EFFECT OF KETAMINE ON INDIVIDUAL SYMPTOMS OF DEPRESSION IN A RANDOMIZED CONTROLLED TRIAL OF KETAMINE FOR TREATMENT-RESISTANT DEPRESSION

Intravenous ketamine has emerged as an effective rapid acting treatment for treatment resistant depression (TRD). However, the precise mechanism of action of ketamine is not yet understood. Moreover, not all patients respond to ketamine, and it remains unclear who is likely to respond to ketamine. The difficulty in understanding the mechanism of action of ketamine and its clinical effects might be related to the heterogeneity of TRD symptom profiles. An understanding of ketamine’s differential effects on depressive symptoms could provide insight into these issues. In this secondary analysis of a randomized controlled trial, 152 individuals with TRD were randomized to receive a single infusion of IV ketamine (n=103) or IV saline (n=49). Clinician-ratings of depressive symptoms were assessed at multiple time points from pre-infusion to 30 days post-infusion using Montgomery–Åsberg Depression Rating Scale (MADRS). We examined the effect sizes of improvements in specific symptoms between ketamine and saline arms in the early post-infusion period (i.e., ≤ 1 day post-infusion), where peak antidepressant effects are observed. Second, we assessed individual symptom improvement related to ketamine in the later treatment period (i.e. postinfusion days 2 – 30), examining individual symptom profiles over time. Also, to study how ketamine affects the relationship between depressive symptoms (the mutual effect of a pair of depressive symptoms on each other), we estimated the depressive symptom networks with Gaussian Graphical Models (GGM) at pre-infusion and 24-hrs-post-infusion. Compared to saline infusion, the greatest effects of ketamine at 24-hr post-infusion were on symptoms of apparent sadness, reported sadness, sleep problems, and numbness (ds > 0.2, p < 0.05). For certain symptoms such as lassitude, the greatest improvement in symptoms was found on day 5 (ds > 0.5, p < 0.05), and suicidality showed maximal improvement later in treatment on day 21 and day 30 (dday21=0.44, p < 0.05). The symptom networks did not reveal an overall network structure change pre-to post-infusion; however, the global strength of the network post infusion increased significantly, suggesting a denser symptom network 24-hrs post infusion. Moreover, the association between apparent sadness and reported sadness increased. Overall, ketamine improved all symptoms, but appeared to improve symptoms of sadness most profoundly, both immediately and in the initial week after treatment. Ketamine also rapidly altered the architecture of symptom networks, appearing to strengthen interrelationships between residual symptoms. Efficacy for ketamine (over saline) in improving suicidal symptoms emerged not rapidly, but only after a greater delay. Our findings suggest potentially divergent efficacy, time course, and mechanisms for different symptoms of depression.