Relationships between pulmonary function testing and high-resolution computed tomography (HRCT) were studied in 39 untreated patients with idiopathic pulmonary fibrosis (IPF) at diagnosis, 23 of whom were followed during 7.5 ± 0.3 mo (mean ± SEM). At diagnosis, the extent of overall lung involvement in the HRCT scans showed a moderate but significant correlation only with FVC (r = − 0.46, p = 0.003) and Dl CO (r = − 0.40, p = 0.03). The extent of ground glass pattern also correlated with FVC (r = − 0.58, p = 0.0001). Arterial Po 2 at peak exercise (n = 13 patients) showed a significant association with the extent of both ground-glass pattern and overall lung involvement in HRCT (r = − 0.60, p = 0.02; and r = − 0.64, p = 0.01, respectively). On multivariate analysis a significant independent correlation between the global disease extent in HRCT and both FVC and Dl CO was observed. Changes over time in the total extent of the disease evaluated with HRCT scans were also related to those observed in Dl CO and in FVC (r = − 0.57, p = 0.01, and r = − 0.51, p = 0.01, respectively). The present study suggests that FVC and Dl CO are the physiological variables that best reflect the global extent of disease in IPF and thus may provide significant information for the assessment of the disease's progression.
Idiopathic pulmonary fibrosis (IPF) is a chronic progressive disorder, characterized by inflammation and fibrosis of the lung parenchyma (1). The management of patients with IPF is difficult, due to their variable clinical evolution and unpredictable response to long-term therapy. Different procedures such as bronchoalveolar lavage, 67Ga lung scans, pulmonary function tests, and high-resolution computed tomography (HRCT) scans have been used in the assessment of these patients to find safe, cost-effective parameters to evaluate the disease's progression (2).
HRCT is now recognized as a sensitive tool to characterize the histopathologic process in patients with IPF (3). Several studies (4-6) have demonstrated that HRCT findings can predict histologic patterns observed in samples obtained by open lung biopsy. Both reticular and honeycombing patterns correlate with fibrosis, whereas a ground-glass pattern identifies zones of alveolar and interstitial inflammation. It has been also shown that HRCT appearances have a prognostic value in IPF (3). Patients with a predominant ground-glass pattern show both the best survival rates and response to therapy. Although HRCT is the best noninvasive method for assessing the pattern (fibrosis versus inflammation) and quantifying the extent of the disease in IPF patients at diagnosis, its frequent repetition is questionable because of the radiation burden and its high economic cost. Conventional pulmonary function tests are routinely performed in the clinical evaluation of patients with IPF, but their correlation with the extent and progression of the disease is poorly known. The main objective of the present study has been to identify the lung function parameters that show an association with the severity and the progression of disease in IPF, using HRCT as the gold standard. With this purpose, we have investigated the relationships between pulmonary function tests and HRCT findings in untreated patients with IPF at diagnosis and we have evaluated in a prospective manner whether the changes over time in HRCT findings correlate with those observed in lung function assessment.
