| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Thoracic Imaging |
1 From the Department of Radiology, King’s College Hospital, Denmark Hill, London SE5 9RS, England (S.R.D.); Interstitial Lung Disease Unit (S.V., A.N., N.S.L.G., R.M.d.B., A.U.W.) and Departments of Radiology (D.M.H.) and Pathology (A.G.N.), Royal Brompton Hospital, London, England; Department of Laboratory Medicine and Pathology, Mayo Clinic, Scottsdale, Ariz (T.V.C.); and Centre for Rheumatology, Royal Free and University College Medical School, London, England (C.P.D., C.M.B.). Received August 1, 2003; revision requested October 14; revision received November 24; accepted January 5, 2004. Address correspondence to S.R.D. (e-mail: sujal.desai@kingsch.nhs.uk).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The CT features of consecutive patients with SSc (n = 225; male patients, 44; female patients, 181; median age, 47 years; age range, 16–78 years), IPF (n = 40; men, 26; women, 14; median age, 54.5 years; age range, 36–77 years) and NSIP (n = 27; men, 18; women, nine; median age, 53 years; age range, 32–68 years) were quantified separately by two observers. The extent of interstitial lung disease, ground-glass opacification, emphysema, and the coarseness of a reticular pattern were quantified. Group comparisons were made nonparametrically with the Wilcoxon rank sum test. Differences in CT features were identified with multiple logistic regression analysis.
RESULTS: The coarseness of fibrosis was similar in patients with SSc and idiopathic NSIP but strikingly different between patients with SSc (median coarseness score, 5.5; range, 0.0–13.3) and IPF (median coarseness score, 8.8; range, 2.5–15.0) (P < .001). The proportion of ground-glass opacification at CT was similar in patients with SSc and idiopathic NSIP but differed significantly between patients with SSc (median proportion, 49.9%; range, 0.0%–100.0%) and IPF (median proportion, 23.5%; range, 0.0%–97.2%) (P < .001). At logistic regression analysis, there were no differences in the CT features between patients with SSc and those with NSIP after controlling for age, disease extent, and the percentage predicted forced vital capacity and carbon monoxide diffusing capacity.
CONCLUSION: Interstitial lung disease in patients with SSc is less extensive, less coarse, and characterized by a greater proportion of ground-glass opacification than that in patients with IPF. The CT features of lung disease in patients with SSc closely resemble those in patients with idiopathic NSIP.
© RSNA, 2004
Index terms: Lung, CT, 68.1211 • Lung, diseases, 60.6113, 60.613, 60.917 • Lung, fibrosis, 60.6113, 60.613 • Lung, interstitial disease, 60.613, 60.917 • Scleroderma, 60.613 • Pneumonia, usual interstitial, 60.6113 • Pneumonia, nonschedule interstitial and fibrosis, 60.6113
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
Histopathologic Evaluation
Specimens obtained with open or video-assisted lung biopsy were reviewed separately by two thoracic pathologists (T.V.C. and A.G.N., both core pathologists for the American Thoracic Society/European Respiratory Society sponsored committee for classification of the idiopathic interstitial pneumonias, with 27 and 8 years of experience, respectively, in pulmonary pathology) who were blinded to clinical data; disagreements were resolved with joint review. All biopsies were performed for clinical reasons at the request of the referring clinician. Histopathologic specimens were routinely stained with hematoxylin-eosin; when necessary, additional staining was performed with elastic van Gieson stains. IPF was diagnosed with histopathologic analysis if there was evidence of temporal heterogeneity with nonuniformity or variability in interstitial changes (the latter comprising intermingled zones of established interstitial fibrosis, fibroblastic foci, inflammation, and honeycombing merging with regions of normal lung parenchyma, all coexisting in variable proportions) (4,5). NSIP was diagnosed if there was apparent temporal uniformity in regions of patchy or diffuse fibrosis and inflammation (4,5).
CT Scanning Technique
All patients underwent CT (Imatron, San Francisco, Calif) with a collimation of 1.5–3.0 mm and an interspace of 10 mm. Images were reconstructed with a high-spatial-frequency algorithm and photographed at window settings appropriate for viewing lung parenchyma (window center, –550 HU; window width, 1,500 HU).
