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) All Versions of this Article: 2231010315v1 223/1/189    most recent 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 Avila, N. A. Articles by Moss, J. Search for Related Content PubMed PubMed Citation Articles by Avila, N. A. Articles by Moss, J.

Lymphangioleiomyomatosis: Correlation of Qualitative and Quantitative Thin-Section CT with Pulmonary Function Tests and Assessment of Dependence on Pleurodesis¹

¹ From the Department of Diagnostic Radiology, Warren Grant Magnuson Clinical Center, National Institutes of Health, Bldg 10, Rm 1C-660, 10 Center Dr, MSC 1182, Bethesda, MD 20892-1182 (N.A.A., A.J.D., D.L.J., E.C.J.); and Pulmonary-Critical Care Medicine Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD (J.A.K., J.M.). Received January 15, 2001; revision requested March 5; final revision received September 24; accepted October 26. Address correspondence to N.A.A. (e-mail: navila@nih.gov).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To explore the relationship between findings at thin-section computed tomography (CT) and pulmonary function tests in lymphangioleiomyomatosis (LAM) and to evaluate the influence of pleurodesis on this relation and the effectiveness of quantitative versus qualitative CT in the assessment of disease severity.

MATERIALS AND METHODS: Thirty-seven patients with LAM (17 with pleurodesis) underwent CT and pulmonary function tests. The severity of pulmonary cystic involvement was graded qualitatively by two independent readers and measured quantitatively at CT with a thresholding technique. Relationships between findings at CT and pulmonary function tests and the influence of pleurodesis on these findings were assessed with regression analysis and analysis of covariance.

RESULTS: Qualitative ratings had good agreement between observers ( = 0.75). Quantitative CT had good repeatability and showed significant correlation with the percent predicted forced expiratory volume in 1 second (FEV₁%) (r = 0.67, P < .001), percent predicted diffusing capacity of lung for carbon monoxide (DLCO%) (r = 0.48, P < .005), percent predicted ratio of residual volume to total lung capacity (RV/TLC%) (r = -0.65, P < .001), and percent predicted TLC (r = 0.34, P < .04). Quantitative CT results were somewhat better than qualitative CT results. The standard error of the FEV₁% for the quantitative CT was about 85% of that for the qualitative CT. Pleurodesis had no statistically significant effect on the slope of the regression line between quantitative CT findings, FEV₁%, and DLCO% (corrected for alveolar volume). The slope between quantitative CT and RV/TLC% was significantly (P = .044) more negative in patients with pleurodesis.

CONCLUSION: Qualitative and quantitative CT findings correlate with pulmonary dysfunction over a wide range of disease severity in patients with LAM. Pleurodesis influences the relationship between CT measurements and pulmonary function test results.

© RSNA, 2002

Index terms: Computed tomography (CT), thin-section, 60.12118 • Lung, CT, 60.12111, 60.12118 • Lymphangiomyomatosis, 60.799, 87.829, 99.829

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Lymphangioleiomyomatosis (LAM) is a rare, idiopathic disease affecting women and is characterized by progressive smooth muscle proliferation in the pulmonary lymphatic vessels, blood vessels, and airways (1). The classic chest radiographic manifestations include interstitial lung disease, pneumothoraces, and chylous pleural effusions (2). Hallmarks of LAM are parenchymal cysts, which are best demonstrated at thin-section computed tomography (CT). The cysts are diffusely distributed throughout the lungs and range in diameter from a few millimeters to several centimeters (3–9). Pulmonary dysfunction is a prevalent complication in patients with LAM. Obstruction of air flow is most common. However, restriction or combined restriction and obstruction are also frequently seen, particularly in patients with history of pleurodesis (10). Therefore, it is possible that pleurodesis could influence the relationship between pulmonary function tests and thin-section CT, a relationship described in previous studies (8,11–13), which suggests that qualitative and quantitative analyses of these images correlate with some aspects of pulmonary function.

