Background/Purpose: Recently, several functional brain imaging techniques including functional magnetic resonance imaging (fMRI) have been proposed to be useful in probing the mechanisms of action of centrally acting analgesics.  In parallel, our group has utilized brain resting state functional connectivity (fcMRI) outcomes to predict clinical pain response in fibromyalgia (FM) to pregabalin (PGB) and milnacipran (MLN), two known efficacious compounds.  While preclinical work suggests that the mechanisms of action of these drugs may differ, a key hurdle in the identification of novel acting central analgesics would be the ability to differentially predict clinical response to different compounds in humans.

Methods: 29 female FM patients were included in this study: 14 underwent a double-blind placebo-controlled crossover study with the administration of PGB, and15 were enrolled in a similar study with MLN. All patients underwent a 6 minute fcMRI scan pre-treatment for both the drug and placebo periods. Patients also reported clinical pain pre- and post-treatment using a 0-10 rating scale.  fcMRI data were analyzed with an a priori regions of interest approach using the Conn toolbox running through SPM8 and Matlab. Regions previously identified as predictive for PGB and MLN in isolation were cross-utilized in both data sets.  Seeds included: the insula, rostral and genual anterior cingulate, dorsal lateral prefrontal cortex (DLPFC), periaqueductal gray, and amygdala. Prediction results were analyzed in SPM8 with a multiple regression that included seed-to-whole brain correlation maps with clinical pain difference scores. Results were significant on the cluster level with a false discovery rate p value < 0.05 derived from an uncorrected voxel level p value < 0.001.

Results: Reduced functional connectivity between the right DLPFC (seed) to both the inferior parietal lobule (r = 0.94 p < 0.001) and inferior/mid temporal gyrus (r = 0.88, p < 0.001) was associated with greater subsequent reduction in clinical pain for MLN, but not PGB. In contrast, greater connectivity between the rostral ACC (seed) to the subgenual ACC (r = – 0.91, p < 0.001) and the left amygdala (seed) to the left insula/sensory motor cortex (r= – 0.96, p < 0.001) and the primary motor cortex (r= – 0.89, p=0.01) was associated with subsequent reduction in pain for PGB, but not MLN. Furthermore, connectivity between the right periaqueductal gray (seed) to the precuneus (r= – 0.89, p = 0.02) and between the right anterior insula (seed) to the precuneus (r= -0.86, p = 0.042) was associated with reduction in pain to PGB, but not MLN.

Conclusion: These data demonstrate that fcMRI can differentially predict clinical response to two efficacious analgesics in FM, thus providing evidence that these compounds may have differing mechanisms of action.  As suspected from preclinical data, it appears that SNRIs such as milnacipran may be acting partly by increasing fcMRI to brain regions involved in anti-nociception, whereas gabapentinoids such as pregabalin may be acting by decreasing fcMRI to pro-nociceptive regions. This provides some rationale for the common clinical practice of using these two classes of drugs in chronic pain states.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/brain-functional-connectivity-differentially-predicts-response-to-two-centrally-acting-analgesics-in-fibromyalgia/