Bronchogenic Carcinoma Presenting as an Endobronchial Mass: Optimal Number of Biopsy Specimens for Diagnosis

Abstract

The purpose of this study was to determine the optimal number of biopsy specimens which should be taken of an endobronchial mass lesion to assure the diagnosis of bronchogenic carcinoma. Five biopsies were performed on each of 18 consecutive mass lesions. Before biopsy, the tumor was well visualized, cleared of blood if possible, and the forceps carefully placed on the mass. Each biopsy specimen was placed in a separate coded container so that the pathologic interpretation was blind to the order in which the samples were taken. Ten out of a total of 90 biopsy specimens were negative for tumor and all of the tumors were diagnosed within the first three biopsy samples. The totals of negative diagnoses for each of the five biopsy specimens were not significantly different by the approximate Cochran Q test (p = 0.441) or by the exact test (p = 0.360), and no evidence was found for dependence of successive biopsies (p = 0.106) or for a patient effect (p = 0.325). The estimated probability of obtaining a diagnosis of malignancy after one biopsy was .889; after two biopsies, .9877; and after three, .9986. The yield is not meaningfully increased with more biopsies being done. We suggest performing three biopsies of endobronchial mass lesions to achieve an optimal diagnostic yield with minimal risk of bleeding.