Background Osteoarthritis (OA) is a major cause of chronic pain; however, the exact pathophysiology of OA pain remains poorly understood. Quantitative sensory testing (QST) is a tool used to assess pain perception. People with OA demonstrate reduced pain thresholds to blunt pressure over affected joints and also at remote body sites, suggesting a contribution of central sensitization to OA pain [1]. Thresholds for thermally induced pain have been less intensively investigated in OA, perhaps because OA pain is traditionally thought to be of mechanical origin. Recent studies have used QST in functional magnetic resonance imaging to further understand the mechanisms of OA pain. For technical reasons, such studies may report thermal rather than pressure pain thresholds; the generalisation of findings to mechanical stimulation is uncertain.

Objectives We hypothesised that thermal and pressure pain thresholds both measure aspects of central sensitisation in people with knee OA, and that findings with one modality may predict thresholds to other stimuli. This study aimed to compare pressure with thermal pain thresholds in people with knee OA.

Methods 22 participants satisfying American College of Rheumatology criteria for knee OA were recruited from a previous larger pain study. Pressure then thermal (heat and cold) pain thresholds were measured using a Sensebox (Somedic) both at affected (lateral and medial joint margins) and remote (ipsilateral anterior tibia, hand and sternum) sites. Participants were given a PainDETECT questionnaire to evaluate pain phenotype.

Results Seven participants scored over the threshold for likely neuropathic pain with PainDETECT. Cold pain thresholds were below -2 °C at joint margins and anterior tibia in >25% of participants.

Significant heterogeneity in pain thresholds was detected between sites. Pressure pain thresholds were lowest at the sternum and anterior tibia; heat pain thresholds were lowest in the hand (median 41 °C), and highest at the anterior tibia and lateral joint margin (medians 45 °C). However, within each modality, pain thresholds were significantly correlated between sites (r = 0.48 to 0.86, all p < 0.025).

There was a significant positive correlation between pressure and heat pain thresholds at all sites (r = 0.46 to 0.58, p = 0.03 to 0.004), except the sternum. Pressure and cold pain thresholds were significantly associated at the hand (r = -0.64, p = 0.001) and medial knee (r = -0.46, p = 0.03).

Conclusions We showed regional differences in pain thresholds, emphasising the need for standardised protocols in QST studies. Strong correlations within each modality between sites suggest that patient rather than local factors predominantly influence measured pain thresholds. Cold pain thresholds did not generate continuous data in the leg, as thresholds were often below -2 C. We show that heat pain thresholds predict pressure pain thresholds in people with OA, both at affected and unaffected sites, suggesting either may be a valid measure of central pain processing. Further research is required to determine whether QST may inform mechanism-based therapeutic stratification of OA pain management.

1. Suokas AK; Walsh DA, McWilliams DF, Condon L, Moreton B, Wylde V, Arendt-Nielsen L, Zhang W. Quantitative sensory testing in painful osteoarthritis: a systematic review and meta-analysis. Osteoarthritis Cartilage. 2012 Oct;20(10):1075-85.

Acknowledgements Arthritis Research UK

Disclosure of Interest None Declared