABD were performed as previously described in detail (6, 8). In brief, a single breath of a monodisperse aerosol (particle diameter about 1 µm) is inhaled by the patients while the particle number concentration is continuously measured at the mouth as a function of the respired air volume. During postinspiratory breathholding periods, particles slowly settle in the air spaces with a constant velocity. The smaller the air spaces, the larger the probability that particles are deposited onto air- space walls. They are therefore not recovered upon expiration. Hence, for a given breathhold period more particles are recovered with the expired air from larger air spaces than from smaller ones. Moreover, the number of particles lost with increasing breathhold period is lower in larger air spaces than in smaller air spaces, so that the decline of the particle recovery with increasing breathhold periods allows the estimation of air space size.

The inspired breath of aerosol can be considered to be composed of infinitesimally small volume elements, which penetrate into different volumetric lung depths (V_p), and hence reach larger or smaller air spaces. For each of these volume elements, particle recovery decreases with increasing breathholding time. From the slope of this relation effective air space dimensions (EAD) can be calculated for each of these volume elements, i.e., as a function of volumetric lung depth. For peripheral lung regions EAD can be considered to be equivalent to the mean linear intercept (18).

All subjects inhaled aerosol to 85% TLC at end-inspiration to ensure the same level of lung inflation for each subject. The V_p was normalized to the end-inspiratory lung volume (VL = 85% TLC) to ensure that EAD from subjects of differing lung sizes are related to comparable anatomic regions: V_pr = V_p /VL where V_pr is relative volumetric lung depth (6). In this study data obtained for the V_pr = 0.20 (830 ml ± 242 ml) were taken into consideration because this lung depth certainly represents peripheral lung regions, and in our study all patients with COPD were able to respire the air volume necessary for its determination.

Aerosol bolus dispersion (ABD). Because particles with diameters of about 1 µm have low intrinsic mobility (i.e., diffusion or sedimentation) these particles can be used to probe intrapulmonary convective gas transport (8, 10, 11, 19, 20). To study ABD a subject inhales a small volume (bolus) of a monodisperse aerosol into a certain V_p and then immediately exhales. During this procedure, particle number concentration is measured at the mouth as a function of the respired volume. During inspiration the bolus is assumed to divide at each airway bifurcation. At end-inspiration these sub-boluses are distributed throughout the lung. During expiration, all sub-boluses move toward the mouth and recombine at the bifurcations. Because recombination of the sub-boluses is not completely reversible, the exhaled particles are distributed over a larger air volume than the inhaled particles. Therefore, a broadened bolus is recovered from the lungs: the bolus is dispersed. To quantify ABD the volumetric half-widths of the inhaled (H_50,i) and exhaled bolus (H_50,e) are considered. They are the inhaled and exhaled volume, in which particle concentration exceeds half-maximum particle concentration. The increase of volumetric bolus width during respiration is called bolus dispersion (ABD). ABD is determined from H_50,e by correcting for the bolus width of the inspired bolus (H_50,i). It is given by

Instrumental setup and inhalation protocol. ADAM and ABD measurements are performed using the Respiratory Aerosol Probe (Pari GmbH, Starnberg, Germany) (21). This computer-controlled device combines laser aerosol photometry with pneumotachography in order to measure the number concentration of respired monodisperse aerosol particles as a function of the respired air volume. Aerosol application is provided by the system of pneumatical valves, which allows the inhalation channel to be switched between particle-free air and an aerosol supply.

All ADAM and ABD measurements were performed at a constant airflow of 250 cm³ · s¹, controlled by use of a visual flow signal. The breathing maneuver for the determination of EAD starts with an exhalation of half the expiratory reserve volume (ERV) followed by an inhalation of test aerosol up to 85% of TLC. After a predetermined breathholding time, exhalation is performed until reserve volume (RV) is reached. The breathing maneuver is repeated for breathholding periods of 0-, 2-, 4-, 6-, 8-, and 10-s duration. The breathing maneuver for studying aerosol bolus dispersion started from FRC at which the subjects inhaled an air volume (Va) with a constant flow rate. Individual values of V_a were chosen in such a way that the end-inspiratory lung volume of all subjects, FRC plus Va, was equal to 85% TLC. During inspiration an aerosol bolus with 25 cm³ half-width was applied to this air volume and inhaled into a V_p of 600 cm³. The subjects then immediately exhaled with a constant flow rate until the aerosol bolus was recovered from the lungs or residual volume (RV) is reached. The lung depth of 600 cm³ was chosen because even in patients with emphysema, boluses inhaled into this lung depth could completely be recovered from the lungs.