The purposes were to explore further the relationship between findings at thin-section CT and pulmonary function studies, to evaluate the influence of pleurodesis on this relation, and to evaluate the effectiveness of quantitative versus qualitative CT in the assessment of disease severity.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The study protocol was approved by the National Heart, Lung, and Blood Institute review board. Written informed consent was obtained from all study participants. Thirty-seven consecutive women (age range, 23–53 years; mean age, 41 years) with pulmonary LAM, who were seen at our institution between March 1996 and March 1998, were examined. The diagnosis was established with open lung biopsy (26 patients), transbronchial biopsy (four patients), or retroperitoneal lymph node biopsy (four patients). Three patients did not have biopsy proof, but all had findings characteristic of LAM on CT scans of the chest. In addition, two patients had a history of repeated pneumothoraces (one also had a history of chylothorax and the other, a biopsy-proven renal angiomyolipoma). Of the 37 patients, 17 had undergone pleurodesis: 15 for recurrent pneumothoraces and two for intractable chylous pleural effusions. Every patient underwent thin-section CT of the chest and pulmonary function tests at our institution within 5 days of each other.

Thin-Section CT
Thin-section CT (HiSpeed Advantage; GE Medical Systems, Milwaukee, Wis) of the chest was performed. The scans were obtained with the patients in a prone position during end inspiration at 120–140 kVp and 170–280 mAs, with 1–2-second scanning times and a 34–38-cm field of view. We used 1.0-mm-thick sections at 3-cm intervals, which resulted in a total of eight to nine images per study. A high-spatial-frequency algorithm was used to reconstruct each image. The images were examined for the extent of lung disease and the presence of pleural abnormalities. Pleural abnormalities included pleural plaques (discreet areas of pleural thickening), pneumothorax, pleural effusions, and pleural-based masses.

Qualitative Analysis
The CT scans were interpreted independently by two board-certified radiologists (N.A.A., E.C.J.), who were aware of the patients’ diagnoses but were blinded to their clinical status and the results of pulmonary function tests; the interobserver variability was determined. The lungs were divided into three equal zones (upper, middle, and lower) by dividing the scans into three equal subsets of images. The severity of the involvement with cysts in each zone was graded, according to the percentage of the area judged abnormal, as follows: grade 0, absent; grade 1, less than 30% abnormal (Fig 1a); grade 2, 30%–60% abnormal; grade 3, more than 60% abnormal (Fig 1b). A global score was obtained for each patient by averaging the grades from all three lung zones and rounding them off to the nearest integer; hence, in the overall assessment, each zone was given equal weight rather than an average value weighted by the volume of each zone. These methods are similar to those used by other authors (8,9,13).

View larger version (111K): [in this window] [in a new window] [Download PPT slide]   Figure 1a. Transverse thin-section CT scans (1-mm-thick sections) of the chest obtained in patients in the prone position show different degrees of the severity of lung involvement with pulmonary cysts. Images have been reversed. (a) CT scan in a 44-year-old woman with LAM shows scattered small pulmonary cysts (white arrows) that involve less than 30% (grade 1) of the lung zones and thickening of the pleura (black arrow) on the right side of the thorax secondary to previous pleurodesis. (b) CT scan in a 52-year-old woman with LAM shows pulmonary cysts (arrows) that involve more than 60% (grade 3) of the lung zones.

View larger version (133K): [in this window] [in a new window] [Download PPT slide]   Figure 1b. Transverse thin-section CT scans (1-mm-thick sections) of the chest obtained in patients in the prone position show different degrees of the severity of lung involvement with pulmonary cysts. Images have been reversed. (a) CT scan in a 44-year-old woman with LAM shows scattered small pulmonary cysts (white arrows) that involve less than 30% (grade 1) of the lung zones and thickening of the pleura (black arrow) on the right side of the thorax secondary to previous pleurodesis. (b) CT scan in a 52-year-old woman with LAM shows pulmonary cysts (arrows) that involve more than 60% (grade 3) of the lung zones.

View larger version (148K): [in this window] [in a new window] [Download PPT slide]   Figure 2a. Transverse thin-section CT scans depict the method in which quantitative assessment of the lung involvement was used in a patient with LAM. Images have been reversed. (a) A region of interest is drawn around both lungs, excluding the trachea and major bronchi. (b) Threshold values are used to choose pixels between -1,000 and -300 HU. This provides the total lung volume. (c) Threshold values are then applied to choose pixels that are between -1,000 and -950 HU. This provides the volume of lung covered by pulmonary cysts. Pulmonary cysts (arrows) are labeled in a and c.