Quantification of CT Patterns
CT scans were reviewed separately by two observers (S.R.D. and A.U.W.) who were blinded to clinical and histopathologic data. Scans were reviewed at the following five levels: 1, origin of great vessels; 2, carina; 3, pulmonary venous confluence; 4, between levels 3 and 5; and 5, 1 cm above the right hemidiaphragm.
The following features were quantified at each level. First, the overall extent of interstitial lung disease, both reticular pattern and ground-glass opacification, was estimated to the nearest 5%. Second, the relative proportions of interstitial lung disease (up to a total of 100%) made up of a reticular pattern and ground-glass opacification were quantified. A reticular pattern was defined as innumerable interlacing line shadows that were fine, intermediate, or coarse, with associated distortion of the lung architecture. Ground-glass opacification was defined as a hazy increase in lung parenchymal attenuation, with preservation of bronchial and vascular markings (16). Third, the coarseness of fibrosis was quantified semiquantitatively as follows: 0, ground-glass opacification alone; 1, fine intralobular fibrosis; 2, microcystic reticular pattern comprising air spaces smaller than or equal to 4 mm in diameter; and 3, a macrocystic reticular pattern comprising air spaces larger than 4 mm in diameter. Fourth, the extent of emphysema, which was defined as areas of decreased attenuation, usually without discrete walls and of nonuniform distribution causing permeative destruction of lung parenchyma (16), was quantified to the nearest 5%. Fifth, observers were asked to review all CT scans and record an overall grade of interstitial lung disease depending on the relative extent of a reticular pattern (denoting established interstitial fibrosis [17]) or ground-glass opacification (indicating an increased likelihood of inflammatory infiltration [18]) on the whole scan as follows: grade 1, predominant ground-glass opacification; grade 2, equal proportions of ground-glass opacification and a reticular pattern; and grade 3, predominant reticular pattern. This method of quantification has clinically acceptable observer variation and has been reported to facilitate prediction of the findings at histopathologic examination (19,20).
Discrepancies in the extent of interstitial lung disease, the relative percentages of a reticular pattern, and ground-glass opacification and emphysema of more than 20% at any level were reviewed jointly and resolved with consensus evaluation. Similarly, a difference of more than one grade at each CT level for the coarseness score and the overall grade of fibrosis were resolved by means of joint review. For each patient, the extent of interstitial lung disease and emphysema was derived by averaging the scores at each level, as assessed by the two observers; the mean value was used in analysis. The proportion of ground-glass opacification was computed by dividing the extent of ground-glass opacification by the extent of interstitial lung disease (extent of ground-glass opacification was equal to the percentage of ground-glass opacification multiplied by the total exent of interstitial lung disease).
The overall coarseness score for each patient was derived by summing the scores at five levels (minimum score, 0; maximum score, 15). To prevent spurious reductions in the coarseness score, it was adjusted proportionately to a five-level score in patients with no disease in one or more CT sections; thus, in a patient with no disease in one section, a coarseness score of 8 would be adjusted by 5/4 and would result in a coarseness score of 10.
Pulmonary Function Tests
In all patients, pulmonary function tests were performed within 1 month of CT. The percent predicted forced vital capacity and the single-breath carbon monoxide diffusing capacity, corrected for hemoglobin concentration, were recorded.
Statistical Analysis
Data are expressed as median values with ranges. A P value of less than .05 indicated statistical significance. A weighted 
coefficient of agreement (w) was used for ranked categorical data to quantify observer variation (21). A value of more than 0.4 was considered to represent moderate but clinically acceptable observer agreement, and a value of more than 0.6 was considered to represent good observer agreement (22). Observer variation for continuous variables was expressed as the single determination SD (23). Group comparisons were made by using the Wilcoxon rank sum test. In preliminary analyses, multiple linear regression models were used to determine whether age, sex, and smoking status were determinants of the coarseness of fibrosis and the proportion of ground-glass opacification. Logistic regression analysis was used to determine the factors that influence CT features in patients with SSc and those with IPF and similarly, in patients with SSc and those with NSIP.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
w = 0.88) and the overall grade (ie, predominant reticular pattern vs predominant ground-glass opacification vs a mixed pattern; w = 0.74) were good.
|
|
|
|
2 analysis was used to quantify the trend).