All air volumes were corrected to BTPS.

Particle production and classification. Monodisperse di-2-ethylhexylsebacate (DEHS) droplets suspended in nitrogen were produced by heterogeneous nucleation of DEHS vapor on NaCl nuclei. The aerosol was then diluted with particle-free air to obtain a particle number concentration of about 2 · 10⁴ cm³. The size of the particles was classified by measuring the terminal settling velocity of the particles in a convection free sedimentation cell. The average settling velocity (mean ± SD) of the particles throughout the study was 25 ± 3 µm · s¹, representing a mean geometrical particle diameter of 0.9 ± 0.07 µm.

HRCT protocol. HRCT scans were performed on a Siemens Somatom Plus (Siemens, Erlangen, Germany) scanner during breathholding at full inspiration according to Gevenois and colleagues (3). None of the patients received contrast medium intravenously. All HRCT scans were obtained within 3 d before ADAM and ABD measurements. A set of scan images consisted of nine HRCT slices from the sternoclavicular joint down to the bottom of the lungs. The 1-mm collimation scans were performed at 2-cm intervals (137 kV_p, 255 mA; scanning time, 1 s).

Visual HRCT-score. Visual assessment of HRCT scans was used as the noninvasive "gold standard" in the calculation of sensitivity and specificity of routine lung function parameters and the aerosol methods. Pulmonary emphysema was visually assessed by an expert chest radiologist unaware of the clinical and lung function data as described by Goddard and colleagues (15). In addition, visual observation was quantified using a semiquantitative visual score according to Sakai and colleagues (22). On three HRCT slices (at the level of the carina, 5 cm above and 5 cm below carina) the lung parenchyma was assessed for two aspects of emphysema: severity and extent. The three levels were graded and scored separately for the left and right lung, giving a total of six lung fields. Severity was graded on a 4-point scale: 0, no emphysema; 1, low HRCT attenuation areas < 5 mm in diameter with or without vascular pruning; 2, circumscribed low HRCT attenuation areas > 5 mm in diameter in addition to those < 5 mm in diameter (vascular pruning is present, but with normal lung intervening); 3, diffuse low attenuation areas without intervening normal lung or confluent larger low attenuation areas with vascular pruning and distortion of the branching pattern of the lung, occupying all or almost all of the involved parenchyma. The extent of emphysema using the direct observation method was on a 4-point scale. The same six lung fields were used and the extent score was 1, < 25% of the lung field involved; 2, 25 to 50% involvement; 3, 50 to 75% involvement; 4, 75 to 100% involvement. For each of the six lung fields, the extent score was multiplied by that for severity to give a degree of emphysema score (Sakai Score). The sum of the product of severity score and extent score for the six lung fields has a potential maximum value of 72.

HRCT lung density measurements. Visual assessment of HRCT scans was further complemented by measurements of mean lung density (MLD) and pixel index. MLD describes the mean value of density in Hounsfield units (HU) of all pixels within a defined area, a so-called region of interest (ROI). The ROI resulted from identification of the lung boundaries by the autocontour algorithm with exclusion of the major blood vessels, mediastinal structures, and chest wall tissue by a manual step before analysis of the lung fields. The pixel index describes the relative value of all pixels lower than a defined threshold in relation to the total of pixels of a slice. The pixel index was evaluated for three threshold: < 950 HU, < 925 HU, < 900 HU.

