Convergent increases in serotonin 1B receptor binding following ketamine and electroconvulsive therapy: a multi-centre bayesian re-analysis of PET data

The serotonin 1B receptor (5-HT1BR) can be imaged in living humans using a PET tracer called [11C]AZ10419369 and is linked to major depressive disorder (MDD) and its treatment. Ketamine and electroconvulsive therapy (ECT) are rapid-acting antidepressants that raise serotonin levels, but whether they directly alter serotonin receptors was unclear. Reanalyzing 222 PET scans from three centers—including MDD patients before and after ketamine (19 completers), saline placebo (10), or ECT (13 completers)—using a hierarchical Bayesian method, the authors demonstrate large increases in 5-HT1BR binding after both ketamine (6.4%, 95% CI: 3.1–9.6%) and ECT (9.3%, 95% CI: 4.3–14.2%).

Study at a glance

Abstract

The serotonin 1B receptor can be studied in vivo with PET using [^11C]AZ10419369, and has been linked to both the pathophysiology and treatment of major depressive disorder (MDD). Ketamine and electroconvulsive therapy (ECT) both exert rapid and potent antidepressive effects, and although these treatments may not act directly on the serotonin system, they both cause dose-dependent increases in serotonin levels, and there is convergent evidence suggesting that the serotonin system may be important for their mechanisms of action. In this study, we re-analysed a multi-centre dataset of 222 [^11C]AZ10419369 PET measurements from three centres, including MDD patients examined with PET both before and after treatment with ketamine ( n = 19 completers), saline placebo ( n = 10), or ECT ( n = 13 completers). Using a hierarchical Bayesian approach (SiMBA) that takes advantage of the full dataset to improve parameter estimation and enable data harmonisation across centres, we demonstrate large increases in 5-HT_1BR binding following both ketamine (6.4%, 95% CI: 3.1–9.6%) and ECT (9.3%, 95% CI: 4.3–14.2%). Ketamine-induced changes were statistically distinguishable from placebo, and an exploratory cross-centre comparison enabled by the data harmonisation within the combined modelling framework, suggests that ECT-induced changes are also distinguishable from placebo. These changes were not associated with individual symptom improvement. These findings suggest that despite differences in the pri-mary target of these rapid acting treatments for depression, they may converge on similar downstream changes to the serotonin system.