The Sensitivity of High-Resolution CT in Detecting Idiopathic Pulmonary Fibrosis Proved by Open Lung Biopsy*

Transcription

1 The Sensitivity of High-Resolution CT in Detecting Idiopathic Pulmonary Fibrosis Proved by Open Lung Biopsy* A Prospective Study jonathan B. Orens, MD; Ella A. Kazerooni, MD, FCCP; Fernando]. Martinez, MD, FCCP; Jeffrey L. Curtis, MD; Barry H. Gross, MD, FCCP; Andrew Flint, MD, FCCP; and Joseph P. Lynch III, MD, FCCP Objectives: To assess the sensitivity of high-resolution chest computed tomography (HRCT) in detecting idiopathic pulmonary fibrosis proved by biopsy specimen. To determine the degree of physiologic and pathologic abnormalities in patients with idiopathic pulmonary fibrosis who have a false-negative HRCT. Design: Prospective 2-year study. Setting: Tertiary care university hospital. Patients: All patients with dyspnea and suspected interstitial lung disease referred to the University of Michigan for enrollment in the Idiopathic Pulmonary Fibrosis Specialized Center of Research (SCOR) protocol were included; 25 underwent open lung biopsy and formed the final study group. Measurements: All patients underwent physiologic (pulmonary function, gas exchange, and exercise testing), radiologic (chest x-ray film and HRCT), and pathologic assessments (bronchoscopic and open lung biopsy). The results of HRCT were prospectively compared with results of standard pulmonary function tests, cardiopulmonary exercise testing, and open lung biopsy. Results: Of 25 patients who had both HRCT and open lung biopsy, 3 patients (12%) had HRCTs that demonstrated no evidence of interstitial lung disease. These three patients had less severe disease based on clinical, radiographic, and physiologic (CRP) scores, gas exchange abnormalities, and pathologic scoring of open lung biopsy specimens, compared with those with an abnormal HRCT. Conclusion: We conclude that in the evaluation of patients with dyspnea and abnormal results of pulmonary function studies, a normal HRCT does not exclude early and clinically significant interstitial lung disease. In our patient population, physiologic testing was more sensitive than HRCT in detecting mild abnormalities in patients with idiopathic pulmonary fibrosis proved by biopsy specimen. (CHEST 1995; 108:109-15) CPET=cardiopulmonary exercise testing; CRP=clinical, radiographic, and physiologic; HRCT=high-resolution computed tomography; PFT=pulmonary function test; SCOR=Specialized Center of Research Key words: high-resolution computed tomography; idiopathic pulmonary fibrosis; pulmonary function tests; sensitivity The evaluation of patients with dyspnea is a complex diagnostic challenge. Dyspnea may arise from pulmonary, cardiac, muscular, or neuropsychiatric abnormalities. However, most dyspneic patients, regardless of the treatment setting, have pulmonary disease. 1 Therefore, excluding the presence of pulmonary parenchymal abnormalities is a critical *From the Departments of Internal Medicine (Drs. Orens, Martinez, Curtis, and Lynch), Radiology (Drs. Kazerooni and Gross), and Pathology (Dr. Flint), The University of Michigan School of Medicine, Ann Arbor. Supported in part by National Institutes of Health NHLBI grant P50HL46487, NIH/ NCRR 3 MOl RR S3, and by funds from the Clinical Research Center of the University of Michigan grant M01RR Manuscript received September 6, 1994; revision accepted November 23. part of the diagnostic workup. Many techniques have been used to accomplish this goal, including radiologic and nuclear imaging studies, pulmonary function testing (PFTs), cardiopulmonary exercise testing (CPET), and various biopsy techniques. Advances in imaging technologies have raised questions about the critical diagnostic pathway for dyspneic patients. Recently, a new method for performing CT scanning of the lung parenchyma, high-resolution CT (HRCT), has been shown to more clearly delineate the anatomy of the lungs when compared with conventional CT. 2 Specifically, because HRCT can image to the anatomic level of the secondary pulmonary lobule, the necessity of performing a biopsy to establish the diagnosis of interstitial lung disease has been questioned.3 If HRCT is to replace open lung biopsy, the CHEST / 108 / 1 / JULY,

