Neuroimaging correlates and predictors of response to repeated-dose intravenous ketamine in PTSD: preliminary evidence

This preregistered randomized double-blind active placebo-controlled between-subjects study (n=21) measured brain activity while viewing facial stimuli in response to repeated and acute administration of ketamine (35mg/70kg) and midazolam (3.15mg/70kg), among patients with severe PTSD. Both midazolam and ketamine improved symptoms, which were most reliably predicted by the reduced excitation of the amygdala and ventromedial prefrontal cortex (vmPFC) during the processing of socio-emotional threat signals, but only ketamine-related improvement was associated with increased top-down inhibition of the amygdala by the vmPFC.


Introduction: Promising initial data indicate that the glutamate N-methyl-D-aspartate (NMDA) receptor antagonist ketamine may be beneficial in post-traumatic stress disorder (PTSD).

Methods: Here, we explore the neural correlates of ketamine-related changes in PTSD symptoms, using a rich battery of functional imaging data (two emotion-processing tasks and one task-free scan), collected from a subset of participants of a randomized clinical trial of repeated-dose intravenous ketamine vs midazolam (total N = 21). In a pre-registered analysis, we tested whether changes in an a priori set of imaging measures from a target neural circuit were predictive of improvement in PTSD symptoms, using leave-one-out cross-validated elastic-net regression models (regions of interest in the target circuit consisted of the dorsal and rostral anterior cingulate cortex, ventromedial prefrontal cortex, anterior hippocampus, anterior insula, and amygdala).

Results: Improvements in PTSD severity were associated with increased functional connectivity between the ventromedial prefrontal cortex (vmPFC) and amygdala during emotional face-viewing (change score retained in model with minimum predictive error in left-out subjects, standardized regression coefficient [β] = 2.90). This effect was stronger in participants who received ketamine compared to midazolam (interaction β = 0.86), and persisted following inclusion of concomitant change in depressive symptoms in the analysis model (β = 0.69). Improvement following ketamine was also predicted by decreased dorsal anterior cingulate activity during emotional conflict regulation, and increased task-free connectivity between the vmPFC and anterior insula (βs = −2.82, 0.60). Exploratory follow-up analysis via dynamic causal modelling revealed that whilst improvement in PTSD symptoms following either drug was associated with decreased excitatory modulation of amygdala→vmPFC connectivity during emotional face-viewing, increased top-down inhibition of the amygdala by the vmPFC was only observed in participants who improved under ketamine. Individuals with low prefrontal inhibition of amygdala responses to faces at baseline also showed greater improvements following ketamine treatment.

Discussion: These preliminary findings suggest that, specifically under ketamine, improvements in PTSD symptoms are accompanied by normalization of hypofrontal control over amygdala responses to social signals of threat.”

Authors: Agnes Norbury, Sarah B. Rutter, Abigail B. Collins, Sara Costi, Manish K. Jha, Sarah R. Horn, Marin Kautz, Morgan Corniquel, Katherine A. Collins, Andrew M. Glasgow, Jess Brallier, Lisa M. Shin, Dennis S. Charney, James W. Murrough & Adriana Feder



Ketamine may improve symptoms of PTSD and depression, over and above effects on comorbid depression. Ketamine may do this by promoting synaptogenesis in the prefrontal cortex and hippocampus, and by facilitating the un- or re-learning of maladaptive associations that contribute to longer-term symptom maintenance.

Here, we analysed neuroimaging data collected during a recent randomized clinical trial of repeated-dose ketamine vs midazolam for chronic, severe PTSD. We identified a circuit feature most reliably related to symptom improvement, and performed exploratory follow-up analyses to confirm this finding.

Although this is an observational study, the data suggest that hypofrontal control over amygdala responses to social signals of threat is normalized following ketamine treatment, which may help promote reconsolidation and extinction of trauma memories.

MATERIALS AND METHODS Pre-registration of analysis

A pre-registered analysis plan was deposited with the Open Science Framework and time-stamped documents are available at the project OSF page. Follow-up analyses are described below.


Participants were individuals taking part in a clinical trial of repeated-dose ketamine for chronic PTSD. They gave written consent and the study received ethical approval.


