Enhancing treatment of osteoarthritis knee pain by boosting expectancy: A functional neuroimaging study

Abstract

Objectives: Expectation can significantly modulate pain and treatment effects. This study aims to investigate if boosting patients' expectancy can enhance the treatment of knee osteoarthritis (KOA), and its underlying brain mechanism.

Methods: Seventy-four KOA patients were recruited and randomized to three groups: boosted acupuncture (with a manipulation to enhance expectation), standard acupuncture, or treatment as usual (TAU). Each patient underwent six treatments before being debriefed, and four additional treatments after being debriefed. The fMRI scans were applied during the first and sixth treatment sessions.

Results: We found significantly decreased knee pain in the boosted acupuncture group compared to the standard acupuncture or TAU groups after both six and ten treatments. Resting state functional connectivity (rsFC) analyses using the nucleus accumbens (NAc) as the seed showed rsFC increases between the NAc and the medial prefrontal cortex (MPFC)/rostral anterior cingulate cortex (rACC) and dorsolateral prefrontal cortex in the boosted group as compared to the standard acupuncture group after multiple treatments. Expectancy scores after the first treatment were significantly associated with increased NAc-rACC/MPFC rsFC and decreased knee pain following treatment.

Conclusions: Our study provides a novel method and mechanism for boosting the treatment of pain in patients with KOA. Our findings may shed light on enhancing outcomes of pharmacological and integrative medicines in clinical settings.

Keywords: Acupuncture; Expectancy; Knee osteoarthritis; Resting state functional connectivity; Reward.

Fig. 1

A: Details of expectancy manipulation paradigm. We first used a marker to draw three boxes identical in size on the medial side of the most affected knee. Then, we placed the thermal probe in one of the boxes at a time. One box received random pain (RP), consisting of four moderate intensity pain stimuli and four low intensity pain stimuli in a random order. RP stimuli were applied to ensure subjects could consistently rate pain stimuli. The other two boxes received identical pain (IP) sequences consisting of 6 identical moderate pain stimuli. After acupuncture treatment (boosted or standard) or waiting for 20 min (TAU group), decreased identical pain sequences (dIP), consisting of 6 identical low intensity pain stimuli, were applied on all the two IP boxes in the same order as above. Those in the boosted acupuncture group were informed they were going to receive identical pain stimuli similar to before the treatment to test the analgesic effect of acupuncture. Those in the other groups (standard and TAU) were informed they were going to receive lowered intensity stimuli to investigate brain responses to different levels of pain. After this manipulation, patients in the boosted acupuncture group felt acupuncture significantly relieved the heat pain. Expectancy scores were collected before and after expectancy manipulation. B: Locations of 6 acupoints. C: Stimulation parameters applied for acupuncture treatment. This procedure was applied twice. S, needle stimulation.

Fig. 2

KOOS pain score changes measured at the midpoint and the end of the study compared to baseline (mean ± SE).

Fig. 3

A. Partial plot showing the positive association between the expectancy scores after the expectancy manipulation in session 2 (x axis) and KOOS pain increases (clinical improvement) after six acupuncture treatments (y axis), including age and gender as covariates (p = 0.04) across the two acupuncture groups (solid dot indicates standard acupuncture group, hollow dot indicates boosted acupuncture group). B. Representative brain regions showed a significant rsFC increase in boosted acupuncture as compared to standard acupuncture (voxel-wise p < 0.005, and cluster-level p < 0.05 FWE corrected). C. The scatter plot represents the partial plot showing the positive association between the expectancy score (x axis) after expectancy manipulation in session 2 and NAc and rACC/MPFC connectivity changes (y axis), including age and gender as covariates (p = 0.002). MCC, middle cingulate cortex; rACC, rostral anterior cingulate cortex; MPFC, medial prefrontal cortex.