• Sarcoidosis masquerading as eosinophilic pneumonia SANDRA K. WILLSIE-EDIGER, DO GARY A. SALZMAN, MD WILLIAM P. SCHAETZEL, DO A 29-year-old woman pre­ nosis without biopsy.1,2 We describe the clini­ sented with progressive dyspnea, fever, cal course of a patient with roentgenographic cough, and weight loss. A chest roent­ and clinical features compatible with CEP in genogram revealed bilateral peripheral in­ whom a trial of corticosteroids did not support filtrates suggestive of chronic eosinophilic the diagnosis. Sarcoidosis was diagnosed by pneumonia. Bronchoscopic evaluation, as findings observed in tissue recovered at open well as a therapeutic trial of cortico­ lung biopsy. steroids, was nondiagnostic. Open lung bi­ opsy revealed findings consistent with a Report of case diagnosis of sarcoidosis. Roentgenographi­ A 29-year-old woman was hospitalized with a 1- cally, differentiating between sarcoidosis month history of cough, progressive dyspnea, chest and chronic eosinophilic pneumonia can tightness, low-grade fevers, generalized malaise, be difficult. A diagnostic approach, as well and an associated 3-kg weight loss. The patient de­ scribed her cough as initially nonproductive, chang­ as the differential diagnosis of bilateral pe­ ing to productive 2 weeks before admission. She ripheral pulmonary infiltrates, is dis­ reported no other significant symptoms. cussed. The patient denied using any prescription medi­ (Key words: Sarcoidosis, eosinophilic cations. She said she had used an over-the-counter pneumonia, pulmonary infiltrates) cold preparation on two occasions before admission. She reported 25 pack-years of tobacco use and the Chronic eosinophilic pneumonia (CEP), as episodic use of free-base cocaine and marijuana. first described by Carrington and associates Her occupational history was unremarkable; the in 1969,1 is associated with characteristic pe­ patient had previously been employed only as a wait­ ress. ripheral pulmonary infiltrates described as the She had a temperature of 99.4°F and a respira­ "photographic negative" of pulmonary edema. tory rate of 24 per minute; chest examination re­ It has been suggested that typical roentgenogra­ vealed diffuse expiratory wheezing. The patient's phic findings in combination with compatible admission chest roentgenogram is shown in Fig­ clinical features and a rapid therapeutic re­ ure 1. Results of an intermediate-strength purified sponse to corticosteroids may permit the diag- protein derivative skin test were negative; delayed hypersensitivity responses to candida and mumps Dr Willsie-Ediger is assistant professor and Dr Salzman antigens were intact. is associate professor of medicine and pulmonary dis­ eases, School of Medicine, University of Missouri-Kan­ The patient's complete blood cell count revealed sas City, Mo, and Dr Schaetzel is assistant professor of a total white blood cell count of 5.9 x 103 cells/ pathology, University of Health Sciences, College of Os­ mm3 with 4 eosinophils, 49 segmented neutrophils, teopathic Medicine, Kansas City, Mo. 42 lymphocytes, 4 monocytes, and 1 basophil. Ar­ Reprint requests to Sandra K. Willsie-Ediger, DO, Uni­ terial blood gases on room air revealed a pH of7 .44, versity of Missouri-Kansas City School of Medicine, 2411 Holmes, Kansas City, Mo 64108. a PaC02 of 35 mm Hg, and a PaC02 of 90 mm Hg. Case report· Willsie-Ediger et a l JAOA • Vol 92 • No 6 • June 1992 • 795 remained stable clinically and by pulmonary func­ tion testing. Discussion Since its description in 1969,1 CEP has been considered diagnosable when the classic roent­ genographic findings described as the "photo­ graphic negative" of pulmonary edema are seen in association with typical clinical fea­ tures,1.3 including fever, dyspnea, productive cough, and weight loss. Recent reviews, how­ ever, have suggested that the typical roent­ genographic features of CEP may be present in only one fourth of the cases, and that, in fact, the chest roentgenogram may be entirely Figure 1. Posteroanterior chest roentgenogram reveal­ normaP ing bilateral peripheral infiltrates (with upper lobe pre­ The differential diagnosis of peripheral pul­ dominance) in the absence of hilar or mediastinal ade­ monary infiltrates (in addition to CEP) in­ nopathy. cludes Loeffler's syndrome, mycobacterial and fungal diseases, periarteritis nodosa, Churg­ The serum angiotensin-converting enzyme level Strauss vasculitis, lymphoma, eosinophilic was 48 U/L (normal, 8 to 53 U/Ll, and sputum ex­ granuloma, desquamative interstitial pneumo­ amination failed to reveal typical or atypical patho­ nitis, pneumonitis or fibrosis complicating ax­ gens. Results of pulmonary function studies were illary radiation, bronchiolitis obliterans organ­ within normal limits, with the exception of a total izingpneumonia (HOOP), and sarcoidosis. 