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 2b. Transverse thin-section CT scans depict the method in which quantitative assessment of the lung involvement was used in a patient with LAM. Images have been reversed. (a) A region of interest is drawn around both lungs, excluding the trachea and major bronchi. (b) Threshold values are used to choose pixels between -1,000 and -300 HU. This provides the total lung volume. (c) Threshold values are then applied to choose pixels that are between -1,000 and -950 HU. This provides the volume of lung covered by pulmonary cysts. Pulmonary cysts (arrows) are labeled in a and c.

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 2c. Transverse thin-section CT scans depict the method in which quantitative assessment of the lung involvement was used in a patient with LAM. Images have been reversed. (a) A region of interest is drawn around both lungs, excluding the trachea and major bronchi. (b) Threshold values are used to choose pixels between -1,000 and -300 HU. This provides the total lung volume. (c) Threshold values are then applied to choose pixels that are between -1,000 and -950 HU. This provides the volume of lung covered by pulmonary cysts. Pulmonary cysts (arrows) are labeled in a and c.

Pulmonary Function Studies
Spirometry and measurements of functional residual capacity and diffusing capacity of lung for carbon monoxide (DLCO) were performed with the patients in a seated position with use of a computerized system (Gold Standard Plus; Warren E. Collins, Braintree, Mass). Spirometry and DLCO measurements were obtained by using standard techniques recommended by the American Thoracic Society (14,15). Functional residual capacity was measured by using the closed-circuit helium equilibration technique in the standard way. Expiratory reserve volume was subtracted from the functional residual capacity to obtain the residual volume. The vital capacity volume was added to the residual volume to obtain the total lung capacity (TLC). The predicted values for all pulmonary functions were derived from published standards (16–18). Single-breath hold DLCO was measured with the method of Ogilvie et al (19) and corrected for hemoglobin (20). Results are expressed as a percentage of the predicted values by using the reference values of Crapo and Morris (21).

Comparison of Quantitative Analysis of CT and Pulmonary Function Tests in Patients with and Those without Pleurodesis
We correlated the ratio of normal lung volume to total lung volume measured with CT and pulmonary function tests (percent predicted forced expiratory volume in 1 second [FEV₁%], percent predicted DLCO [DLCO%], and DLCO corrected for alveolar volume) for patients with and those without pleurodesis. We studied the effect of pleurodesis on this relationship.

Comparison between Qualitative and Quantitative Data Analyses
The relative abilities of qualitative and quantitative CT to help predict pulmonary function test results were assessed by means of comparison of a scatterplot of qualitative readings by reader 1 versus FEV₁% measurements and a scatterplot of the quantitative ratio of cyst volume to total lung volume versus FEV₁ measurements. Specifically, the spread (the standard errors) of the FEV₁% values above or below the regression line for the quantitative CT was compared with the spread of the FEV₁% values at each of the qualitative CT ratings.

Statistical Methods
The agreement of repeated CT measurements was assessed as described by Bland and Altman (22) and Fleiss (23,24). The degree of agreement between the measurements was assessed in terms of the coefficient of reliability and the SD and the mean absolute value of the differences. The agreement between the qualitative scoring of the CT studies by the two raters was assessed with a statistic. Spearman rank correlation coefficient was used to assess the correlation between the qualitative CT and the pulmonary function test results (25).

The Student t test was used to assess differences in CT and pulmonary function test results between patients with and those without pleurodesis. Linear regression analysis was used to assess the relationships between the quantitative CT assessments and the pulmonary function results. Analysis of covariance was used to assess the statistical significance of the influence of the presence or absence of pleurodesis on the slopes of the regression lines.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Thin-Section CT
Pleural abnormalities were identified on CT scans in all patients (n = 17) with pleurodesis and much less frequently in those (six of 20) without pleurodesis. The abnormalities included pleural plaques (thickening), pneumothoraces, and pleural effusions (Table 1). Some patients had more than one pleural abnormality.

The qualitative extent of pulmonary disease and its frequency distribution in patients with LAM were grade 0 (no patients), grade 1 (10 patients), grade 2 (13 patients), and grade 3 (14 patients). Negative linear correlations were found between qualitative CT scores and FEV₁ ( = -0.55, P < .001) and DLCO ( = -0.68, P < .001).