The population studied consisted of 39 untreated patients (28 men) (age, 66 ± 1 yr, mean ± SEM, range 49 to 88) with IPF. Twelve were smokers (41 ± 4 pack-years, range 15 to 70), five had ceased smoking at least 5 yr before entry into the study, and the remaining 22 had never smoked. All patients had cough and 34 had dyspnea, with a duration of symptoms of 11 ± 1 mo (range, 0 to 36). All of them showed bilateral widespread crackles and bilateral interstitial infiltrates in the chest radiography. The diagnosis of IPF was established by open lung biopsy in 17 cases (44%). In the remaining 22 patients without histologic confirmation of the disease, two conditions were required for the diagnosis: fulfillment of the clinical criteria described by Turner-Warwick (7, 8), and findings compatible with IPF in the HRCT scan (5). In seven of these 22 cases without open lung biopsy, the diagnosis of IPF was also supported by the findings in transbronchial lung biopsy. Furthermore, bronchoalveolar lavage cell analysis showed alterations compatible with IPF, that is neutrophilia (11 ± 2%; normal values < 3%) with or without eosinophilia (5 ± 1%; normal values < 1%) in the 20 cases in which fiberoptic bronchoscopy was performed (8). None of the patients had associated collagen vascular disease. A follow-up study was performed on 23 of these patients during 7.5 ± 0.3 mo (range, 4 to 11) after the initial assessment. Of the remaining 16 IPF patients, 10 died 2 to 9 mo after the diagnosis (mean, 5 mo) and in six cases, contact was lost. The causes of death were progression of the disease in seven cases, pulmonary infections in two, and bronchogenic carcinoma in one. Eleven of the 23 patients received treatment with glucocorticoids after the diagnosis (1 mg/kg daily), while the remaining 12 patients were not treated. The indication for treatment was based on the presence of progressive clinical disease, defined as: (1) progressive increase in the degree of dyspnea; or (2) fall in FVC and/ or in single-breath diffusing capacity (Dl CO) > 15% with respect to previous pulmonary functional testing.
Slow and forced spirometry, plethysmographic lung volumes, single-breath carbon monoxide transfer factor, and arterial blood gases at rest while breathing room air were all measured, as previously described. Reference values from our own laboratory were used (9, 10). The lower limit of reference (LLR) for each physiologic variable was defined as: the predicted value − (1.645 × RSD). RSD stands for the residual standard deviation of the corresponding prediction equation. Dl CO and TLC were not performed in 11 and eight patients, respectively, because of a severe reduction in lung volumes and/or lack of cooperation. Dl CO values were corrected to hemoglobin (10). Symptom-limited exercise capacity was assessed in 13 patients using a cycloergometer (E. Jaeger, Würzburg, Germany). We performed a progressive incremental protocol, which consisted of a 20-W work rate increase every minute (9). Predicted values used in this technique were those reported by Jones and colleagues (11). Exercise testing was not performed on the remaining 26 patients due to severe respiratory compromise, concurrent cardiac disease, and/or an inability to cycle properly.
Computed tomography scans were performed using either a Somaton HiQ or a Somaton Plus scanner (Siemens, Erlanger, Germany). All patients underwent conventional computed tomographic scanning of the chest, using a 10-mm section thickness at 12-mm intervals. HRCT scans were obtained at six predetermined levels: the great vessels, the aortic arch, the tracheal carina, the pulmonary hilae, the pulmonary venous confluence, and 1 cm above the right diaphragm. The scans were performed with a 1- to 2-mm section thickness and a 1- to 2-s scanning time during breath holding at the end of inspiration. These scans were reconstructed with a high spatial frequency algorithm and viewed at window levels appropriate for pulmonary parenchyma (mean −500 to −600 Hounsfield units; width 1,400 to 1,600 Hounsfield units) (12).
Two radiologists, without knowledge of any of the clinical, functional and radiographic findings, examined the HRCT scans. The overall extent of parenchymal abnormalities was scored at the six predetermined levels where thin sections had been carried out, and a semiquantitative analysis of the relative proportion (to within 10%) of both the ground-glass and reticular patterns was performed by consensus. The scores of the six lung zones were then averaged out to obtain one mean score. The extent of emphysema was determined using the same methodology as that of the scoring for the extent of IPF. The follow-up HRCT scans were evaluated independently of the initial ones. However, before reading the films, HRCTs performed at diagnosis were also examined to calculate the semiquantitative score at the same scan sections. Pulmonary function testing and HRCT scans were always performed less than a week apart.
Results are expressed as the mean ± SEM. Spearman's rank correlation was used to examine the correlation between HRCT and pulmonary function tests. Stepwise multiple regression analysis was used to determine the relationships between HRCT abnormalities (dependent variable) and lung function parameters adjusted by age, sex, height, pack-years smoked, and body mass index (covariates). Moreover, multivariate analysis was also performed to evaluate the influence of emphysema on the relationship between the extent of IPF on HRCT and pulmonary function tests. Wilcoxon's test was used for the comparison of paired data and Mann-Whitney's test for unpaired data. Statistical significance was established at a p value ⩽ 0.05.