Data Evaluation

All statistical calculations were performed using the Statgraphics Plus for Windows 2.0 software package (Manugistics, Inc., Rockville, MD). The significance of differences between group averages was tested using the t-test for independent samples. The required level of significance was 0.05. Pearson's product-moment correlations were performed to evaluate whether ADAM and ABD measurements in patients with emphysema correlate with the visual HRCT score and HRCT lung density measurements. The receiver operating characteristics (ROC) (23) was used to investigate the diagnostic value of aerosol methods. The sensitivity and specificity were quantified using the area under the ROC curve, which represents the probability (p_ROC) that a randomly selected pair of patients with pulmonary emphysema and without pulmonary emphysema is correctly ranked (24). A p_ROC value of 1.0 represents maximal sensitivity and specificity. Visual assessment of HRCT scans was used as the noninvasive "gold standard" in the calculation of sensitivity and specificity of the aerosol methods and routine lung function parameters.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anthropometric and lung function data of the study groups are given in Table 2. Patients with emphysema and patients without emphysema did not differ significantly in their age, lifetime cigarette consumption, airway resistance (Raw), VC (% predicted value), and transfer factor for carbon monoxide (DL_CO,sb). However, DL_CO,sb normalized for lung volume (DL_CO/VA) was significantly lower in patients with emphysema. The severity of airflow limitation based on values of FEV₁ (% predicted value) was mild to moderate in the group of patients without emphysema and moderate to severe in the group of patients with emphysema according to the guidelines of the American Thoracic Society (25).

Aerosol-derived Airway Morphometry (ADAM)

The peripheral EAD at a relative lung depth of V_p = 0.20 was found in patients with emphysema to be increased by a factor of two compared with that of patients without emphysema (0.84 ± 0.53 mm versus 0.33 ± 0.10 mm, p < 0.0001) (Figure 1). The peripheral EAD of patients without emphysema did not differ significantly from that previously reported in healthy subjects (0.34 ± 0.05 mm) (6).

View larger version (18K): [in this window] [in a new window]   Figure 1.   Effective peripheral airspace dimensions (EAD) and aerosol bolus dispersion in patients with HRCT-confirmed pulmonary emphysema (E) and in patients with various lung diseases (non-E) (box-and-whisker-graphics). The central box covers the middle 50% of the data. The bottom and the top of the boxes are the lower and upper quartiles, and the horizontal line drawn through the box is the median.

Aerosol Bolus Dispersion (ABD)

Patients with emphysema showed a significantly enhanced bolus dispersion (ABD) compared with that of patients without emphysema (706 ± 154 cm³ versus 462 ± 109 cm³; p < 0.0001) and previously reported data of healthy subjects (346 ± 53 cm³) (8). ABD of patients without emphysema did not differ significantly from that measured in healthy subjects (8).

High-resolution Computed Tomography (HRCT)

HRCT data are given in Table 3. Compared with patients with other lung diseases, patients with emphysema showed a visual HRCT score of 49 ± 20 with a significantly higher pixel index for the thresholds of < 950 HU, < 925 HU, < 900 HU (p < 0.001). The mean lung density (MLD) calculated from five HRCT slices was significantly more negative in patients with emphysema than in the group of patients without emphysema (876 ± 37 versus 805 ± 62).

Sensitivity and Specificity

The p_ROC for conventional lung function parameters, EAD measured by ADAM, and aerosol bolus dispersion (ABD) for the discrimination between patients with emphysema and those without emphysema are shown in Table 2. The p_ROC for HRCT parameters (visual score, pixel index, mean lung density) are shown in Table 3. From all parameters the HRCT visual score showed the highest value (0.97) followed by EAD (0.92). Aerosol bolus dispersion and pixel index < 950 HU had the same sensitivity and specifity with p_ROC values of 0.90. From all conventional lung function tests the most sensitive were Vtg (p_ROC, 0.87) and MEF₅₀ (p_ROC, 0.83). The lowest p_ROC showed DL_COsb (p_ROC, 0.50).

Analysis of Correlations of HRCT Parameters with Effective Air-space Dimensions (EAD) and Aerosol Bolus Dispersion (ABD)