Quantitative Analysis
The mean total lung volume was 125,393 mm² ± 26,984 (range, 78,156–194,181 mm²), the ratio of the mean cyst volume to total lung volume was 0.1101 ± 0.0580 (range, 0.0029–0.2247), and the ratio of the mean normal lung volume to total lung volume was 0.8899 ± 0.0580 (range, 0.729–0.997) (Table 2).

Pulmonary Function Tests
The patients with pleurodesis had lower mean pulmonary function test results (FEV₁%, percent predicted TLC, percent predicted ratio of residual volume to TLC [RV/TLC%], and DLCO%) than the patients without pleurodesis (Tables 2–5). However, only the results for FEV₁% were statistically significant at P .05 (Student t test, P = .007).

Figures 3–6 illustrate the influence of the absence or presence of pleurodesis on the relationships of the pulmonary function test results to the ratio of normal lung volume to total lung volume. As shown in Figure 3, the absence or presence of pleurodesis had no significant effect on the relationship between the ratio of normal lung volume to total lung volume and the FEV₁ (analysis of covariance, P = .9); in both situations, the linear regression lines had similar slopes. The intercept for the regression line in the presence of pleurodesis was slightly lower than that in the absence of pleurodesis. This shift in the regression lines is consistent with pleurodesis, which reduces FEV₁, but the difference was not statistically significant, P = .9. However, the absence or presence of pleurodesis did influence the slopes of the regression lines for RV/TLC% and the DLCO% at about the P = .05 level (Figs 4–6).

View larger version (21K): [in this window] [in a new window] [Download PPT slide]   Figure 3a. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus FEV1%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. Both groups of patients exhibited statistically significant positive correlations between the ratio of normal lung volume to total lung volume and the FEV1% (r = 0.59, P < .005 for patients without pleurodesis; r = 0.69, P < .002 for patients with pleurodesis). The presence or absence of pleurodesis had no statistically significant effect on the slope between the ratio of normal lung volume to total lung volume and the FEV1% (analysis of covariance, P = .9).

View larger version (19K): [in this window] [in a new window] [Download PPT slide]   Figure 3b. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus FEV1%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. Both groups of patients exhibited statistically significant positive correlations between the ratio of normal lung volume to total lung volume and the FEV1% (r = 0.59, P < .005 for patients without pleurodesis; r = 0.69, P < .002 for patients with pleurodesis). The presence or absence of pleurodesis had no statistically significant effect on the slope between the ratio of normal lung volume to total lung volume and the FEV1% (analysis of covariance, P = .9).

View larger version (16K): [in this window] [in a new window] [Download PPT slide]   Figure 3c. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus FEV1%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. Both groups of patients exhibited statistically significant positive correlations between the ratio of normal lung volume to total lung volume and the FEV1% (r = 0.59, P < .005 for patients without pleurodesis; r = 0.69, P < .002 for patients with pleurodesis). The presence or absence of pleurodesis had no statistically significant effect on the slope between the ratio of normal lung volume to total lung volume and the FEV1% (analysis of covariance, P = .9).

View larger version (22K): [in this window] [in a new window] [Download PPT slide]   Figure 4a. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (21K): [in this window] [in a new window] [Download PPT slide]   Figure 4b. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 4c. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (22K): [in this window] [in a new window] [Download PPT slide]   Figure 5a. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (21K): [in this window] [in a new window] [Download PPT slide]   Figure 5b. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (16K): [in this window] [in a new window] [Download PPT slide]   Figure 5c. Scatterplots and regression analysis of quantitative CT ratio of normal lung volume to total lung volume versus DLCO%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for the patients without (solid line) and those with (broken line) pleurodesis. A positive correlation observed between the ratio of normal lung volume to total lung volume and the DLCO% in patients without pleurodesis (slope = 284, r = 0.66, P = .001) was reduced in patients with pleurodesis (slope = 50, r = 0.4, P = .6); the reduction in slope of the regression line was significant at P = .058.

View larger version (23K): [in this window] [in a new window] [Download PPT slide]   Figure 6a. Scatterplots and regression analysis of quantitative CT measurements of ratios of normal lung volume to total lung volume versus RV/TLC%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for patients without (solid line) and those with (broken line) pleurodesis. A negative linear regression in the patients without pleurodesis (slope = -156 , r = -0.35, P = .13) was more negative in the patients with pleurodesis (slope = -457, r = -0.75, P < .001), a change in slope of the regression line was significant at P = .044 (analysis of covariance).

