Blood-based biomarkers of antidepressant response to ketamine and esketamine: A systematic review and meta-analysis

This meta-analysis (n=2,801) explored the association between baseline levels and longitudinal changes in blood-based biomarkers and response to ketamine/esketamine. Over 460 individual biomarkers were examined and there were no consistent associations between baseline levels of blood-based biomarkers and response to ketamine. However, a longitudinal analysis revealed ketamine responders had statistically significant increases in brain-derived neurotrophic factor (BDNF) when compared to pre-treatment.

Abstract

“(R,S)-ketamine (ketamine) and its enantiomer (S)-ketamine (esketamine) can produce rapid and substantial antidepressant effects. However, individual response to ketamine/esketamine is variable, and there are no well-accepted methods to differentiate persons who are more likely to benefit. Numerous potential peripheral biomarkers have been reported, but their current utility is unclear. We conducted a systematic review/meta-analysis examining the association between baseline levels and longitudinal changes in blood-based biomarkers, and response to ketamine/esketamine. Of the 5611 citations identified, 56 manuscripts were included (N = 2801 participants), and 26 were compatible with meta-analytical calculations. Random-effect models were used, and effect sizes were reported as standardized mean differences (SMD). Our assessments revealed that more than 460 individual biomarkers were examined. Frequently studied groups included neurotrophic factors (n = 15), levels of ketamine and ketamine metabolites (n = 13), and inflammatory markers (n = 12). There were no consistent associations between baseline levels of blood-based biomarkers, and response to ketamine. However, in a longitudinal analysis, ketamine responders had statistically significant increases in brain-derived neurotrophic factor (BDNF) when compared to pre-treatment levels (SMD [95% CI] = 0.26 [0.03, 0.48], p = 0.02), whereas non-responders showed no significant changes in BDNF levels (SMD [95% CI] = 0.05 [−0.19, 0.28], p = 0.70). There was no consistent evidence to support any additional longitudinal biomarkers. Findings were inconclusive for esketamine due to the small number of studies (n = 2). Despite a diverse and substantial literature, there is limited evidence that blood-based biomarkers are associated with response to ketamine and no current evidence of clinical utility.”

Authors: Gustavo C. Medeiros, Todd D. Gould, William L. Prueitt, Julie Nanavati, Michael F. Grunebaum, Nuri B. Farber, Balwinder Singh, Sudhakar Selvaraj, Rodrigo Machado-Vieira, Eric D. Achtyes, Sagar V. Parikh, Mark A. Frye, Carlos A. Zarate & Fernando S. Goes

Summary

INTRODUCTION

Major depressive disorder and bipolar depression are common, disabling, and often difficult to treat. Conventional antidepressant treatment may take weeks to provide significant symptomatic relief, increasing the risk of negative consequences such as suicide.

Ketamine and esketamine are NMDA receptor antagonists that have been shown to have rapid and substantial antidepressant effects in major depressive episodes in MDD or bipolar disorder, including in subjects previously unresponsive to conventional treatments. However, ketamine and esketamine can be associated with potentially problematic side effects.

Precision medicine aims to identify which treatments are most beneficial for a specific individual. However, there is no well replicated means to identify which depressed individuals are more likely to benefit from specific therapeutic interventions.

Biomarkers are characteristics that can be measured to predict response to treatment. They can be baseline markers or dynamic markers that reflect biological changes potentially related to therapeutic mechanisms of action.

Several biomarker modalities have been studied in psychiatric disorders, including depression and its treatment. Blood-based predictors are more easily obtained, more widely accepted by patients, and more likely to be more cost-effective than brain-based predictors.

Search strategy

The search strategy was run in the electronic databases MEDLINE, Embase, The Cochrane Library, PsycINFO, and Web of Science. Results were limited to human studies in PubMed and Embase.

The electronic searches were complemented by manual searches. Personal communication and Google Scholar were used to evaluate the references.

