HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fain, S. B. Articles by Grist, T. M. Search for Related Content PubMed PubMed Citation Articles by Fain, S. B. Articles by Grist, T. M. Related Collections Related Article

Early Emphysematous Changes in Asymptomatic Smokers: Detection with ³He MR Imaging¹

¹ From the Departments of Radiology (S.B.F., F.R.K., T.M.G.), Medical Physics (S.B.F., J.H.H., T.M.G.), Biomedical Engineering (S.B.F., S.R.P., A.L.W., T.M.G.), Biostatistics and Medical Informatics (M.D.E.), and Pulmonary Function (M.J.O.), University of Wisconsin, J3/110 CSC Medical Physics, 600 Highland Ave, Madison, WI 53792; and GE Healthcare, Princeton, NJ (H.F.). Received January 22, 2005; revision requested March 21; revision received June 10; accepted July 1; final version accepted August 24. Supported by National Institutes of Health grant 2P50-HL056396-06, GE Healthcare, and an award to S.B.F from the Sandler Program for Asthma Research. Address correspondence to S.B.F. (e-mail: sfain{at}wisc.edu).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Purpose: To prospectively compare apparent diffusion coefficient (ADC) measurements derived from diffusion-weighted hyperpolarized helium 3 (³He) magnetic resonance (MR) imaging with functional and structural findings using spirometric tests and thin-section computed tomography (CT) of the lungs in asymptomatic smokers and healthy nonsmokers of similar age.

Materials and Methods: All studies were HIPAA compliant and were approved by the institutional review board. Informed consent was obtained. Ventilation and diffusion-weighted ³He MR images were obtained in healthy subjects: 11 smokers (five women, six men; mean age, 47 years ± 18 [standard deviation]; range, 23–73 years) and eight nonsmokers (<100 cigarettes in lifetime) (four women, four men; mean age, 46 years ± 16; range, 23–69 years). Mean ADC values for smokers and nonsmokers were compared with spirometric values, diffusing capacity of the lung for carbon monoxide (DLCO), age, and pack-years with Spearman rank correlation coefficient (r_s) and multiple linear regression analysis. Mean ADC value and thin-section CT emphysema index of relative area less than –950 HU (RA₉₅₀) were compared on a regional basis by using linear mixed-effect models.

Results: Mean ADC values and number of pack-years were significantly correlated (r_s = 0.60; 95% confidence interval (CI): 0.21, 1.00; P = .007); relationship remained significant after adjustment for age (P = .003). DLCO was strongly correlated with pack-years (r_s = –0.63; 95% CI: –0.97, –0.29; P = .004). Negative correlations between mean ADC values and percentage predicted DLCO (r_s = –0.79; 95% CI: –0.93, –0.64; P < .001) and the ratio of forced expiratory volume in 1 second to forced vital capacity (r_s = –0.72; 95% CI: –0.92, –0.52; P = .001) were statistically significant. Correlations between spirometric values or RA₉₅₀ and number of pack-years were not significant (.05 level).

Conclusion: Correlations between mean ADC values and pulmonary function test measurements for diagnosing emphysema, especially DLCO, were statistically significant.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Emphysema is characterized by a breakdown in the alveolar walls of the lung. The diagnosis of emphysema is typically determined with whole-lung pulmonary function tests and is characterized by increases in airway obstruction and diffusional abnormalities (1). To quantify regional emphysematous changes in the lung, thin-section computed tomographic (CT) images typically are used to measure the fraction of the lung involved below a given threshold value in Hounsfield units. The voxels below this threshold value contain mostly air and, thus, are likely to be regions of disease. Several threshold values have been described in the literature (2–6). One specific index that is commonly used and has been positively correlated with histologic findings is relative area less than –950 HU (RA₉₅₀) (2,5,6).

Hyperpolarized gases also have been used to probe the microstructural changes due to emphysema through measurement of the apparent diffusion coefficient (ADC) of inhaled hyperpolarized gas within the airspaces of the lungs by using diffusion-weighted magnetic resonance (MR) imaging (7,8). Data in prior studies have shown that there is an increase in the ADC in animal models (9) and in patients with emphysema, compared with normal control subjects (8,10,11). Results of studies in animal models have confirmed that the increase in ADC is caused by less restricted diffusion of gas within the airspaces as the disease progresses (9). Given this understanding of the relationship between ADC and emphysematous changes, it is expected that ADC measurements might correlate with other functional measurements that are dependent on structural and functional changes in the lung.