All patients had a restrictive ventilatory impairment according to forced spirometry, but seven of the 31 patients on whom TLC was performed, showed values above the LLR. Dl CO was below the LLR in all but two patients. Arterial blood gases at rest were analyzed in all but one patient. Arterial hypoxemia (PaO2 < 80 mm Hg [10.7 kPa]) was shown in 25 cases and an increase in alveolar-arterial oxygen tension difference (aaPo 2) (> 20 mm Hg [2.7 kPa]) was observed in 30 patients. At peak exercise, a marked decrease was seen in PaO2 (from 78 ± 2 mm Hg [10.5 ± 0.3 kPa] to 63 ± 3 mm Hg [8.5 ± 0.5 kPa], p = 0.01), with a simultaneous increase in aaPo 2 (from 24 ± 2 mm Hg [3.2 ± 0.3 kPa] to 45 ± 2 mm Hg [6.1 ± 0.5 kPa], p = 0.003) in the 13 patients who underwent cycloergometry (Table 1).
|Actual Values||% Predicted|
|Pulmonary function tests|
|FVC, L||2.3 ± 0.1||62 ± 2|
|FEV1 L||1.9 ± 0.1||71 ± 3|
|FEV1/FVC, %||83 ± 1|
|TLC, L||4.4 ± 0.2||74 ± 3|
|Dl CO, ml/min/mm Hg||13.3 ± 0.8||54 ± 3|
|Kco *, ml/min/mm Hg/L||3.9 ± 0.2||77 ± 5|
|PaO2 at rest, kPa||9.5 ± 0.3|
|(71 ± 2 mm Hg)|
|aaPo2 at rest, kPa||4.4 ± 0.3|
|(33 ± 2 mm Hg)|
|At Rest||At Peak Exercise|
|PaO2 , kPa (n = 13)||10.5 ± 0.3||8.5 ± 0.5|
|(78 ± 2 mm Hg)||(66 ± 3 mm Hg)|
|aaPo2, kPa (n = 13)||3.2 ± 0.3||6.1 ± 0.5|
|(24 ± 2 mm Hg)||(45 ± 3 mm Hg)|
|HRCT scan score, %|
|Ground glass extent||12.3 ± 2.4|
|Reticular extent||29.9 ± 2.2|
|Overall lung involvement||42.2 ± 2.5|
HRCT showed only either reticular abnormalities or ground-glass pattern in four and one patients, respectively. In the remaining 34 cases, HRCT scan showed a mixed pattern with both ground-glass and reticular abnormalities. In 30 of them, the extent of reticular pattern was higher than the extent of ground-glass opacification (Table 1). A significant correlation between the global disease extent in the HRCT and both FVC (r = −0.46, p = 0.003) and Dl CO (r = −0.40, p = 0.03) was observed (Figure 1). In addition, the extent of ground-glass pattern also correlated significantly with FVC (r = −0.58, p = 0.0001) (Figure 2). Arterial Po 2 at peak exercise showed a significant correlation with both the total extent of pulmonary disease and the extent of ground-glass pattern (r = −0.60, p = 0.02, and r = −0.64, p = 0.01, respectively). A multivariate analysis was performed to determine the influence of age, sex, height, body mass index, and pack-years smoked on the relationship between HRCT and pulmonary function tests. A significant independent correlation was observed between the global disease extent as assessed by the HRCT and both FVC and Dl CO (Table 2).