Participants received six intravenous infusions of ketamine or midazolam three times per week for two weeks. Imaging data were collected during a baseline (pre-infusion) scan session and a second post-infusion scan session.

Clinical measures

The Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) was used to measure total symptom severity over the past week, and the Montgomery-sberg Depression Rating Scale was used to measure depression.

Drug side-effect measures

Clinician-rated dissociative, psychotomimetic, and manic effects and patient-rated somatic effects were included in the analysis of symptom change data.

Neuroimaging data

The emotional face-processing task and the emotional conflict regulation (face Stroop) paradigm were used to study how people process emotions. Stimuli were presented in pseudorandom order with jittered presentation timing, in an event-related design (60 congruent, 60 incongruent, and 60 baseline trials).

Imaging data was acquired using a 3 T Siemens Biograph mMR scanner with 64-channel head coil. T1-weighted structural images were collected using an MPRAGE sequence.

Imaging data were pre-processed using fMRIprep, version 1.1.4, and then denoised using non-aggressive independent components analysis (ICA-AROMA).

For the emotional face-processing and emotional conflict regulation tasks, first level models were specified using SPM12, run in MATLAB, version R2019a, and included erroneous responses as regressors of no interest.

Due to the small sample size, we focused on a target circuit consisting of the ventromedial prefrontal cortex, rostral and dorsal anterior cingulate cortices, anterior insula, amygdala, and anterior hippocampus.

For each functional scan, a pre-defined set of imaging measures were extracted, including mean regional BOLD signal, covariation in time series or functional connectivity between regions, and/or multivariate representational similarity across voxels within a region.

All imaging measures were first regressed against mean framewise displacement (head movement) recorded during the relevant functional scan. The residuals were then passed on to further analysis.

Statistical analysis

Further statistical analysis was carried out in R, version 3.6.1 (R Core Team, 2019). Elastic net regression was used to analyze imaging measures and sociodemographic variables related to PTSD symptom improvement following treatment.

A penalized regression model, called an elastic net, balances LASSO penalization with ridge penalization, and deals better with multicollinearity among predictors than LASSO alone. The parameters governing the balance were chosen via grid search and a leave-one-out cross-validation procedure.

We used imaging measures to assess changes in CAPS-5 total scores following treatment, and to identify predictors of treatment response. We also used MADRS total scores to assess changes in depressive symptoms following treatment.

The above models included several demographic and clinical measures that may potentially confound imaging or symptom change metrics, and were scaled prior to entering into the prediction models. The baseline prediction model also included several other variables that have previously been related to clinical response to ketamine or resilience to trauma-related psychopathology.

We conducted exploratory correlation analyses to examine how strongly changes in imaging metrics were related to changes in different PTSD symptom dimensions, and found that some dimensions shared conceptual overlap with symptoms of depression.

To gain insight into the directionality of functional connectivity effects identified in the main analysis, follow-up effective connectivity analysis was carried out via dynamic causal modelling.

Study participants and clinical measures

Participants had severe chronic PTSD and the most common primary trauma was sexual violence. Ketamine provided greater improvement in PTSD and depressive symptoms than midazolam, but accuracy on emotion-processing tasks did not vary according to imaging session or received drug.

Neuroimaging correlates of PTSD symptom improvement

Across all study participants, increased functional connectivity between the vmPFC and amygdala during emotional face-viewing was the strongest correlate of improvement in PTSD symptom severity. Increases in rACC BOLD during negative emotional conflict regulation, and greater task-free vmPFC-anterior insula connectivity were also retained in the minimum MSE model.

A follow-up analysis was carried out to investigate how changes in face-related vmPFC-amygdala connectivity were related to changes in symptoms across different PTSD symptom dimensions. There was no evidence of a relationship between pre-post change in vmPFC-amygdala connectivity and number of infusions received.

Exploratory follow-up analysis via dynamic causal modelling revealed that the onset of emotional face stimuli modulated the amygdalavmPFC pathway towards an excitatory connection, with weaker evidence for face-induced inhibition of the amygdala by the vmPFC. Using PEB analysis, researchers found that individuals who received ketamine had a stronger effect of increased top-down inhibition of the amygdala on PTSD symptoms than individuals who received midazolam.

