Jump to Main Content
Shaved margin histopathology and imprint cytology for assessment of excision in canine mast cell tumors and soft tissue sarcomas
- Milovancev, Milan, Townsend, Kaitlin L., Gorman, Elena, Bracha, Shay, Curran, Katie, Russell, Duncan S.
- Veterinary surgery 2017 v.46 no.6 pp. 879-885
- dogs, excision, glass, histopathology, mast cells, neoplasm cells, patients, sarcoma, statistical models, tissues
- OBJECTIVE: To determine the feasibility and agreement of margin assessment by imprint cytology, shaved margin histopathology, and radial section histopathology in canine cutaneous and subcutaneous mast cell tumors (MCT) and soft tissue sarcomas (STS). STUDY DESIGN: Prospective clinical study. SAMPLE POPULATION: Three hundred and forty margins from 72 excised tumors (52 MCT and 20 STS) in 54 client‐owned dogs. METHODS: Imprint cytology samples were acquired by pressing glass slides to the cut surgical margin of the freshly excised surgical specimen. Shaved margin samples were obtained from the patient wound bed using a scalpel immediately prior to closure. Radial section histopathology was performed as part of routine histopathologic processing. All margins were assessed as either positive or negative for presence of tumor cells at the surgical margin. Agreement among methods was calculated using Fleiss Kappa coefficients and an association of method, margin direction, and tumor type with positive margin status was evaluated using a general linear mixed model. RESULTS: Positive margin detection rates differed for MCT (imprint cytology 21%, radial section histopathology 9%, and shaved margin histopathology 3%; P < .0001) but not for STS. Intermethod agreement was poor (Fleiss Kappa = 0.051 and 0.176 for MCT and STS, respectively). Margin direction did not influence margin status for either tumor type. CONCLUSION: Imprint cytology and shaved margin histopathology are feasible, but their results are frequently disparate from routine radial section histopathology. Future studies are needed to evaluate the correlation of each method with local recurrence rates.