One-Year Durability of the Clinical Effect of Accelerated Compared to Once-Daily Deep TMS for Depression- Long-Term Follow-Up After a Multisite Randomized Trial

Background: rTMS to left DLPFC is an effective treatment for depression, but optimal targeting approaches remain undetermined.We used surface-based brain reconstruction based upon the Human Connectome Project Multimodal Parcellation atlas (HCP-MMP1.0)to determine individualized treatment targets for open-label rTMS.Methods: Data was obtained from 15 individuals with moderate to severe depression resistant to at least two adequate antidepressant trials, of whom 11 received rTMS targeting the group-mean peak anticorrelated region to sgACC refined as the junction of parcels 46, 9-46d and 9-46v, with 4 receiving treatment targeting the middle of parcel 46.Pre-treatment MADRS scores (mean SEM) were 30.9 1.7 and 33.32.5, respectively.All individuals received standard 10-Hz rTMS for 6-weeks.Resting-state fMRI was obtained prior to and following treatment.Results: Among patients receiving junction-targeted rTMS, there was a 57.89.6% reduction in MADRS total score.Response and remission rates were 72.7% and 54.5%, respectively.Pre-treatment anti-correlation between parcel 46 and sgACC significantly predicted treatment response with area under the ROC curve of.875.110and sensitivity/specificity of.75 and 1.0 at maximal discrimination.For remission, the AUC was 1.0.Targeting the midpoint of parcel 46 produced a mean reduction if 38.219.3%, with a response rate of 25%.Discussion: These findings demonstrate feasibility of individual, parcel-guided rTMS and suggest response/remission rates similar to that observed using individualized localization of the peak anticorrelated region.Parcel-based RSFC calculations between DLPFC (parcel 46) and sgACC strongly predicted treatment response, suggesting that baseline patient characteristics may be as critical as targeting approaches for predicting treatment response.