THE EFFECT OF KETAMINE ON BRAIN FUNCTIONAL DYNAMICS

1745 

Abstract Submission 

Nooshin Javaheripour¹, Zümrüt Duygu Sen¹, Frederick Lange², Meng Li³, Lena Danyeli¹, Carina Heller⁴, Lejla Colic⁵, Habib Ganjgahi⁶, Martin Walter⁷

¹Department of Psychiatry and Psychotherapy, Jena University Hospital, Jena, Thuringen, ²Wellcome Centre for Integrative Neuroimaging University of Oxford, Oxford, Oxfordshire, ³Jena University Hospital, Jena, Thuringia, ⁴Friedrich Schiller University Jena, Jena, Germany, ⁵Jena University Hospital, Jena, Germany, ⁶University of Oxford, Oxford, United Kingdom, ⁷Department of Psychiatry and Psychotherapy, Jena University Hospital, Jena, Germany

Nooshin Javaheripour  
Department of Psychiatry and Psychotherapy, Jena University Hospital
Jena, Thuringen

Zümrüt Duygu Sen  
Department of Psychiatry and Psychotherapy, Jena University Hospital
Jena, Thuringen

Frederick Lange  
Wellcome Centre for Integrative Neuroimaging University of Oxford
Oxford, Oxfordshire

Meng Li  
Jena University Hospital
Jena, Thuringia

Lena Danyeli  
Department of Psychiatry and Psychotherapy, Jena University Hospital
Jena, Thuringen

Carina Heller  
Friedrich Schiller University Jena
Jena, Germany

Lejla Colic  
Jena University Hospital
Jena, Germany

Habib Ganjgahi  
University of Oxford
Oxford, United Kingdom

Martin Walter  
Department of Psychiatry and Psychotherapy, Jena University Hospital
Jena, Germany

A single sub-anaesthetic dose of ketamine induces transient alterations in cognitive functions and sensory perception. This study aims to comprehensively examine ketamine's impact on brain dynamics, particularly its brief dissociative effects, using the hidden Markov model (HMM). Functional dynamics during intravenous esketamine and 24 hours post-infusion in resting-state fMRI are investigated, along with the association of dynamic features with subjective experiences measured by the Altered States of Consciousness Scale (5D-ASC).

A double-blinded randomized placebo-controlled crossover study was designed on 34 healthy participants (mean ± SD = 25.1 ± 4.2 years). The participants were randomized into two groups by treatment order (ketamine-placebo and placebo-ketamine) for two periods (day 1: baseline and infusion and day 2: 24h post-infusion) in each treatment arm (ketamine and placebo) with a three-week washout period. On day 1, participants underwent MR scanning (ultra-high field 7T scanner, Siemens) before and during the infusion of either esketamine (at a dose of 0.33 mg/kg body weight, brand name "Ketanest S Pfizer") or 0.9% saline solution. On day 2, the same neuroimaging protocol was used. MRI data were preprocessed by the FSL pipeline, and artefacts were removed using ICA-FIX. To obtain time series for cortical regions, we used Yeo 2011 parcellation scheme with seven functional networks (DMN, FPN, SN, DAN, SMN, VN, and LN). The HMM-Multivariate Autoregressive (HMM-MAR) toolbox was employed for variational Bayes inversion in dynamic brain analysis. The HMM ran for 500 cycles with 4 to 16 states, and the optimal number was determined based on the minimum free energy criterion, balancing accuracy and complexity. Temporal measures included fractional occupancy (FO), averaged lifetime (ALT), and switching rate (SR). we compared dynamic features between ketamine and placebo arms using the linear mixed-effect model (LME), adjusting for within-subject variability and treatment order. We defined the contrast to examine changes from baselines in infusion and 24-hour post-infusion (ketamine arm [infusion – baseline] vs. placebo arm [infusion – baseline]) and (ketamine arm [24-hour post-infusion - baseline] - placebo arm [24-hour post-infusion-baseline]), respectively, using "emmeans" in R.

The LME results showed significant differences of FO and ALT occur between two arms in infusion sessions, where the FO and ALT of state #2 are lower during ketamine compared to placebo and higher in state #7 (state #2 ketamine infusion < placebo infusion: FO: estimate = -0.05, p = 0.002; ALT: estimate = -1.62, p-value (p) = 0.005; state #7 ketamine infusion > placebo infusion: FO: estimate = 0.05, p = 0.001; ALT: estimate = 1.37, p = 0.002) (Fig2). The findings on contrasts indicate that in state #2 (DAN+), the FO and ALT of Δ (Infusion – baseline) ketamine were significantly lower compared to Δ placebo (state #2; FO: ketamine < placebo; t-value = -3.19, p = 0.0017; ALT: ketamine < placebo; t-value = -2.81, p = 0.0054). In contrast, in state #7 (SN+), the FO and ALT of Δ ketamine exhibited higher values compared to Δ placebo (state #7; FO: ketamine > placebo; t-value =3.45, p < 0.001; ALT: ketamine > placebo; t-value = 3.09, p = 0.002). In state #3 (FPN+), the ALT of Δ ketamine is lower than Δ placebo (state #3: ALT: ketamine < placebo; t-value = -2.33, p = 0.02). For partial correlations between ASC-5D subscales and the dynamic metrics, we calculated the relative changes (ketamine [infusion – baseline] - placebo [infusion – baseline]). The association between disembodiment and Δ FO of state #7 was significant (t-value = 2.5, p = 0.017) but not with FO or ALT of state #2.

These findings highlight the potential of brain dynamic models to investigate the neuropharmacological impacts of ketamine on brain functional activity dynamics associated with subjective experiences.

Connectivity (eg. functional, effective, structural)

fMRI Connectivity and Network Modeling ¹

Pharmacology and Neurotransmission ²

Pharmacotherapy