2 false-negative rate of HRCT must be established. To test adequately the sensitivity of HRCT, it is essential to evaluate a homogeneous population of patients with a diagnosis confirmed by open lung biopsy specimen. We chose to address this issue in patients with idiopathic pulmonary fibrosis because open lung biopsy specimens are still considered the standard for establishing the diagnosis of this important disease. 4 5 Given the morbidity and high cost of open lung biopsy, it is essential to establish whether open lung biopsy specimens are still necessary to confirm this disease. Because the anatomic abnormalities of idiopathic pulmonary fibrosis are progressive, beginning with microscopic inflammation that precedes fibrosis, we hypothesized that some patients may have microscopic disease undetectable by HRCT. Accordingly, we prospectively evaluated dyspneic patients with suspected interstitial lung disease using plain chest radiographs, PFTs, CPET, HRCT, and when indicated, trans bronchial and open lung biopsy, to establish the cause of their complaint. To establish the sensitivity of HRCT in detecting.idiopathic pulmonary fibrosis, results of HRCT were compared with open lung biopsy specimens in 25 patients with confirmed idiopathic pulmonary fibrosis. Additionally, the severity of the disease in each patient was graded by an assessment of the degree of abnormalities of physiologic test results and pathologic findings. Patient Recruitment METHODS Patients were entered into the study protocol in a prospective fashion between January 1, 1992 and December 31, The FIGURE l. Representative HRCT scan of case 1 reveals normal lung parenchyma. There is no evidence of ground-glass opacification, thickening of septal lines, or cystic changes. initial study group included all patients referred to the University of Michigan for enrollment in the Idiopathic Pulmonary Fibrosis Specialized Center of Research (SCOR) protocol (n = 65). Patients were eligible if there was a clinical suspicion of idiopathic pulmonary fibrosis based on symptoms, physiologic abnormalities, or radiographic findings, and neither previous biopsy nor therapy. Patients were excluded if during the enrollment workup they were found to have a disease other than idiopathic pulmonary fibrosis (n=23). Specifically, excluded patients had the following entities: collagen vascular disease, pneumoconiosis, sarcoidosis, cancer, lymphoma, eosinophilic granuloma, and lymphangioleiomyomatosis. Additionally, patients were excluded if they had medically unstable conditions, or were unwilling to undergo, open or video-assisted thoracoscopic lung biopsy (n=17). HRCT Protocol All HRCTs were performed with 1.0- or 1.5-mm-thick sections taken at 1-cm intervals throughout the entire thorax and were reconstructed using a bone algorithm. No intravenous contrast was administered. The scans were performed using a CT scanner (General Electric HiLite Advantage CT/ T scanner, GE Medical Systems, Milwaukee). For scoring, HRCT scans of patients with Table 1-CRP Scoring Results Normal Abnormal Parameter HRCT HRCT p Value Subjective score 10.7 ± ± 1.3 NS* Radiographic score 1.3 ± ±0.3 Physiologic score 10.5± ± 1.5 NS FVC 64.7± ±3.7 NS FEV ± ±4.2 NS Thoracic gas volume 59.3± ± DCO/ VA 110.7± ± Resting P(A-a)02 3.9± ± Exercise score 8.3± ±2.0 NS Total CRP 24.6± ± *NS=not significant. 110 Clinical Investigations

