Given uncertainties in the literature regarding the diagnostic yield of cytologic sampling in bronchial cancer, we decided to investigate this issue using conventional bronchoscopic instrumentation.Objectives:
The purpose of the study was to validate bronchoscopy-guided cytology sampling against forceps and surgical resection biopsies in central and peripheral bronchial cancer.Methods:
All patients presenting with potentially malignant pulmonary or mediastinal lesions on a chest x-ray or a chest computed tomography scan were investigated prospectively from July 1999 through July 2000. A total of 156 patients underwent flexible bronchoscopy and conscious sedation and local anesthesia. Samples were obtained using transbronchial or endobronchial biopsies (TBB/EBB = histology) and needle aspiration or endobronchial brushing (TBNA/SBB = cytology).Results:
Central lesions were found in 95 and peripheral lesions in 61 patients. Altogether, there were 154 malignancies and 2 benign lesions. Malignancy was confirmed by TBB/EBB alone in 98 of 156 patients (sensitivity 62.8%), and by TBNA/SBB alone in 108 of 156 patients (sensitivity 69.2%). Both methods combined assessed tumor tissue in 136 of 156 patients (sensitivity 87.2%). Forty-four of 156 patients underwent surgical resection. The sensitivities for tumor detection in this subgroup were comparable to the entire group. However, as a result of the frequently necessary postoperative adjustment of preoperative cell typing, the preoperative diagnosis was only 60% specific for tumor cell types.Conclusions:
In case of suspected lung cancer, bronchoscopic samples must be obtained for both histologic and cytologic analysis to achieve optimal diagnostic yield. The bronchoscopic diagnosis reliably distinguishes small-cell (SCLC) from nonsmall-cell lung cancer (NSCLC), and therefore determines the conventional therapeutic procedure (chemotherapy vs. surgery). On the other hand, given a low specificity for the particular tumor cell types as confirmed from the resected material, bronchoscopic biopsies or aspirates are unsuitable for database for tumor-cell-type incidence in the population at risk.