View larger version (21K): [in this window] [in a new window] [Download PPT slide]   Figure 6b. Scatterplots and regression analysis of quantitative CT measurements of ratios of normal lung volume to total lung volume versus RV/TLC%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for patients without (solid line) and those with (broken line) pleurodesis. A negative linear regression in the patients without pleurodesis (slope = -156 , r = -0.35, P = .13) was more negative in the patients with pleurodesis (slope = -457, r = -0.75, P < .001), a change in slope of the regression line was significant at P = .044 (analysis of covariance).

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 6c. Scatterplots and regression analysis of quantitative CT measurements of ratios of normal lung volume to total lung volume versus RV/TLC%. (a) Scatterplot of the data and regression line for the 20 cases without pleurodesis. (b) Scatterplot of the data and regression line for the 17 patients with pleurodesis. (c) Regression lines for patients without (solid line) and those with (broken line) pleurodesis. A negative linear regression in the patients without pleurodesis (slope = -156 , r = -0.35, P = .13) was more negative in the patients with pleurodesis (slope = -457, r = -0.75, P < .001), a change in slope of the regression line was significant at P = .044 (analysis of covariance).

Comparison between Qualitative and Quantitative Data Analyses
There was a statistically significant correlation (P < .01) between the qualitative ratings and the quantitative measurements for each reader ( = 0.55 for reader 1, = 0.58 for reader 2). However, the scatterplots showed overlap among the quantitative measurements of the three categories of the extent of disease (Fig 7).

View larger version (21K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Scatterplot of the qualitative subjective CT gradings versus the quantitative ratio of cyst volume to total lung volume. Ratings of each of the two readers are graphed side by side in each rating category. There is a positive trend for both readers; cases with higher ratings tend to have greater ratio of cyst volume to total lung volume ( = 0.55 for reader 1, = 0.59 for reader 2).

View larger version (15K): [in this window] [in a new window] [Download PPT slide]   Figure 8a. Comparison of qualitative gradings and quantitative CT measurements in predicting pulmonary function. (a) Scatterplot of the reader 1 qualitative ratings versus FEV1%. (b) Scatterplot and linear regression line of the ratio of quantitative cyst volume to total lung volume versus FEV1%.

View larger version (20K): [in this window] [in a new window] [Download PPT slide]   Figure 8b. Comparison of qualitative gradings and quantitative CT measurements in predicting pulmonary function. (a) Scatterplot of the reader 1 qualitative ratings versus FEV1%. (b) Scatterplot and linear regression line of the ratio of quantitative cyst volume to total lung volume versus FEV1%.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Aberle et al (9) applied a subjective scoring method based on visual assessment of cystic replacement of lung parenchyma and found good correlation with measures of airflow limitation and diffusing capacity in eight patients. Using a similar method in 14 patients, Muller et al (8) showed a correlation with DLCO, although not with other pulmonary function measurements. Our qualitative CT analysis demonstrated similar positive correlation between the grade of lung severity and the decrease in FEV₁%. Our qualitative analysis demonstrated overlap of the FEV₁% measurements between adjacent severity categories, a consequence in part of the limited number of severity categories used. Also, the precision of the estimates of pulmonary function values from the qualitative assessment of lung severity was somewhat lower than that in the estimates from the quantitative assessment; the standard error of the latter was about 85% of the former.

The quantitative thin-section CT results of this study confirm and extend the published results. In 10 patients with LAM, Crausman et al (12) applied a quantitative method of thin-section CT analysis and found correlation between the CT findings and pulmonary function test results. We also found good correlation between quantitative CT measurements of the disease extent—across a spectrum from mild to severe pulmonary disease in 37 patients—and measurements of air flow (FEV₁%), diffusing capacity (DLCO%), and air trapping (RV/TLC%). However, unlike Crausman et al, we also found good correlation between quantitative CT measurements of the total lung volume and TLC. This disparity may result from the larger number of patients in our study or differences in techniques. In the study of Crausman et al, only two CT images were obtained through the lungs. In our study, images were obtained every 3 cm through the lungs; the total number varied with the height of the lungs. Thus, our method provides a more complete assessment of the lung cross-sections; hence, the sum of the cross-sectional lung areas on the CT images in our study is more reflective of the total lung volume than are the findings in the study of Crausman et al. Also in our study, patients underwent imaging at end inspiration, whereas in the other study, patients underwent imaging at end expiration. Since TLC is a measure of the total volume of air in the chest at end inspiration, end-inspiration CT measurements are likely more reflective of TLC than are the end-expiration CT measurements.

