Analysis of Ki67 Density Maps Reveals Variation of Spatial Proliferation Patterns in Different Canine Neoplasms

Mitotic count (MC) is a well-established prognostic factor in many canine malignancies. While standardisation efforts have improved inter-pathologist agreement regarding the morphology of mitotic figures and the size of the counting area, the selection of the tumour region for MC assessment remains to be standardised. This study aimed to evaluate the spatial distribution of the most proliferative areas in selected canine tumour types, using Ki67 immunohistochemistry, to identify optimal candidate regions for MC assessment. Tumour types analysed included melanomas, cutaneous mast cell tumours (cMCT), canine mammary carcinomas (CMC) and soft tissue sarcomas (STS). Using image analysis, Ki67 density maps were generated from digital slides and classified according to their distribution pattern (focal/multifocal or diffuse) and location within the tumour (central, peripheral or scattered). A total of 202 cases were included: 43 melanomas, 30 cMCTs, 42 CMCs and 87 cSTSs. The vast majority of tumours (92.6%) exhibited a multifocal hotspot distribution. Peripheral hotspot localisation was predominant in 55% of cases, particularly in cMCTs (73.3%) and melanomas (76.7%). In contrast, cSTSs more frequently showed a scattered hotspot pattern (60.9%) (χ2 = 41.9; p < 0.001). CMCs had a higher proportion of centrally located hotspots (16.7%). These findings suggest that pathologists should focus on peripheral tumour regions when assessing MC in cMCTs and melanomas. Both central and peripheral regions should be considered in CMCs, while a more extensive, comprehensive evaluation may be required in STSs. The observed association between tumour histotype and proliferation pattern likely reflects inherent biological differences among the tumour types studied.