The results of the correlation analysis are listed in Table 3. In patients with emphysema, EAD showed a strong significant correlation with the HRCT visual score (r = 0.78, p = 0.01), but it did not correlate with pixel index and mean lung density. Bolus dispersion (ABD) showed significant correlations with all HRCT parameters under consideration (visual score, pixel density, mean lung density) (r = 0.45 to 0.66; p < 0.05). The highest correlation between ABD and HRCT parameters was found for pixel index analysis < 900 HU (r = 0.62) and the mean lung density (r = 0.66). In the heterogeneous group of patients with other lung diseases (no emphysema group) EAD showed only a moderate correlation with the pixel indices (r = 0.47 to 0.54, p < 0.05). ABD correlated only with MLD (r = 0.50, p = 0.005) (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In patients with HRCT-confirmed macroscopic emphysema, peripheral air-space dimensions (EAD) were 155% larger and aerosol bolus dispersion (ABD) was 53% larger than those observed in patients with other lung diseases whose EAD and ABD values were similar to those measured in healthy subjects (8, 13). Among all lung function parameters under consideration, EAD (p_ROC, 0.92) and ABD (p_ROC, 0.92) had the highest sensitivity and specificity for the discrimination of patients with emphysema from those suffering from other lung diseases (i.e., lung fibrosis, bronchiectasis, pneumonia, chronic obstructive bronchitis, asthma) of all lung function parameters under consideration. The diagnostic power of the aerosol methods was close to that of the "gold standard" HRCT, which by definition has a p_ROC of 1 in the ROC analysis. The low sensitivity and specificity of DL_CO is due to the fact that diffusing capacity can be decreased in both restrictive disorders such as lung fibrosis and pulmonary emphysema. DL_CO may also be affected by a variety of conditions that cause deviation in the ratio of alveolar perfusion to ventilation (26). Further, recent studies suggest that DL_CO lacks sensitivity to depict emphysema in patients with COPD or healthy smokers by using new generations of CT scanners: Loubeyre and colleagues (27) reported a normal gas transfer (DL_CO/VA: 93 ± 4% pred) in patients with HRCT-confirmed emphysema with more severe airflow limitation (mean FEV₁, 45% pred) than that of the patients with emphysema in the present study. Yoshioka and colleagues (28), reported in smokers with HRCT-confirmed emphysema mean FEV₁/FVC of 70 ± 3% and DL_CO/Va of 81 ± 7% pred. Gurney and colleagues (29) defined functional emphysema as a DL_CO less than 75% pred and a FEV₁ less than 80% pred: 40% of patients with emphysema at HRCT had no functional abnormalities consistent with emphysema, i.e., a normal DL_CO.

Only two of 52 subjects were unable to complete the protocol of the aerosol methods. The total time required for testing was about 20 min. However, before the aerosol testing begins, additional time is needed to measure the absolute lung volumes. In patients with lung diseases, an advantage is that the desired flow rates are comparable to tidal flow rates, thus avoiding forced exhalations. The entire system for ADAM and ABD tests has become commercially available (Respiratory Aerosol Probe, Pari GmbH, Germany) and includes software for data collection and analysis.

EAD: A Marker for Peripheral Morphometric Changes of the Lung

Previous studies indicated that EAD reflects air-space dimensions in the human lung periphery. A convincing approach to validate ADAM measurements by comparing them to actual morphometric measurements of the same lungs was reported by Spektor and associates (30) in lungs from donkeys, by Rosenthal (30) in emphysematous dog lungs, and Nikoforov and associates (31) in excised human lungs. Spektor and associates (30) found that in excised lungs from donkeys the mean EAD was 0.16 mm, which was only 0.03 mm less than the mean alveolar dimension derived from the histologic sections. Rosenthal (32) reported similar findings in emphysematous canine lungs probed with the aerosol bolus recovery technique. Nikiforov and coworkers (31) examined a total of 17 excised human lungs from nonsmokers and smokers. EAD values agreed closely with morphometric data obtained from the same lungs. In addition, Brand and colleagues (6) have shown that EAD measured in 79 healthy subjects was similar to that measured by Thurlbeck using histologic techniques and that in both studies air-space dimensions increased with age to the same extent. Bennett and Smaldone (33) estimated the mean diameter of pulmonary air-spaces in eight patients with COPD by using in vivo measures of aerosol recovery as a function of breathhold time: The mean diameter of pulmonary spaces was measured twice as large as that of healthy subjects and correlated negatively with DL_CO (r = 0.95; p < 0.01). These results are confirmed by the present study, which showed both a marked increase of EAD and a negative correlation of EAD with DL_CO/Va in patients with emphysema (r = 0.85; p = 0.02). A recently published study (10) showed that in patients with chronic obstructive bronchitis the presence of macroscopic emphysema could be excluded by measurement of normal peripheral EAD with a higher sensitivity and specificity than by conventional lung function parameters. This observation may contribute to the high diagnostic power found in the current study in patients with various lung diseases in comparison with HRCT parameters.