|All Patients||Patients without Emphysema|
|Changes in FVC|
|Changes in Dl CO|
Changes in pulmonary function parameters and in HRCT findings are expressed as percentage of variation from diagnosis (Table 3). As a whole, there was a moderate deterioration in all the functional parameters. The extent of ground-glass opacification in HRCT increased in seven cases, decreased in seven, and remained unchanged in nine. A decrease in the extent of reticular pattern was not observed in any patient. A significant correlation was found between changes in HRCT overall lung involvement and changes in both Dl CO and FVC (r = −0.57, p = 0.01, and r = −0.51, p = 0.01, respectively) (Figure 3). A multivariate analysis was performed to examine the influence of potential confounders (age, sex, height, body mass index, and pack-years smoked) in these correlations. A significant independent relationship was shown between changes in global disease extent in HRCT and changes in both FVC and Dl CO (Table 2). It is worth noting that no differences in pulmonary function tests and in HRCT were observed during the follow-up between treated and nontreated patients. There were no differences in the results obtained when data analysis was restricted to patients with open lung biopsy (data not shown).
|At Diagnosis||Variation (%)|
|Pulmonary function tests|
|FVC, L||2.5 ± 0.1||−4.2 ± 2|
|FEV1, L||2 ± 0.1||−6.2 ± 1.9|
|FEV1/FVC, %||83 ± 1||−1.6 ± 1.9|
|TLC, L||4.6 ± 0.2||−3.3 ± 2|
|Dl CO, ml/min/mm Hg||15 ± 0.8||−6.8 ± 3.8|
|Kco, ml/min/mm Hg/L||4.2 ± 0.2||−22 ± 3|
|PaO2 at rest, kPa||10.3 ± 0.1||−1.5 ± 2|
|(77 ± 0.7 mm Hg)|
|aaPo2, at rest, kPa||3.5 ± 0.1||+22 ± 13|
|(26.2 ± 0.7 mm Hg)|
|HRCT scan score|
|Ground glass extent||8 ± 2||56 ± 4|
|Reticular extent||29 ± 7||4.4 ± 3|
|Overall lung involvement||37.7 ± 2||6.6 ± 7|
HRCT scans showed that emphysema was present in four cases (score 15%, 25%, 31%, and 15%, respectively). Patients with emphysema had a significantly higher FVC and TLC with respect to patients without emphysema (90 ± 1% versus 59 ± 2%, p = 0.001, and 99 ± 2% versus 69 ± 2%, p = 0.001, respectively), although no differences in Dl CO (48 ± 3% versus 55 ± 3%) were observed. A significant and independent relationship between FVC and the extent of emphysema in multivariate analysis was shown (regression coefficient = 0.4; p = 0.005). As the presence of emphysema on HRCT was correlated significantly with FVC values, we performed the analysis of the relationship between HRCT findings and pulmonary function parameters in patients without emphysema. There were no differences between the results obtained when all patients were considered together and those obtained when analysis was restricted to patients without emphysema (Table 2).
The present study suggest that FVC and Dl CO are the pulmonary functional parameters that best reflect the global extent of parenchymal abnormalities in IPF, as shown by their independent correlation with HRCT findings. Thus, our results indicate that both FVC and Dl CO can provide useful information both at baseline and in the assessment of the disease's progression in IPF.
Several studies (13-18) have investigated the role of pulmonary function tests in the management of patients with IPF, but the results obtained have been rather inconclusive. Although discrepancies among these studies are difficult to explain, they may be partly due to the variable course of the disease, and also to the heterogeneity of the study population. Several studies (15, 18) have pooled data from patients with lone IPF and with pulmonary fibrosis associated with collagen disorders or have examined patients under treatment together with others without treatment. Finally, Erbes and coworkers (13) only included patients with open lung biopsy, which may introduce a selection bias. In the present study we have evaluated only patients with lone IPF and the results obtained were based on HRCT findings, which represent a sensitive method for determining the extent of parenchymal alterations in IPF. In addition, the use of HRCT allowed us to include patients too compromised to undergo an open lung biopsy. Thus, our study population is homogeneous and it can be considered representative of all the stages of the disease. Both factors markedly strengthen the results of the present study.