|
|
Because there was a significant (negative) correlation between the coarseness of fibrosis and the proportion of ground-glass opacification (Spearman correlation coefficient, rs = –0.67; P < .001), the following multivariate analyses, which are used to evaluate differences in coarseness and the proportion of ground-glass opacification between disease subgroups, were considered in separate regression models. At logistic regression analysis, after controlling for disease extent and age, there were no differences between patients with SSc and those with idiopathic NSIP for either the coarseness of fibrosis or the proportion of ground-glass opacification. Contrary to these findings, when compared with SSc after controlling for disease extent and age, IPF was characterized by a lower proportion of ground-glass opacification (odds ratio, 0.97; 95% CI: 0.95, 0.99; P < .001) and, in a separate multivariate model, coarser fibrosis (odds ratio, 1.33; 95% CI: 1.14, 1.55; P < .001).
Comparisons between Subgroups Matched for Disease Extent on CT Scans
To exclude possible systematic bias due to differences in disease extent, subgroup analyses were performed in which CT variables were compared between patients with IPF, idiopathic NSIP, and an equivalent number of patients with SSc (n = 67) with the most extensive disease. The CT features and pulmonary function abnormalities in this cohort are given in Table 4. There was no significant difference in the median extent of interstitial lung disease between the combined group of patients with IPF and idiopathic NSIP (median extent, 32%; range, 3.0%–91.5%) and patients with SSc (median extent, 30.5%; range, 20.5%–84.0%). The coarseness of fibrosis and the proportion of ground-glass opacification did not differ significantly between patients with SSc and those with idiopathic NSIP (Fig 3). Compared with IPF, however, fibrosis was less coarse (P < .001), and there was a higher proportion of ground-glass opacification (P < .001) in patients with SSc.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The view that the histopathologic changes in patients with SSc and IPF are indistinguishable is no longer tenable. At microscopy, NSIP is increasingly regarded as the more prevalent histologic pattern in patients with SSc (6–8), as well as in patients with other connective tissue diseases (28–30). However, surgical biopsy is performed infrequently in patients with SSc; therefore, histopathologic series apply to a minority of patients and are subject to major selection biases. Moreover, there is considerable variation in the reported prevalence of NSIP and UIP in patients with SSc. In the first report, UIP was seen in nearly half the patients (6); however, in a subsequent larger cohort, UIP was rare and occurred in only 8% of the patients, whereas NSIP was present in 78% (8). The variation in the reported prevalence of NSIP may reflect low patient numbers (6,7), selection biases in patients who underwent biopsy, observer variation between histopathologists, and differently applied diagnostic criteria (8,10,11).
Thus, current histopathologic series are imprecise with regard to the relative frequency of UIP and NSIP in the larger subgroup of patients with SSc who did not undergo biopsy. Because of this uncertainty, we evaluated CT appearances in a large cohort of consecutive patients, as CT is routinely performed at our institution in patients with SSc in whom pulmonary involvement is suspected. Furthermore, the CT appearances of IPF (ie, idiopathic UIP) and idiopathic NSIP are now well documented (9,12,17); therefore, it is valid to make comparisons between these entities on the basis of CT appearances. Our results indicate that the high prevalence of NSIP and the low prevalence of UIP reported by Bouros et al (8) can be extrapolated to the wider population of patients with SSc who did not undergo biopsy. In essence, the CT features, specifically the coarseness of pulmonary fibrosis and the proportion of ground-glass opacification, were virtually identical in patients with SSc and those with idiopathic NSIP. When compared with patients with IPF, however, fibrosis in patients with SSc was less coarse, and there was a greater proportion of ground-glass opacification.
Our findings were not materially influenced by substantial disparities in the extent of lung diseases between the SSc and idiopathic subgroups. Because interstitial involvement occurs so commonly, clinicians tend to evaluate the lungs at an early stage of SSc and detect pulmonary fibrosis at a less extensive stage (31) than in patients with idiopathic disease, as was the case in the present study. Thus, in theory, our findings might merely have represented earlier morphologic evaluation of patients with SSc rather than disease-specific differences. The possibility of such a bias prompted us to evaluate a large subgroup of patients with SSc who had the most extensive lung involvement (virtually identical to that in the idiopathic subgroups), thereby removing disease extent as a potential major confounding variable. After adjustment for functional severity, another potential confounding factor—the range of CT appearances in patients with SSc—remained identical to the CT appearance in patients with idiopathic NSIP and strikingly different from that in patients with IPF.