Because pleural disease is well known to cause restrictive lung disease with changes of pulmonary function (TLC, RV/TLC, and DLCO), in this study, we sought to determine the influence of pleurodesis on the relationship between pulmonary function tests and quantitative CT (27,28). On average, patients with pleurodesis had worse pulmonary function than those without pleurodesis. Moreover, analysis of covariance of the CT measurements and pulmonary function test results demonstrated that the presence of pleurodesis influenced the relationship between the ratio of total cyst volume to total lung volume. This observation points out the value of multivariate methods, such as analysis of covariance that include the influence of additional factors in the analysis of the relationship of pulmonary structure—as reflected on CT scans—with pulmonary function.

The influence of pleurodesis on the relationship between CT normal and/or total measurements and pulmonary function (reflected in the change of the slope of the regression line) varied with the pulmonary function test. No change was seen with FEV₁%. With the RV/TLC%, the relationship was stronger and the slope more negative in patients with pleurodesis. With DLCO, the relationship was stronger and the slope more positive in patients without pleurodesis. However, when DLCO was corrected for alveolar volume, no statistically significant difference in slopes was observed between the patients with and those without pleurodesis.

The observation that uncorrected DLCO% is correlated with pleurodesis in our patient population is not surprising. Pleurodesis is a well-known cause of restrictive ventilatory impairment, with reduction in lung volumes (TLC) and increase in RV/TLC% (10). DLCO% is volume dependent, and in patients with pleural disease, it is reduced in accordance with the reduction in TLC. However, this is not the case when DLCO% is corrected for alveolar volume. For example, a reduction in TLC without change in DLCO corrected for alveolar volume has been shown in an experimental animal model by using unilateral pleural symphysis and in patients with asbestos-induced benign pleural conditions (29,30). Moreover, other authors have demonstrated elevated ratio of maximum diffusing capacity to alveolar gas volume in patients with pleural disease (31). In concordance with those findings, we found that—when corrected for alveolar volume—there was no significant difference in the relationship between DLCO% and the ratio of normal lung volume to total lung volume in patients with and those without pleurodesis. This underscores the importance of correcting for alveolar volume when DLCO% is measured in patients with LAM, particularly those with pleurodesis.

CT scans provide a detailed visual depiction of the structural abnormalities that underlie the global limitations of air flow, lung volumes, and gas exchange measured with pulmonary function tests. More detailed analysis of the CT features, such as the distribution of cyst sizes and the assessment of pulmonary compliance by comparing ratios of inspiratory cyst volume to total lung volume, allows refined understanding of the structure and function correlations and regional analysis of lung function. For example, Brown et al (32) used the knowledge-based segmentation method to derive indirect quantitative measures of single lung function. Similarly, Goldin et al (33) used functional thin-section CT to assess regional air trapping as an indirect measure of small airways obstruction and in regional hyperreactivity. Thus, CT may have a role in following disease progression and monitoring response to therapy.

Both qualitative and quantitative CT provide measures of abnormal lung structure in LAM that correlate directly with pulmonary function studies; the presence or absence of pleurodesis influences these relationships. Quantitative CT is more precise than qualitative CT. However, a current major drawback of quantitative CT, which may potentially be remedied with automated analysis, is the time and effort needed to obtain and analyze the data. Subjective qualitative assessment of the extent of lung abnormality on CT scans is a less burdensome alternative. Its adequacy as a substitute for the quantitative method depends on the acceptability of the tradeoff between precision and/or accuracy and time and/or effort. For imaging research and other applications that require maximal precision and a detailed interval numeric scale, the quantitative method appears warranted. However, for clinical purposes, it is unclear that the apparently modest increase in precision justifies the added effort associated with quantitative CT. As CT volumetric techniques become more sophisticated and practical, they may eventually complement routine pulmonary function tests.