The relationship between HRCT findings and pulmonary function tests in IPF has been investigated by several authors (19-23). Wells and coworkers (19) analyzed the changes in serial CT scans and showed that improvement in pulmonary function tests was associated with regression of ground-glass pattern. Other studies (20-22) have shown that HRCT findings significantly correlate with several functional parameters, such as static lung volumes, FEV1, or Dl CO. However, some of these studies (21, 22) evaluated only the significance of ground-glass opacification or reticular pattern, but not the global extent of the disese in HRCT. The findings of the present study confirm the results of Staples and coworkers and Wells and coworkers who observed that Dl CO was the pulmonary parameter that best correlated with the global extent of disease in HRCT (20, 23). It should be noted that both Staples and coworkers (20) and Wells coworkers (23) only made their analyses at the diagnosis, but did not investigate the relationships between HRCT and functional parameters during the disease's progression, as in the present study. The moderate but significant correlation between functional deterioration and changes over time in HRCT can most likely be explained by the small range of variation in both functional parameters and HRCT findings during the follow-up, and it should not be attributed to a low signal of these relationships. According to our results, conventional functional tests can offer significant information about the extent of parenchymal abnormalities, and thus, indicate that it is probably not necessary to perform HRCT as a routine test to evaluate the progression of the disease. In addition, pulmonary function tests can be measured in almost all patients and they have an acceptable reproducibility (24). Although HRCT scan represent a reproducible method for determining the extent of disease in IPF, their frequent repetition is questionable owing to the radiation burden and their high economic cost.
The degree of gas exchange impairment during incremental exercise has proved to be a good marker of the severity of morphologic findings observed in biopsy specimens of patients with IPF (25). In the present study we have found that the extent of both ground-glass pattern and overall lung abnormalities in HRCT scans also correlated significantly with PaO2 measured at peak exercise. These findings reinforce our previous observation that PaO2 fall during exercise seems to predict the disease progression in IPF (8). Moreover, our results agree with those of Wells and coworkers (23) which showed a strong correlation between the amount of O2 desaturation during exercise and the overall extent of the disease in HRCT scan. However, exercise tests were only performed on 13 patients, and thus, the clinical implications of these findings have to be viewed with caution.
We have also examined the significance of concurrent emphysema in the evaluation of the relationships between lung function indices and HRCT findings. It has been shown that the presence of emphysema can influence both FVC and Dl CO values in patients with IPF (23). In our study, FVC was significantly higher in patients with emphysema and there was a significant relationship between this lung function parameter and the presence of emphysematous changes in HRCT scans. However, there were no differences between the results obtained when all patients were considered together and those obtained when analysis was restricted to patients without emphysema. In spite of these results, the present study emphasizes the importance of taking into account the presence of concurrent emphysema for the interpretation of pulmonary function tests in IPF.
A potential limitation of our study is that the diagnosis of IPF was established by open lung biopsy in only 17 (44%) cases. In the remaining 22 patients, the diagnosis of IPF was done by clinical data and the HRCT findings, and supported by transbronchial lung biopsy in seven of them. Furthermore, bronchoalveolar lavage was performed in 20 of these patients and showed an increase in the percentage of both neutrophils and eosinophils. The role of bronchoalveolar lavage in the diagnosis of interstitial lung diseases has been reported in a recent publication by Drent and coworkers (26) in which bronchoalveolar data from 277 patients with IPF, extrinsic allergic alveolitis, and sarcoidosis were evaluated. In this study, patients with IPF had an increase in the percentage of neutrophils and eosinophils, and the diagnostic effectiveness of bronchoalveolar cell profiles for this disease was 84%. These results confirm previous findings (27) and strongly support the concept that the information provided by bronchoalveolar lavage cell analysis, in the context of typical clinical and radiologic features and HRCT findings, may be of diagnostic value in patients suspected of having IPF. On the other hand, some points have to be considered with respect to the interpretation of HRCT findings. In some cases it is not possible to distinguish clearly between the thickening of alveolar walls, due to inflammatory cell infiltration and/or edema, and enhanced deposition of connective tissue; under these circumstances, areas with very fine fibrosis could not be differentiated from alveolitis. However, it has been shown (28, 29) that thin sections (1– 2 mm) allow the ground-glass pattern in HRCT to be accurately identified and it can be fully distinguished from a fine reticular pattern. In the current study we have used thin sections, 1–2 mm in thickness, which allowed us to properly evaluate the extent of ground-glass attenuation. In addition, although ground-glass opacification may also be the result of respiratory motion, giving rise to an overestimate of the disease's activity (30), such an event was unlikely in our patients because they had no difficulty in holding their breath for 2 s during HRCT scanning.
