Study On Baseline Voxel-mirrored Homotopic Connectivity Characteristics In Predicting The Treatment Outcome Of SSRI In Major Depressive Disorder

Objectives:Neuroimaging studies showed that functional changes of key brain regions and networks might be involved in the occurrence,development and prognosis of major depressive disorder,and functional characteristics of the brain at baseline may be closely related to the later treatment response to antidepressant medications.The purposc of this study was:1.To explore the differences of baseline voxel-mirrored homotopic connectivity(VMHC)in patients with good treatment response and poor treatment response,as well as their mechanisms in the improvement of clinical symptoms.2.To investigate the predictive value of baseline VMHC characteristics based on the relationship between VMHC and the treatment response of SSRI,trying to establish the neuroimaging predictors of the efficacy of SSRI antidepressants.Methods:The present study used a longitude follow-up study design.Major depressive disorder patients who met DSM-IV criteria went through resting-state functional magnetic resonance imaging scans at baseline and then were given SSRI antidepressant medications for eight weeks.The 17-item Hamilton Depression Scale(IIAMD-17)was used to assess the outcome of medications.Based on the baseline VMHC,biomarkers were explored to predict the treatment response of eight-week acute medications.Participants included sixty patients in total,and four of them were excluded due to the further quality control(1.5mm and 1.5°).Analysis 1:Based on the reduction rate of HAMD-17,subjects whose 8-week HAMD-17 score dropped greater than or equal to 50%from baseline were defined as responders.Others were defined as non-responders.The VMHC was calculated and analyzed to compare differences between responders and non-responders.Partial correlation analysis was performed between the treatment outcome and baseline VMHC of brain regions where there were obvious differences.Receiver operating characteristic(ROC)analysis was applied to distinguish non-responders from responders according to the altered VMHC.Analysis 2:Given that there were lots of voxels in the brain investigated in the functional magnetic resonance imaging,we discussed as following to make sure of the validity.In clinical practices,some patients had a distinct improvement of clinical symptoms and achieved remission within 8 weeks of acute treatment.So based on the endpoint HAMD-17 scores after eight-week SSRI medications,subjects whose endpoint HAMD-17 score was less than or equal to seven were defined as remitters.Others were defined as non-remitters.The VMHC was calculated and analyzed to compare differences between remitters and non-remitters.Partial correlation analysis was performed between the treatment outcome and VMHC of brain regions where there were obvious differences.ROC was applied to distinguish non-remitters from remitters according to the altered VMHC.Results:There were statistically significant baseline VMHC differences as following(GRF corrected,voxel P<0.01,cluster P<0.05):1.Compared with responders(n=35),non-responders(n=21)showed a significantly greater VMHC in the bilateral posterior cingulate cortex,precuneus,lingual gyrus and calcarine(t=4.64)and it showed a negative correlation between VMHC values and the reduction rate of HAMD-17 scores(r=-0.710,P<0.001)and a positive correlation between VMIIC values and the endpoint HAMD-17 scores(r=0.708,P<0.001).The ROC curve indicated that VMHC in these brain regions mentioned above can distinguish non-responders from responders[Area under the curve(AUC)=0.963,P<0.001,95%CI=0.921-1.000,Sensitivity=85.7%,Specificity=94.3%].2.Compared with remitters(n=29),non-remitters(n=27)showed a significantly greater VMHC in the bilateral posterior cingulate cortex,lingual gyrus and calcarine(t=4.83)and there was a negative correlation between VMHC values and the reduction rate of HAMD-17 scores(r=-0.622,P<0.001)and a positive correlation between VMHC values and the endpoint HAMD-17 scores(r=0.659,P<0.001).The ROC curve indicated that VMHC in these brain regions mentioned above can distinguish non-remitters from remitters[AUC=0.858,P<0.001,95%CI=0.761-0.955,Sensitivity=81.5%,Specificity=82.8%].Conclusions:1.There were significant differences in certain brain regions at baseline in patients with different response statuses.Compared to the patients with good responses,patients with poor responses had a greater VMHC in the bilateral posterior cingulate cortex/precuneus/lingual gyrus/calcarine.The characteristics of VMHC might be a potential tool to predict treatment responses at baseline.2.The baseline VMHC in the bilateral posterior cingulate cortex/precuneus/lingual gyrus/calcarine might be associated with the later symptom improvement,providing a theoretical basis for further elucidate the neuropathological mechanism of the symptom improvement in first-episode major depressive disorder.3.Two analyses showed a concordant trend,and ROC indicated that the higher VMHC in the posterior cingulate cortex/precuneus/lingual gyrus/calcarine appears to be a predictor of poor response to SSRI medication.