Abstract
OBJECTIVE: Synovial pathology has been linked to osteoarthritis (OA) pain in patients. Microscopic grading systems for synovial changes in human OA have been described, but a standardized approach for murine models of OA is needed. We sought to develop a reproducible approach and set of minimum recommendations for reporting of synovial histopathology in mouse models of OA. METHODS: Coronal and sagittal sections from male mouse knee joints subjected to destabilization of medial meniscus (DMM) or partial meniscectomy (PMX) were collected as part of other studies. Stains included Hematoxylin and Eosin (H&E), Toluidine Blue (T-Blue), and Safranin O/Fast Green (Saf-O). Four blinded readers graded pathological features (hyperplasia, cellularity, and fibrosis) at specific anatomic locations. Inter-reader agreement of each feature score was determined. RESULTS: There was acceptable to very good agreement when using 3-4 individual readers. After DMM and PMX, expected medial predominant changes in hyperplasia and cellularity were observed, with fibrosis noted at 12 weeks post-PMX. Synovial changes were consistent from section to section in the mid-joint area. When comparing stains, H&E and T-blue resulted in better agreement compared to Saf-O stain. CONCLUSIONS: To account for the pathologic and anatomic variability in synovial pathology and allow for a more standardized evaluation that can be compared across studies, we recommend evaluating a minimum set of 3 pathological features at standardized anatomic areas. Further, we suggest reporting individual feature scores separately before relying on a single summed "synovitis" score. H&E or T-blue are preferred, inter-reader agreement for each feature should be considered.