In conclusion, the present study shows that both forced vital capacity and single-breath diffusing capacity for carbon monoxide are the physiologic variables that best reflect the global extent of disease in IPF, as shown by HRCT. Thus, our findings reinforce the value of conventional pulmonary function tests in the management of IPF patients. Although we also found that gas exchange parameters during exercise are related to the extent of disease in HRCT scans, the small number of patients evaluated precludes any definitive conclusions.
Supported by Grants 94/998 from the FIS (Fondo de Investigación Sanitaria), 1995 SGR 00446 from the Comissionat per a Universitats i Recerca (Generalitat de Catalunya), and Fundació Catalana de Pneumologia (FUCAP/94).
|1.||Crystal R. G., Fulmer J. D., Roberts W. C., Moss M. L., Line B. R., Reynolds H. Y.Idiopathic pulmonary fibrosis: clinical, histologic, radiographic, physiologic, scintigraphic, cytologic, and biochemical aspects. Ann. Intern. Med.851976769788|
|2.||Raghu G.Idiopathic pulmonary fibrosis: a rational clinical approach. Chest921987148154|
|3.||Wells A. U., Hansell D. M., Rubens M. B., Culliman P., Black C. M., du Bois R. M.The predictive value of appearances on thin-section computed tomography in fibrosing alveolitis. Am. Rev. Respir. Dis.148199310761082|
|4.||Müller N. L., Stables C. A., Miller R. R., Vedal S., Thurbeck W. M., Ostrow D. N.Disease activity in idiopathic pulmonary fibrosis: CT and pathologic correlation. Radiology1651987731734|
|5.||Hansell D. M., Kerr I. H.The role of high resolution computed tomography in the diagnosis of interstitial lung disease. Thorax4619917784|
|6.||Müller N. L., Miller R. R., Webb W. R., Evans K. G., Ostrow D. N.Fibrosing alveolitis: CT–pathologic correlation. Radiology1601986585588|
|7.||Turner-Warwick M., Burrows B., Johnson A.Cryptogenic fibrosing alveolitis: clinical features and their influence on survival. Thorax351980171180|
|8.||Agusti C., Xaubet A., Agusti A. G. N., Roca J., Ramirez J., Rodriguez-Roisin R.Clinical and functional assessment of patients with idiopathic pulmonary fibrosis: results of a 3-year follow-up. Eur. Respir. J.71994643650|
|9.||Roca J., Sanchis J., Agusti-Vidal A., Segarra F., Navajas D., Rodriguez-Roisin R.Spirometric reference values from a Mediterranean population. Bull. Eur. Physiopathol. Respir.221986217224|
|10.||Roca J., Rodriguez-Roisin R., Cobo E., Burgos F., Perez J., Clausen J. L.Single-breath carbon monoxide diffusing capacity (DlCO) prediction equations from a Mediterranean population. Am. Rev. Respir. Dis.140199010261032|
|11.||Jones N. L., Makrides L., Hitchcock C., Chypchar T., McCartney N.Normal standards for an incremental progressive cycle ergometer test. Am. Rev. Respir. Dis.1311985700708|
|12.||Agusti C., Xaubet A., Luburich P., Ayuso M. C., Roca J., Rodriguez-Roisin R.Computed tomography–guided bronchoalveolar lavage in idiopathic pulmonary fibrosis. Thorax511996841845|
|13.||Erbes R., Schaberg T., Loddenkemper R.Lung function tests in patients with idiopathic pulmonary fibrosis: are they helpful for predicting outcome? Chest11119975157|
|14.||Schwartz D. A., Helmers R. A., Galvin J. R., Van Fossen D. S., Fress K. L., Dayton C. S., Burmeister L. F., Hunninghake G. W.Determinants of survival in idiopathic pulmonary fibrosis. Am. J. Respir. Crit. Care Med.1491994450454|