3 Table 2-Pathologic Scoring Results Parameter Normal HRCT Inflammatory/ exudative changes Alveolar wall Cell infiltrate (total 3.8 ± 0.3 extent) Cell infiltrate (average l.l ± 0.1 severity) Bronchiolar metaplasia 0.0±0.0 Hyperplasia 0.0 ± 0.0 Alveolar space Cellularity (total 2.1 ±0.6 extent) Cellularity (average 1.1 ± 0.1 severity) Fibrotic/ reparative changes Alveolar wall Interstitial young 0.0±0.0 connective tissue Interstitial fibrosis (% 2.0±0.3 alveolar wall) Honeycombing 0.2±0.1 alone (%) Smooth muscle (% ) 0.0±0.0 low power Vessel myointimal 0.7±0.2 changes (extent) Vessel myointimal 0.6±0.1 changes (severity) Abnormal HRCT 4.4± ± ± ± ± ± ± ± ± ± ± ±0.2 p Value idiopathic pulmonary fibrosis (proved by open lung biopsy specimens after HRCT obtained) (n=25) were coded and mixed with HRCT scans of patients (n=27) with a variety of alternative diagnoses (including pulmonary hypertension, sarcoidosis, lymphoma/ cancer, eosinophilic granuloma, pulmonary fibrosis associated with collagen vascular disease, and suspected but not biopsy specimen-proved idiopathic pulmonary fibrosis). Two experienced thoracic radiologists (B.H.G. and E.A.K.) independently evaluated all HRCTs for evidence of interstitial lung disease. At the time of CT interpretation, the two radiologists were not aware of the patient's history, physiologic assessment, or lung biopsy results. Scans were graded as either abnormal or normal for interstitial lung disease based on previously published criteria 6 The criteria for a positive interpretation included one or more of the following: irregular interlobular septal thickening, intralobular interstitial thickening or irregular interfaces, visible intralobular bronchioles, honeycombing, traction bronchiectasis, or ground-glass opacity. All HRCTs in the 25 patients with idiopathic pulmonary fibrosis were obtained 1 to 4 weeks before open lung biopsy. Physiologic Assessment The physiologic assessment was performed before open lung biopsy and initiation of therapy. Pulmonary function tests, including spirometry, lung volumes, and DCO, were performed on the same day but before CPET. All spirometric studies were performed on a calibrated pneumotachograph (Medical Graphics Co, St. Paul, Minn), and values were expressed as a percent of the predicted values published by Morris et ap Lung volumes were measured in a whole-body plethysmograph, using the technique of Dubois et al; 8 data were expressed as a percent of the predicted values published by Goldman and Becklake. 9 Diffusion capacity was measured using the technique of Ogilvie et al 10 and was expressed as a percent of the predicted values published by Crapo and Morris. 11 CPETs were performed on an electronically braked, calibrated cycle ergometer. Work load was increased by 20 W /min until maximal symptom-limited exercise was achieved. Expired gases and ventilation were measured using a calibrated metabolic cart (2001 System, Medical Graphics Co, St. Paul, Minn, or Collins CPXII, Warren E. Collins, Inc, Braintree, Mass). Arterial blood gas values were obtained at rest and every 2 min of exercise via an indwelling radial artery catheter. P(A-a)Oz was calculated for each sample by the alveolar gas equation, using the measured R value during the same time-point of exercise. Oxygen saturation was measured by co-oximetry of the blood gas sample. Twelvelead electrocardiograms and noninvasive measurements of blood pressure were recorded every minute of exercise. The values of Jones et ajl 2 were used to predict maximal exercise oxygen consumption (Voz). CRP Scoring System Overall clinical severity was assessed for each patient using a previously developed composite clinical, radiographic, and physiologic (CRP) score for idiopathic pulmonary fibrosis. 13 The total CRP score ranges from 0 to 100 points (100 being the most severe disease) based on seven variables that include the following: level of dyspnea, 0 to 20 points; chest radiograph, 0 to 10 points; spirometry (FVC, 0 to 12 points; FEY[, 0 to 3 points); lung volume, 0 to 10 points; diffusion capacity (corrected for alveolar volume), 0 to 5 points; resting P(A-a)Oz, 0 to 10 points; Oz saturation corrected for maximal achieved Vozmax, 0 to 30 points. Biopsy Technique All patients underwent bronchoscopy with bronchoalveolar lavage and transbronchial biopsies before open lung biopsy. Patients were referred for open lung biopsy when the results of transbronchial biopsy did not reveal a clear-cut alternative diagnosis. Open lung biopsy was performed by formal thoracotomy (n=7) or video-assisted thoracoscopy (n= 18). Similar-sized biopsy specimens were obtained by either technique. Biopsy specimens were obtained from all three lobes on the right side or from the upper and lower lobe on the left with exclusion of the lingula. Pathologic Scoring To correlate the severity of disease with the presence or absence of abnormalities on HRCT, all open lung biopsy specimens were reviewed by a specialist in lung pathology (A. F.). The pathologist was not aware of the radiographic scoring for individual patients at the time of pathologic scoring. Each specimen was processed in routine fashion. 