	FOOTNOTES

 
Abbreviations: DLCO% = percent predicted diffusing capacity of lung for carbon monoxide, FEV₁% = percent predicted forced expiratory volume in 1 second, LAM = lymphangioleiomyomatosis, RV/TLC% = percent predicted ratio of residual volume to TLC, TLC = total lung capacity

Author contributions: Guarantor of integrity of entire study, N.A.A.; study concepts, N.A.A., J.A.K.; study design, N.A.A.; literature research, N.A.A., J.A.K.; clinical studies, J.A.K., J.M.; data acquisition, N.A.A., J.A.K., A.J.D., E.C.J., D.L.J.; data analysis/interpretation, N.A.A., E.C.J., J.A.K.; statistical analysis, A.J.D.; manuscript preparation and definition of intellectual content, N.A.A., A.J.D., J.A.K.; manuscript editing and revision/review, all authors; manuscript final version approval, N.A.A., J.A.K., A.J.D.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Corrin B, Liebow AA, Friedman PJ. Pulmonary lymphangiomyomatosis: a review. Am J Pathol 1975; 79:348-367.
2. Carrington CB, Cugell DW, Gaensler EA, et al. Lymphangiomyomatosis: physiologic-pathologic-radiologic correlations. Am Rev Respir Dis 1977; 116:977-995.
3. Berger JL, Shaff MI. Pulmonary lymphangioleiomyomatosis. J Comput Assist Tomogr 1981; 5:565-567.
4. Merchant RN, Pearson MG, Rankin RN, Morgan WKC. Computerized tomography in the diagnosis of lymphangioleiomyomatosis. Am Rev Respir Dis 1985; 131:295-297.
5. Templeton PA, McLoud TC, Muller NL, Shepard JO, Moore EH. Pulmonary lymphangioleiomyomatosis: CT and pathologic findings. J Comput Assist Tomogr 1989; 13:54-57.
6. Rappaport DC, Weisbrod GL, Herman SJ, Chamberlain DW. Pulmonary lymphangioleiomyomatosis: high-resolution CT findings in four cases. AJR Am J Roentgenol 1989; 152:961-964.
7. Sherrier RH, Chiles C, Roggli V. Pulmonary lymphangioleiomyomatosis: CT findings. AJR Am J Roentgenol 1989; 153:937-940.
8. Muller NL, Chiles C, Kullnig P. Pulmonary lymphangiomyomatosis: correlation of CT with radiographic and functional findings. Radiology 1990; 175:335-339.
9. Aberle DR, Hansell DM, Brown K, Tashkin DP. Lymphangiomyomatosis: CT, chest radiographic, and functional correlations. Radiology 1990; 176:381-387.
10. Chu SC, Horiba K, Usuki J, et al. Comprehensive evaluation of 35 patients with lymphangioleiomyomatosis. Chest 1999; 115:1041-1052.
11. Lenoir S, Grenier P, Brauner MW, et al. Pulmonary lymphangiomyomatosis and tuberous sclerosis: comparison of radiographic and thin-section CT findings. Radiology 1990; 175:329-334.
12. Crausman RS, Lynch DA, Mortenson RL, et al. Quantitative CT predicts the severity of physiologic dysfunction in patients with lymphangioleiomyomatosis. Chest 1996; 109:131-137.
13. Avila NA, Chen CC, Chu SC, et al. Pulmonary lymphangioleiomyomatosis: correlation of ventilation-perfusion scintigraphy, chest radiography, and CT with pulmonary function tests. Radiology 2000; 214:441-446.
14. American Thoracic Society. Standardization of Spirometry, 1994 Update. Am J Respir Crit Care Med 1995; 152:1107-1136.
15. American Thoracic Society. Single-breath carbon monoxide diffusing capacity (transfer factor). Recommendations for a standard technique—1995 update. Am J Respir Crit Care Med 1995; 152:2185-2198.
16. Morris JF, Koski A, Johnson LC. Spirometric standards for healthy nonsmoking adults. Am Rev Respir Dis 1971; 103:57-67.
17. Goldman HI, Becklake MR. Respiratory function tests: normal values at median altitudes and the prediction of normal results. Am Rev Tuberc Pulm Dis 1959; 79:457-467.
18. Gaensler EA, Wright GW. Evaluation of respiratory impairment. Arch Environ Health 1966; 12:146-189.
19. Ogilvie C, Foster R, Blakemore W, et al. A standardized breathholding technique for the clinical measurement of diffusing capacity of the lung for carbon monoxide. J Clin Invest 1957; 36:1-17.
20. Dinakara P, Blumenthal W, Johnston R, et al. The effect of anemia on pulmonary diffusing capacity with derivation of correction equation. Am Rev Respir Dis 1970; 102:965-969.
21. Crapo RO, Morris AH. Standardized single breath values for carbon monoxide diffusing capacity. Am Rev Respir Dis 1981; 123:185-189.
22. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307-310.
23. Fleiss JL. Reliability of measurements: The design and analysis of clinical experiments New York, NY: Wiley, 1986; 8-12.
24. Fleiss JL. Statistical methods for rates and proportions 2nd ed. New York, NY: Wiley, 1981.
25. Siegel S, Castellan NJ. Nonparametric statistics for the behavioral sciences 2nd ed. New York, NY: McGraw-Hill, 1988.
26. Kitaichi M, Nishimura K, Itoh H, et al. Pulmonary lymphangioleiomyomatosis: a report of 46 patients including a clinicopathologic study of prognostic features. Am J Respir Crit Care Med 1995; 151:527-533.
27. Colp C, Reichel J, Park SS. Severe pleural restriction: static pulmonary recoil pressure as an aid in diagnosis. Chest 1975; 67:658-664.
28. Yoo OH, Ting EY. The effects of pleural effusion on pulmonary function. Am Rev Respir Dis 1964; 89:55-63.
29. DesChamps C, Rodarte JR. Effects of unilateral pleural symphysis on respiratory system mechanics and gas exchange in anesthetized dogs. Am Rev Respir Dis 1988; 137:1385-1389.
30. Jones RN, McLoud T, Rockoff SD. The radiographic pleural abnormalities in asbestos exposure: relationship to physiologic abnormalities. J Thorac Imaging 1988; 3:57-66.
31. Baylor P, Goebal P. Clinical correlates of an elevated diffusing capacity for carbon monoxide corrected for alveolar volume. Am J Med Sci 1996; 311:266-271.
32. Brown MS, Goldin JG, McNitt-Gray MF, et al. Knowledge-based segmentation of thoracic computed tomography images for assessment of split lung function. Med Phys 2000; 27:592-598.
33. Goldin JG, Tashkin DP, Kleerup EC, et al. Comparative effects of hydrofluoroalkane and chlorofluorocarbon beclomethasone dipropionate inhalation on small airways: assessment with functional helical thin-section computed tomography. J Allergy Clin Immunol 1999; 104:S258-S267.

