Repetitive transcranial magnetic stimulation induces oscillatory power changes in chronic tinnitus

Chronic tinnitus is associated with neuroplastic changes in auditory and non-auditory cortical areas. About 10 years ago, repetitive transcranial magnetic stimulation (rTMS) of auditory and prefrontal cortex was introduced as potential treatment for tinnitus. The resulting changes in tinnitus loudness are interpreted in the context of rTMS induced activity changes (neuroplasticity). Here, we investigate the effect of single rTMS sessions on oscillatory power to probe the capacity of rTMS to interfere with tinnitus-specific cortical plasticity. We measured 20 patients with bilateral chronic tinnitus and 20 healthy controls comparable for age, sex, handedness, and hearing level with a 63-channel electroencephalography (EEG) system. Educational level, intelligence, depressivity and hyperacusis were controlled for by analysis of covariance. Different rTMS protocols were tested: Left and right temporal and left and right prefrontal cortices were each stimulated with 200 pulses at 1 Hz and with an intensity of 60% stimulator output. Stimulation of central parietal cortex with 6-fold reduced intensity (inverted passive-cooled coil) served as sham condition. Before and after each rTMS protocol 5 min of resting state EEG were recorded. The order of rTMS protocols was randomized over two sessions with 1 week interval in between. Analyses on electrode level showed that people with and without tinnitus differed in their response to left temporal and right frontal stimulation. In tinnitus patients left temporal rTMS decreased frontal theta and delta and increased beta2 power, whereas right frontal rTMS decreased right temporal beta3 and gamma power. No changes or increases were observed in the control group. Only non-systematic changes in tinnitus loudness were induced by single sessions of rTMS. This is the first study to show tinnitus-related alterations of neuroplasticity that were specific to stimulation site and oscillatory frequency. The observed effects can be interpreted within the thalamocortical dysrhythmia model assuming that slow waves represent processes of deafferentiation and that high frequencies might be indicators for tinnitus loudness. Moreover our findings confirm the role of the left temporal and the right frontal areas as relevant hubs in tinnitus related neuronal network. Our results underscore the value of combined TMS-EEG measurements for investigating disease related changes in neuroplasticity.