Combining transbronchial aspiration with endobronchial and transbronchial biopsy in sarcoidosis.

This study was carried out to evaluate the diagnostic yield and safety of flexible transbronchial needle aspiration (TBNA), endobronchial biopsy (EBB) and transbronchial lung biopsy (TBLB) combinations in stages I-III sarcoidosis (SA). Between 1989 and 1997, 74 patients suspected of having SA underwent fibreoptic bronchoscopy along with TBNA + EBB + TBLB or EBB + TBLB. During the same fibreoptic bronchoscopy, TBNA (using a 19-gauge histological needle and contrast-enhanced computed tomography (CT) guidance), EBB (from abnormal or normal bronchial mucosa) and TBLB were performed in stages I (n = 33) and II (n = 25), and EBB and TBLB in stage III (n = 16). The diagnosis of SA required the presence of noncaseating granulomas and the absence of "allergic granulomatosis and angiitis with eosinophilic infiltration" or foreign body reaction, with negative Ziehl-Neelson and methenamine silver stains, as well as negative cultures for acid-fast bacilli, fungi and other organisms. TBNA was diagnostic in 20 (61%) and 10 (42%) cases of stages I and II, whereas EBB was diagnostic in 15 (45%), 12 (50%) and seven (58%) cases, of stages I, II and III, respectively, and TBLB in 17 (52%), 15 (63%) and 10 (83%). By means of TBNA, EBB and TBLB alone, the diagnostic yields were nine (27%), four (12%) and six (18%) cases in stage I, two (8%), four (17%) and seven (29%) in stage II, and two (17%) and five (42%) by EBB and TBLB in stage III, respectively. The diagnostic yield of TBNA + EBB + TBLB was 30 (91%) in stage I and 21 (88%) in stage II, and that of EBB + TBLB was 12 (100%) in stage III. Overall, 63 (91%) cases of SA were diagnosed by TBNA + EBB + TBLB and EBB + TBLB. In six of the eleven bronchoscopically-negative cases, mediastinoscopy (four) or thoracotomy (two) established the diagnosis of SA, whereas the remaining five were diagnosed along with non-sarcoidosis diseases (one case in stage II and four in stage III) by various tissue biopsies in the follow-up. The combination of TBNA + EBB + TBLB and EBB + TBLB provided an overall sensitivity and accuracy of 90% and a specificity of 100%. All six (9%) significant complications--pneumothorax (four) and 40-100 mL haemorrhage (two)--were attributable to TBLB. It is, therefore, inferred that the combination of transbronchial needle aspiration, endobronchial biopsy and transbronchial lung biopsy in stages I and II, and that of endobronchial biopsy and transbronchial lung biopsy in stage III, is safe and cost-effective as well as increasing the diagnostic yield, and should therefore be performed routinely in the diagnosis of sarcoidosis.