This double-blind, placebo-controlled, cross-over, within-subjects study (n=22) investigated the effects of ketamine (30mg/70kg) on whole-brain functional connectivity in healthy male participants while attenuating pre-synaptic glutamate release directly via pretreatments with the sodium-channel modulator lamotrigine (300 mg), and indirectly via pretreatment with the 5-HT2A receptor antagonist risperidone (2mg). Ketamine induced robust changes in the functional connectivity pattern and produced a shift from a cortically-centered to a sub-cortically-centered brain state. Pre-treatment with risperidone, but not lamotrigine, resulted in a strong modulation of the ketamine-induced hub changes, which suggests that these changes are likely a result of NMDA blockade and possible serotonergic modulation rather than purely modulation of glutamate release.
Abstract
“Introduction: Ketamine, an N-methyl-D-aspartate receptor (NMDAR) antagonist, has been studied in relation to the glutamate hypothesis of schizophrenia and increases dissociation, positive and negative symptom ratings. Ketamine effects brain function through changes in brain activity; these activity patterns can be modulated by pre-treatment of compounds known to attenuate the effects of ketamine on glutamate release. Ketamine also has marked effects on brain connectivity; we predicted that these changes would also be modulated by compounds known to attenuate glutamate release.
Methods: Here, we perform task-free pharmacological magnetic resonance imaging (phMRI) to investigate the functional connectivity effects of ketamine in the brain and the potential modulation of these effects by pre-treatment of the compounds lamotrigine and risperidone, compounds hypothesised to differentially modulate glutamate release. Connectivity patterns were assessed by combining windowing, graph theory and multivariate Gaussian process classification.
Results: We demonstrate that ketamine has a robust effect on the functional connectivity of the human brain compared to saline (87.5 % accuracy). Ketamine produced a shift from a cortically centred, to a subcortically centred pattern of connections. This effect is strongly modulated by pre-treatment with risperidone (81.25 %) but not lamotrigine (43.75 %). Based on the differential effect of these compounds on ketamine response, we suggest the observed connectivity effects are primarily due to NMDAR blockade rather than downstream glutamatergic effects.
Discussion: The connectivity changes contrast with amplitude of response for which no differential effect between pre-treatments was detected, highlighting the necessity of these techniques in forming an informed view of the mechanistic effects of pharmacological compounds in the human brain.”
Authors: R. Joules, O. M. Doyle, A. J. Schwarz, O. G. O’daly, M. Brammer, S. C. Williams & M. A. Mehta
Summary
Ketamine, an N-methyl-D-aspartate receptor (NMDAR) antagonist, has been shown to increase dissociation, positive and negative symptom ratings, and change brain activity patterns. It also has marked effects on brain connectivity, which can be modulated by pre-treatment of compounds known to attenuate glutamate release.
Based on the differential effect of these compounds on ketamine response, we suggest the observed connectivity effects are primarily due to NMDAR blockade.
Introduction
Ketamine, an NMDAR antagonist, causes glutamatergic dysfunction in healthy humans and can be used to study the role of glutamatergic dysfunction in relation to schizophrenia.
Ketamine has been shown to evoke robust and reliable effects on the blood oxygen level-dependent (BOLD) signal in healthy volunteers. Pre-treatment with compounds expected to reduce glutamate release has been demonstrated to attenuate the ketamine-induced BOLD response.
Acute NMDAR blockade alters the functional connectivity of several neural systems in rodents and humans. This article discusses how ketamine affects whole-brain connectivity and how this is altered through different modulatory mechanisms using risperidone and lamotrigine as pharmacological probes.
Pattern recognition techniques have been used to model disease states, cognitive states, and age groups. The current study uses whole-brain measures of functional connectivity with PR techniques to identify spatial patterns of whole-brain connectivity underlying the effects of acute ketamine.
Participants
Twenty right handed male volunteers were recruited for this double blind, placebo controlled, cross-over design study, of which 16 completed all sessions. Four participants withdrew due to illness, and two participants were withdrawn due to violating the lifestyle guidelines.
Experimental design
Data were collected over four sessions, each separated by a period of at least 10 days. Participants received a single oral dose of either lamotrigine, risperidone or placebo, followed by an intravenous infusion of either saline or ketamine.
