Digital Image Comparison of Cellular Yield in Bronchial Brushing: Pre- and Post-Biopsy Lavage Cytology

Abstract

Introduction: Bronchoscopy is a useful diagnostic tool capable of performing core biopsy, forceps biopsy, bronchoalveolar lavage, and bronchial brushing. This study compares the cellularity of bronchial cytology including pre- and post-biopsy lavage by digital image analysis, aiming to increase diagnostic and tumor yield by optimizing the sequence and combination of bronchial biopsy and cytology. Methods: Alveolar macrophage, bronchial epithelium, and tumor cell cellularity from liquid-based cytology preparations of bronchial brushing and pre-biopsy and post-biopsy bronchoalveolar lavage were annotated on digitized whole-slide images and compared. Secondary analysis on the relationship of tumor cell and non-lesional cell yield was performed. Results: Overall, 118 cytology specimens from 43 patients were retrieved in total. Bronchial epithelium count was higher in pre-biopsy than post-biopsy lavage (p < 0.01) but not for alveolar macrophages nor tumor cell (p > 0.05). Tumor cell count was higher for bronchial brushing cytology samples than lavage (p = 0.018). The alveolar macrophage count was higher in post-biopsy lavage than bronchial brushing (p = 0.033); otherwise, brushing showed consistently higher bronchial epithelium and tumor cell counts. There were 33 false negative (tumor cell absent) specimens, and the combination of bronchial brushing and pre-biopsy lavage yielded the lowest false negative cases. Correlation between bronchial epithelium and alveolar macrophage counts with tumor cell count was weak (correlation coefficient = −0.168–0.203) except for post-biopsy lavage (correlation coefficient = 0.412–0.479, p < 0.05). Conclusion: Bronchial brushing yields a greater amount of tumor cell than lavage, and timing lavage before or after core biopsy does not affect tumor cell yield. Combining bronchial brushing and pre-biopsy lavage results in the lowest false negative rate.