14 Four histologic sections were prepared from each paraffin block. Individual slides were stained with hematoxylin-eosin, pentachrome stain (which demonstrates differential staining of elastic tissue, collagenized connective tissue, and mucopolysaccharide-rich stroma), Prussian blue (iron stain), and a trichrome stain (which demonstrates differential staining of collagenized connective tissue and smooth muscle). The alveolar septa, alveolar spaces, airways, and blood vessels were evaluated semiquantitatively (on a scale of 0 to 5) for cellular infiltrates (extent and severity), granulation tissue (absence or presence, and degree), honeycombing, interstitial fibrosis, interstitial young connective tissue, vessel myointimal changes, airway (terminal and respiratory bronchioles) mural inflammation and fibrosis, and luminal granulation tissue. These scores were tabulated separately and incorporated into inflammatory, fibrotic, and total pathology scores. Statistical Analysis All data were compiled into a total CRP score and a total CHEST/108/1/JULY,

4 pathologic score, 13 which were compared with the results of chest radiographs and HRCTs. Comparisons between patients with a normal HRCT vs those with an abnormal HRCT were performed with a two-tailed Student's t test; a p value of <0.05 was considered statistically significant. Patient Characteristics RESULTS The study group consisted of 25 patients with abnormal open lung biopsy specimens compatible with idiopathic pulmonary fibrosis. An approximately equal number of men (n=12) and women (n=13) were present. The mean age for the total group was 55.7 years (range, 28 to 74 years). There was a significant difference between the ages of the group with normal HRCTs (see below) (41.3 years; range, 38 to 43 years) and the group with abnormal HRCTs (56.9 years; range, 28 to 74 years; p=). Radiologic Imaging An abnormal HRCT compatible with interstitial lung disease was found in 22 of the 25 patients with biopsy specimen-proven idiopathic pulmonary fibrosis (88.0%). Importantly, three patients had HRCT scans that were not indicative of interstitial lung disease (Fig 1). Clinical features of these three patients are summarized in the Appendix. Plain chest radiographs were normal in all three patients with normal HRCTs, as well as in one patient with an abnormal HRCT. Thus, 3 of the 25 patients (12.0%) had HRCTs and 4 of the 25 patients (16%) had chest radiographs that did not reveal evidence of interstitial lung disease, when there was open lung biopsy evidence of idiopathic pulmonary fibrosis. The two radiologists agreed on the interpretation of the chest radiographs and HRCTs in all cases. Physiologic Assessment Patients with a normal HRCT had a similar perceived level of breathlessness (subjective score) compared with those whose scans were abnormal (Table 1). Neither the overall physiologic score nor spirometry (FEV 1 and FVC) differed significantly between those patients with a normal HRCT and those with an abnormal HRCT (Table 1). Additionally, total lung capacity (TLC) as a percent of predicted did not differ between the two groups (68.6 ± 9.5% for those with an abnormal HRCT vs 77.0 ± 3.6% for those with a normal HRCT, p=0.48). However, there was a significant difference between the two groups of patients in measurements that reflect abnormalities of gas exchange. Thus, DCO as a percent of predicted (46.1 ± 2.3% for those with an abnormal HRCT vs 65.7 ± 3.8% for those with a normal