Pharmacodynamic interactions between ketamine and psychiatric medications used in the treatment of depression: a systematic review

Since 2000, repeatedly reported rapid antidepressant effects of ketamine have led to increasing off-label use for depression and to regulatory approval of intranasal S-ketamine for treatment-resistant depression (TRD) in conjunction with an oral antidepressant. Many patients considered for ketamine are already taking other psychiatric medications, yet clinical guidance about pharmacodynamic interactions is limited. Ketamine’s principal action is as a non-competitive NMDA (N-methyl-d-aspartate) glutamate receptor antagonist, but it has affinity at multiple other targets (including monoamine transporters and several receptor classes), and its putative antidepressant mechanism involves disinhibition of prefrontal glutamate release, AMPA receptor stimulation and downstream synaptic plasticity changes in regions such as the anterior cingulate, prefrontal cortex and hippocampus. Veraart and colleagues set out to systematically review human studies reporting pharmacodynamic interactions between ketamine and commonly co-prescribed psychiatric drugs relevant to depression treatment. The review focused on mood stabilizers, benzodiazepines, monoamine oxidase inhibitors (MAOIs), antipsychotics and psychostimulants, with the aim of summarising available clinical, neuroimaging and electrophysiological evidence that might indicate potentiation or attenuation of ketamine’s effects and highlight safety signals or gaps in knowledge important for clinical practice and research planning.

The investigators searched MEDLINE and Web of Science on 9 July 2020 for human studies examining pharmacodynamic interactions between ketamine and psychiatric drugs (antidepressants, mood stabilizers, antipsychotics, benzodiazepine agonists or psychostimulants). English and Dutch filters were applied. The search combined drug generic and brand names with the term 'ketamine'; the detailed search strings were reported in supplementary material. Grey literature, including conference abstracts, was eligible. Inclusion criteria prioritised randomized controlled trials (RCTs). Where RCT data were sparse (one or no RCTs), the team included lower levels of evidence such as post hoc analyses, open-label studies and case reports if they provided relevant interaction information. Studies of ketamine given at anaesthetic doses or trials simply adding ketamine to a routine antidepressant (except MAOIs) were excluded, since safety and additive antidepressant effects of ketamine on regular antidepressants had been addressed elsewhere. Two reviewers independently screened titles, abstracts and full texts, resolving disagreements by discussion and involving a third or fourth reviewer when necessary. Data extracted included study design, sample size, population characteristics, intervention details, clinical and mechanistic outcomes and authors’ hypotheses about mechanisms. Quality assessment used Joanna Briggs Institute tools appropriate to RCTs, quasi-experimental designs and case reports. The review adhered to PRISMA reporting guidance. The extracted text notes that details of the selection process and quality assessments are available in supplementary figures and appendices.

The search identified 4,887 records; after deduplication and screening, 24 studies met the eligibility criteria and were included. Included evidence comprised a mix of study types: two RCTs on lithium, seven studies on lamotrigine, several analyses and a placebo-controlled study relating to benzodiazepines, two case reports for tranylcypromine (an MAOI), three double-blind placebo-controlled studies for haloperidol, four articles on risperidone (three using neuroimaging outcomes), three studies on clozapine and one study on olanzapine. No studies were identified that examined ketamine combined with psychostimulants. Mood stabilizers: Two RCTs examining lithium found no significant potentiation of ketamine’s antidepressant effects when lithium was co-administered. Lamotrigine (typically 300 mg administered about 2.5 hours before IV ketamine in the included studies) produced mixed results across seven reports: several studies in healthy subjects and one in patients with major depressive disorder showed attenuation of ketamine-induced effects on dissociation (CADSS), psychotomimetic symptoms (BPRS) and blood-oxygen-level-dependent (BOLD) responses in frontal and thalamic regions, whereas other studies found no interaction on resting perfusion or degree-centrality measures. The authors note heterogeneity of dose, infusion protocols and subject populations as possible explanations for inconsistent findings. Benzodiazepines: Evidence consistently indicated that benzodiazepines can diminish ketamine’s antidepressant benefits. Case reports and post hoc analyses showed that higher benzodiazepine doses were associated with delayed time to response and remission and with shorter duration of antidepressant effect after ketamine. One pooled analysis reported that benzodiazepine medication at 10 mg or more diazepam equivalent after a 0.54 mg/kg ketamine infusion predicted nonresponse within one week (odds ratio = 1.5, p = 0.0445). A healthy volunteer study showed that lorazepam 2 mg reduced some subjective distress and sensory distortions during ketamine infusion but did not fully block psychotomimetic, cognitive, neuroendocrine or physiological ketamine responses. MAOIs: Two case reports describing concurrent use of tranylcypromine (10–80 mg/day) with S-ketamine (IV doses 12.5–75 mg or intranasal doses 28–56 mg) did not report clinically relevant cardiovascular or haemodynamic changes, nor serotonergic syndrome; however, clinical experience remains very limited. Antipsychotics: Findings varied by drug and study population. Haloperidol pretreatment (oral or parenteral doses reported in the included studies) reduced certain acute ketamine effects in healthy subjects — for example, reductions in ketamine-induced impairments in executive function and anxiogenic effects — but a study in nine patients with schizophrenia on high-dose haloperidol found no blunting of ketamine-induced psychotic symptoms and, in that sample, greater increases in psychosis ratings during haloperidol treatment. For risperidone, three pharmacological MRI studies in healthy males reported that 2 mg risperidone attenuated ketamine-induced BOLD/perfusion changes and modulated degree-centrality increases; a separate study found no effect of risperidone on ketamine-induced eye-movement changes. No clinical trial data established whether risperidone attenuates ketamine’s antidepressant efficacy. Clozapine results were inconsistent: low single doses (30–50 mg) produced limited or variable attenuation of S-ketamine responses on neuroimaging or psychotomimetic measures in healthy subjects, whereas a study in patients with schizophrenia receiving higher maintenance clozapine doses (mean ~430 mg/day) showed significant blunting of ketamine-induced positive symptoms. One small study of olanzapine (5–10 mg) found no difference versus placebo in blocking ketamine-induced psychotic symptoms. Safety and other findings: Across studies, sample sizes were small (range 1–72), and several included conference abstracts. Many experiments used ketamine as a model of psychosis in healthy volunteers, limiting direct translation to antidepressant treatment contexts. Heterogeneity in ketamine formulations (racemic versus S-ketamine), doses and administration schedules, and variation in outcome measures, complicated comparisons. The authors judged the overall evidence quality as low and called for further targeted trials.

