Characterization of canine smooth muscle tumour: pilot study on 68 cases.

Introduction: Canine smooth muscle tumours (SMTs) usually arise in the gastrointestinal and urogenital tracts, but rarely in soft tissues. While the criteria used to identify smooth muscle differentiation are well established, the distinction between leiomyomas and leiomyosarcomas is based on mitotic activity, cellular atypia and amount of necrosis, but reliable cut-off values are lacking. Materials and Methods: Cases of canine SMTs were collected retrospectively and examined histologically. Differentiation, cellular atypia, necrosis, mitotic count (MC) and MIB1-based labelling index (LI) were assessed and analyzed statistically. Results: Sixty-eight SMTs were collected from 67 dogs (male:female ratio 0.5). Twenty-seven SMTs were uterine/vaginal, 18 gastrointestinal, 10 soft tissue, nine urinary and four splenic. Twenty-five SMTs were leiomyomas and 25 were leiomyosarcomas. Eighteen SMTs were not classifiable based on veterinary criteria and were diagnosed as SMTs with uncertain malignant potential according to human criteria. MC ranged between 0 and 59 (mean 13.6; median 4) and LI between 0 and 30.1 (mean 14.6; median 4.7). Diagnosis was associated with gender (P !0.01), with leiomyomas more frequent in females and leiomyosarcomas in males, and with LI (P !0.01), which was O5 in 72% of leiomyosarcomas and _5 in 76% of leiomyomas. Conclusions: Relying on veterinary criteria, the differential diagnosis between leiomyomas and leiomyosarcomas was not possible in 18 cases that, despite being well-differentiated, had a slightly higher MC than leiomyomas and some degree of cellular atypia and necrosis. We suggest using the LI with a cut-off of 5 to distinguish leiomyomas from leiomyosarcomas when histology alone is not sufficient.