Recently, the normal increase in ADC that is expected as a result of alveolar growth during development in children has been investigated (12). Researchers in other studies (13) have measured changes in ADC relative to spirometric measurements in patients with a diagnosis of emphysema and relative to measurements with thin-section CT in healthy smokers (14). Still, little is known about ADC changes in smokers prior to the onset of symptomatic emphysema. In particular, investigators in previous studies (13,14) have not explored the early onset of disease and the time when smokers start to show changes in lung structure prior to development of the overt symptoms of emphysema with the use of more established measurements. The objective of our study, therefore, was to prospectively compare ADC measurements derived from diffusion-weighted hyperpolarized helium 3 (³He) MR imaging with functional and structural findings obtained with spirometric tests and thin-section CT of the lungs in asymptomatic smokers and healthy nonsmokers of similar age.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Subject Groups
This study was supported by GE Healthcare, Milwaukee, Wis, through the provision of a commercial gas polarizer for hyperpolarization of ³He, ³He gas, and calibration equipment for determining the dose of hyperpolarized ³He delivered to the subjects. The authors who are not employed by or a consultant for GE Healthcare were given full control of data and authority over inclusion of any information that might present a conflict of interest to those who are.

All studies were Health Insurance Portability and Accountability Act compliant and were approved by the institutional review board of University of Wisconsin, Madison, Wis. Informed consent was obtained. Two independent groups were evaluated: one group of eight healthy nonsmokers, defined as individuals who had smoked less than 100 cigarettes in their lifetime (four women, four men; mean age, 46 years ± 16 [standard deviation]; range, 23–69 years) and one group of 11 healthy smokers (five women, six men; mean age, 47 years ± 18; range, 23–73 years). Inclusion criteria were forced expiratory volume in 1 second (FEV₁) greater than 75% predicted and normal chest findings at physical examination or radiography. Subjects were excluded if they had a history of asthma or allergies or they had received any medications within 3 days of the study.

Of the healthy nonsmokers, subjects were recruited in the following age categories: 18–30 years (three subjects); 31–59 years (three subjects), and 60 years or older (three subjects). One volunteer who was a healthy nonsmoker was unable to complete the MR imaging examination because of claustrophobia and was not included in the study. Therefore, subsequent analyses were calculated with the remaining eight nonsmokers. Of the healthy smokers, subjects were recruited with smoking histories of 5–10 pack-years (three subjects), 11–19 pack-years (three subjects), 20–30 pack-years (one subject), and longer than 30 pack-years (four subjects) on the basis of self-reported smoking histories. As defined for this study, 1 pack-year refers to the equivalent of smoking a pack of cigarettes a day for 1 year. For example, a person who has smoked 1 pack a day for 12 years and a person who has smoked a half a pack a day for 24 years both have a smoking history of 12 pack-years.

Pulmonary Function Tests
Subjects who were smokers were asked to refrain from smoking 4 hours prior to imaging and pulmonary function tests. Pulmonary function tests, including spirometric measurements and diffusing capacity of the lung for carbon monoxide (DLCO) (Jaeger Masterscreen Body Plethysmograph; Viasys Healthcare, Yorba Linda, Calif) measurements, were performed by a pulmonary function technologist (M.J.O.) with 19 years of experience. Spirometric and DLCO measurements were performed 18–24 hours prior to imaging, and spirometric measurements were repeated within 30–90 minutes before imaging. The DLCO was measured with the single-breath method (15). Specifically, the DLCO measurement was based on a simplification of the Fick equation, as follows:

(1)

where _CO is the diffusion rate of CO into the bloodstream and PACO is the alveolar partial pressure of CO. The PACO was measured by using a mouthpiece shutter valve after inhalation of normoxic gas (21% O₂) containing 2.95% CO and 9.68% helium. The helium was included as a tracer gas to determine alveolar volume.

The percentage predicted measurements of forced vital capacity (FVC), FEV₁, and DLCO were determined by using the reference tables of Miller et al (16) published in 1983.

MR Imaging and Evaluation
With a helium polarizer (IGI.9600; GE Healthcare), spin exchange optical pumping was used to polarize ³He to 30%–40%. One-liter doses of hyperpolarized ³He with a net activity of 4.5 mmol/L were prepared by mixing 200–300 mL of ³He with nitrogen gas added to reach a 1-L total volume in a bag made of polyvinyl fluoride (Tedlar; DuPont Chemical, Wilmington, Del) that was purged and rinsed with nitrogen to remove oxygen. The volunteer inhaled the polarized gas through an attached plastic tube (Tygon; United States Plastic, Lima, Ohio) with -inch inner diameter and starting with the lung volume at functional reserve capacity. Administration of ³He was approved by the U.S. Food and Drug Administration (investigational new drug approval no. 60311).