Veraart and colleagues interpret the assembled evidence as providing preliminary but clinically relevant signals about pharmacodynamic interactions with ketamine. The most consistent finding is that benzodiazepines are likely to shorten ketamine’s antidepressant duration and, at higher doses, predict nonresponse; the authors explain this mechanistically by benzodiazepine agonism at GABA receptors increasing inhibitory interneuron tone, thereby opposing ketamine’s glutamate-driven synaptic potentiation. Animal studies cited in the review support the concept that benzodiazepines (for example diazepam) can block ketamine-induced limbic metabolic and dopamine-release effects. For lamotrigine, several human studies suggest antagonism of acute ketamine effects, including reductions in dissociation, psychotomimetic scores and ketamine-induced BOLD responses, consistent with lamotrigine’s capacity to reduce glutamate transmission via sodium-channel blockade. Nevertheless, other investigations failed to detect an interaction, and the authors highlight differences in ketamine dose, infusion speed, subject population (healthy volunteers versus TRD patients) and metabolic factors as plausible explanations for discrepant results. Regarding antipsychotics, neuroimaging studies indicate that risperidone attenuates ketamine-induced perfusion and connectivity changes, but clinical data on antidepressant outcome are absent. Haloperidol reduced some cognitive and anxiogenic effects of ketamine in healthy volunteers, yet did not blunt ketamine-induced psychosis in a small sample of patients with schizophrenia. Clozapine produced inconsistent results across studies; in patients on therapeutic clozapine doses positive symptoms induced by ketamine were blunted, whereas low single doses had variable or limited effects in healthy subjects. For MAOIs (tranylcypromine) the two case reports included did not show hypertensive crises or serotonergic toxicity when combined with S-ketamine, but the authors caution that clinical experience is insufficient to exclude rare adverse events. The authors acknowledge several key limitations: inclusion of unpublished conference abstracts, small and heterogeneous samples, diverse ketamine regimens and outcome measures, and a predominance of studies using ketamine as a psychosis model in healthy subjects rather than probing antidepressant outcomes in clinical populations. They conclude these factors weaken confidence in causal inferences and generalisability. Consequently, Veraart and colleagues recommend minimising benzodiazepine (and Z-drug) use in patients receiving ketamine for depression, considering potential lamotrigine interactions when patients fail to respond, and cautious co-prescription of low-dose MAOIs with close monitoring. For future research they urge prospective trials that systematically record concomitant medications, explore dose-dependence (for example temporary benzodiazepine withdrawal or lower benzodiazepine doses), and assess both efficacy and safety of relevant combinations in patients with depression.

The systematic review found no evidence of an interaction between lithium and ketamine. Some but not all studies indicate lamotrigine can attenuate ketamine-induced effects. Available evidence suggests benzodiazepines very likely shorten the duration of ketamine’s antidepressant effects, which should inform clinical decision-making. Limited case-report evidence did not reveal acute haemodynamic or serotonergic crises when S-ketamine was combined with the MAOI tranylcypromine, but data are sparse. Overall, the evidence base is small, heterogeneous and of low quality, and the authors call for further rigorous studies to clarify pharmacodynamic interactions and their clinical implications.