1·3,6,7 lung capacity of 4.0 L (69% of predicted values) and Common chest roentgenographic features of a diffusion capacity for carbon monoxide of 55% sarcoidosis have been described, including tho­ of predicted values. racic lymphadenopathy with or without par­ Bronchoscopy with bronchoalveolar lavage and enchymal infiltrates.8,g Uncommonly, periph­ transbronchial biopsies was performed. The bron­ eral pulmonary infiltrates may be seen in sar­ choalveolar lavage was performed in the posterior coidosis which resemble those seen typically segment of the right upper lobe with less than 25% in CEP.7,lO Glazer and associates7 delineated return of instilled fluid. The bronchoalveolar lav­ age differential revealed 84% macrophages, 13% several roentgenographic features useful in dis­ lymphocytes, and 3% eosinophils. Results of spe­ tinguishing between these two diseases. These cial stains for acid fast, fungal, and parasitic mi­ features include adenopathy and a nodular qual­ croorganisms (as subsequent culture results) were ity of the infiltrates, the presence of which is negative. The transbronchial biopsies were remark­ more suggestive of a diagnosis of sarcoidosis able only for the finding of a single noncaseating than of CEP.7 granuloma. Our patient had clinical and roentgenogra­ Therapy was instituted with prednisone, 40 mg phic features compatible with a diagnosis of daily, and maintained for 3 weeks. The patient's CEP. Results of the transbronchial biopsies, symptoms and roentgenographic abnormalities per­ which revealed a single noncaseating gran­ sisted despite this therapy. The prednisone ther­ uloma, and the lack of peripheral eosinophilia, apy was discontinued, and an open lung biopsy of did not initially dissuade us from this diagno­ the left upper lobe was performed. Grossly, diffuse parenchymal nodularity extending to the pleura sis because a lack of peripheral eosino­ was found. Histopathologic findings included mul­ philia,1,3,4 as well as the presence of noncase­ tiple non caseating granulomas, asteroid bodies, and ating granulomas,1,4 has been described in interstitial fibrosis (Figure 2). These findings were CEP. consistent with a diagnosis of sarcoidosis. In addition, the bronchoalveolar lavage dif­ At long term follow-up (36 months), the patient ferential was made suspect by the poor return 796 • JAOA • Vol 92 • No 6 • June 1992 Case report· Willsie·Ediger et a l Figure 2. Open lung biopsy specimen revealing a noncaseating granuloma and an aster­ oid body (arrow), consistent with a diagnosis of sarcoidosis (original magnification x 1000). of instilled fluid. The absence of adenopathy has been described.13.14 Bronchoalveolar lav­ and the nodularity of the infiltrates seen on age eosinophil counts i.n five untreated CEP chest roentgenogram also suggested a diagno­ patients ranged from 14% to 75% as reported sis of CEP as opposed to sarcoidosis when by Dejaegher and Demedts.!4 In sarcoidosis, Glazer's criteria were used.7 The absence of bronchoalveolar lavage findings are extremely a response to corticosteroids led us to perform variable and are dependent on both the activ­ an open lung biopsy; hence, a diagnosis of sar­ ity and stage of the disease.15 Bronchoalveo­ coidosis was confirmed. lar lavage differentials from patients with sar­ The use of free-base cocaine has been asso­ coidosis, as opposed to those from patients with ciated with significant respiratory complica­ CEP, rarely include significant numbers of tions and has been the subject of a recent re­ eosinophils. view.!l Hypersensitivity pneumonitis has been Certainly, when bronchoalveolar lavage or reported as a complication of this form of sub­ transbronchial biopsy or both are nondiagnos­ stance abuse12; typical features include fever, tic for CEP or when the disease course does transient pulmonary infiltrates, eosinophilia, not respond to a trial of corticosteroids, open and bronchospasm temporally linked to the in­ lung biopsy is indicated to rule out other pa­ halation of cocaine. In addition, BOOP has thologic entities that may require alternative been associated with free-base cocaine use. As modes of therapy. Additional information pro­ previously indicated, BOOP has also been re­ vided by open lung biopsy is often related to ported to exhibit peripheral pulmonary infil­ the larger tissue sampling available from trates.6 It is unlikely, however, that either of grossly abnormal regions oflung parenchyma, these reported complications was a factor in as opposed to the random small samples re­ this case because the clinical and pathologic trieved by transbronchial biopsy. findings are not supportive of either diagnosis. The utility of bronchoscopy and, in particu­ Comment lar, bronchoalveolar