Association between sensitisation and pain-related behaviours in an experimental canine model of osteoarthritis.

Evaluation of nociceptive sensitisation in canine osteoarthritis studies has been poorly reported, or even related to other clinical symptoms. In 16 dogs, peak vertical force (PVF), subjective pain assessment using 3 scales, sympathetic stress response with electrodermal activity (EDA) measurement, and behavioural changes with video analysis and telemetered motor activity were quantified at baseline (D-7), and 28 and 56 days post transection of the cranial cruciate ligament. As markers of central sensitisation, selected spinal cord biomarkers (substance P and transthyretin) were quantified at D56. Electrical withdrawal thresholds on the stifle and the tail were measured as indicative of peripheral and central quantitative sensory testing (QST) sensitisation, respectively. The effects of vehicle administration (n=8) were compared with tiludronate (2mg/kg subcutaneously, q2 week, starting at D0) administration. Generalized estimated equations tested the association between the behavioural and physiological methods and QST sensitisation, and therefore the sensitivity of the methods for detecting treatment efficacy. Compared to tiludronate, at D56, vehicle-treated dogs had increased spinal substance P (P=0.01), concomitant decreased transthyretin (P=0.02), and (compared to baseline) demonstrated peripheral and central QST sensitisation, which was not present for tiludronate. Only PVF, the spontaneous behaviour "walking with full weight-bearing," and EDA were associated with occurrence of QST sensitisation and indicated significant tiludronate analgesic efficacy after inclusion of central QST sensitisation as a predictor variable in the statistical model. This study establishes the strong interest to implement QST as a predictor of canine osteoarthritis pain symptoms explained by pain sensitisation.