Inclusion and exclusion criteria

This systematic review and meta-analysis included manuscripts that examined adult human subjects with MDD or BD, and assessed the association between levels of blood-based biomarkers and improvement of depressive symptoms after ketamine/esketamine treatment.

We excluded articles that investigated individuals with current psychotic symptoms, depressive symptoms due to other disorders besides MDD and BD, allowed concurrent neuromodulatory treatments, or assessed individuals in surgical/perioperative settings.

Screening and selection

Two investigators independently screened the titles and abstracts, then the full-text manuscripts that seemed eligible for inclusion. Discrepancies between the two reviewers were solved by consensus discussion, and the Covidence software was used to facilitate the screening process.

Risk of bias assessment

The risk of bias of included studies was determined by the Quality in Prognostic Studies (QUIPS) tool. Two reviewers independently assessed the risk of bias, and disagreements were solved by consensus discussion.

Outcomes

In order to address our research questions, we reported associations between baseline levels of blood-based biomarkers and improvement of depressive symptoms in individuals who were treated with ketamine or esketamine.

We reported on the association between blood-based biomarkers and improvement in depression scores, and compared the results using response status as the primary outcome.

Statistical analyses

Meta-analytical calculations were conducted if at least three comparable studies on the same blood-based biomarker were found. Heterogeneity between studies was examined using I2, with values between 0 and 40% considered trivial, 30 to 60% considered moderate, 50 to 90% considered substantial, and 75 to 100% considered considerable.

Publication bias was visually assessed by funnel plots, and objectively measured by Egger’s tests. Meta-regressions were performed to identify potential effect modifiers, and sensitivity analyses were conducted by serially excluding individual studies.

RESULTS

The initial search identified 5611 citations, of which 103 were potentially eligible for inclusion. 56 manuscripts met criteria for inclusion in the qualitative evaluation. Analyses of 56 included studies revealed that the research base on blood-based biomarkers of response to ketamine/esketamine is progressively growing, with the number of publications per year increasing over time. The most frequently studied groups of biomarkers were neurotrophic factors, ketamine related metabolites, inflammatory markers, metabolites of the tryptophan-kynurenine pathway, genetic markers, and amino acids.

Forty-three studies examined a single infusion of ketamine, and the response rate was 46% at 1 day post-infusion. There were no statistically significant differences in response between individuals with MDD and bipolar depression.

The risk of bias was moderate in 21 studies (37%), and high in 35 studies (63%), due to factors such as insufficient assessment and control of important confounders, insufficient description of study participants, and variable measurement of prognostic factors.

ASSOCIATION BETWEEN BLOOD-BASED BIOMARKERS AND ANTIDEPRESSANT RESPONSE TO KETAMINE Neurotrophic factors

Ketamine increases synaptic plasticity and production of neurotrophic factors, which may be reflected in changes in blood. Fifteen studies reported on the relationship between baseline levels of neurotrophic factors and antidepres- sant response to ketamine. The meta-analyses showed no statistically significant associations between baseline levels of neurotrophic factors and response status.

However, quantitative analyses revealed that responders had increased post-treatment levels of BDNF but not non-responders. There was no evidence for between-study heterogeneity or publication bias.

Further analyses revealed that the magnitude of the increase in BDNF levels was greater in studies with multiple infusions than in studies with a single infusion, but the difference was not statistically significant.

Levels of ketamine and ketamine metabolites

Ketamine is rapidly metabolized by liver p450 enzymes to a number of metabolites, including norketamine, dehydronorketamine, and one of the 12 hydroxynorketamines. Thirteen studies investigated the association between ketamine and antidepressant response, but two metabolites showed similar evidence of association.

Inflammatory markers

Ketamine has anti-inflammatory effects, which motivated several studies to examine the association between pro-inflammatory markers and antidepressant response. However, no consistent pattern in the findings was found among the eleven manuscripts that reported on the association between baseline levels of inflammatory markers and response to ketamine.

