932 - Variability in Transcranial Magnetic Stimulation for the Treatment of Chronic Pain – An Illustrative Review

Objectives: The objective of this review is to identify the variability in transcranial magnetic stimulation (TMS) parameters and the effect on the treatment of chronic pain.

Design: We conducted an in-depth analysis of randomized controlled trials, parallel or crossover studies, and sham-controlled, peer-reviewed studies on adults diagnosed with chronic pain, primarily focusing on pain intensity as measured by visual analog scales or numeric rating scales.

Results: The studies demonstrated inconsistent results with wide-ranging differences in the stimulation parameters of TMS and the estimated energy available at the brain. Seven studies of stimulation targeting the motor cortex resulted in improvement in reported pain levels. The degree of improvement and stimulation parameters varied with a median frequency of 20 Hz, mean pulse width 357 µs, mean number of pulses per session was 1,457, and the minimum and maximum estimated energy delivered to the brain was 10 V/cm and 130 V/cm per session respectively. Stimulation targeting the dorsolateral prefrontal cortex (DLPFC) yielded less consistent results, with only one of four studies showing a difference in pain scores following treatment. The median frequency was 10 Hz, a mean pulse width of 180 µs, mean number of pulses per session 2,450, and the minimum and maximum estimated energy delivered to the brain was 14 V/cm and 180 V/cm per session respectively.

Conclusions: Based on these findings, our review agrees with Turi et al. who described inconsistent reporting with over 90% of TMS studies failing to report sufficient information to quantify the stimulation intensity. There is a need for standardization in reporting stimulation parameters, including the percentage of total stimulator output for the motor threshold, and the coil and stimulator used. Further research focused on these wide-ranging outcomes may allow analysis of intra-study variability to enhance the effectiveness of TMS as a treatment for chronic pain.

Scientific Paper Best Paper Award Abstract: Non-invasive stimulation of the brain with transcranial magnetic stimulation (TMS) provides a promising treatment option for patients with chronic pain who may not be good candidates for interventional treatments. However, variability in the results of treatments has made conclusive statements regarding the effectiveness of this treatment quite difficult. In this illustrative review, we included randomized controlled clinical trials, parallel or crossover studies, of repetitive TMS. Either sham-controlled, peer-reviewed studies on adults older than 18 years with a diagnosis of chronic pain published in English, a clear primary outcome of pain intensity by visual analog scales (VAS) or numeric rating scales (0-10) were included. We reviewed these publications for variability within the stimulation parameters of TMS, which produced wide-ranging differences of estimated energy available at the brain. The chronic pain diagnoses included in these studies ranged from post-stroke central pain, chronic widespread pain, chronic phantom pain, chronic neuropathic pain, post-stroke hand pain, and trigeminal neuralgia. A separate study reported a range of electrical energy estimated at the brain of 0.14 – 1.82 V/cm with a single pulse of stimulation at the motor threshold. The papers in this review did not report energy delivered but rather reported stimulation intensity as a percentage of the motor threshold. Therefore, we used the range of estimated energy from Turi et. al (2021) to make these calculations and provide a range for context. Calculations of the root mean square (RMS) were made by multiplying the amplitude of the intensity (minimum and maximum estimated as reported above) by percentage of the motor threshold indicated by the study, and then by the sqare-root of the frequency multiplied by the pulse width. The RMS value was then multiplied by the number of pulses per session to get an estimate of the energy delivered to the cortical surface for each TMS session reported.

Stimulation targeting the motor cortex with TMS for the treatment of chronic pain resulted in consistent improvement in reported pain levels across studies with seven out of seven studies reporting a significant improvement in pain scores, but the amount of improvement and stimulation parameters varied widely. For stimulation of the motor cortex, the median frequency was 20 Hz, with a mean pulse width of 357 µs, mean number of pulses per session of 1,457, and we used these to calculate the minimum and maximum estimated energy delivered to the brain. For this, we used the reported percentage of motor threshold and the above-mentioned estimated amounts of energy at the brain with motor threshold stimulation to calculate a range of 10 - 130 V/cm per session respectively. The median number of sessions for studies targeting the motor cortex was five.

Stimulation targeting the dorsolateral prefrontal cortex (DLPFC) with TMS for the treatment of chronic pain resulted in only one of four studies demonstrating a difference in pain scores following treatment with TMS. Stimulation of this brain region utilized a median frequency of 10 Hz, a mean pulse width of 180 µs, mean number of pulses per session of 2,450, and we used these to calculate the minimum and maximum estimated energy delivered to the brain. For this, we used the reported percentage of motor threshold and the above-mentioned estimated amounts of energy at the brain with motor threshold stimulation to calculate a range of 14 - 180 V/cm per session respectively. The median number of sessions for studies targeting the DLPFC was 17.5.

These studies demonstrate the lack of uniformity between studies and the challenge of determining proper protocols to use in further studying TMS in the treatment of chronic pain. Bungert et al. found significant differences in energy delivered to the anatomic regions of the brain by modeling the differences in stimulation with the coil in the anterior-posterior versus the posterior-anterior orientations. They also suggested the variability could be determined by identifying changes at the neural level, but that for now the maximum electric field strength was the most beneficial in field calculations for quantitative dosing. A study by Turi et al. systematically reviewed TMS studies between over a nearly 30-year period and found more than 90% described stimulation intensity in their study as a percentage of motor threshold rather than a quantifiable value. They also noted that over 90% of studies did not provide enough information to calculate the stimulation intensity. They suggested standardization of reporting with consistent detail regarding the method for determining motor threshold including whether resting or active. They also note the importance of reporting the measure of muscle activation with electromyography and in addition to noting the percentage of the motor threshold for stimulation, they recommend reporting the percentage of device output required to get the motor threshold. They also recommend reporting both the coil and stimulator used in the study to be able to replicate study parameters more completely. We agree that these factors will allow rigorous study of TMS in the treatment of specific diagnoses. However, we also recommend further research of the intra-study variability in outcomes to determine how best to evaluate and compare outcomes across studies.