Different hierarchical reconfigurations in the brain by psilocybin and escitalopram for depression

Two serotonergic interventions—psilocybin therapy and the antidepressant escitalopram—rebalance brain dynamics in major depressive disorder through opposite hierarchical reconfigurations. In a double-blind phase II trial, 22 patients received two 25 mg doses of psilocybin plus daily placebo, while 20 patients received two 1 mg doses of psilocybin plus daily escitalopram. Resting-state fMRI scans before and after treatment, analyzed with generative effective connectivity models, showed that the two treatments produced significantly different and opposite changes in whole-brain hierarchy. Machine learning predicted treatment response with 85% accuracy. The findings suggest that depression may involve disrupted function of brain regions that orchestrate dynamics from the top of the hierarchy.

Study at a glance

Abstract

Abstract Effective interventions for neuropsychiatric disorders may work by rebalancing the brain’s functional hierarchical organization. Here we directly investigated the effects of two different serotonergic pharmacological interventions on functional brain hierarchy in major depressive disorder in a two-arm double-blind phase II randomized controlled trial comparing psilocybin therapy (22 patients) with escitalopram (20 patients). Patients with major depressive disorder received either 2 × 25 mg of oral psilocybin, three weeks apart, plus six weeks of daily placebo (‘psilocybin arm’) or 2 × 1 mg of oral psilocybin, three weeks apart, plus six weeks of daily escitalopram (10–20 mg; ‘escitalopram arm’). Resting-state functional magnetic resonance imaging scans were acquired at baseline and three weeks after the second psilocybin dose ( NCT03429075 ). The brain mechanisms were captured by generative effective connectivity, estimated from whole-brain modeling of resting state for each session and patient. Hierarchy was determined for each of these sessions using measures of directedness and trophic levels on the effective connectivity, which captures cycle structure, stability and percolation. The results showed that the two pharmacological interventions created significantly different hierarchical reconfigurations of whole-brain dynamics with differential, opposite statistical effect responses. Furthermore, the use of machine learning revealed significant differential reorganization of brain hierarchy before and after the two treatments. Machine learning was also able to predict treatment response with an accuracy of 0.85 ± 0.04. Overall, the results demonstrate that psilocybin and escitalopram work in different ways for rebalancing brain dynamics in depression. This suggests the hypothesis that neuropsychiatric disorders could be closely linked to the breakdown in regions orchestrating brain dynamics from the top of the hierarchy.