In all subjects, MR imaging was performed in a room that was adjacent to the room with the polarizer by using a 1.5-T MR imager with broadband capabilities (Signa LX; GE Medical Systems, Milwaukee, Wis). A vest radiofrequency coil (IGC-Medical Advances, Milwaukee, Wis) tuned to receive at the resonant frequency of ³He was used. Each MR imaging session consisted of localization, conventional proton-density weighted fast spin-echo imaging, ³He flip-angle calibration, breath-hold ³He ventilation MR imaging, and diffusion-weighted ³He MR imaging. For ventilation MR imaging, we used a spoiled gradient-echo sequence, with repetition time msec/echo time msec of 6/2.5, a flip angle of 10°, ±15.63-kHz readout bandwidth, 128 x 128 image matrix, 14–16 sections, and a 1.0-cm section thickness. For diffusion-weighted MR imaging, we used the same sequence, bandwidth, and matrix as were used for ventilation imaging but with 8/4.5, a flip angle of 7°, 10 sections, and 1.5-cm section thickness. Images were acquired in the coronal plane. Bipolar diffusion-weighted gradients (trapezoidal pulses with 500-µsec ramp times, 460-µsec peak pulse width, and 1.94848 G/cm [19.4848 mT/m] peak pulse amplitude) were added to the section-encoding axis (anteroposterior direction), and phase-encoded views were acquired alternately with and without diffusion weighting in an interleaved order. For all studies, a field of view within the ranges of 32–38 x 24–29 cm was used to include the lung anatomy. Electrocardiographic and oxygen saturation signals were monitored throughout the imaging session.

Several postprocessing steps were used to calculate the ADC values on a pixel-by-pixel basis, as follows:

1. The lungs were segmented from background noise by applying a lower threshold value that eliminated pixels with a signal intensity less than 3.0 , where is the standard deviation of the background noise.

2. The lung parenchyma was manually segmented from the major airways to eliminate bias from ADC values in the trachea and primary bronchi by an image scientist with 1 year of experience (S.R.P.).

3. After segmentation, the ADC was estimated by applying a log-linear fit to Equation (2), as follows:

(2)

where S₀ and S₁ are the signal intensity values without and with diffusion weighting, respectively, and b is the b value, which was 1.6 sec/cm² and was associated with the diffusion-weighted gradients used.

4. The mean (mean ADC) and standard deviation (standard deviation ADC) of the ADC values were then calculated over both lungs for whole-lung comparisons.

5. Regional analysis was performed with segmentation of the ADC maps into three regions each for right and left lungs as follows: apical (apex of the lung to the carina), middle (carina to the bifurcation of the left lobar bronchi), and basal (bifurcation of the left lobar bronchi to the base of the lung) regions. The mean ADC for the smokers and nonsmokers (mean ADC of the groups) for voxels within each region were calculated and compared with RA₉₅₀ derived from thin-section CT images obtained in similar locations. Specifically, the corresponding boundaries were found on the transverse CT images, and the position of each CT section was confirmed with direct measurement to be within the three regions. In all cases, two of five CT sections were in the range for the apical region, one of five sections was in the range for the middle region, and two of five sections were in the range for the basal region.

Ventilation images were assessed by a radiologist (T.M.G.) by using a four-point qualitative scoring system consisting of the following categories: score 1, no defects; score 2, small peripheral defects with a maximum diameter of 3 cm; score 3, large peripheral defects with a maximum diameter of 10 cm; and score 4, large peripheral to medial defects with a maximum diameter of more than 10 cm.

Thin-Section CT and Evaluation
Thin-section CT was performed with a 16–detector row CT scanner (LightSpeed; GE Medical Systems). Five 1-mm-thick transverse thin-section CT sections were acquired at specific anatomic locations corresponding to the following: the apex of the lung, the level of the aortic arch, the carina, the level intermediate to the carina and diaphragm, and the basal (at the dome of the diaphragm) region. Typical in-plane spatial resolution achieved was 0.25 x 0.25 mm by using 120-kV energy and 40-mAs tube current–time product.

The RA₉₅₀ was calculated by using the full-inspiration breath-hold data for each section by using tools in a software package (Analyze; Biomedical Imaging Resource, Mayo Clinic, Rochester Minn), with application of the following procedures: (a) Manual segmentation of the right and left lungs from the chest wall and back was performed by an image scientist with 2 years of experience (A.L.W.). (b) Threshold limits were applied to the segmented image to include only voxels within the range of –1024 to –950 HU. (c) The ratio of the area of the threshold image to that of the entire area of the segmented right and left lungs for each section was then calculated.

These threshold limits were chosen on the basis of data from the study of Gevenois et al (3) that showed that the –950-HU threshold limit best correlated with histologic quantification of emphysema. Evaluation of thin-section CT image quality and postprocessing analyses were performed by an experienced radiologist (T.M.G.) with 15 years of experience and a medical physicist (S.B.F.) with 8 years of experience in image processing and image quantification, including 2 years of experience specific to quantitative CT imaging of the lung.

