Revealing the Neural Mechanism Underlying the Effects of Acupuncture on Migraine: A Systematic Review

This systematic review represents an important milestone in understanding the neurologic mechanisms by which acupuncture exerts its therapeutic effects on migraine. Conducted by researchers at the Chengdu University of Traditional Chinese Medicine, the analysis systematically examined the neuroimaging literature on acupuncture for migraine published between 2009 and 2020. The comprehensive search in four major databases resulted in identification of 15 eligible studies, including 634 patients with migraine without aura and 234 healthy controls. The rigorous methodology followed PRISMA guidelines and included both randomized controlled trials and observational studies that used neuroimaging techniques to investigate the effects of acupuncture on the brain.

The included studies employed various neuroimaging techniques, with functional magnetic resonance imaging (fMRI) being the most used (80% of the studies), followed by PET-CT (13.3%) and magnetic resonance spectroscopy (6.7%). Data analyses included sophisticated methods such as regional homogeneity analysis (ReHo), amplitude of low-frequency fluctuations (ALFF), independent component analysis (ICA), and functional connectivity. The results revealed two distinct patterns of neural effects of acupuncture: immediate effects during acute migraine treatment and sustained effects after prolonged treatment. For immediate effects, increased brain metabolism was observed in regions involved in cognitive and affective-emotional processing of pain, including the orbitofrontal cortex, parahippocampal gyrus, and insula.

Resting-state networks, such as the default mode network and the limbic system, also showed significant alterations. As for sustained effects, which are fundamental to migraine prevention, acupuncture demonstrated modulation of multiple pain-related brain networks. Specifically, there was increased activity in regions of the affective-emotional processing of pain (insula, cerebellum, brainstem) and the cognitive processing of pain (orbitofrontal cortex), in addition to changes in the descending pain modulation system. Simultaneously, decreased activity was observed in regions of the cognitive processing of pain (hippocampus) and in resting-state networks such as the default mode network and the frontoparietal network.

A particularly interesting aspect was the analysis of the correlations between neural changes and clinical outcomes. Eight studies performed these correlations, revealing that improvement in pain intensity was negatively correlated with activity in specific regions such as anterior cingulate cortex, insula, thalamus, and cerebellum. These findings suggest that the prophylactic effect of acupuncture on migraine may occur through regulation of the visual network, default mode network, sensorimotor network, frontoparietal network, limbic system, and descending pain modulation system. The study also identified significant limitations in the current literature.

Most studies had small samples (ranging from 10 to 72 participants), only 53.3% assessed the psychological state of the participants (anxiety and depression), and only 20% avoided performing imaging during the menstrual period in women. In addition, there was great variability in acupuncture protocols, with 21 different points used, the most frequent being TE-5 (Waiguan), GB-34 (Yanglingquan), and GB-20 (Fengchi). The clinical implications of these findings are substantial. First, they provide a solid neurobiologic basis for the therapeutic effects of acupuncture on migraine, validating its efficacy through objective brain markers.

Second, the identification of specific neural networks may guide the development of more precise and personalized treatment protocols. Third, neural markers may serve as biomarkers to predict treatment response and identify patients who would benefit most from acupuncture. The authors emphasize the need for future studies with more rigorous designs, including larger samples, standardized acupuncture protocols, systematic psychological assessment, and multimodal studies that integrate different neuroimaging techniques. Particularly promising is the application of machine learning methods to develop predictive models of acupuncture efficacy based on pretreatment brain patterns.