196. Resting State Functional MRI Reveals Connectivity Changes in the Pain Matrix Following Prophylactic RPNI Surgery

PURPOSE: Painful neuromas lead to significant changes across the brain. A promising treatment approach for preventing both neuroma and phantom limb pain (PLP) is the Regenerative Peripheral Nerve Interface (RPNI), which involves grafting muscle to a transected nerve. In humans, prophylactic RPNIs created at the time of amputation effectively prevent the development of symptomatic neuromas and PLP. However, the changes in brain connectivity following RPNI surgery, specifically in pain-related regions, have not been extensively investigated. In this study, resting state-functional magnetic resonance imaging (rs-fMRI) was utilized to quantify brain connectivity in rats with prophylactic RPNIs. METHODS: Nine male Fischer rats were randomized into three groups: neuroma, RPNI, and sham. In the neuroma group, the right tibial nerve was transected and sutured to the overlying skin fascia. In the RPNI group, the right tibial nerve was transected and implanted into a muscle graft. No intervention was performed in the sham group. Three months following surgery all animals underwent rs-fMRI scans using a 7-Tesla small animal scanner. The data from all three groups was analyzed using parcel-based correlation analysis (z-scores) on the following regions of interest in the pain matrix: insula cortex (IC), retrosplenial cortex (RSC), amygdala (Amy), thalamus (Thal), cingulate cortex (Cg), and primary somatosensory cortex-hind limb region (S1HL). Behavioral testing for pain tolerance was conducted by delivering electrical stimulation with increasing stimulation amplitudes to the naïve muscle, neuroma, or RPNI using implanted electrodes. RESULTS: Functional connectivity between regions of the pain matrix of the neuroma group was higher than that of RPNI for most brain region correlation pairs. Z scores, which quantify the strength of correlation, were as follows for neuroma and RPNI groups, respectively: ZIC-RSC = 0.35 and 0.30, ZIC-Thal = 0.19 and 0.16, ZIC-Cg= 0.59 and 0.49, ZCg-S1HL= 0.71 and 0.55, ZS1HL-RSC = 0.65 and 0.59, ZS1HL-Thal = 0.50 and 0.37. With other pairs, however, this pattern was reversed, that is, the neuroma group had slightly lower (e.g., ZIC-Amy = 0.28 and 0.31) or similar (e.g., ZCg-Amy = 0.28 and 0.26) connectivity compared to the RPNI group, The functional connectivity for the sham group was generally lower than (e.g., ZIC-RSC = 0.23, ZIC-Cg= 0.25, ZS1HL-Thal = 0.30), or similar to (e.g., ZIC-Thal = 0.16, ZS1HL-RSC = 0.59, ZIC-Amy= 0.33) that of the RPNI group. From the behavioral study for all groups, the neuroma group experienced evident pain response at the threshold with a much lower stimulation amplitude on average (0.37 mA) compared to the RPNI group (1.1 mA), suggesting a higher pain tolerance in the RPNI group compared to the neuroma group. CONCLUSION: In this study, we demonstrate differential functional connectivity patterns in the pain matrix of sham, neuroma, and RPNI animals. The RPNI group had attenuated connectivity between multiple regions of the pain matrix, as well as higher average behavioral pain tolerance. Studying cortical reorganization after prophylactic RPNI repair could help inform treatment for neuroma and PLP.