Meta-analysis included six studies (N = 260). Responders had non-significantly lower levels of pro-inflammatory factors at baseline compared to non-responders, but there were no longitudinal changes in inflammatory markers.

Tryptophan-kynurenine pathway

Ketamine may be protective against potentially neurotoxic effects of kynurenine and its metabolites, possibly via its action as an NMDA glutamate receptor antagonist. However, only one study found a statistically significant association between baseline levels of metabolites of the tryptophan-kynurenine pathway and response to ketamine.

Genetic markers

Seven studies have tested the association between common genetic markers and response to ketamine. Two studies examined the BDNF single nucleotide polymorphism (SNP), one studied the -opioid receptor 1 SNP, one studied several polymorphisms in the CYP450 enzymes, and three studies used a genome-wide approach.

Grunebaum and colleagues (2020) failed to find an association between response to ketamine and a common loss-of-function SNP in the OPRM1 gene. There were also no significant findings in genome-wide association studies.

Amino acids and derivates

Amino acids can have varied functional effects on the CNS. Five studies examined the association between levels of amino acids and derivates, and response to ketamine.

ASSOCIATION BETWEEN BLOOD-BASED BIOMARKERS AND ANTIDEPRESSANT RESPONSE TO ESKETAMINE

Two studies have tested the association between blood-based biomarkers and antidepressant response to esketamine. Rotroff and colleagues (2016) found no statistically significant correlations with response to esketamine, and Li and collaborators (2020) found an association between response to esketamine and the BDNF Val66Met polymorphism.

DISCUSSION

We conducted a comprehensive qualitative and quantitative synthesis of blood-based biomarkers of response to ketamine and esketamine. We found no consistent associations between baseline levels of blood-based biomarkers and response to ketamine.

Ketamine may promote synaptogenesis by activating post-synaptic AMPA receptors, which in turn triggers intracellular pathways ultimately resulting in increased BDNF release. This increase may reverse impaired neuroplasticity and atrophic changes associated with MDD.

We conducted the largest analysis to date of nine clinical trials, and did not observe a statistically significant association between blood levels of ketamine or norketamine and antidepressant response. However, future studies should examine the minimum effective concentration or therapeutic range required to exert significant antidepressant effects.

This systematic review and meta-analysis should be interpreted in the context of several limitations, including small sample sizes, lack of account for potential confounding factors, and heterogeneity in study design, sample characteristics, treatment schedule, timepoints assessing antidepressant response and laboratory procedures.

CONCLUSIONS

Ketamine’s rapid antidepressant effect has been described as one of the most impactful recent discoveries in biological psychiatry. However, there is currently limited evidence that blood-based biomarkers may serve as predictors of ketamine treatment efficacy.

AUTHOR CONTRIBUTIONS

GCM, TDG, WLP, MFG, NBF, BS, SS, RMV, EDA, SVP, MAF, and CAZJ conceptualized the study, designed the study, selected the manuscripts, extracted the data, conducted the statistical analysis, conducted the analysis of bias, drafted the manuscript, revised the manuscript, and edited the manuscript for critical intellectual content.

FUNDING

This systematic review/meta-analysis was supported by the Johns Hopkins School of Medicine.

COMPETING INTERESTS

TDG is listed as co-author on patents related to the use of (2 R,6 R)-hydroxynorketamine in the treatment of depression, anxiety, anhedonia, suicidal ideation, and post-traumatic stress disorder. EDA, BS, SS, RMV, MAF and SVP report having received grants from NIMH, SAMHSA, Eurofarma Pharmaceuticals, Abbott, BioStrategies group, Janssen, Lundbeck/Otsuka, Roche, Sunovion and Teva, and have received speaking fees from British Medical Journal Publishing Group, Janssen and LivaNova. CAZ is a full-time U.S. government employee and is listed as a coinventor on several patents for the use of ketamine in the treatment of depression and neuropathic pain. FSG has received research grant support from Janssen Therapeutics.