HRCT; p=0.01), DCO/ VA as a percent of predicted (Table 1), and resting P(A-a)Oz (Table 1) were all more abnormal in patients with an abnormal HRCT compared with those with a normal HRCT. During CPET, there was no significant difference between the two groups in exercise score (Table 1), in maximal achieved Vo2 as a percent of predicted (64.03 ± 4.26% for those with an abnormal HRCT vs 67.5 ± 2.0% for those with a normal HRCT; p=0.47), or in maximal achieved work rate (100.0 ± 0.0 W for those with an abnormal HRCT vs ± 3.9 W for those with a normal HRCT; p=0.33). However, despite this similarity in achieved Voz, there was a significant difference in 02 saturation at maximal exercise (87.6± 1.1% for those with an abnormal HRCT vs 96.3± 1.1 % for those with a normal HRCT; p=0.005). This result also indicates less severe abnormalities of gas exchange. Importantly, the total CRP score was lower (less abnormal) for the group with a normal HRCT (Table 1). Pathologic Scoring Similar pathologic features were evident in the lung biopsy samples from the three patients with HRCT not suggesting idiopathic pulmonary fibrosis. Figure 1 shows thin alveolar septa that were delicate in most areas of the parenchyma. A usually subtle alveolar wall infiltrate of lymphocytes and rare histiocytes was present in many areas. Small numbers of intra-alveolar macrophages were evident as was focal alveolar wall fibrosis. The airways appeared normal. Pneumocyte hyperplasia and alveolar wall metaplasia were only rarely observed. Extensive fibrosis or honeycomb formation was not identified. These mild features were confirmed by the low scores for all pathologic criteria shown in Table 2. In contrast, those patients with HRCT indicating interstitial lung disease showed significantly worse pathologic changes in all areas independently scored (Table 2). DISCUSSION Three salient findings emerge from our study. First, as previously documented for normal plain chest radiographs, 15 a normal HRCT does not exclude clinically significant idiopathic pulmonary fibrosis in symptomatic patients. Second, in our group of patients, physiologic studies were more sensitive than HRCT in detecting idiopathic pulmonary fibrosis. Third, patients with an HRCT that did not indicate idiopathic pulmonary fibrosis had milder disease, as demonstrated by the results of pathologic and CRP scoring. Although our study shows that HRCT is, in fact, a good tool for detecting idiopathic pulmonary fibrosis with a sensitivity of 88%, it is not sensitive enough to detect all cases, particularly when mild disease exists. With the powerful imaging capabilities of 112 Clir.ical Investigations

5 HRCT, many clinicians have come to rely on this test to exclude such diseases as idiopathic pulmonary fibrosis when evaluating patients with dyspnea. However, previous studies have not documented a falsenegative rate for HRCT in detecting idiopathic pulmonary fibrosis. The findings from our study suggest that reliance on HRCT alone is not appropriate. Additional studies in patients with interstitial lung disease because other confirmed etiologies are needed. Despite the less-than-perfect sensitivity of HRCT, we believe it is a highly useful part of the workup of the dyspneic patient with decreased DCO. Numerous radiologic studies have documented a high correlation between pathologic findings and radiologic changes with HRCT.l 6 17 HRCT complements physiologic assessment by pointing toward likely specific etiologies. HRCT can reliably detect macroscopic emphysema, an alternative cause of altered gas exchange, even in patients without evidence of airflow obstruction.l 8 In patients with focal abnormalities, HRCT is very helpful in selecting sites for biopsy. HRCT is also superior to plain radiographs or conventional CT scanning in following potential changes during therapy of idiopathic pulmonary fibrosis, including resolution, stabilization, or deterioration It is important to recognize that PFT results were abnormal in every patient with idiopathic pulmonary fibrosis proved by biopsy specimen, including