The presence of coarser fibrosis and a lower proportion of ground-glass opacification in roughly one-third of patients with SSc in the present study was more suggestive of a diagnosis of IPF. This observation is not, however, in itself indicative of a 30% prevalence of UIP in patients with SSc, as the same overlap was apparent when CT features were compared between the idiopathic UIP and NSIP subgroups; it appears that microcystic disease occurs in a subset of patients with NSIP. The more compelling observation was the identical population distribution of CT abnormalities when SSc and biopsy-proved idiopathic NSIP were compared. Thus, our findings strongly suggest that the 5%–10% prevalence of UIP reported in the largest histopathologic SSc series (8) of which we are aware also applies to the larger population of patients with SSc who do not undergo surgical biopsy.
The CT features of IPF and the fibrosing alveolitis of SSc have been the subject of previous studies (32–35). In an earlier series, there was little difference in the extent of honeycombing (likely to reflect the coarseness score in the present study), and surprisingly, the extent of ground-glass opacification was higher in patients with IPF than in those with SSc (32). However, the number of patients with IPF and SSc in the earlier study was small (nine and 14 patients, respectively); furthermore, the authors had no biopsy proof of IPF in four of nine patients. In a larger study, Chan et al (34) compared the CT features of IPF and SSc. The major difference to emerge was the greater proportion of fibrosis in the upper zones in patients with IPF. Unlike the present study, differences in the proportion of ground-glass opacification were not evaluated. However, the study of Chan et al (34) was undoubtedly confounded by selection bias and classification issues. The authors excluded patients with predominant ground-glass opacification at CT from analysis (34). Furthermore, the IPF population almost certainly comprised a mixture of patients with IPF and NSIP (36); thus, differences at CT, specifically with regard to the proportion of ground-glass opacification and the coarseness of fibrosis, may have been obscured.
There is a broad consensus that patients with idiopathic NSIP survive longer than those with IPF (36–39). Whether the better outcome of patients with SSc can be attributed wholly to the presence of an NSIP pattern at biopsy, however, is debatable (8). Indeed, when evaluating the influence of an NSIP pattern at biopsy on prognosis, the practice of lumping together variable histopathologic characteristics to describe the overall severity of lung disease may mask true relationships. In subjects with IPF for example, only some histopathologic features—namely, the extent of alveolar granulation tissue and interstitial young connective tissue—have a bearing on survival (40). In the recent series of 80 patients with fibrosing alveolitis secondary to SSc, Bouros et al (8) found that there was no survival difference between patients with SSc and NSIP and those with UIP/end-stage lung disease. It should be noted, however, that the number of patients with UIP and end-stage lung disease in this study was small (six in each group), and it is possible that there would be a difference in outcome if a larger population had been evaluated.
The limitations of this study include the retrospective design and the possibility of selection bias at the time of referral to our institution. However, neither is considered likely to have distorted our findings to a major degree. Pulmonary function tests and CT are performed with a prospective protocol in all patients with SSc (and, indeed, all patients with IPF) at the time of presentation to our unit. Thus, there is no apparent bias as a result of variability in clinical evaluation. Because of the pattern of tertiary referral pattern to our institution, there is also a potential for patients with SSc with unusually severe or atypical disease to have been included in the study. This source of selection bias is minimized, however, by the fact that the majority of patients were referred from a large rheumatology center, where it is the usual practice to send all patients with SSc to our unit for evaluation of suspected or established interstitial lung disease. Furthermore, our patients with SSc exhibited a wide range of disease severity; both minimally extensive and normal thin-section CT appearances were well represented in the referred population. Thus, we believe that our findings are applicable to patients with SSc that may be encountered in routine clinical practice.
