Bronchial brushing in bronchogenic carcinoma. Factors influencing cellular yield and diagnostic accuracy.

In a prospective study, varying the bronchial brushing technique through the flexible fiberoptic bronchoscope was examined for the effects of cellular yield and diagnostic accuracy. The yield of cells obtained from the 1.7-mm brush was increased by more than twofold when the brush and bronchoscope were withdrawn as a unit through the pharynx and nose, (nonwithdrawn brushing), rather than withdrawing the brush alone through the aspiration channel of the bronchoscope (withdrawn brushing). The greater number of total cells in the nonwithdrawn brushing specimen was largely due to squamous cells, and the yield of tumor cells by the two methods was similar. The yield of cells from the same brush withdrawn through a new suction adapter available on the bronchoscope (Olympus BF-B2) was significantly decreased, as compared with the withdrawn brushing. There was a further decrease using a shielded 1-mm brush. In 23 of 30 cases of proven carcinoma, the diagnosis was made by the single withdrawn brushing, and in 26 of the 30 by the single nonwithdrawn brushing, a difference which was not statistically significant. We conclude that diagnostic accuracy is not significantly enhanced by withdrawing the brush and bronchoscope as a unit through the pharynx and nose; use of a small shielded brush offers no advantage over a larger unshielded brush; modifications of flexible bronchoscopes, such as the adapter to improve efficacy of suction, should be evaluated for effects on harvests of cells from bronchial brushes before being put into general use.