those in whom HRCT failed to detect the abnormality. The most common PFT abnormality was a reduced DCO. Thus, our study supports previous reports that have shown that abnormal pulmonary function (particularly a low DCO) in symptomatic patients requires further assessment, including lung biopsy.l 5 21 However, physiologic tests, although highly sensitive, are nonspecific and cannot be relied on solely to establish a firm diagnosis in interstitial lung disease. All patients in this study were referred to our tertiary care center, which has a widely recognized interest in interstitial lung disease. Moreover, all patients were symptomatic and had physiologic abnormalities. Hence, our data should not be extrapolated to the treatment of asymptomatic patients. Although our study documented only three cases in which HRCT failed to detect idiopathic pulmonary fibrosis, such patients are perhaps the most important to identify, as they may be the single group with the most likely chance of responding to therapy. Idiopathic pulmonary fibrosis has been postulated to progress from an early inflammatory stage, which may respond to therapy, to fibrosis, which is seldom reversible This hypothesis is supported by the observation that patients with idiopathic pulmonary fibrosis with predominantly ground-glass opacification (the radiologic correlate of early disease-alveolitis) on initial HRCT rather than reticular patterns (the radiologic correlate of late disease-fibrosis) have improved response to therapy and better survivaj Determining whether patients with initially milder, radiologically inapparent disease have a better response to therapy will require further studies with long-term follow-up. It is highly unlikely that the cases not detected by HRCT were due to technical inadequacy of imaging. In our study, the HRCT was performed with 1.0- to 1.5-mm sections at 1-cm intervals throughout the entire chest. This procedure differs from that of some centers, where thin sections are obtained at only three or four levels throughout the chest. Since idiopathic pulmonary fibrosis may cause patchy abnormalities with normal-appearing lung adjacent to highly distorted parenchyma, the latter protocol may miss areas of diseased lung. Furthermore, in our study, two highly experienced CT radiologists agreed on the interpretation of the scans in all cases; thus, there was no interobserver variability. To establish a standard to compare prospectively individual patients with idiopathic pulmonary fibrosis, we used previously published scoring methods for pathologic and physiologic assessments (CRP and pathologic score).l 3.l 4 Such a standardized scoring system is necessary for assessing a disease such as idiopathic pulmonary fibrosis, which is relatively rare and which has a heterogeneous clinical course. Data from our study support previous findings of studies using the CRP score, in that individual components of the CRP may not be significantly abnormal until added together into a total composite score. Furthermore, this total score correlates closely with the degree of anatomic abnormalities found at open lung biopsy.l 3 Based on the results of this study, we conclude that in the evaluation of patients with dyspnea, abnormal pulmonary function, and suspected interstitial lung disease, a normal HRCT does not exclude the presence of idiopathic pulmonary fibrosis. Physiologic testing, including standard PFT and CPET, should be pursued early in the evaluation of patients with unexplained dyspnea, even if imaging procedures are normal. If the clinical suspicion for interstitial lung disease is high, a normal HRCT should not exclude further evaluation with definition of etiology by lung biopsy specimen. ACKNOWLEDGMENTS: The authors wish to acknowledge Janet Hampton, RN, who coordinated patient evaluations and collected data for this study. We also wish to acknowledge M. Anthony Schork, PhD, and Molly Young for their assistance with statistical analysis. CASE 1 APPENDIX A 38-year-old black woman was evaluated at an outside hospital for chronic dyspnea and cough. She CHEST / 108/1 I JULY,

