Brain Connectivity Correlates Of Accelerated RTMS Treatment In Major Depression

Major depression disorder(MDD)or clinical depression is a common but serious mental health problem,and also one of the leading causes of the overall global burden of disease.Unfortunately,approximately 30% of MDD patients are nonresponsive to the regular pharmacological and psychothetrapeutic treatments.For these patients with treatment-resistant depression(TRD),neuromodulation therapy such as repetitive Transcranial Magnetic Stimulation(rTMS)is gradually recognized as an effective treatment option.Due to its non-invasive nature and very low risk of lasting side effects,rTMS has been approved by US Food and Drug Administration(FDA)in 2008 to alleviate symptoms of depressed patients.Major depression has been characterized as a neural network-level disorder.Large-scale network dysfunction has been tightly linked to core deficits in cognitive and affective functioning in MDD,as reflected by resting-state functional connectivity(rsFC)abnormalities.Among the major functionally relevant network organizations,three may underlie major depression: the default mode network(DMN),the frontoparietal network(FPN)and the salience network(SN).Each of these three networks,as well as the central regions of them-specifically the rostral anterior cingulate(rACC),the dorsolateral prefrontal cortex(DLPFC)and the dorsal anterior cingulate cortex(dACC)-have shown promise for the assessment of active antidepressant response,non-invasive brain stimulation methods(e.g.,rTMS)included.Recent meta-analytic findings suggested that the clinical efficacy of rTMS for depressed patients could be decomposed into treatment-specific(i.e.,neuro-electrophysiological effects)and non-specific factors effects(i.e.,placebo responses).It has been suggested that the placebo effect is actually a neurobiological phenomenon and a core network of brain regions associated within the DMN may play a critical role in this phenomenon,by influencing cognitive appraisals about the significance of treatment context for the self.However most of the aforementioned neuroimaging studies focusing on rTMS in the treatment of depression,due to the lack of sham control arms,cannot exclude placebo responses.Therefore,we aimed to distinguish treatment-specific effects and placebo effects on the neural level and to develop functional connectivity biomarkers for such treatment responses in the first study.Neuroimaging studies focusing on rTMS treatment for depression have also highlighted structural changes in limbic system and forebrain regions,such as the hippocampus,dACC,rACC and medial prefrontal cortex.In particular,several structural measures derived from T1-weighted images(e.g.,gray matter density,gray matter volume,cerebral cortical thickness)may be potential biomarkers that predict antidepressant response to rTMS.However,these studies have yielded inconsistent results or even opposite conclusions to date,and this could be due to the lack of investigation on structural covariance/interactions between distinct brain regions.It is also important to note that depressive symptoms are associated with abnormal structural covariance between regions that form functional networks,such as the DMN,SN,FPN and the frontal-limbic system.Consequently,the second study conducted cortical thickness analyses to examine whether the rTMS responses are predictable from structural covariance profiles.In the first study,a specific form of rTMS parameters,called accelerated intermittent theta burst stimulation was used in the treatment of majorly depressed patients,which is thought to result in equal if not greater effects than standard rTMS.Specifically,we applied a sham-controlled,cross-over experimental design covering two weeks.The 17-item Hamilton Depression Rating Scale(HAMD)was used to assess depression severity at pre-stimulation,and after each week of treatment.To have an idea of the magnitude of the real and sham rTMS responses,the percent reduction in HAMD scores from baseline to the end of first-week treatment period was calculated.Importantly,in order to avoid a priori selection of analysis regions of interest,we performed the resting-state functional connectivity(FC)analysis based on a whole-brain voxel-wise connectivity metric denoting the average strength of global brain connectivity.Both of the conjunction analysis and contrast analysis were then performed to investigate the relationship between the antidepressant effects of real and sham rTMS.Paired T-tests showed that HAMD scores pre-and post-stimulation were statistically different in both real and sham group.Sham rTMS response was lower than actual effects,but the effect size for the differences between the two groups was small.In contrast to the sham condition,the treatment-specific effects of rTMS were associated with the FC of FPN regions(inferior parietal lobule,DLPFC)and SN areas(dACC).Meanwhile,the conjunction analysis results revealed that baseline resting-state FC strength of the rACC/vmPFC region(part of the DMN)significantly predicted depression severity symptom improvement to both real and sham rTMS.In the second study,we mainly focused on the rTMS treatment responses of the entire two-week period.T1-weighted anatomical brain scans were used to estimate vertex-wise cortical thickness and then for seed-based structural covariance network analysis.In order to increase our insight into all the components of the therapeutic response to rTMS,seed ROIs(regions of interest)were selected based on the findings of the first study(key regions of the DMN,SN and FPN).As expected,clinical effects of rTMS treatment on depression severity remained statistically significant after the second-week period,and the order of stimulation had no significant impact on the clinical result.The cortical thickness analysis results revealed that patterns of structural correlations from the seed in DLPFC and dACC were identified to predict the beneficial effects of rTMS treatment.Specifically,the seed in DLPFC co-varied strongly with inferior parietal lobule,precentral gyrus,superior and transverse temporal gyrus,while the pattern of structural covariance from the seed in dACC encompassed regions in lateral orbitofrontal cortex,rostral and caudal middle frontal cortex,postcentral gyrus,inferior and superior parietal lobule,supramarginal gyrus,superior and transverse temporal gyrus.Moreover,the structural covariance network of DMN regions was not modulated by rTMS responses.There were also no significant associations between structural covariance and placebo effects.Overall,the brain imaging study combined voxel-wise functional connectivity and vertex-wise structural covariance metrics to investigate the brain network correlates of the antidepressant responses to accelerated rTMS treatment,and to distinguish placebo responses from actual effects of rTMS at the systems level.The clinical results confirmed the effectiveness and efficiency of rTMS in treating patients with depression,which was useful in increasing the treatment capacity of clinics.The brain connectivity analysis results support the assumption that different components of the therapeutic effect of rTMS might be associated with different parts of the neurocircuit implicated in major depression.Specifically,the DMN(especially its key region,vmPFC)may play a central role in placebo response,while pretreatment individual connectivity profiles of the FPN and SN may be important predictors of treatment-specific effects for rTMS depression trials.Our findings also suggest that rTMS could exert its effects by affecting deeper brain structures related to emotion regulation if connectivity(whether structural or functional)exists between its stimulation target and these regions.Furthermore,the current study provides brain imaging evidence in support of the dACC as a promising alternative target for rTMS in depression.These findings shed light on the brain connectivity correlates of rTMS treatment response in treatment-resistant depression,and might be helpful to identify predictive biomarkers of rTMS response,to optimize the design of clinical trails and to guide more individualized treatment approaches.