Aerosol Bolus Dispersion: Pathophysiologic Aspects

Patients with emphysema showed a marked increase in aerosol bolus dispersion (ABD) compared with patients with various lung diseases, which was a parameter with a high sensitivity and specificity (p_ROC, 0.90). The pathophysiologic mechanisms resulting in increased ABD in patients with emphysema are still under discussion (19). However, Kohlhäufl and colleagues (11) showed that ABD was normal in patients with chronic obstructive bronchitis without emphysema, but enhanced in patients with chronic obstructive bronchitis combined with CT-confirmed emphysema. This result excluded central airway changes as a major factor of increased dispersion. The main histopathologic difference between patients with emphysema and those with other lung diseases is the presence of destruction of the alveolar region, which causes loss of alveolar attachments (radial traction) to the bronchioles. Functional consequences of emphysematous lung injury are ventilation inhomogeneities caused by changes of local lung compliance, peripheral expiratory airway closure, and enhanced collateral ventilation between adjacent bronchopulmonary compartments (1, 34). Each of these pathophysiologic mechanisms may lead to a considerably delayed aerosol particle recovery and may have a marked impact on bolus dispersion in contrast to other lung diseases. The results of the present study support this view. The high variability in bolus dispersion observed among patients with emphysema suggests that the contribution of these different mechanisms varies in these patients from case to case.

Comparison of HRCT Parameters with Aerosol-derived Airway Morphometry and Aerosol Bolus Dispersion

We could demonstrate a strong correlation between EAD and the HRCT visual score, which assesses severity and extent of emphysematous destruction. However, EAD did not correlate with parameters provided by measurement of HRCT lung density. This result can be explained by the fact that EAD measured by the single-breath maneuver is regarded as a volume-weighted average of air-space caliber of all ventilated lung regions (5, 6). However, previous reports (37, 38) showed that bullae are usually not significantly ventilated. In the present study, 65% of the patients with emphysema (13 of 20 patients) had obvious bullous disease combined with varying degrees of parenchymal emphysema. Thus, the presence of localized bullous disease in patients with emphysema may be a reasonable explanation for the fact that we failed to find a significant correlation between EAD and density measurements. This explanation is further supported by the fact that in two patients in the emphysema group in this study, the same EAD value was measured after bullectomy (unpublished data). Consequently, EAD reflects emphysematous lung injury in the ventilated nonbullous lung areas rather than in bullous areas. Bullous areas have recently been shown to be of less functional importance (39). Also, coincidental areas of lung fibrosis may also influence the measurement of CT lung density. The results of the current study are therefore in contrast to those of Beinert and colleagues (9): These investigators found in a small group of patients with COPD without bullous emphysematous lung disease (n = 10) a high correlation of EAD with MLD (r = 0.82). Thus, the presence of noncommunicating bullae weakens the correlation between EAD and CT lung density measurements.

Aerosol bolus dispersion (ABD), a marker of ventilation inhomogeneities, weakly correlates with measurements of CT lungs density and the visual score. This result is supported by the work of Biernacki and colleagues (40) who demonstrated that the severity and extent of emphysema on the computed tomogram does not predict gas exchange disturbances that characterize respiratory alterations in emphysema. In discussing our findings, it has to be noted that the issue of detection of early microscopic emphysema was not addressed because there is still controversy as to whether HRCT is sufficient to exclude adequately such histologically defined alterations (41), which may be present in patients without macroscopic emphysema. Also, cross-sectional data cannot resolve the question of whether these methods are suitable for detecting subclinical emphysematous disease. Therefore the value of these new techniques warrants longitudinal evaluation.

Conclusion

Aerosol-derived airway morphometry (ADAM) and aerosol bolus dispersion (ABD) are powerful tools for detecting or corroborating macroscopic emphysema. The sensitivity and specificity of ADAM and ABD are comparable to HRCT parameters for diagnosis of macroscopic emphysema in vivo. Because ADAM and ABD are noninvasive and rapid to perform, they are suitable for screening for macroscopic lung emphysema in epidemiologic or occupational studies.