Statistical Analysis
Pairwise associations among mean ADC values, RA₉₅₀ values, spirometric DLCO measurements, age, and pack-years were assessed by using the Spearman rank correlation coefficient (r_s) and Spearman test (S.B.F. and M.D.E.). These relationships were further assessed in the presence of other explanatory variables, such as age and pack-years of smoking, by using multiple linear regression analysis. Regional comparisons of the mean ADC and RA₉₅₀ for each group were examined by using linear mixed-effect models. Analyses were conducted (M.D.E.) by using the R environment for statistical computing (17) that included the mixed-effects models package (nlme, version 3.1-60, 2005; R Foundation for Statistical Computing, Vienna, Austria), authored by Jose Pinheiro, Douglas Bates, Saikat DebRoy, and Deepayan Sarkar. All tests for significance were two-tailed, and a difference with a P value of less than .05 was considered significant. Each correlation coefficient is reported with the 95% confidence interval (CI). Residual diagnostics and influence diagnostics, which included homoscedasticity of residuals, quantile-quantile plots, and Cook distance plots, were examined and found satisfactory for all models reported.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Ventilation Images
Ventilation images showed no apparent defects for 10 (53%) of 19 subjects; six were smokers and four were nonsmokers (Fig 1a, Table 1). A total of six (32%) of 19 subjects had findings suggestive of small peripheral ventilation defects less than 3 cm in maximum diameter (Fig 1b). Defects, including large peripheral defects, were similar in smokers and nonsmokers for all categories (Fig 1c). In no subjects did defects affect the medial regions of the lung. Only in three (16%) of 19 subjects were there severe enough defects to cause an absence of signal intensity, but these areas were of limited extent.

View larger version (147K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Coronal sections from ventilation 3He MR images (6/2.5, 128 x 128 matrix, 1.0-cm section thickness, 10° flip angle) demonstrate typical defects (arrows) that were classified in each descriptive category in Table 1. (a) Subject 13. No defects in a smoker with 14 pack-years of smoking. (b) Subject 1. Small peripheral defects in a nonsmoker. (c) Subject 19. Large peripheral defects in a smoker with 39 pack-years of smoking.

View larger version (147K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Coronal sections from ventilation 3He MR images (6/2.5, 128 x 128 matrix, 1.0-cm section thickness, 10° flip angle) demonstrate typical defects (arrows) that were classified in each descriptive category in Table 1. (a) Subject 13. No defects in a smoker with 14 pack-years of smoking. (b) Subject 1. Small peripheral defects in a nonsmoker. (c) Subject 19. Large peripheral defects in a smoker with 39 pack-years of smoking.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 1c: Coronal sections from ventilation 3He MR images (6/2.5, 128 x 128 matrix, 1.0-cm section thickness, 10° flip angle) demonstrate typical defects (arrows) that were classified in each descriptive category in Table 1. (a) Subject 13. No defects in a smoker with 14 pack-years of smoking. (b) Subject 1. Small peripheral defects in a nonsmoker. (c) Subject 19. Large peripheral defects in a smoker with 39 pack-years of smoking.

ADC Maps and Thin-Section CT RA₉₅₀
Whole-lung comparisons.—The pulmonary function test measurements for the two study groups (Table 2) showed that pack-years of smoking was associated with the percentage predicted values for DLCO (r_s = –0.63; 95% CI: –0.97, –0.29; P = .004) but not with the FEV₁/FVC ratio (r_s = –0.27; 95% CI: –0.74, 0.19; P = .24) or percentage predicted values for FEV₁ (r_s = 0.08; 95% CI: –0.45, 0.61; P = .73). In our thin-section CT data, whole-lung mean RA₉₅₀ had a near significant correlation (r_s = 0.41; 95% CI: 0.06, 0.77; P = .08) with percentage predicted values for DLCO, but RA₉₅₀ was not significantly associated with pack-years of smoking (r_s = –0.35; 95% CI: –0.75, 0.05; P = .14).

View larger version (18K): [in this window] [in a new window] [Download PPT slide]   Figure 2: Whole-lung mean ADC values according to FEV1/FVC ratio and smoking status.

View larger version (20K): [in this window] [in a new window] [Download PPT slide]   Figure 3: Whole-lung mean ADC values according to DLCO and smoking status.

View larger version (18K): [in this window] [in a new window] [Download PPT slide]   Figure 4: Whole-lung mean ADC values according to age and smoking status.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 5: Whole-lung mean ADC values according to pack-years of smoking.

View larger version (73K): [in this window] [in a new window] [Download PPT slide]   Figure 6a: ADC maps derived from coronal sections acquired with diffusion-weighted 3He MR imaging (8/4.5, 128 x 128 matrix, 1.5-cm section thickness, 7° flip angle). (a) Subject 1. ADC (0.166 cm2/sec) map in nonsmoker (Table 2) with (b) histogram for the typical section. (c) Subject 14. ADC map (0.268 cm2/sec) in smoker (Table 2) with (d) histogram for the typical section. Whole-lung mean ADC values are reported in Table 2. Graduated color bars represent values in squared millimeters per second.