6 FIGURE 2. Open lung biopsy specimen. Small numbers of macrophages lie within alveolar spaces. Rare alveolar septae are thickened by collagen deposits (arrows) (hematoxylin-eosin, original magnification Xl65). had been exposed to toluene diisocyanate (TDI) during a brief period at her work place and had been treated for TDI-induced asthma. Subsequent pulmonary function studies revealed an FVC of 1.97 L (51 % of predicted), FEY 1 of L (57% of predicted), TLC of 2.76 (51% of predicted), and DCO of ml/ min/ mm Hg (52% of predicted), supporting a restrictive ventilatory defect with abnormal gas exchange. Medical history was unremarkable. The patient was a nonsmoker with no other chemical exposures. Results of physical examination were normal. A CPET confirmed an aerobic limitation (Vozmax=68% of predicted), normal anaerobic threshold, and a reduced ventilatory reserve (16% [normal >30%]). Pressure-volume curves demonstrated marked reductions in lung compliance at all lung volumes. HRCT was interpreted as normal by two thoracic CT radiology specialists (Fig 1). Transbronchial bronchoscopic lung biopsies were normal. A representative sample from the open lung biopsy is shown in Figure 2. She was placed on a regimen of oral prednisone at 60 mg/ d for 3 months that was subsequently tapered to 10 mg/ d of prednisone. Her pulmonary function and symptoms have stabilized after 6 months of therapy. CASE 2 A 43-year-old white woman presented with a 1-year history of progressive dyspnea. She had been hospitalized 1 year earlier with chest pain and breathlessness at which time myocardial infarction was ruled out. Cardiac catheterization revealed normal coronary arteries, ventricular function, and pulmonary artery pressure. Ventilation perfusion lung 114 scanning was low probability for pulmonary embolism, and lower extremity venous Doppler studies were negative for venous thrombosis. Over the ensuing year, she noted progressive dyspnea limiting her ability to perform moderate degrees of exercise. Medical history was notable for a partial thyroidectomy at age 16 years, a total abdominal hysterectomy and oophorectomy at age 24 years, and a cholecystectomy at age 36 years. She was a nonsmoker without significant occupational exposure. Physical examination revealed well-healed surgical scars. Chest radiograph was normal. PFTs included a FVC of 2.81 L (78% of predicted), FEY 1 of 2.18 L (78% of predicted), TLC of 4.01 L (75% of predicted), and DCO of (57% of predicted) confirming a mild restrictive ventilatory defect. CPET confirmed ventilatory limitation with abnormal widening of the P(A-a)Oz. HRCT was not indicative of interstitial lung disease. Bronchoscopic transbronchial biopsy specimens were normal. Open lung biopsy identified findings suggestive of idiopathic pulmonary fibrosis. Three months treatment with 1 mg/ kg/ d of prednisone resulted in no clinical or physiologic benefit. Subsequent therapy with cyclophosphamide (2 mg/ kg/ d) resulted in stabilization of symptoms and pulmonary function. CASE 3 A 43-year-old black woman with Madan's syndrome was referred with breathlessness for 1 year. She denied fever, chills, night sweats, sputum production, or cough. Medical history was notable for thyroidectomy for hyperthyroidism and pericarditis with surgical drainage of a pericardia! effusion at age 40 years. She was a nonsmoker and had no significant occupational exposure. Physical examination was notable only for marfanoid features. Chest radiograph was normal. PFTs demonstrated a FVC of 3.32 L (78% of predicted), FEY 1 of 2.71 L (82% of predicted), TLC of 5.26 L (82%), and a DCO of ml/ min/ mm Hg (68% of predicted). CPET confirmed aerobic limitation with a Vozmax of 1.16 L (63.7%) and reduced ventilatory reserve (6.1%; normal > 30%). HRCT did not reveal evidence of interstitial lung disease. Bronchoscopy with transbronchial biopsy specimens showed patchy interstitial fibrosis confirmed by trichrome staining. Open lung biopsy specimens identified findings suggestive of idiopathic pulmonary fibrosis. Prednisone (1 mg/ kg/ d) was given for 3 months with stability in physiologic parameters but no significant improvement. A subsequent trial of cyclophosphamide (2 mg/ kg/ d) was discontinued due to severe side effects. She has not been receiving immunosuppressive therapy for 8 months with no evidence of deterioration. Clinical Investigations