|15.||Rudd R. M., Haslam P. L., Turner-Warwick M.Cryptogenic fibrosing alveolitis: relationship of pulmonary physiology and bronchoalveolar lavage to response treatment and prognosis. Am. Rev. Respir. Dis.124198118|
|16.||van Oortegem K., Wallaert B., Marquette C. H., Ramon P., Perez T., Lafitte J. J., Tonnel A. B.Determinants of response to immunosuppressive therapy in idiopathic pulmonary fibrosis. Eur. Respir. J.7199419501957|
|17.||Hanson D., Winterbauer R. H., Kirtland S. H., Wu R.Changes in pulmonary function test results after one year of therapy as predictors of survival in patients with idiopathic pulmonary fibrosis. Chest1081995305310|
|18.||Tukianen P., Taskinen E., Holsti P., Korhola O., Valle M.Prognosis of cryptogenic fibrosing alveolitis. Thorax381983349355|
|19.||Wells A. U., Rubens M. B., du Bois R. M., Hansell D. M.Serial CT in fibrosing alveolits: prognostic significance of the initial pattern. A.J.R.161199311591165|
|20.||Staples C. A., Müller N. L., Vedal S., Abboud R., Ostrow D., Miller R. R.Usual interstitial pneumonia: correlation of CT with clinical, functional, and radiologic findings. Radiology1621987377381|
|21.||Lee J. S., Im J. G., Ahn J. M., Kim Y. M., Han M. C.Fibrosing alveolitis: prognostic implication of ground-glass attenuation at high-resolution CT. Radiology1841992451454|
|22.||Terrif B. A., Kwan S. Y., Chan-Yeung M. M., Müller N. L.Fibrosing alveolitis: chest radiography and CT as predictors of clinical and functional impairment at follow-up in 26 patients. Radiology1841992445449|
|23.||Wells A. U., King A. D., Rubens M. B., Cramer D., du Bois R. M., Hansell D. M.Lone cryptogenic fibrosing alveolitis: a functional–morphologic correlation based on extent of disease on thin-section computed tomography. Am. J. Respir. Crit. Care Med.155199713671375|
|24.||Standardized lung function testing: official statement of the European Respiratory Society. 1993. Eur. Respir. J. 6(Suppl. 16):1–100.|
|25.||Fulmer J. D., Roberts W. C., von Gal E. R., Crystal R. G.Morphologic–physiologic correlates of the severity of fibrosis and degree of cellularity in idiopathic pulmonary fibrosis. J. Clin. Invest.631979665676|
|26.||Drent M., van Nierop M. A. F. M., Gerritsen F. A., Wouters E. F. M., Mulder P. G. H.A computer program using BALF-analysis results as a diagnostic tool in interstitial lung diseases. Am. J. Respir. Crit. Care Med.1531996736741|
|27.||Drent M., Mulder P. G. H., Sj, Sc, Wagenaar, Hoogsteden H. C., van Velzen-Blad H., van den Bosch J. M. M.Differences in BAL fluid variables in interstitial lung diseases evaluated by discriminant analysis. Eur. Respir. J.61993803810|
|28.||Nishimura K., Kitaichi M., Izumi T., Nagai S., Kanaoka M., Itoh H.Usual interstitial pneumonia: histologic correlation with high-resolution CT. Radiology1821992337342|
|29.||Remy Jardin, M., J. Remy, C. Deffontaines, and A. DuhamelAssessment of diffuse infiltrative lung disease: comparison of conventional CT and high resolution CT. Radiology1811991157162|
|30.||Leung A. N., Mille R. R., Müller N. L.Parenchymal opacification in chronic infiltrative lung diseases: CT–pathologic correlation. Radiology1881993209214|