Stress induced brain changes are increasingly implicated in the development of mental disorders (Averill 2020). Research indicates that the stress response alters neuronal function in brain structures including the hippocampus (Cattaneo and Riva 2016), a key structure within the limbic system that subserves learning, memory and emotional regulation (McEwen and Morrison 2013). Furthermore, the hippocampus is highly vulnerable to stress and strongly associated with the pathogenesis of depression and suicidality (Kempton et al., 2011; Colle et al., 2015). Chronic stress related hippocampal volume reductions, mediated by glucocorticoids and excitatory amino acids such as glutamate, may be an important element in the development of stress-related mental disorders (Egeland et al., 2015; Sheline et al., 2019). Though the mechanisms are complex, in addition to cellular atrophy and reduced neurotrophic factor support (Du Preez et al., 2021), reduced hippocampal volume may be further sustained by altered glucocorticoid sensitivity (Anacker et al., 2013b; McEwen and Morrison 2013; Kott et al., 2016) and stress related glial cell functional abnormalities (Ardalan et al., 2017; Kreisel et al., 2019; Du Preez et al., 2021).

While prolonged stress has been linked to cellular atrophy, recent evidence points to reduced neurogenesis, particularly cellular proliferation, as the primary contributor to hippocampal volume reductions (Snyder et al., 2011; Anacker 2014). Several studies show both acute and chronic stress to be implicated in impaired neurogenesis (Schoenfeld and Cameron 2015; Planchez et al., 2020; Du Preez et al., 2021), culminating in synaptic loss and dysconnectivity (Abdallah 2020; Averill et al. 2020a, 2020b). Additionally, aberrant glutamatergic changes have been shown to induce excitotoxicity, neuronal atrophy and synaptic loss, particularly in the prefrontal cortex, hippocampus and limbic circuitries (Abdallah et al., 2017a). Moreover, research also shows the effects of stress to be cumulative, involving hippocampal dendritic atrophy, postsynaptic dendritic spine loss, reduced size and complexity of dendritic arborisation, as well as effects on synaptic signalling and microcircuit functionality (Anacker et al., 2013a; Kim et al., 2016; Sheline et al., 2019).

Preclinical studies demonstrate that chronic stress induces damage to hippocampal CA3 pyramidal neurons and suppresses long term potentiation within the CA1 region (Stein-Behrens et al., 1994; Joëls et al., 2004; Phillips et al., 2006) and is further associated with reduced neurogenesis in the dentate gyrus - which is enhanced by adrenalectomy (Gould and Tanapat 1999). Roddy et al., (2019), reported in depressed patients, that the CA2, CA3 and CA4 subfields of the hippocampus are most sensitive to this effect, with volumetric reductions observed as early as first episode depression. While volume changes in the dentate gyrus, CA1 and subiculum were reported to emerge after illness progression. Additionally, the subfield methodology utilised by Roddy et al., (2019), showed that the degree of volume loss observed depended on the defined boundary of the hippocampus, with more restrictive definitions showing greater loss. Thus, subfield measures may have higher sensitivity to detect subtle anatomical changes than whole hippocampal volumes, with left CA1 emerging as a potential biomarker of severe depressive illness (Roddy et al., 2019). Also, though hippocampal volume change has been widely reported in depression, it is still uncertain whether the volume change is a cause or an effect of the condition (Belleau et al., 2019). Furthermore, findings have been inconsistent in studies of those with suicidality, with volumetric reduction (Colle et al., 2015), no change (Gifuni et al., 2016), or enlargements all reported, which may be explained by severity of illness, comorbidities, concomitant medication use, as well as methodological differences (Zhang et al., 2021).

Neuroimaging has been commonly used in psychiatry research and is particularly useful to elucidate structural changes (Kempton et al., 2011). Meta-analyses of studies examining magnetic resonance imaging (MRI) of the hippocampus have shown volume loss of between 4 and 10% in depression (Videbech and Ravnkilde 2004; Cole et al., 2011), whilst voxel-based morphometry has also shown decreased grey matter hippocampal volume in those with depression (Zhao et al., 2014), and mechanistic studies have linked network dysconnectivity to perturbed glutamatergic neurotransmission (Abdallah et al., 2017a). Such findings provide strong evidence for structural and functional brain alterations in individuals experiencing prolonged stress, and suicidality, and these changes may constitute the neurobiological basis for stress-related pathological conditions (Averill et al., 2020a). Furthermore, the characterisation of such abnormalities provide opportunities for new therapeutic treatment targets that may counteract stress-induced neurobiological changes (McEwen 2017).

Ketamine, a rapid acting glutamatergic agent, well established in anaesthesia and pain medicine, is emerging as a robust option for treating stress related disorders (Averill et al., 2020a; Alnefeesi et al., 2022). Though its mode of action is still being investigated, the predominant hypothesis is that subanaesthetic doses of ketamine preferentially inhibit N-methyl-D-aspartate receptors (NMDAR) on interneurons, whilst disinhibiting pyramidal neuron glutamate release and provoking a transient surge in post-synaptic alpha-amino-3-hydroxy5-methyl-4-isoxazolepropionic acid (AMPA) receptor and NMDAR activity (Abdallah et al., 2018). This rapidly facilitates synaptic plasticity and modulates neuronal function within prefrontal and hippocampal circuits, through enhanced brain derived neurotrophic factor (BDNF) release, with resulting increased protein synthesis and synaptic strength (Averill et al., 2019). However, these effects have been characterised preclinically (Li et al., 2011; Ardalan et al., 2017) and human studies are lacking.

Several studies of adults with major depressive disorder (MDD) have demonstrated structural brain changes in response to ketamine treatment. First, (Zhou et al., 2020b) showed that six ketamine infusions increased right side hippocampal volume, whilst (Abdallah et al., 2017b) reported left hippocampal volume increases after ketamine infusion, in treatment responders with MDD. Furthermore, Zhou et al., (2020a) reported subcortical structural changes following ketamine treatment and noted baseline enlargement of the right thalamus and left subiculum head may be predictive of treatment response in those with MDD.

We have previously shown oral ketamine to be a promising, feasible and well tolerated option in the treatment of chronic suicidality (Can et al., 2021; Dutton et al., 2022). Furthermore, we have reported supportive neurobiological evidence showing that oral ketamine appears to increase striato-limbic GMVs (Gallay et al., 2021) and ameliorates functional connectivity in aberrant brain networks (Can et al., 2023a). Thus, in the current study we sought to determine whether oral ketamine treatment specifically affects hippocampal (whole and subfield) volumes in patients with chronic suicidality and MDD. We hypothesised that oral ketamine treatment would differentially affect hippocampal volumes in trial participants categorised as responders, versus those who were non-responders (defined according to changes in suicidality symptoms at post-ketamine [week 6] and follow-up [week 10] timepoints). This builds on existing evidence by focusing on specific hippocampal subfields previously shown to be of particular relevance in chronic stress pathology, that is, the cornu ammonis (CA) 1–4 (Roddy et al., 2019), and granular cell molecular layer of dentate gyrus (GC ML DG), in addition to ‘whole hippocampal’ volumes (Schoenfeld et al., 2017; Belleau et al., 2019).