Background
At the group level, antidepressant efficacy of rTMS targets is inversely related to their normative connectivity with subgenual anterior cingulate cortex (sgACC). Individualized connectivity may yield better targets, particularly in patients with neuropsychiatric disorders who may have aberrant connectivity. However, sgACC connectivity shows poor test-retest reliability at the individual level. Individualized resting-state network mapping (RSNM) can reliably map inter-individual variability in brain network organization.
Objective
To identify individualized RSNM-based rTMS targets that reliably target the sgACC connectivity profile.
Methods
We used RSNM to identify network-based rTMS targets in 10 healthy controls and 13 individuals with traumatic brain injury-associated depression (TBI-D). These “RSNM targets” were compared with consensus structural targets and targets based on individualized anti-correlation with a group-mean-derived sgACC region (“anti-group-mean sgACC targets”). The TBI-D cohort was randomized to receive active (n=9) or sham (n=4) rTMS to RSNM targets.
Results
The group-mean sgACC connectivity profile was reliably estimated by individualized correlation with default mode network (DMN) and anti-correlation with dorsal attention network (DAN). Individualized RSNM targets were then identified based on DAN anti-correlation and DMN correlation. Counterintuitively, anti-correlation with the group-mean sgACC connectivity profile was stronger and more reliable for RSNM-derived targets than for “anti-group-mean sgACC targets”. Improvement in depression after RSNM-targeted rTMS was predicted by target anti-correlation with the portions of sgACC. Active treatment led to increased connectivity within and between several relevant regions.
Conclusions
RSNM may enable reliable individualized rTMS targeting, although further research is needed to determine whether this personalized approach can improve clinical outcomes.