This pooled individual data (PID)-based meta-analysis collectively assessed the analgesic effect of repetitive transcranial magnetic stimulation (rTMS) on various neuropathic pain states based on their neuroanatomical hierarchy. Available randomized controlled trials (RCTs) were screened. PID was coded for age, gender, pain neuroanatomical origins, pain duration, and treatment parameters analyses. Coded pain neuroanatomical origins consist of peripheral nerve (PN); nerve root (NR); spinal cord (SC); trigeminal nerve or ganglion (TGN); and post-stroke supraspinal related pain (PSP). Raw data of 149 patients were extracted from 5 (1 parallel, 4 cross-over) selected (from 235 articles) RCTs. A significant (P < .001) overall analgesic effect (mean percent difference in pain visual analog scale (VAS) score reduction with 95% confidence interval) was detected with greater reduction in VAS with rTMS in comparison to sham. Including the parallel study (Khedr et al), the TGN subgroup was found to have the greatest analgesic effect (28.8%), followed by PSP (16.7%), SC (14.7%), NR (10.0%), and PN (1.5%). The results were similar when we excluded the parallel study with the greatest analgesic effect observed in TGN (33.0%), followed by SC (14.7%), PSP (10.5%), NR (10.0%), and PN (1.5%). In addition, multiple (vs single, P = .003) sessions and lower (>1 and < or =10 Hz) treatment frequency range (vs >10 Hz) appears to generate better analgesic outcome. In short, rTMS appears to be more effective in suppressing centrally than peripherally originated neuropathic pain states.
Perspective: This is the first PID-based meta-analysis to assess the differential analgesic effect of rTMS on neuropathic pain based on the neuroanatomical origins of the pain pathophysiology and treatment parameters. The derived information serves as a useful resource in regards to treatment parameters and patient population selection for future rTMS-pain studies.