View larger version (29K): [in this window] [in a new window] [Download PPT slide]   Figure 6b: ADC maps derived from coronal sections acquired with diffusion-weighted 3He MR imaging (8/4.5, 128 x 128 matrix, 1.5-cm section thickness, 7° flip angle). (a) Subject 1. ADC (0.166 cm2/sec) map in nonsmoker (Table 2) with (b) histogram for the typical section. (c) Subject 14. ADC map (0.268 cm2/sec) in smoker (Table 2) with (d) histogram for the typical section. Whole-lung mean ADC values are reported in Table 2. Graduated color bars represent values in squared millimeters per second.

View larger version (74K): [in this window] [in a new window] [Download PPT slide]   Figure 6c: ADC maps derived from coronal sections acquired with diffusion-weighted 3He MR imaging (8/4.5, 128 x 128 matrix, 1.5-cm section thickness, 7° flip angle). (a) Subject 1. ADC (0.166 cm2/sec) map in nonsmoker (Table 2) with (b) histogram for the typical section. (c) Subject 14. ADC map (0.268 cm2/sec) in smoker (Table 2) with (d) histogram for the typical section. Whole-lung mean ADC values are reported in Table 2. Graduated color bars represent values in squared millimeters per second.

View larger version (31K): [in this window] [in a new window] [Download PPT slide]   Figure 6d: ADC maps derived from coronal sections acquired with diffusion-weighted 3He MR imaging (8/4.5, 128 x 128 matrix, 1.5-cm section thickness, 7° flip angle). (a) Subject 1. ADC (0.166 cm2/sec) map in nonsmoker (Table 2) with (b) histogram for the typical section. (c) Subject 14. ADC map (0.268 cm2/sec) in smoker (Table 2) with (d) histogram for the typical section. Whole-lung mean ADC values are reported in Table 2. Graduated color bars represent values in squared millimeters per second.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

Salerno et al (13) demonstrated statistically significant dependence of mean ADC values on pulmonary function test measurements in patients with a diagnosis of emphysema compared with healthy control subjects. The results presented here extend the application of the results of Salerno et al (14) to asymptomatic smokers and corroborate preliminary data that showed differences in regional ADC for a cohort of healthy smokers compared with a cohort of nonsmoking control subjects with similar inclusion criteria. The present study is distinguished by the comparison of ADC measurements in asymptomatic smokers with respect to DLCO and quantitative CT.

In our study population, the DLCO measurement was most highly correlated with changes in mean ADC values in smokers. This relationship might be expected because DLCO is associated with decreased surface area and affinity for gas exchange into the bloodstream and is an established indicator of emphysematous changes in lung structure (15). Furthermore, DLCO was the only conventional measurement that could be used a priori to distinguish smokers from nonsmokers and correlated significantly with smoking history.

Results in several studies (18–20) have indicated that decreased DLCO is a more sensitive indicator of early smoking-related changes in lung function than are other pulmonary function tests and that it is not always associated with significant increases in airway obstruction in cases of mild chronic obstructive pulmonary disease (21). Also, the correlation between CT indexes, including RA₉₅₀, and DLCO is well established (22,23).

Correlations between thin-section CT measurements and changes in pulmonary function in smokers (5) and emphysema (2–4,6) also have been shown. Because of these findings, it was anticipated that regional ADC changes would parallel those observed by using the RA₉₅₀ index derived from thin-section CT. In our data, whole-lung mean RA₉₅₀ had a near significant correlation (r_s = 0.41; 95% CI: 0.06, 0.77; P = .08) with DLCO, but RA₉₅₀ was not significantly associated with pack-years of smoking (r_s = –0.35; 95% CI: –0.75, 0.05; P = .14). The lack of such a correlation in this study may be caused by a reduced sensitivity of quantitative CT compared with mean ADC values or possibly by a regional bias imposed by the CT sections that were selected. The five CT sections were chosen to provide consistent positioning with respect to anatomy and inclusion from the apical to basal regions and with minimal patient dose. Data from a study (24) by other researchers suggest, however, that whole-lung quantitative CT is superior to techniques in which individual sections are used.

With respect to regional measurements of ADC, the elevated ADC values in the apical versus basal regions of the lungs for smokers is consistent with the pattern of disease observed in symptomatic individuals (13). The elevated ADC values in the apical versus basal regions in nonsmokers, however, though less pronounced than those observed in smokers, was unexpected. When the supine position is used for imaging, the gravity-dependent intrapleural pressure gradients are in the anteroposterior direction and do not explain potential apical to basal regional variations in alveolar size. Regional variation in ADC measurements in normal subjects is an area for further study. In particular, the potential of a method for detection of nonclinically evidenced emphysema opens up new areas of study and opportunities for treatment of disease. A method capable of depiction of changes or stasis of lung microstructure on an individual basis would provide an imaging end point for therapies for emphysema and may guide the identification of genetic markers of disease that would be inexpensive and effective as screening tools for identification of individuals who are at risk for emphysema.

