Osteoarthritis and Central Pain
Traditionally, osteoarthritis (OA) has been considered to be a peripheral pain disorder, related to progressive cartilage and bone damage, with little evidence for tissue inflammation. During the last decade, however, there has been greater appreciation of the inflammatory aspects of OA, including contributions from various cytokines and nociceptors, such as nerve growth factor.1 There is now a growing appreciation of the role of peripheral and central pain sensitization in OA, which will be the focus of this review.
OA and Pain Sensitivity
Pain sensitivity and intensity are magnified in OA. In fact, patients with OA, compared with normal controls, report increased pain intensity in widespread areas, including referred and radiating pain. In a study by Kosek et al, the researchers found an increased sensitivity to pressure, ischemia, and innocuous warm stimuli at the affected OA hip and at the contralateral hip in patients with OA.2 In addition, hyperalgesia has been reported in both the affected knee and the tibialis anterior muscle in patients with knee OA.3
To see how pain affects the brain, 12 patients with knee OA underwent positron emission tomography (PET) of the brain, using 18F-fluorodeoxyglucose.4 Scanning was performed during 3 different states: arthritic knee pain, experimental knee pain, and pain-free. Both pain conditions activated the pain matrix as seen on PET, but arthritic pain was associated with increased activity in the cingulate cortex, the thalamus, and the amygdala. These areas of the brain are involved in the processing of fear and other emotions, and in aversive conditioning. The study authors proposed that although arthritic pain and experimental pain activate similar areas of the brain, arthritic pain is also associated with areas of the brain implicated in affect, aversive conditioning, and motivation.
Researchers have found that there were significantly lower pressure pain thresholds (PPTs) over multiple joint, muscle, and soft tissue locations in patients with knee OA than in healthy controls.5 The lower PPTs correlated with higher pain intensity, higher disability, and poor quality of life scores.5-6
For all locations (knee, leg, and arm), significantly negative correlations between pain sensitivity and PPT were found (more pain, more sensitization) in a study by Arendt-Nielsen et al.7 The OA patients studied showed a significant facilitation of temporal summation from the affected knee and had significantly less diffuse noxious inhibitory control as compared with controls. Moreover, the clinical/experimental pain parameters correlated poorly with standard radiological findings.
A systematic evaluation of studies in painful OA found that compared with controls, OA subjects had lower PPTs both at the affected joint and at remote sites.8 Using quantitative sensory testing (QST), assessing PPTs had “good ability” to differentiate between patients with OA and healthy controls.9 In 168 adults with symptomatic knee OA, QST was used to measure sensitivity to heat pain, pressure pain, and cold pain, as well as to the temporal summation of heat pain, a marker of central sensitization. Pain hypervigilance was associated with greater clinical pain severity, as well as greater pressure pain. Pain hypervigilance was also a significant predictor of temporal summation of heat pain.
Compared to controls and a low-symptom group, OA patients with increased pain severity were more sensitive to supra-threshold heat stimuli, blunt pressure, punctuate mechanical, and cold stimuli.10 Individuals in the low symptomatic OA group exhibited experimental pain responses similar to the pain-free group on most measures. Mechanical knee stimulation in patients with OA was associated with bilateral activity in the thalamus, as well as the secondary somatosensory, insular, and cingulate cortices, with unilateral activity in the putamen and amygdala. These data suggest that painful stimulation in subjects with OA of the knee engages many brain regions commonly observed in acute pain.
Why Joint Images Don’t Always Match OA Pain
In OA, the pain intensity often correlates poorly with the severity of peripheral joint damage. For example, 30% to 50% of individuals with moderate-to-severe radiographic changes of OA are asymptomatic, and 10% to 20% of individuals with moderate-to-severe knee pain have normal findings on radiography.11
In one study that examined this paradox (Table 1), the investigators found significantly heightened pain sensitivity in high pain/low knee OA (Kellgren x-ray) grade subjects, while the low pain/high knee OA Kellgren grade group were less pain-sensitive.12 The results suggest that central sensitization in knee OA is especially apparent among patients who report high levels of clinical pain in the absence of moderate-to-severe radiographic evidence of pathologic changes of knee osteoarthritis.
In addition, PPT and temporal summation were associated with OA-related pain, but not with radiographic evidence of OA.13 A study of functional magnetic resonance imaging (fMRI) of the brain compared images taken while 11 participants with moderate/severe right OA pain performed motor tasks involving isolated isometric muscle contractions of quadriceps (knee), tibialis anterior (ankle), and finger/thumb flexor (hand) muscles and compared them with images from
7 asymptomatic controls.14 Differences in the organization of the motor cortex in knee OA were demonstrated in relation to performance of knee and ankle motor tasks and were related to quality of performance of the knee motor task. These results highlight the possible mechanistic link between cortical changes and modified motor behavior in people with knee OA.
In another study, structural MRI data were acquired from 26 patients with knee OA and 31 healthy individuals (controls).15 The data showed that the normalized volumes of bilateral caudate nucleus were significantly smaller in the group with knee OA than in the control group; and there was a trend toward smaller volume of the hippocampus in OA patients compared to controls. Detailed surface analyses further localized these differences with a greater involvement of the left hemisphere for the caudate nucleus.
Twenty-three subjects with knee osteoarthritis and 23 healthy controls underwent studies to measure pain thresholds to pressure, cold, and heat at the knee, ipsilateral heel, and ipsilateral elbow.16 Osteoarthritic subjects demonstrated significantly increased sensitivity to both pressure (Table 2) and cold stimuli, compared with controls. A similar pattern was noted at the pain-free ipsilateral ankle and elbow, indicating widespread pressure and cold hyperalgesia.