Anterior cingulate desynchronization and functional connectivity with the amygdala during a working memory task predict rapid antidepressant response to ketamine.

Pregenual anterior cingulate cortex (pgACC) hyperactivity differentiates treatment responders from non-responders to various pharmacological antidepressant interventions, including ketamine, an N-methyl-D-aspartate receptor antagonist. Evidence of pgACC hyperactivition during non-emotional working memory tasks in patients with major depressive disorder (MDD) highlights the importance of this region for processing both emotionally salient and cognitive stimuli. However, it is unclear whether pgACC activity might serve as a potential biomarker of antidepressant response during working memory tasks as well, in line with previous research with emotionally arousing tasks. This study tested the hypothesis that during the N-back task, a widely used working memory paradigm, low pretreatment pgACC activity, as well as coherence between the pgACC and the amygdala, would be correlated with the clinical improvement after ketamine. Magnetoencephalography (MEG) recordings were obtained from 15 drug-free patients with MDD during working memory performance 1 to 3 days before receiving a single ketamine infusion. Functional activation patterns were analyzed using advanced MEG source analysis. Source coherence analyses were conducted to quantify the degree of long-range functional connectivity between the pgACC and the amygdala. Patients who showed the least engagement of the pgACC in response to increased working memory load showed the greatest symptomatic improvement within 4 h of ketamine administration (r=0.82, p=0.0002, false discovery rate (FDR) <0.05). Pretreatment functional connectivity between the pgACC and the left amygdala was negatively correlated with antidepressant symptom change (r=-0.73, p=0.0021, FDR <0.05).These data implicate the pgACC and its putative interaction with the amygdala in predicting antidepressant response to ketamine in a working memory task context.