There were several limitations of the present study. First and foremost was the small sample size. Although a large proportion of the total variation in mean ADC values is explained by pack-years of smoking and age (R² = 0.70), the sample size makes any specific statements about the relationship between the ADC measurements and smoking history difficult. Qualitatively, a clear separation of smokers and nonsmokers is not observed until the level of 15–20 pack-years of smoking has been reached. Moreover, in this study, we were not able to determine the power of the technique for detection of onset of disease for a given individual because no follow-up or confirmation of results with histologic analysis was performed.

Second, no effort was made to normalize the volume of gas inhaled to total lung volume across subjects. Each subject inhaled a fixed 1-L volume of gas from functional reserve capacity. Consequently, differences in alveolar volume that are related to body size and lung volume rather than to disease may be depicted on diffusion-weighted ³He MR images. This possibility may degrade the precision of the mean ADC measurements because of different alveolar inflation volumes across subjects.

Third, although diffusion-weighted ³He MR imaging is robust to small signal intensity variations that result from radiofrequency coil sensitivity, lung parenchyma with large ventilation defects may yield inadequate signal for accurate calculation of the diffusion coefficient. This problem, however, did not appear to be a marked in the present study on the basis of visual comparisons of the ventilation maps with the ADC maps.

Fourth, the reproducibility of the ADC measurements was not explored in this study. Subjective steps in the postprocessing of the MR imaging data include assessment of the ventilation defect score, segmentation of the lungs, and segmentation of the major airways prior to ADC calculation. Despite these steps, it is important to note that the quantitative nature of the ADC measurement makes it less vulnerable to interobserver variability. For example, in our study, the statistical significance of the comparisons involving mean ADC values did not depend on the segmentation of the airways; in fact, this processing step was implemented to more accurately measure the standard deviation of the ADC and had little, if any, effect on the calculated mean ADC values. Larger studies directed at determining more precisely the relationship among age, smoking history, and measured ADC values are warranted.

An important step toward establishing clinical utility of this technique is to develop improved descriptors of the ADC distribution that are related to structural changes proved with histologic findings. The mean ADC for the whole segmented lung parenchyma was used in favor of the standard deviation of the ADC because the mean ADC appeared to be more robust than the standard deviation, even after removing bias introduced by extreme ADC values by segmenting out the major airways. Specifically, the mean ADC correlated more strongly with pulmonary function test results and history of smoking. One explanation is that the standard deviation is more sensitive to reduced signal-to-noise ratio than are the mean ADC measurements and is thus more variable and vulnerable to systematic error (25), particularly given that there are inhomogeneities in radiofrequency coil sensitivity. An alternative to the standard deviation would be a threshold index similar to that used for quantitative CT, although the appropriate index has yet to be determined and would probably best be established by a careful exploration of the dependence of the ADC histogram on the signal-to-noise ratio and correlation to histologic sections similar to the whole-lung thin-section CT analysis in the study of Gevenois et al (3). Whole-lung thin-section CT images and selection of transverse sections on ³He MR images would facilitate the direct comparison of regions of ventilation defect with changes on thin-section CT images. With the current study, these relationships were impossible to assess because the selected CT sections did not often intersect regions of ventilation defect on MR images.

In conclusion, in the study presented here, we found a statistically significant correlation between mean ADC values and common pulmonary function test measurements used to diagnose emphysema, especially DLCO. In addition, mean ADC values had a positive correlation with pack-years of smoking and age, both globally and regionally. No such trends were observed in these same individuals for spirometric measurements of airway obstruction associated with emphysema or for a common quantitative CT index, RA₉₅₀. Studies in larger patient groups, however, will be required to confirm these results, ideally with direct comparisons of diffusion-weighted hyperpolarized gas MR images with identical section positions reformatted from whole-lung thin-section CT images.

	ADVANCES IN KNOWLEDGE

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

• The ADC derived from ³He MR imaging is a sensitive regional measurement of early changes in lung microstructure that correlates with pulmonary function test results.
  
• The data suggest that the ADC may be more sensitive to early disease onset than established quantitative CT measurements.
  

	ACKNOWLEDGMENTS

 
The authors thank Kelli Hellenbrand, BS, and Sandy Fuller, RN, for their valuable assistance.