7 REFERENCES Martinez FJ, Stanopoulos I, Acero R, et al. Graded comprehensive cardiopulmonary exercise testing in the evaluation of dyspnea unexplained by routine evaluation. Chest 1994; 105: Webb WR. High resolution lung computed tomography: normal anatomic and pathologic findings. Radio! Clin North Am 1991; 29: Tung KT, Wells AU, Rubens MB, et al. Accuracy of the typical computed tomographic appearance of fibrosing alveolitis. Thorax 1993; 48: Panos RJ, King TE Jr. Idiopathic pulmonary fibrosis. In: Lynch JP III, DeRemee R, eds. Immunologically mediated pulmonary disease. Philadelphia: JB Lippincott; 1991: Lynch JP III, Chavis AD. Chronic interstitial lung disorders. In: Victor L, ed. Clinical pulmonary medicine. Boston: Little Brown; 1992: Webb WR. High-resolution computed tomography of the lung: normal and abnormal anatomy. Semin Roentgenol 1991; 26: ll Morris J, Koski A, Johnson L. Spirometric standards for healthy nonsmoking adults. Am Rev Respir Dis 1971; 103: Dubois A, Bothelho S, Bedell G, et al. A rapid plethysmographic method for measuring thoracic gas volume: a comparison with a nitrogen washout method for measuring functional residual capacity in normal subjects. J Clin Invest 1956; 35: Goldman H, Becklake M. Respiratory function tests: normal values at median altitudes and the prediction of normal results. Am Rev Tuberc 1959; 79: Ogilvie C, Forster R, Blakemore W, et al. A standard breathholding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Clin Invest 1957; 36: Crapo R, Morris A. Standardized single breath normal values for carbon monoxide diffusing capacity. Am Rev Respir Dis 1981; 123: Jones NL, Makrides L, Hitchcock C, et al. Normal standards for an incremental progressive cycle ergometer test. Am Rev Respir Dis 1985; 131: Watters LC, King TE, Schwarz Ml, et al. A clinical, radiographic and physiologic scoring system for the longitudinal assessment of patients with idiopathic pulmonary fibrosis. Am Rev Respir Dis 1986; 133: Cherniack RM, Colby TV, Flint A, et al. Quantitative assessment of lung pathology in idiopathic pulmonary fibrosis. Am Rev Respir Dis 1991; 144: Epler GR, McCloud TC, Gaensler EA, et al. Normal chest roentgenograms in chronic diffuse infiltrative lung disease. N Eng! J Med 1978; 298: Mathieson JR, Mayo JR, Staples CA, et al. Chronic diffuse infiltrative lung disease: comparison of diagnostic accuracy of CT and chest radiography. Radiology 1989; l7l:ll Padley SPG, Hansell DM, Flower CDR, et al. Comparative accuracy of high resolution computed tomography and chest radiography in the diagnosis of chronic diffuse infiltrative lung disease. Clin Radiol1991; 44: Sanders C, Nath PH, Bailey WC. Detection of emphysema with computed tomography: correlation with pulmonary function tests and chest radiography. Invest Radiol1988; 23: Veda! S, Welsh EV, Miller RR, et al. Desquamative interstitial pneumonia: computed tomographic findings before and after treatment with corticosteroids. Chest 1988; 93: Akira M, Sakatani M, Veda E. Idiopathic pulmonary fibrosis: progression of honeycombing at thin-section CT. Radiology 1993; 189: Wright PE, Sheller JR. What is the meaning of an isolated reduction in single breath diffusing capacity for carbon monoxide? Respir Care 1988; 33: Carrington C, Gaensler E, Coutu R, et al. Natural history and treated course of usual and desquamative interstitial pneumonia. N Eng] J Med 1978; 298: Lee JS, lm J, Ahn JM, et al. Fibrosing alveolitis: prognostic implications of ground glass attenuation at high-resolution CT. Radiology 1992; 184: Wells AU, Rubens MB, dubois RM, et al. Serial CT in fibrosing alveoli tis: prognostic significance of the initial pattern. AJR 1993; 16l:ll59-65 CHEST/108/1 /JULY,