91 A Prospective Study of Loose Tissue Fragments in resected NSCLC are found in many malignancies. For example, tumor cell displacement is seen in up to one third of breast cancer core biopsies119, 121. This prospective study proved that the presence of loose fragments is an inducible and reproducible phenomenon in lung cancer specimens. Tumor islands or loose tumor cells, tting the criteria of STAKS were identi ed in 73% of the cases and 93% of these malignant loose fragments can be attributed to mechanical artifacts related to surgical resection and gross room specimen processing. Furthermore, benign loose fragments within alveolar spaces were also found in 61% of the cases. It is likely that intrapulmonary spread of benign and malignant fragments is due to mechanical forces imposed on the tissue during surgery and/or specimen prosection. As lung tissue largely consists of empty spaces, that is air spaces, there are many opportunities to “trap” displaced tissue fragments. Such explains the relative low frequency of intravascular and intrabronchial as compared with intra-alveolar fragments. Tumor cell dissociation is a recognized characteristic in malignancy related to adhesion molecule expression122 123 124 125 126. There was a tendency for more loose fragments when comparing all AdC to SqCC and comparing the AdC with a predominant micropapillary growth pattern to the other types of AdC. In addition, there were also more SqCC cases without loose fragments than AdC cases. One could suggest that SqCC are more cohesive than AdC, and that the AdC micropapillary subtype is the most discohesive AdC subtype. Interestingly, circulating tumor cells are detected in patients after undergoing lung resection, some of which are likely dislodged by the surgical procedure127 128. Although we assume lung is handled “rougher” during VATS operations, where collapsed lung is removed through a small hole in the thoracic wall, our data did not demonstrate different incidences of loose fragments in the different procedure groups. Nevertheless, part of the 7% loose fragments in block 1 may be explained by intra- operative manipulations or other mechanical forces during specimen handling. Three cases featured intravascular tumor fragments, but were not considered true angioinvasion. Artifactual presence of loose fragments in a vascular lumen mimics vascular invasion. Perhaps pathologists’ inability to discern artifactual vascular tumor from true angiolymphatic invasion muddies our ability to de nitively comment on the prognostic signi cance of lymphovascular invasion in lung and other cancers129 130 131. Morphologic clues favoring artifactual presence of loose fragments over true vascular invasion are single fragments without erythrocytes in an otherwise “empty” vascular lumen. Alternatively, clues favoring true vascular invasion are (i) tumor fragments with brin and or erythrocytes in the vascular lumen or (ii) multiple tumor fragments in dilated lymph vessels surrounded by clear lymph with or without occasional lymphocytes. STAS has diagnostic criteria. The morphologic criteria for STAS are (i) micropapillary structures including ring-like structures compatible with the WHO micro- papillary pattern; (ii) solid nests or “tumor islands” consisting of solid collections of tumor cells 6
RkJQdWJsaXNoZXIy MTk4NDMw