Baseline prediction of PTSD symptom improvement

The strongest predictor of clinical response was lower baseline vmPFC-amygdala connectivity during the emotional face-viewing task. Moderate-to-large effects were also observed for lower baseline rACC BOLD during both emotional face-viewing and emotional conflict regulation tasks, and in individuals with more distinct representation of fearful vs neutral faces across rACC voxels.

Values represent mean (SD) unless otherwise specified. p values represent the result of statistical comparisons for each variable between drug groups.

The interaction between drug and face-related vmPFC-amygdala connectivity was retained when change in MADRS score was included in the model, indicating some specificity in predicting improvements in cardinal PTSD symptoms.

Exploratory follow-up analysis of baseline effective connectivity data revealed that individuals who responded better to ketamine tended to have lower baseline inhibition of the amygdala by the vmPFC and greater baseline excitement of the vmPFC by the amygdala, during viewing of emotional face stimuli.


In a sample of individuals with severe, chronic PTSD, increased functional connectivity between the amygdala and vmPFC during viewing of emotional face stimuli was the strongest correlate of symptom change across all subjects, with some evidence of a stronger effect in individuals who received ketamine. Exploratory follow-up analysis indicated that responders to ketamine treatment showed greater top-down vmPFC inhibition of the amygdala during emotional face viewing than to midazolam treatment, perhaps representing decreased threat-responsivity in responders via both common and drug-specific pathways.

Ketamine has been shown to activate brain regions associated with extinction learning, and may help remediate mood and stress-related symptoms by opening a window of plasticity.

Ketamine may facilitate the extinction of maladaptive fear responses related to trauma memories in individuals who respond to ketamine. This may explain why ketamine is most strongly related to improvements in prototypical PTSD symptoms.

The baseline prediction analysis indicated that lower baseline vmPFC-amygdala connectivity during emotional face-viewing was associated with improvement in PTSD symptoms following ketamine.

Ketamine treatment was associated with lower baseline vmPFC inhibition of amygdala during viewing of emotional stimuli, and with greater disparate representation of fearful vs neutral faces in the rACC. These findings may be reflective of a type of PTSD that is more likely to respond to this form of treatment.

The work presented here includes highly novel data collected during a rigorously controlled randomized clinical trial for PTSD. The directionality of effects identified here converges with those identified during previous PTSD treatment studies.

This is a small pilot study of the effects of ketamine on PTSD symptoms. Although we should be adequately powered to detect moderate-to-large effects, we urge caution when interpreting results from the emotion regulation (face Stroop) task. Recent work has highlighted that mean univariate BOLD measures derived from emotional processing tasks may exhibit poor within-subject reliability, which limits the ability to detect true between-subjects effects.

Improvement in PTSD symptom severity was related to connectivity changes between regions previously identified as showing abnormal activity in PTSD. These changes were primarily observed during processing of socio-emotional stimuli. We propose that down-regulation of threat responses and/or enhancement of fear extinction learning may contribute to decreases in the severity of prototypical PTSD symptoms following ketamine, and that this might help explain the greater rate of response and longer time to relapse in individuals who received ketamine.

Drs. Feder, Charney, Jha, Collins, and Murrough are named co-inventors on patents related to the use of ketamine for the treatment of PTSD, and Dr. Murrough is named on a patent related to the use of KCNQ channel openers to treat depression and related conditions.

Dr. Murrough is not named on any patents related to ketamine use in the treatment of depression or PTSD. Dr. Charney is named as co-inventor on patents filed by the ISMMS for the treatment of treatment-resistant depression and suicidal ideation. The medical school has entered into a licensing agreement with Janssen Pharmaceuticals, Inc. Dr. Charney is a named co-inventor on several patents filed by ISMMS and has received payments related to the use of a cognitive training intervention for the treatment of psychiatric disorders. The other authors declare no conflicts of interest.


Contributions to the conception and design of the work were made by MKJ, LMS, DSC, JWM, AF, and AN, SBR,ABC,SC,SRH,MK,MC,KAC,AMG,JB,JMW,AF.

Study details

Compounds studied

Topics studied

Study characteristics
Placebo-Controlled Active Placebo Double-Blind