Filtered Lung Specimens Imaged with Proprietary Label-Free Imaging Platform for ROSE

Summary: Introduction: Currently lung biopsies produce aliquots of cytological specimen, and for procedures utilizing ROSE, these cytology samples are smeared on slides, air dried, and typically Diff-Quik stained to provide a rapid differential between nucleated cells to inform the proceduralist on each biopsy location/pass for specimen adequacy. Materials and Methods: Unique samples from three cases were processed through the F50 System (Cell Solution) to filter out blood and produce a monolayer specimen. The unstained samples were imaged on a proprietary label-free imaging system (Pathware). These slides were subsequently stained with Diff-Quik and scanned on a digital microscope. The three lung biopsy procedures were performed at Grady Memorial Hospital: 1. Percutaneous needle biopsy of LUL mass 2. Bronchoscopic forceps biopsy of LUL nodule 3. Bronchoscopic FNA of LUL nodule Results: For the F50 filtered samples, both the label-free image and the stained brightfield image for each case were reviewed and determined to be adequate samples for their respective nucleated cells (Figures 1-4). Final pathology results for procedures 1 and 2 were malignant, while procedure 3 was benign (reactive). Utilizing a smaller volume sample for ROSE can enable more specimen to be sent to the lab for diagnostic processing and review. Filtering out blood cells may make rapid evaluation faster. Using a label-free microscope not only saves staining time but also reduces staining variability, eliminates added staining artifacts, and prevents cell wash-off from the staining process. The label-free imaging enables the unstained sample to later be sent for diagnostic staining. Conclusions: Both the F50 filtering of lung specimens and label-free imaging show promise for improving ROSE procedures. In future work we will review the label-free images of the unstained monolayer slides for adequacy determination and then send to the lab for diagnostic staining.