Compound infusion
Ketamine was administered intravenously to achieve a target plasma level of 75 ngml 1, adjusted for the participant’s weight and height, and a pseudo-continuous infusion of approximately 0.31 mgkg1 h1 was administered.
MRI acquisition
MR images were acquired using a 3T GE HDx scanner. A high-resolution gradient-echo scan was performed 5 min into the 15 min scan to obtain 43 near-axial slices.
Experimental conditions
The 450 volumes of EPI time series acquired at each session were divided into pre- and post-infusion conditions, using the naming scheme described in Table 1.
Pre-processing
The imaging data were pre-processed using SPM5, and the structural images were segmented into grey matter, white matter and cerebrospinal fluid tissue types.
Linear regression was used to remove nuisance signal parameters from the imaging time-series data, and a band pass filter was applied to reduce high-frequency physiological signals and low-frequency noise such as scanner drift.
Parcellation
The pre-processed data were parcellated into regions to facilitate network analysis. The entire brain was represented as a graph, and node time series were calculated as the mean of voxel time courses in each atlas region.
Task-free spatiotemporal dynamics
Resting state networks (RSNs) display complex spatiotemporal dynamics, which can be accounted for by applying temporal windowing. A tapered window was preferred to a box car in order to minimise overlap between windows and reduce the contribution of any outliers not central to the window.
Network formulation
We calculated edges between regions across the whole brain using linear correlation and generated adjacency matrices for each window position.
Centrality measure
A graph theory centrality measure, weighted degree centrality, was used to quantify the importance of each node within a brain network and to determine how the mean connectivity of a region with the rest of the brain was affected by glutamatergic dysfunction.
Pattern recognition
To compare regional network connectedness between session conditions, Gaussian process classification was employed. This technique has been demonstrated to provide robust performance in several neuroimaging studies.
In order to assess the performance of the classifier, leave-one-out cross validation (LOOCV) was used. The classification accuracy was calculated as the mean accuracy value obtained across all LOOCV folds.
A distribution of permuted accuracies was generated and the number of times the permuted accuracy was greater than the true accuracy was counted to provide an estimate of the p value.
Classification was performed using two configurations: (1) between PLA+SAL and PLA+KET sessions, and (2) between lamotrigine and risperidone pre-treated sessions.
Ordinal regression using Gaussian processes (ORGP) was applied to investigate the relationship between experimental conditions, identifying an ordinal trend across classes.
Classification maps
We visualised the distribution of the classes relative to one another using MNI space and spheres representing the contribution of the node in driving the discriminatory performance. We confirmed the direction of change in node strength between classification groups using paired t tests.
Ketamine robustly alters the whole-brain functional connectivity pattern
Ketamine had a significant effect on the pattern of degree centrality (DC), but the classifier was unable to distinguish between pre-infusion placebo (PLA+KET) and saline conditions, or between pre-infusion placebo conditions and saline conditions.
Figure 3c illustrates the pattern of weights driving the classification between ketamine and placebo conditions. The pattern of weights favouring ketamine is strongly represented in sub-cortical and cerebellar regions, whereas the pattern of weights favouring placebo is predominantly cortical.
Pre-treatment with risperidone alters the ketamine-induced functional connectivity pattern
Risperidone modulates the ketamine response by increasing DC in the frontal and temporal cortices and decreasing DC in the basal ganglia, occipital and parietal regions, when compared to placebo pre-treated ketamine.
We observed an interaction effect between risperidone and the ketamine response, as measured by DC, and the classifier was unable to significantly separate the conditions for the risperidone pre-treated session.
The GPC model was trained using saline and ketamine classes and reported chance level accuracies for both conditions.
Pre-treatment with lamotrigine does not alter the ketamine-induced functional connectivity pattern
Lamotrigine pre-treatment did not have an effect on ketamine-induced changes in functional connectivity, and similar trends were observed when comparing the pre-infusion and post-infusion ketamine states for the placebo sessions and lamotrigine pre-treatment.
We were unable to distinguish between lamotrigine and risperidone sessions, but could distinguish between post-ketamine conditions following a pre-treatment of lamotrigine and risperidone.
