Exploratory Study of rTMS Neuromodulation Effects on Electrocortical Functional Measures of Performance in an Oddball Test and Behavioral Symptoms in Autism

Abstract

There is no accepted pathology to autism spectrum disorders (ASD) but research suggests the presence of an altered excitatory/inhibitory (E/I) bias in the cerebral cortex. Repetitive transcranial magnetic stimulation (rTMS) offers a non-invasive means of modulating the E/I cortical bias with little in terms of side effects. In this study, 124 high functioning ASD children (IQ > 80, <18 years of age) were recruited and assigned using randomization to either a waitlist group or one of three different number of weekly rTMS sessions (i.e., 6, 12, and 18). TMS consisted of trains of 1.0 Hz frequency pulses applied over the dorsolateral prefrontal cortex (DLPFC). The experimental task was a visual oddball with illusory Kanizsa figures. Behavioral response variables included reaction time and error rate along with such neurophysiological indices such as stimulus and response-locked event-related potentials (ERP). One hundred and twelve patients completed the assigned number of TMS sessions. Results showed significant changes from baseline to posttest period in the following measures: motor responses accuracy [lower percentage of committed errors, slower latency of commission errors and restored normative post-error reaction time slowing in both early and later-stage ERP indices, enhanced magnitude of error-related negativity (ERN), improved error monitoring and post-error correction functions]. In addition, screening surveys showed significant reductions in aberrant behavior ratings and in both repetitive and stereotypic behaviors. These differences increased with the total number of treatment sessions. Our results suggest that rTMS, particularly after 18 sessions, facilitates cognitive control, attention and target stimuli recognition by improving discrimination between task-relevant and task-irrelevant illusory figures in an oddball test. The noted improvement in executive functions of behavioral performance monitoring further suggests that TMS has the potential to target core features of ASD.

Keywords: ERP; TMS; aberrant and repetitive behaviors; autism spectrum disorder; executive functions; oddball task; reaction time.

FIGURE 1

Flowchart of the study. After clinical evaluation at intake stage, subjects were tested in the lab using visual oddball task, while parents rated behavioral symptoms using ABC and RBS-R questionnaires. Then participants were randomized either to waitlist, 6 rTMS, 12 rTMS, or 18 rTMS treatment sessions and after completion of the assigned course of treatment were tested again in the lab using the same ERP test, whereas parents completed post-treatment behavioral ratings using ABC and RBS-R questionnaires.

FIGURE 2

Schematic representation of TMS coil location for the stimulation of the dorsolateral prefrontal cortex. Stimulations is administered first over the left motor cortex (motor strip) to determine the optimal area for stimulation of the first dorsal interossei (FDI) muscle of the right hand. The output of the TMS machine is increased until the least amount of machine power that induces an EMG response or a visible twitch is identified in 4 out of 5 trials over the motor cortical area controlling the contralateral FDI. The site for rTMS treatment location is then placed 5 cm anterior to, and in a parasagittal plane to the site of maximal FDI stimulation.

FIGURE 3

(A) Total error rate (in %) in oddball test at baseline, post-wait period, and post-treatment in 6 TMS, 12 TMS, and 18 TMS groups of children with ASD. Accuracy gradually improved in all TMS groups. Most significant difference was between the 18 TMS group as compared to baseline and waitlist (p < 0.05). Accuracy difference between three TMS groups was not significant. (B) Post-error reaction time (RT), calculated as first RT post-error minus mean RT, in visual oddball test at baseline and post treatment in waitlist, 6 TMS, 12 TMS, and 18 TMS groups. Most significant differences were noted between 18 TMS and baseline (p < 0.001), 18 TMS and waitlist (p < 0.001), as well as between 12 TMS and baseline (p = 0.004). Both 12 TMS and 18 TMS groups showed normative post-error slowing.

FIGURE 4

(A) Amplitude of P100 component at the parietal sites (ROI of five parietal channels) in response to non-target Kanizsa stimuli was significantly attenuated post-TMS, and is illustrated by comparing ERP waveforms in 18 TMS and waitlist groups. Post-TMS treatment differences (as compared to baseline and waitlist) were statistically different only for standard (p = 0.023) and non-target Kanizsa (p < 0.001) stimuli. (B) Latency of the frontal P3a component to non-target Kanizsa stimuli was shorter in the 18 TMS group as compared to the waitlist (p < 0.01), though amplitude differences were not reaching statistical significance level.

FIGURE 5

(A) Latency of P3b component in response to three types of stimuli (target, rare non-target, and standard) in visual oddball test in the waitlist and 18 TMS groups. Note longer latency of P3b to all stimuli, more significant in response to targets, in the 18 TMS group as compared to the waitlist group. (B) Latency of the frontal N100 component to non-target and target Kanizsa figures show Stimulus ×Group interaction with prolonged latencies of N100 in response to target Kanizsa as compared to non-target Kanizsa distracters in TMS groups. In particular, 12 TMS and 18 TMS groups had longer latencies to targets as compared to the waitlist group (p = 0.001 and p = 0.012, respectively).

FIGURE 6

(A) Left parieto-occipital (PO3) and parietal (P1) ERPs to target Kanizsa, non-target Kanizsa and non-Kanizsa standard stimuli at baseline, and in waitlist, 12 TMS and 18 TMS groups. Both 12 TMS and 18 TMS groups show lower amplitude of the P3b components (marked by blue line) to non-target stimuli. In the waitlist group ERPs to all three types of stimuli are comparable by their P3b amplitude, especially at the P1 site. (B) Frontal (F1, F2) and fronto-central (FC1) ERPs to three types of stimuli in oddball task at baseline and in waitlist, 12 TMS, and 18 TMS groups. Amplitude to all types of stimuli post-TMS treatment was decreased, more in the 18 TMS group. Note delayed latency and higher amplitude of the P3a component in response to non-target Kanizsa distracter in the waitlist group.

FIGURE 7

(A) Ritualistic/Sameness behavior (left) and Stereotype behavior (right) rating scores of RBS-R questionnaire at baseline, post-waiting period, and post 12 and 18 sessions of rTMS. Most dramatic decrease of scores was observed in the 18TSM group. (B) Hyperactivity (left) and Irritability (right) rating scores of ABC questionnaire at baseline, post-waiting period, and post 12 and 18 sessions of rTMS. The Irritability scores in the 18 TMS group decreased practically in half (–6.01, p = 0.029) as compared to the baseline.