	FOOTNOTES

 

Abbreviations: ADC = apparent diffusion coefficient • CI = confidence interval • DLCO = diffusing capacity of the lung for CO • FEV₁ = forced expiratory volume in 1 second • FVC = forced vital capacity • RA₉₅₀ = relative area less than –950 HU

J.M.G. is a consultant for GE Healthcare. See Materials and Methods for pertinent disclosures.

See also Science to Practice in this issue.

Author contributions: Guarantors of integrity of entire study, S.B.F., A.L.W., J.H.H., T.M.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.B.F., J.H.H.; clinical studies, S.B.F., A.L.W., J.H.H., F.R.K., M.J.O., H.F., T.M.G.; statistical analysis, S.B.F., S.R.P., M.D.E., T.M.G.; and manuscript editing, S.B.F., S.R.P., M.D.E., A.L.W., J.H.H., F.R.K., M.J.O., T.M.G.

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

1. Petty T. Chronic obstructive pulmonary disease. In: Hanley M, Welsh C, eds. Current diagnosis and treatment in pulmonary medicine. San Francisco, Calif: McGraw-Hill, 2003; 82–90.
2. Hayhurst MD, MacNee W, Flenley DC, et al. Diagnosis of pulmonary emphysema by computerised tomography. Lancet 1984;2:320–322.[Medline]
3. Gevenois PA, Zanen J, de Maertelaer V, De Vuyst P, Dumortier P, Yernault JC. Macroscopic assessment of pulmonary emphysema by image analysis. J Clin Pathol 1995;48:318–322.[Abstract/Free Full Text]
4. Uppaluri R, Mitsa T, Sonka M, Hoffman E, McLennan G. Quantification of pulmonary emphysema from lung computed tomography images. Am J Respir Crit Care Med 1997;156:248–254.[Abstract/Free Full Text]
5. Soejima K, Yamaguchi K, Kohda E, et al. Longitudinal follow-up study of smoking-induced lung density changes by high-resolution computed tomography. Am J Respir Crit Care Med 2000;161:1264–1273.[Abstract/Free Full Text]
6. Cederlund K, Tylen U, Jorfeldt L, Aspelin P. Classification of emphysema in candidates for lung volume reduction surgery. Chest 2002;122:590–596.[Abstract/Free Full Text]
7. Chen XJ, Möller HE, Chawla MS, et al. Spatially resolved measurements of hyperpolarized gas properties in the lung in vivo. I. Diffusion coefficient. Magn Reson Med 1999;42:721–728.
8. Saam BT, Yablonskiy DA, Kodibagkar VD, et al. MR imaging of diffusion of (3)He gas in healthy and diseased lungs. Magn Reson Med 2000;44:174–179.[CrossRef][Medline]
9. Chen XJ, Hedlund LW, Möller HE, Chawla MS, Maronpot RR, Johnson GA. Detection of emphysema in rat lungs by using magnetic resonance measurements of 3He diffusion. Proc Natl Acad Sci U S A 2000;97:11478–11481.[Abstract/Free Full Text]
10. Salerno M, Altes TA, Brookeman JR, de Lange EE, Mugler JP III. Dynamic spiral MRI of pulmonary gas flow using hyperpolarized (3)He: preliminary studies in healthy and diseased lungs. Magn Reson Med 2001;46:667–677.[CrossRef][Medline]
11. Yablonskiy DA, Sukstanskii AL, Leawoods JC, et al. Quantitative in vivo assessment of lung microstructure at the alveolar level with hyperpolarized 3He diffusion MRI. Proc Natl Acad Sci U S A 2002;99:3111–3116.[Abstract/Free Full Text]
12. Altes T, Mugler J III, Mata J, et al. Assessment of lung development using hyperpolarized helium-3 diffusion MR imaging (abstr). In: Proceedings of the twelfth meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2004; 765.
13. Salerno M, de Lange E, Altes T, Truwit J, Brookeman J, Mugler J III. Emphysema: hyperpolarized helium 3 diffusion MR imaging of the lungs compared with spirometric indexes—initial experience. Radiology 2002;222:252–260.[Abstract/Free Full Text]
14. Salerno M, Mugler J III, de Lange E, Brookeman J, Altes T. Detection of early smoking related lung diseases with diffusion-weighted hyperpolarized helium-3 MR lung imaging (abstr). In: Radiological Society of North America scientific assembly and annual meeting program. Oak Brook, Ill: Radiological Society of North America, 2003; 526.
15. Crapo RO, Jensen RL, Wanger JS. Single-breath carbon monoxide diffusing capacity. Clin Chest Med 2001;22:637–655.[CrossRef][Medline]
16. Miller A, Thornton JC, Warshaw R, Anderson H, Teirstein AS, Selikoff IJ. Single breath diffusing capacity in a representative sample of the population of Michigan, a large industrial state: predicted values, lower limits of normal, and frequencies of abnormality by smoking history. Am Rev Respir Dis 1983;127:270–277.[Medline]
17. R Development Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2004.
18. Viegi G, Paoletti P, Prediletto R, et al. Carbon monoxide diffusing capacity, other indices of lung function, and respiratory symptoms in a general population sample. Am Rev Respir Dis 1990;141:1033–1039.[Medline]
19. Rizzi M, Sergi M, Andreoli A, Pecis M, Bruschi C, Fanfulla F. Environmental tobacco smoke may induce early lung damage in healthy male adolescents. Chest 2004;125:1387–1393.[Abstract/Free Full Text]
20. Neas LM, Schwartz J. The determinants of pulmonary diffusing capacity in a national sample of U.S. adults. Am J Respir Crit Care Med 1996;153:656–664.
21. Scichilone N, Marchese R, Catalano F, Togias A, Vignola AM, Bellia V. The bronchodilatory effect of deep inspiration diminishes with aging. Respir Med 2004;98:838–843.[CrossRef][Medline]
22. Morrison NJ, Abboud RT, Ramadan F, et al. Comparison of single breath carbon monoxide diffusing capacity and pressure-volume curves in detecting emphysema. Am Rev Respir Dis 1989;139:1179–1187.[Medline]
23. Bankier AA, Madani A, Gevenois PA. CT quantification of pulmonary emphysema: assessment of lung structure and function. Crit Rev Comput Tomogr 2002;43:399–417.[Medline]
24. Park KJ, Bergin CJ, Clausen JL. Quantitation of emphysema with three-dimensional CT densitometry: comparison with two-dimensional analysis, visual emphysema scores, and pulmonary function test results. Radiology 1999;211:541–547.[Abstract/Free Full Text]
25. Salerno M, Brookeman J, de Lange E, Mugler J III. The effect of SNR on helium-3 ADC values in the lung (abstr). In: Proceedings of the eleventh meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2003; 1386.

