An evaluation of computed tomography as an aid to diagnosis in patients undergoing bronchoscopy for suspected bronchial carcinoma

AIM: To determine whether computed tomography (CT) can predict the likeliho od of obtaining a positive tissue diagnosis at fibreoptic bronchoscopy (FOB ), or demonstrate an alternative means of achieving a tissue diagnosis, in patients presenting with a high clinical suspicion of primary bronchogenic carcinoma and an abnormal chest radiograph (CXR). MATERIALS AND METHODS: Sixty-two patients presenting with a high clinical s uspicion of carcinoma and an abnormal CXR had chest and liver CT and FOE pe rformed, All patients subsequently had histocytological confirmation of mal ignancy. Features recorded from the CTs included: the site and characterist ics of a mass if present, and its relationship to adjacent airways; the pre sence of presumed metastatic disease; and a CT prediction of the likelihood of positive FOE was made, RESULTS: Of the patients, 41/62 (66%) had inoperable stage IIIb/IV disease. Fibreoptic biopsy yielded positive tissue diagnoses in 38/62 (61%), Comput ed tomography features predicting a positive FOE in this group included: il l-definition of the mass (12/15, 80%); a mass <4cm from the origin of the n earest lobar bronchus (36/53, 68%); an endobronchial component of mass (22/ 24, 92%); a segmental or larger airway leading to the mass (30/35, 86%). Ov erall, CT had positive and negative predictive values for positive FOE of 8 5% and 78% respectively, The accuracy of the overall CT prediction of posit ive FOB was better than the accuracy of any of the individual factors. Seve nteen of 62 (27%) patients had presumed metastatic disease suitable for per cutaneous biopsy. CONCLUSION: Computed tomography is useful in predicting the likelihood of a chieving positive histocytology at FOE. The overall CT prediction is superi or to any of the individual CT features taken alone, (C) 2000 The Royal Col lege of Radiologists.