DIAGNOSTIC-VALUE OF BRONCHOALVEOLAR LAVAGE IN PERIPHERAL LUNG-CANCER

A prospective study has been performed to assess the value of the addition of bronchoalveolar lavage (BAL) to the routine bronchoscopic exploration with bronchial washing (BW) and postbronchoscopy sputum (PBS) procedures in the diagnosis of peripheral primary lung cancer not visible through bronchoscope when fluoroscopic guidance is not available. BW, BAL, and PBS were performed in 67 patients with suspected primary lung cancer by peripheral lung lesion on chest radiograph (39 nodules and 28 infiltrates) and nonendoscopically visible lesion. The sequence of procedures was in all cases BW, BAL, and postbronchoalveolar lavage bronchoaspirate (PBBA). An attempt was made to collect early morning postbronchoscopy samples of sputum (PBS) on 3 consecutive days. BW and PBBA were collected in the same test tube, and the cytologic result was considered as BW diagnostic yield. If there were negative bronchoscopic results, either percutaneous fine-needle aspiration or open-lung biopsy were performed for diagnosis. Fifty-five patients were found to have malignant disease (23 adenocarcinomas, 22 squamous cell carcinomas, six small cell carcinomas, and four bronchioloalveolar cell carcinomas). BAL was positive in 18 of the 55 (33%) carcinomas, and it gave the only positive result in six (11%). BW was also positive in 18 of the 55 (33%), but it gave positive results in only 3 (5%). PBS was positive in 13 of the 43 (30%) patients from whom samples could be spontaneously obtained and were suitable for cytologic examination (not consisting of saliva), and gave the only positive result in three (7%). The combination of cytologic study of the three samples produced a significantly greater number of diagnoses (31 of 55, 56%) than each procedure alone (p < 0.02). The diagnostic sensitivity was higher for infiltrates (15 of 20, 75%) than for nodular lesions (16 of 35, 46%) without regard to histologic cell type (p = 0.035). Of 12 patients with nonmalignant diseases (five pulmonary tuberculosis and seven with other benign diseases), BAL was diagnostic in four of five patients with pulmonary tuberculosis, and it gave the only positive result in two. In short, in the diagnosis of lung cancer presenting as a peripheral nodule or infiltrate, when fluoroscopic control is not available, at least a 50% yield can be obtained from cytologic examination of BAL, aspiration of all bronchoscopic secretions before and after BAL (BW and PBBA), and 3-day collection of early-morning sputum after bronchoscopy (PBS). Additional advantages are its low morbidity and the high diagnostic value in other diseases such as pulmonary tuberculosis.