In summary, despite previous reports about the similarity of histopathologic features, the present study has shown that in patients with SSc, the pattern of lung involvement at CT is different from that in patients with IPF. In patients with SSc, fibrosis is less coarse, and the proportion of ground-glass opacification is greater than that in patients with IPF. The CT features of lung disease in patients with SSc closely resemble those in patients with NSIP and are in keeping with recent observations that NSIP is the most prevalent pattern at histopathologic examination.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Devaraj, A. U. Wells, M. G. Meister, T. J. Corte, and D. M. Hansell The Effect of Diffuse Pulmonary Fibrosis on the Reliability of CT Signs of Pulmonary Hypertension Radiology, December 1, 2008; 249(3): 1042 - 1049. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. U. Wells High-resolution computed tomography and scleroderma lung disease Rheumatology, October 1, 2008; 47(suppl_5): v59 - v61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A U Wells, N Hirani, and on behalf of the BTS Interstitial Lung Disease Gui Interstitial lung disease guideline Thorax, September 1, 2008; 63(Suppl_5): v1 - v58. [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Goldin, D. A. Lynch, D. C. Strollo, R. D. Suh, D. E. Schraufnagel, P. J. Clements, R. M. Elashoff, D. E. Furst, S. Vasunilashorn, M. F. McNitt-Gray, et al. High-Resolution CT Scan Findings in Patients With Symptomatic Scleroderma-Related Interstitial Lung Disease Chest, August 1, 2008; 134(2): 358 - 367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. D. Travis, G. Hunninghake, T. E. King Jr., D. A. Lynch, T. V. Colby, J. R. Galvin, K. K. Brown, M. P. Chung, J.-F. Cordier, R. M. du Bois, et al. Idiopathic Nonspecific Interstitial Pneumonia: Report of an American Thoracic Society Project Am. J. Respir. Crit. Care Med., June 15, 2008; 177(12): 1338 - 1347. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. L. Goh, S. R. Desai, S. Veeraraghavan, D. M. Hansell, S. J. Copley, T. M. Maher, T. J. Corte, C. R. Sander, J. Ratoff, A. Devaraj, et al. Interstitial Lung Disease in Systemic Sclerosis: A Simple Staging System Am. J. Respir. Crit. Care Med., June 1, 2008; 177(11): 1248 - 1254. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Antoniou, D. M. Hansell, M. B. Rubens, K. Marten, S. R. Desai, N. M. Siafakas, A. G. Nicholson, R. M. du Bois, and A. U. Wells Idiopathic Pulmonary Fibrosis: Outcome in Relation to Smoking Status Am. J. Respir. Crit. Care Med., January 15, 2008; 177(2): 190 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Malerba, A. Radaeli, B. Ragnoli, P. Airo', M. Corradi, A. Ponticiello, A. Zambruni, and V. Grassi Exhaled Nitric Oxide Levels in Systemic Sclerosis With and Without Pulmonary Involvement Chest, August 1, 2007; 132(2): 575 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Mueller-Mang, C. Grosse, K. Schmid, L. Stiebellehner, and A. A. Bankier What Every Radiologist Should Know about Idiopathic Interstitial Pneumonias RadioGraphics, May 1, 2007; 27(3): 595 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Camiciottoli, I. Orlandi, M. Bartolucci, E. Meoni, F. Nacci, S. Diciotti, C. Barcaroli, M. L. Conforti, M. Pistolesi, M. Matucci-Cerinic, et al. Lung CT Densitometry in Systemic Sclerosis: Correlation With Lung Function, Exercise Testing, and Quality of Life Chest, March 1, 2007; 131(3): 672 - 681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. O. Villalba, P. D. Sampaio-Barros, M. C. Pereira, E. M. F. P. Cerqueira, C. A. Leme Jr, J. F. Marques-Neto, and I. A. Paschoal Six-Minute Walk Test for the Evaluation of Pulmonary Disease Severity in Scleroderma Patients Chest, January 1, 2007; 131(1): 217 - 222. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lebtahi, S. Moreau, S. Marchand-Adam, M.-P. Debray, M. Brauner, P. Soler, J. Marchal, O. Raguin, A. Gruaz-Guyon, J.-C. Reubi, et al. Increased Uptake of 111In-Octreotide in Idiopathic Pulmonary Fibrosis J. Nucl. Med., August 1, 2006; 47(8): 1281 - 1287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Souza, N. L. Muller, K. S. Lee, T. Johkoh, H. Mitsuhiro, and S. Chong Idiopathic interstitial pneumonias: prevalence of mediastinal lymph node enlargement in 206 patients. Am. J. Roentgenol., April 1, 2006; 186(4): 995 - 999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Misumi and D. A. Lynch Idiopathic pulmonary fibrosis/usual interstitial pneumonia: imaging diagnosis, spectrum of abnormalities, and temporal progression. Proceedings of the ATS, January 1, 2006; 3(4): 307 - 314. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||
