Electrocortical and Behavioral Outcomes of Novel Experimental Trial of Repetitive Transcranial Magnetic Stimulation in Autism

Background: Previous studies by our group suggest that the neuropathology of autism is characterized by a disturbance of cortical modularity manifested as a “minicolumnar” abnormality. In this model a decrease in the peripheral neuropil space of affected minicolumns provides for an inhibitory deficit and a decrease in their signal to noise ratio.

Objectives: Given the geometric orientation of double bouquet cells within the peripheral neuropil space we proposed using low frequency transcranial magnetic stimulation (rTMS) as a way of selectively increasing the surround inhibition of minicolumns. TMS was applied over the dorsolateral prefrontal cortex of autistic patients in an attempt to strengthen the inhibitory surround of minicolumns in this area. Due to the connectivity of this brain region we expected the intervention to generalize throughout the brain, i.e., a diaschisis effect.

Methods: Eighteen patients (DSM-IV, ADI-R diagnosed) and fifteen age-matched controls participated in the study. Twelve participants with autism were assigned to the TMS group, whereas six patients formed a waiting-list group. Repetitive TMS was delivered at 0.5 Hz, 90 % of motor threshold, twice per week, with 150 pulses per day. We used an oddball task with target and non-target Kanizsa illusory figures, and non-Kanizsa standards at pre- and post rTMS treatment stages. Participants in the waiting-list group were tested twice within a month.

Results: Outcome measures based on event related potentials (ERP), induced gamma band (30 Hz–80 Hz) EEG activity, and behavioral measures pre- and post-TMS showed significant improvement. Autistic subjects as compared to controls had higher magnitude of ERP components of interest to non-target rather than target stimuli. Similar between group differences was typical for gamma. Autistic patients showed higher gamma density in response to non-target stimuli. In all autistic subjects the difference of gamma density for target and non-target Kanizsa figures was negative. TMS resulted in decrease of the amplitude in the frontal ERPs to non-target stimuli, but not to the target Kanizsa stimuli in TMS group. Effects of rTMS on the posterior ERP were significant for the latency of P3b, which decreased to non-targets, but not to targets. TMS affected the power of gamma in response to non-target stimuli on the ipsilateral frontal and parietal sites. The power of induced gamma to non-targets dramatically decreased, and difference between responses to the target and non-target items became less negative. Reaction time to targets did not change after rTMS, but response accuracy significantly increased. Results of the clinical evaluations showed that following rTMS patients with autism were reported to have reduced repetitive-ritualistic behavior as measured by the Repetitive Behavior Scales. This change was due to reduced obsessive-compulsive behaviors reported by caregivers. There was a trend toward reduction of hyperactivity following rTMS.

Conclusions: Selected electrocortical functional outcome measures were shown as sensitive markers of functional connectivity changes and improved excitatory/inhibitory balance after rTMS trial. The results suggest that the brains of autistic patients are often inappropriately activated and that rTMS offers a potential innovative therapeutic intervention.