Background/Purpose: Alterations in the peripheral and central nervous systems including sensitization are thought to play an important role in the pain experience in knee OA. While sensitization could occur due to an underlying predisposition, it is hypothesized that joint inflammation and/or tissue injury in OA could provide sufficient peripheral nociceptive input to cause sensitization. We previously reported that radiographic knee OA severity or duration do not appear to be related to sensitization. However, whether specific MRI lesions related to inflammation (e.g., synovitis, effusion), or mechanical load or remodeling related to noninflammatory tissue injury (e.g., bone marrow lesions (BMLs)) are risk factors for development of sensitization is not yet known.

Methods: The Multicenter Osteoarthritis (MOST) Study is a NIH-funded longitudinal cohort of persons with or at risk of knee OA. Subjects had x-rays and MRIs (1.0 T) of each knee obtained at each study visit, and a standardized somatosensory evaluation of mechanical temporal summation and pressure pain thresholds (PPT) at the patella at 60- and 84-months. Temporal summation was defined by increased pain during repeated mechanical stimulation (1 Hz x 30-sec) with a 60g monofilament. PPT was assessed with an algometer (1 cm²  tip, 0.5 Kg/sec) as the point at which the subject felt the pressure change to slight pain. Lower PPT indicates more sensitivity. Synovitis, effusion and BMLs on MRIs were scored using WORMS (one knee per person); these lesions were considered to be present if their score was ≥1 in any subregion. In sensitivity analyses, we assessed the sum of BML scores across all knee subregions as a measure of BML burden. We assessed the relation of presence of synovitis, effusion, and BMLs at 60-mo to incident temporal summation in the same knee at 84-mo among those who did not have temporal summation at 60-mo, and to change in PPT in the same knee between 60- and 84-mo in the whole sample using logistic and linear regression, respectively, adjusted for relevant potential confounders, including OA severity.

Results: There were 1111 subjects (mean age 66.9, mean BMI 29.7, 62% female) in the whole sample. 22.6% developed incident temporal summation at the 84-mo visit, and the range in change of PPT between the 60- and 84-mo visits was -7.35 to 7.15 kg/cm². Presence of synovitis was associated with a significant decrease in PPT (i.e., more sensitized) over 24 mo, while effusion was significantly associated with incident temporal summation (Table). BML presence or burden were not associated with temporal summation or PPT.

Conclusion: Inflammation, as evidenced by synovitis or effusion, may drive the occurrence of sensitization in knee OA. In contrast, BMLs do not appear to contribute to sensitization in knee OA. Early targeting of inflammation in knee OA may be a reasonable strategy to test for its ability to prevent occurrence of sensitization, thereby reducing pain severity in knee OA.