This article has been cited by other articles:

				W. K. Steagall, C. G. Glasgow, O. M. Hathaway, N. A. Avila, A. M. Taveira-DaSilva, A. Rabel, M. P. Stylianou, J.-P. Lin, X. Chen, and J. Moss Genetic and morphologic determinants of pneumothorax in lymphangioleiomyomatosis Am J Physiol Lung Cell Mol Physiol, September 1, 2007; 293(3): L800 - L808. [Abstract] [Full Text] [PDF]

				N. A. Avila, A. J. Dwyer, A. Rabel, and J. Moss Sporadic Lymphangioleiomyomatosis and Tuberous Sclerosis Complex with Lymphangioleiomyomatosis: Comparison of CT Features Radiology, December 1, 2006; 242(1): 277 - 285. [Abstract] [Full Text] [PDF]

				N. A. Avila, A. J. Dwyer, A. Rabel, R. M. DeCastro, and J. Moss CT of Pleural Abnormalities in Lymphangioleiomyomatosis and Comparison of Pleural Findings After Different Types of Pleurodesis. Am. J. Roentgenol., April 1, 2006; 186(4): 1007 - 1012. [Abstract] [Full Text] [PDF]

				G. F. Abbott, M. L. Rosado-de-Christenson, A. A. Frazier, T. J. Franks, R. D. Pugatch, and J. R. Galvin From the Archives of the AFIP: Lymphangioleiomyomatosis: Radiologic-Pathologic Correlation RadioGraphics, May 1, 2005; 25(3): 803 - 828. [Abstract] [Full Text] [PDF]

				G. Paciocco, E. Uslenghi, A. Bianchi, G. Mazzarella, G. C. Roviaro, G. Vecchi, and S. Harari Diffuse Cystic Lung Diseases: Correlation Between Radiologic and Functional Status Chest, January 1, 2004; 125(1): 135 - 142. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2231010315v1 223/1/189    most recent 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