Ordinal regression
Ordinal regression was performed to investigate the effects of lamotrigine and risperidone on the ketamine-induced change in connectivity. No ordinal trend was observed between saline and ketamine for (ris+ketamine) and (lam+ketamine) conditions.
Discussion
Ketamine induces a shift from a cortically centred to a sub-cortically centred brain state. Risperidone and lamotrigine modulate the ketamine-induced DC changes, but not the BOLD amplitude changes.
Ketamine infusion
During a task-free acute ketamine challenge, a distributed pattern of functional connectivity changes were observed across the brain. The results reveal decreasing cortical centrality and an increase in centrality of the cerebellum and basal ganglia.
Ketamine infusion was shown to increase global connectivity in healthy volunteers, with the thalamus, parietal regions and cerebellum exhibiting the greatest increase. Ketamine was also shown to affect the connectivity of cerebellar, visual, auditory, somatosensory and subcortical regions in relation to pre-defined networks of interest.
The multivariate results presented here provide an alternative perspective on the effect of acute ketamine on whole-brain connectivity. Ketamine causes both increases and decreases in the centrality pattern of the brain, and this shifts the pattern of connectivity from a cortically centred state to a sub-cortically centred state.
The observed pattern of disconnectivity of the cortical hubs is consistent with the experiential effects observed with ketamine, which includes a reduction in contextual processing, impaired memory, spatial representation and sensorimotor processing.
Ketamine effects connectivity changes through disruption of NMDA receptor-mediated transmission, namely the known cortical microcircuit effects where parvalbumin-positive GABAergic interneurons disinhibited by ketamine, leading to decreased connectivity of cortical nodes and the observed pattern of centrality.
Ketamine altered the pattern of centrality in the brain during a task-free infusion, suggesting a complex network reorganisation underpinning the effect of ketamine. The study did not observe any strong and consistent subjective effects, however.
Lamotrigine treatment
Lamotrigine attenuates the ketamine BOLD effect by reducing glutamate release. Ketamine’s effect on functional connectivity may reflect glutamatergic modulation. Lamotrigine does not attenuate the ketamine-induced pattern of DC in the brain, as shown by the lack of ordinal progression between saline, lamotrigine+ketamine and ketamine conditions, and the strong confusion between lamotrigine+ketamine and ketamine conditions.
These results support the conclusion that lamotrigine pre-treatment has a similar effect to placebo on ketamine-induced DC, and that ketamine-induced DC is caused by NMDA receptor antagonism.
Risperidone treatment
Risperidone (2 mg) attenuates the effect of ketamine by reducing glutamate levels downstream through 5-HT2A antagonism, and potentiating NMDAR function directly. The results indicate that risperidone may have an opposing effect on the ketamine response.
We theorise that risperidone acts on the NMDA receptor in opposition to ketamine, resulting in a pattern of DC effects dissimilar to that of ketamine. This may be due to serotonergic effects, but further work is required using a compound with selective 5-HT2A binding.
Stability of pattern recognition method in the absence of pharmacological intervention
Our methodology did not detect differences in connectivity pattern when no pharmacological stimulus was applied. This provides confidence in our interpretation of the effects of ketamine and its modulation by risperidone and lamotrigine.
Limitations
The results presented here provide an informative insight into the mechanisms of effect of ketamine on the human brain, but the use of centrality is insensitive to subtle changes in regional coupling which may be induced by the administered compounds.
A single connectivity metric and temporal scale is used, which is naive to phase-delayed and non-linear relationships between regions, which may exist in a complex biological system such as the brain.
The use of degree centrality provides an initial investigation into the connectivity effects of ketamine in the brain, but further investigation is required to identify an optimal set of orthonormal connectivity measures.
Conclusions
We have demonstrated that ketamine administration results in reduced connectivity of cortical nodes and increased DC of nodes in the basal ganglia and cerebellum. Furthermore, pre-treatment with risperidone but not lamotrigine results in a strong modulation of the ketamine-induced hub changes.
Study details
Topics studied
Neuroscience
Study characteristics
Placebo-Controlled
Double-Blind
Within-Subject
Participants
22