Features of dissociation differentially predict antidepressant response to ketamine in treatment-resistant depression

This follow-up study (n=126) investigated whether the antidepressant effects of ketamine (35mg/70kg) were related to dissociative symptoms experienced by patients with (bipolar) depression, and found that the subjective effects of depersonalization were most closely related to the antidepressant response.

Abstract

“Background: Ketamine induces rapid and robust antidepressant effects, and many patients also describe dissociation, which is associated with antidepressant response. This follow-up study investigated whether antidepressant efficacy is uniquely related to dissociative symptom clusters.

Methods: Treatment-resistant patients with major depressive disorder (MDD) or bipolar disorder (BD) (n = 126) drawn from three studies received a single subanesthetic (0.5 mg/kg) ketamine infusion. Dissociative effects were measured using the Clinician-Administered Dissociative States Scale (CADSS). Antidepressant response was measured using the 17-item Hamilton Depression Rating Scale (HAM-D). A confirmatory factor analysis established the validity of CADSS subscales (derealization, depersonalization, amnesia), and a general linear model with repeated measures was fitted to test whether subscale scores were associated with antidepressant response.

Results: Factor validity was supported, with a root mean square error of approximation of .06, a comparative fit index of .97, and a Tucker-Lewis index of .96. Across all studies and timepoints, the depersonalization subscale was positively related to HAM-D percent change. A significant effect of derealization on HAM-D percent change was observed at one timepoint (Day 7) in one study. The amnesia subscale was unrelated to HAM-D percent change.

Limitations: Possible inadequate blinding; combined MDD/BD datasets might have underrepresented ketamine’s antidepressant efficacy; the possibility of Type I errors in secondary analyses.

Conclusions: From a psychometric perspective, researchers may elect to administer only the CADSS depersonalization subscale, given that it was most closely related to antidepressant response. From a neurobiological perspective, mechanistic similarities may exist between ketamine-induced depersonalization and antidepressant response, although off-target effects cannot be excluded.”

Authors: Mark J. Niciu, Bridget J. Shovestul, Brittany A. Jaso, Cristan Farmer, David A. Luckenbaugh, Nancy E. Brutsche, Lawrence T. Park, Elizabeth D. Ballard & Carlos A. Zarate Jr.

Summary

Ketamine induces rapid and robust antidepressant effects, and many patients also describe dissociation.

1. Introduction

The glutamatergic modulator ketamine is an FDA-approved anesthetic that causes mild to moderate dissociation. Dissociation can range in severity from amnesia to fugue states and dissociative identity disorder. Ketamine is thought to exert its anesthetic, psychotomimetic, and dissociative properties through noncompetitive, high-affinity antagonism at the N-methyl-D-aspartate (NMDA) receptor, although recent preclinical evidence supports non-NMDA receptor-mediated antidepressant properties of the ketamine metabolite (2R,6R)-hydroxynorketamine.

Ketamine’s psychoactive side effects have been explored in several studies, with mixed results. One study found no association between maximum change in dissociative symptoms and change in depressive symptom scores, while another study found a correlation between antidepressant response and psychotomimetic symptoms.

This follow-up study uses a larger sample and explores empirically validated subdimensions of dissociation for the first time to determine the relationship between ketamine-induced dissociative symptoms and antidepressant efficacy.

2.1. Participants and procedures

126 treatment-resistant depressed patients were analyzed from three studies, Ketamine-Bipolar, Ketamine-Riluzole, and Ketamine-MOA. All subjects provided written informed consent as approved by the NIH Combined Neuroscience Institutional Review Board.

All patients were currently experiencing a major depressive episode without psychotic features, and had failed to respond to at least one adequate antidepressant trial. Ketamine was administered over 40 min in either double-blind, placebo-controlled (Ket-MOA and Ket-BD) or open-label (Ket-Riluzole) designs.

2.2. Instruments

Depressive and dissociative symptoms were assessed using the HAM-D and CADSS, and the change in CADSS scores were calculated as a change from baseline to 40 min post-infusion.

2.3. Statistical analysis

The factor validity of the CADSS subscales described by Bremner and colleagues was established using confirmatory factor analysis.

We fitted a general linear model with maximum likelihood estimation and repeated measures with a compound symmetry covariance structure to HAM-D percent change scores to evaluate whether subdimensions of dissociation were uniquely related to antidepressant response.

3.2. Factor validity of the CADSS subscales

The CADSS subscales were factor valid and had a good fit to the data (RMSEA: .06, CFI: .97, Tucker-Lewis Index: .96).

3.3. Ketamine-induced dissociation and antidepressant response

The results of the four mixed models indicate that the CADSS total score was the strongest predictor of change in HAM-D score, and that the relationship between CADSS total score and change in HAM-D score was negative but nonsignificant at 230 min and Day 1.

In all studies, the Derealization subscale was negative but non-significant. In the Ket-MOA study, the Derealization subscale was significantly associated with percent change in HAM-D score at Day 7, but not in the Ket-BD or Ket-Riluzole studies.

The Depersonalization subscale was negatively related to per-cent change in HAM-D score, but this effect did not vary by time or study.

4. Discussion

The present study extends previous work from our laboratory investigating ketamine-induced dissociation and antidepressant response, and found that increased Derealization and Depersonalization were modest predictors of antidepressant response to subanesthetic-dose ketamine infusion.

Ket-BD subjects were maintained on therapeutic doses of mood stabilizers, had a longer course of illness, and a shorter duration of antidepressant response than Ket-MOA subjects. In addition, BD subjects were the least likely to experience dissociation.

The Depersonalization subscale of the CADSS was associated with greater antidepressant response across all time points and studies, suggesting that the underlying neurobiological constructs measured by this subscale may relate to ketamine’s antidepressant mechanism(s) of action.

This study validated the CADSS subscales for use in dissociative disorders, including post-traumatic stress disorder, and healthy volunteers, and found that the Depersonalization subscale was the most predictive of antidepressant response.

This study has several important limitations, including the potential for inadequate blinding, increased levels of dissociation, and Type I (false positive) errors. It also uses a combined MDD and BD dataset, which may underrepresent ketamine’s antidepressant efficacy.

The data support the factor validity of the CADSS dissociative subscales in an actively depressed population with MDD and BD, and suggest a possible mechanistic overlap between ketamine-induced depersonalization and antidepressant response.

Role of funding source

This work was supported by the Intramural Research Program at the National Institute of Mental Health, National Institutes of Health, and by NARSAD and a Brain and Behavior Mood Disorders Research Award.

Declaration of interest

This work was supported by the Intramural Research Program at the National Institute of Mental Health, National Institutes of Health, a NARSAD Independent Investigator Award, and a Brain and Behavior Mood Disorders Research Award. The authors have no conflict of interest to disclose.