Abstract
OBJECTIVES: This study aimed to (1) determine the feasibility of calculating percentage myocardial collagen (PMC) in formalin-fixed, paraffin-embedded canine tissue samples, (2) explore relationships between PMC, mast cell (MC) count and myocardial arteriosclerosis/arteriolosclerosis (MA) and (3) calculate the sample size required to compare PMC between dogs with and without myxomatous mitral valve disease (MMVD). METHODS: Histological sections were prepared from retrospective samples of formalin-fixed, wax-embedded ventricular myocardial tissue from 15 dogs with, and seven dogs without, MMVD. Sections from each sample were stained with Masson's trichrome, haematoxylin and eosin (H&E) or toluidine blue. In each Masson's trichrome-stained section, digital images of 12 random fields were captured and the total image area and total collagen area were measured using computer-assisted morphometric analysis. Mean collagen area, mean total image area and mean PMC ([mean collagen area/mean total area] × 100) were calculated per sample. MCs were counted manually in toluidine blue-stained sections and the mean MC count was calculated from 10 fields. MA was identified by visual inspection of transversely cut vessels in each H&E-stained section. RESULTS: No significant difference was detected between MMVD and control dogs in PMC (p = 0.063), MC (p = 0.476) or MA (p = 0.172). A sample size calculation suggests that data from at least 26 MMVD dogs and 26 controls are required to detect a difference in PMC. No significant association was detected between PMC and MC count (p = 0.606) in cases of MMVD. CLINICAL SIGNIFICANCE: The methods described are feasible and the sample size for a definitive study has been estimated.