This article has been cited by other articles:

				Y. Ohno, T. Iwasawa, J. B. Seo, H. Koyama, H. Takahashi, Y.-M. Oh, Y. Nishimura, and K. Sugimura Oxygen-enhanced Magnetic Resonance Imaging versus Computed Tomography: Multicenter Study for Clinical Stage Classification of Smoking-related Chronic Obstructive Pulmonary Disease Am. J. Respir. Crit. Care Med., May 15, 2008; 177(10): 1095 - 1102. [Abstract] [Full Text] [PDF]

				K. Emami, R. V. Cadman, J. M. Woodburn, M. C. Fischer, S. J. Kadlecek, J. Zhu, S. Pickup, R. A. Guyer, M. Law, V. Vahdat, et al. Early changes of lung function and structure in an elastase model of emphysema--a hyperpolarized 3He MRI study J Appl Physiol, March 1, 2008; 104(3): 773 - 786. [Abstract] [Full Text] [PDF]

				Y. Ohno, H. Koyama, M. Nogami, D. Takenaka, S. Matsumoto, M. Obara, and K. Sugimura Dynamic Oxygen-Enhanced MRI Versus Quantitative CT: Pulmonary Functional Loss Assessment and Clinical Stage Classification of Smoking-Related COPD Am. J. Roentgenol., February 1, 2008; 190(2): W93 - W99. [Abstract] [Full Text] [PDF]

				G. Tino, L. B. Ware, and M. Moss Clinical Year in Review IV: Chronic Obstructive Pulmonary Disease, Nonpulmonary Critical Care, Diagnostic Imaging, and Mycobacterial Disease Proceedings of the ATS, September 15, 2007; 4(6): 494 - 498. [Full Text] [PDF]

				S. Verbanck and M. Paiva Simulation of the apparent diffusion of helium-3 in the human acinus J Appl Physiol, July 1, 2007; 103(1): 249 - 254. [Abstract] [Full Text] [PDF]

				K. F. Rabe, B. Beghe, F. Luppi, and L. M. Fabbri Update in Chronic Obstructive Pulmonary Disease 2006 Am. J. Respir. Crit. Care Med., June 15, 2007; 175(12): 1222 - 1232. [Full Text] [PDF]

				S. R. Hopkins, D. L. Levin, K. Emami, S. Kadlecek, J. Yu, M. Ishii, and R. R. Rizi Advances in magnetic resonance imaging of lung physiology J Appl Physiol, March 1, 2007; 102(3): 1244 - 1254. [Abstract] [Full Text] [PDF]

				B. Driehuys and L. W. Hedlund Imaging Techniques for Small Animal Models of Pulmonary Disease: MR Microscopy Toxicol Pathol, January 1, 2007; 35(1): 49 - 58. [Abstract] [PDF]

				E. E. de Lange Science to Practice: What Is New about Detecting Emphysema? Radiology, June 1, 2006; 239(3): 619 - 620. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar