Open Lung Biopsy for Diffuse Infiltrative Lung Disease

Open Lung Biopsy for Diffuse Infiltrative Lung Disease Vijay Chechani, MD, Rodney J. Landreneau, MD, and Shabbir S. Shaikh, MD Division of Pulmonary Medicine, Department of Medicine, and Division of Cardiothoracic Surgery, Department of Surgery, Harry 5. Truman Memorial Veterans Hospital, University of Missouri-Columbia School of Medicine, Columbia, Missouri The purpose of this study was to determine the most appropriate site of biopsy in patients with diffuse infiltrative lung disease (DILD). Twenty patients were evaluated. During open lung biopsy, a representative region (inflamed and least fibrotic) of the radiographically most involved lobe was identified and a biopsy done. A second biopsy specimen was obtained from an adjacent accessible lobe. The biopsy specimen from each lobe was processed separately, and a descriptive histologic report was generated for each of the two lobes. Tissue was cultured for bacteria, fungi, and mycobacteria. In 17 patients, the second biopsy site had fewer roentgenographic abnormalities than the radiographically most involved lobe. In 3 patients, the number of radiographic abnormalities was equal in all lobes. The same patho- logic diagnosis was reached by histologic examination of the two biopsy specimens in 8 of 10 patients with acute DILD and in 10 of 10 patients with chronic DILD. In 2 patients, the biopsy specimen from the radiographically most involved lobe showed pathologic features not seen in the other specimen. Cultures from both biopsy specimens were grown separately in 6 patients. Fungi (n = 2) and bacteria (n = 1) were isolated from the radiographically most involved lobe (n = 2) or from the lobe of initial lung infiltration (n = 1) in 3 patients. When biopsy specimens are obtained from a representative region of the radiographically most involved lobe in patients with DILD, biopsy of other lobes is unnecessary. (Ann Tkorac Surg 1992;54:296-300) pen lung biopsy is often performed for the diagnosis 0 of diffuse lung disease. Biopsy of the most involved area of the lung is generally discouraged, as it usually shows end-stage disease of unrecognizable origin, whereas biopsy specimens from less involved but still diseased areas of the lung are most likely to show an active and recognizable process [l]. About 10% of patients with histologically confirmed chronic diffuse infiltrative lung disease (DILD) may have a normal chest radiograph [2]. In most instances, the site of thoracotomy and lung biopsy is guided by the abnormalities noted on the chest roentgenogram. To ensure adequate sampling of the pathological process, biopsy of more than one area of the lung (with varying degrees of radiographic abnormalities) has been suggested. This study looks at the diagnostic benefits of obtaining multiple samples at the time of open lung biopsy. Material and Methods Pa tien t Selection Medical records and chest roentgenograms of all patients who underwent thoracotomy and open lung biopsy at the University of Missouri-Columbia Hospitals and Clinics between July 1988 and June 1990 were reviewed to identify patients with DILD who had undergone biopsy of two or more lobes and whose biopsy specimens had been processed separately for histologic examination. If the Accepted for publication Jan 9, 1992. Address reprint requests to Dr Chechani, 1600 SE Main, Roswell, NM 88201. specimens from the two lobes were not examined separately, the patient was not included in the study. Open Lung Biopsy According to standard practice, a representative region (inflamed and least fibrotic) of the radiographically most involved lobe was selected for biopsy. A second biopsy specimen was obtained from an adjacent lobe accessible through the thoracotomy incision. Samples from both biopsy sites were processed separately for frozen-section and permanent-section histologic examinations. Bacterial, fungal, and mycobacterial cultures were grown from tissue from one or both lobes from all patients. Viral cultures were obtained when clinically indicated. The pathologist was not aware of the severity of the disease in the two lobes from which the specimens were obtained. The pathologist was not asked to comment on the severity of different pathologic changes in the two biopsy specimens, but specific diagnoses were assigned to each specimen. Descriptive histologic reports on each biopsy specimen from each patient were analyzed. Chest Roentgenogram Preoperative abnormalities on the chest roentgenogram were noted, and the severity of infiltrates in the second biopsy site was assessed to be equal to or less than that in the radiographically most involved lobe. All previous chest roentgenograms were reviewed to determine the initial site of involvement and the duration of the radiographic abnormality. 0 1992 by The Society of Thoracic Surgeons 0003-4975/92/$5.00

Ann Thorac Surg 1992;54296-300 CHECHANI ET AL 297 Data Collection The following information was tabulated: duration of pulmonary symptoms, reasons for open lung biopsy, tissue culture results, duration of chest tube drainage postoperatively, and changes in therapy, if any, as a result of open lung biopsy. Based on the duration of pulmonary symptoms, radiographic abnormality, or both, DILD was classified as acute (530 days) or chronic (>30 days). Pathologic reports on all biopsy specimens were reviewed to identify features present in only one biopsy site. The results of frozen-section examination were also noted and compared with the results of permanentsection examination. Results Forty-one open lung biopsies were performed during the 2-year period. Twenty patients with DILD qualified for the study. In each of these patients, the open lung biopsy specimen from each of the two lobes was processed individually. Ten patients had acute DILD and 10, chronic DILD. In 6 patients, cultures from both of the biopsy specimens were grown separately, whereas in the other 14 patients, cultures from a single biopsy specimen, usually from the radiographically most involved lobe, were grown. Thoracotomy incisions employed for wedge biopsies were limited axillary (n = 13), limited anterolateral (n = 5), and limited "true" lateral (n = 2). The radiographically most involved lobes were the lower (n = ll), upper (n = 3), and lingula (n = 3). The radiographically less involved lobes from which biopsy specimens were taken were the lower (n = 5), upper (n = 9), lingula (n = 3), and middle (n = 1). In 3 patients, the radiographic abnormalities noted were equal in all lobes. In all, open lung biopsy samples were obtained from 17 lower lobes, 15 upper lobes, 7 lingulas, and 2 middle lobes. Standard computed tomography of the chest was performed preoperatively in 5 patients. It confirmed the sites, severity, and type of radiographic abnormalities seen on the chest roentgenograms of these patients. Table 1 and Table 2 summarize the clinical features, the radiographic and pathologic abnormalities, and the culture results for the 20 patients with acute and chronic DILD. All patients had benign disease. The same pathologic diagnosis was reached by histologic examination of the two biopsy specimens in 8 of 10 patients with acute DILD and in all 10 patients with chronic DILD. Additional pathologic abnormalities not seen in the biopsy specimen from the radiographically less involved lobe were noted in the biopsy specimen from the radiographically most involved lobe in 2 patients with acute DILD (patients 4 and 5; see Table 1). Although there is not a significant difference between the yields from the two biopsy sites (x2 analysis), biopsy of the radiographically most involved lobe in 1 patient (patient 5; see Table 1) led to the administration of correct antibiotics with resolution of pneumonia. Bacterial, fungal, and mycobacterial cultures were sterile in 17 patients. The two positive isolates of fungi were Aspergillus fumigatus (patient 9; see Table 2) and Penicillium sp (patient 8; see Table 2). At postmortem examination, patient 9 was found to have invasive pulmonary aspergillosis. Although equal consolidation of the two lobes from which biopsy specimens were taken was noted on the preoperative chest roentgenogram, only the culture from the right upper lobe was positive for Aspergillus furnigatus. Previous chest roentgenograms revealed that the infiltrates had started in the right upper lobe 6 weeks earlier. Patient 8 had end-stage lung disease with diffuse fibrosis and acute constrictive bronchiolitis obliterans. Penicillium organisms were isolated only from the lingula, which was the radiographically most involved lobe. The exact etiologic role of Penicillium species in the production of constrictive bronchiolitis obliterans could not be determined. Pseudomonas aeruginosa was isolated from the radiographically most invoived lobe in 1 patient (patient 5; see Table 1). Cultures from the radiographically less involved lobe in this patient were not grown. The diagnosis from the frozen-section examination was inaccurate in 5 of 7 patients with chronic lung disease and in 2 of 8 patients with acute lung disease. After open lung biopsy, chest tube drainage was required for 6.5 t 5.7 days in patients with acute DILD and for 11.6 * 9.9 days in patients with chronic DILD (p = not significant by analysis of variance). In 12 of the 20 patients, open lung biopsy did not lead to a change in therapy. It caused initiation of corticosteroid therapy with or without cytotoxic drugs in 7 patients and a change in antibiotic therapy in 1 patient. Improvement in pulmonary disease was noted in 6 of these 7 patients. Comment Etiologic diagnosis of DILD often requires open lung biopsy. The question then is, from which lobe should the biopsy specimen be obtained? Often the site selection is guided by the abnormalities noted on the chest roentgenogram. Biopsy of the moderately diseased region of the lung (as determined by inspection and palpation) is more likely to result in meaningful histologic findings [l]. When a biopsy specimen is obtained from such an "average" region of the radiographically most involved lobe, is there any advantage to obtaining additional specimens from the radiologically less involved lobes? Our study was designed to answer this question. The results of this study do not suggest a need for multiple biopsy specimens during open lung biopsy. Although there was a close correlation between the results obtained from both biopsy sites in these patients with DILD, the biopsy specimens from the radiographically most involved lobe had the better diagnostic yield. It is best to plan the thoracotomy approach and biopsy so as to sample moderately diseased tissue from the radiographically most involved lobe. Our recommendations may appear to be contradictory to those of Gaensler and Carrington [l], but they are not. In a large study of 502 patients with chronic DILD, Gaensler and Carrington recommended that an average

298 CHECHANI ET AL Ann Thorac Surg 1992;54:29&300 Table 1. Patients With Acute Difuse lnfiltrative Lung Disease Patient Age History/Reasons for Chest Roentgenogram No. (y) Sex Open Lung Biopsy RMIL RLIL Pathology Comments 1 62 M Lobectomy (RLL) for squamous cell carcinoma followed by irradiation and chemotherapy 4 y ago; rapidly progressive lung infiltrates; duration, 12 d 35 F SLE; LLL pneumonia 57 F with rapid worsening; FOB duration, 2 d Breast cancer 3 y ago; cough, dyspnea and bilateral lower lobe infiltrates; FOB LLL Lingula Organizing DAD, NOS Cultures: LLL and lingula, sterile FSD: carcinomatosis CTD: 5; Rx: NC RLL LLL RUL,RML 4 76 F Asthma, idiopathic Lingula LLL Emphysema, old hypertrophic subaortic pulmonary hemorrhage, stenosis,chf, CABG; tumorlets," focal cough, dyspnea, lung bronchiectasis" infiltrates; duration, 7d 5 51 M COPD (on prednisone); LLL LUL Focal necrotizing worsening patchy bronchopneumonia with alveolar infiltrates in NOS", emphysemab both lungs despite antibiotics; duration, 7d LUL Exudative DAD, NOS Exudative DAD, large arterial thrombi, NOS 6 17 F Pregnant; fever and Lingula LUL Acute interstitial rapidly progressive lymphohistiocytic infiltrates while on pneumonia, NOS antibiotics; FOB duration, 3 d Cultures: RLL, sterile FSD: DAD; CD: lupus pneumonitis; CTD: 3; Rx: IS with FSD: DAD CTD: 8; Rx: NC Cultures: lingula and LLL, sterile FSD: interstitial process CTD: 5: Rx: NC Cultures: LLL, Pseudomonas FSD: organizing pneumonia CTD: 5; Rx: antibiotics to cover Psseudomonas with FSD: interstitial pneumonitis Rapid clearing of infiltrates postoperatively CTD: 1; Rx: NC 7 54 F Myelodysplasia; fever, LUL LLL Patchy interstitial fibrosis Cultures: LUL and LLL, sterile hypoxia, and bilateral and lymphocytic FSD: mild mononuclear infiltrates more in infiltrates, NOS interstitial infiltrates upper lobes; FOB CTD: 5; Rx: IS with duration, 5 d 8 66 F Fever, cough LUL LLL Focal organizing Cultures: LUL, sterile unresponsive to pneumonia, mild FS: not done antibiotics, respiratory interstitial fibrosis, Serology positive for mycoplasma failure, bilateral pulmonary edema, NOS pneumonia interstitial infiltrates; CTD: 6; Rx: NC 9 50 M Alcoholic; fever, cough, A11 lobes equally Exudative DAD, focal Cultures: RUL and RLL, sterile vomiting, RUL involved (biopsy bronchopneumonia FSD: hyaline membranes and pneumonia leading to of RUL and RLL) bronchopneumonia rapidly worsening CTD: 21; Rx: NC diffuse infiltrates and respiratory failure; 10 43 M SLE; acute new LLL LUL Diffuse old intraalveolar interstitial nodular hemorrhage, NOS FS: not done; CD: lupus infiltrates pneumonia predominantly in Rx: IS with lower lobes; duration, 6d a These were seen only in biopsy specimen from RMIL. This was seen only in biopsy specimen from RLIL. CABG = coronary artery bypass grafting; CD = clinical diagnosis; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; CTD = chest tube drainage in days; DAD = diffuse alveolar damage; FOB = fiberoptic bronchoscopy; FS = frozen section; FSD = frozen-section diagnosis; IS = immunosuppression with corticosteriods 2 cytotoxic drug; LLL == left lower lobe; LUL = left upper lobe; NC = no change; NOS = no organism seen; RLIL = radiographically less involved lobe; RLL = right lower lobe; RMIL = radiographically most involved lobe; RML = right middle lobe; RUL = right upper lobe; Rx = treatment after open lung biopsy; SLE = systemic lupus erythematosus.

Ann Thorac Surg 1992;54:29f5-300 CHECHANI ET AL 299 Table 2. Patients With Chronic Difftlse Infiltrative Lung Disease Patient Age History/Reasons for Open Chest Roentgenogram No. (y) Sex Lung Biopsy MIL RLIL Pathology Comments 1 30 F 2 3 4 5 6 7 8 9 10 74 M 74 M 70 M 59 M 77 M 70 F 64 F 73 M 77 M Scleroderma, bilateral LLL Lingula BOOP. NOS nodular infiltrates most prominent in lower lobes; FOB duration, 2 mo Silica and dioxin exposure, LLL LUL Mixed-dust fibrosis, NOS dyspnea, home oxygen; bilateral interstitial infiltrates most prominent in both lower lobes; duration 6 mo Farmer with dyspnea and RLL RUL Moderate interstitial dense alveolar infiltrates in RLL, LUL, lingula; worsening of infiltrates in R lung from interstitial to alveolar; duration, 1 y Dyspnea and cough after LLL LUL fibrosis, noncaseating granuloma, NOS Focal organizing pneumonia 7 mo ago; bilateral interstitial-alveolar infiltrates worst in LLL with loss of volume in L lung; duration, 7 mo Fever, cough, dyspnea, LUL LLL pneumonitis, focal mild interstitial fibrosis and inflammation, NOS Hyalinized granulomas, weight loss, diffuse pulmonary calcification, localized consolidation in LUL, lingula, RUL and RML; cavitation in RUL and LUL; duration, 3 mo Dyspnea, home oxygen; LLL LUL multifocal fibrosis, NOS Interstitial fibrosis bilateral interstitial nodular infiltrates worst in both lower lobes with hilar adenopathy; duration, 1 y Hiatal hernia, restrictive lung All lobes equally End-stage lung with disease, dyspnea, dry involved honeycombing cough, home oxygen; (biopsy of bilateral small, fibrotic lingula and lungs for 6 y with further LUL) loss of volume 3-y history of pulmonary Lingula LUL End-stage lung with fibrosis, dyspnea, cough; exposure to chicken and duck coops; progressive interstitial fibrosis, loss of volume and honeycombing over 3 y worse in lower lobes; duration, 3 y diffuse fibrosis, acute constrictive bronchiolitis Agnogenic myeloid All equally Emphysema, interstitial ke&plasia,- RUL nodular invblved fibrosis, NOS infiltrate progressed to diffuse consolidation of R lung and infiltrates in L lung; duration, 45 d (biopsy of RUL and RML) Dyspnea and progressive Lingula LLL Giant cell vasculitis of bilateral infiltrates to white out RUL, alveolar infiltrates in LUL and interstitial infiltrates in both lower lobes, FOB duration, 40 d arteries and veins, interstitial inflammation and thickening BOOP = bronchiolitis obliterans organizing pneumonia; all other abbreviation are the same as in Table 1. FSD: active interstitial process CTD: 36; Rx: IS with FSD: silicosis CTD: 4; Rx: NC Cultures: RLL, sterile FSD: patchy interstitial acute and chronic inflammation CD: hypersensitivity pneumonia CTD: 9; Rx: IS without FSD: diffuse mild interstitial fibrosis, focal mild chronic inflammation CTD: 5; Rx: NC Cultures: LUL, sterile FSD: interstitial fibrosis Marked clearing of infiltrates in 1 mo; CTD: 15; Rx: NC FSD: interstitial fibrosis and chronic inflammation CTD: 3; Rx: IS without FS: not done CTD: 3; Rx: NC Died 2 mo later of respiratory failure Cultures: lingula, Penicilliurn sp; LUL, sterile FSD: focal fibrosis with chronic inflammation CTD: 10: Rx: NC Cultures: RUL, Aspergillus fumigatus; RML, sterile FSD: touch preparations: NOS CTD: 15; Rx: NC Autopsy: invasive pulmonary aspergillosis FSD: not done CD: Wegener s granulomatosis CTD: 16; Rx: IS with

300 CHECHANI ET AL Ann Thorac Surg 1992:54:29f%300 region of the lung be selected as the biopsy site, not the radiographically most involved area. After the incision, all lobes were inspected and palpated, less severely diseased regions were identified, and an average region was selected for biopsy. Most often, this was the inferior margin of the anterior segment of the upper lobe (a small anterior approach was used in the majority of patients). A specific diagnosis was made in 495 patients. Overall, severity of roentgenographic opacities correlated highly with pathologic severity. Our method of selection of the biopsy site from the radiographically most involved lobe was not much different, as efforts were made to avoid fibrotic regions of that lobe. Gaensler and Carrington [l] stated that biopsy of the roentgenographically most involved region (usually the lower lobes in patients with chronic DILD) is likely to show meaningless histological findings and that biopsy of lingular and middle lobes should be avoided. However, their study was not designed to make such a statement. They did not obtain specimens from two different lobes or examine them separately. Other investigators [3, 41 have found that lingular and middle lobe biopsy specimens can be useful in the diagnosis of DILD. Like the study of Gaensler and Carrington [l], our study confirms that a single appropriately selected site for open lung biopsy will yield the correct diagnosis in most patients. This site may be from the radiographically most involved lobe. Biopsy of other involved lobes may be necessary intraoperatively if extensive fibrosis is detected in the radiographically most involved lobe. The diagnostic efficacy of open lung biopsy in a patient with DILD is not perfect whether the specimen is obtained from the radiographically most involved lobe or a radiographically less involved lobe. No specific diagnosis was reached in 4 of our patients with chronic DILD (patients 4, 6, 7, and 8; see Table 2), despite biopsy specimens from two different lobes. Two of these patients with end-stage honeycomb lungs did have extensive fibrosis on the chest roentgenogram, and open lung biopsy should not have been done. Computed tomography of the chest in 5 of our patients did not help in the site selection. It may help define areas of honeycombing within the lung that can then be avoided during open lung biopsy. The role of frozen-section pathologic analysis is suspect in the diagnosis and management of patients with chronic DILD. Even when similar histologic diagnoses were reached by frozen-section and permanent-section examinations in 2 patients, no therapeutic benefits were realized from the information obtained from frozen section. Thus, in patients with chronic DILD, frozen-section examination should not be requested routinely. The exact role of frozen sections in acute DILD cannot be determined from this study, as none of our patients had opportunistic lung infections. Our study has a small number of patients and does not encompass all the disease processes that can lead to DILD. A larger prospective study is justified to verify the validity of our suggested approach. It is our impression that multiple pulmonary parenchymal samples need not be obtained during open lung biopsy. It appears that a single biopsy specimen from a representative area (an area with average disease) of the radiographically most involved lobe is more expedient and equally efficacious. References 1. Gaensler EA, Carrington CB. Open biopsy for chronic diffuse infiltrative lung disease: clinical, roentgenographic, and physiological correlations in 502 patients. Ann Thorac Surg 1980; 30:411-26. 2. Epler GR, McLoud TC, Gaensler EA, Mikus JP, Carrington CB. Normal chest roentgenograms in chronic diffuse infiltrative lung disease. N Engl J Med 1978;298:934-9. 3. Miller RR, Nelems B, Miiller NL, Evans KG, Ostrow DN. Lingular and right middle lobe biopsy in the assessment of diffuse lung disease. Ann Thorac Surg 1987;44:269-73. 4. Wetstein L. Sensitivity and specificity of lingular segmental biopsies of the lung. Chest 1986;90:383-6.