Brain-derived neurotrophic factor serum levels following ketamine and esketamine intervention for treatment-resistant depression: secondary analysis from a randomized trial

Abstract Objectives Evidence suggests that ketamine’s influence on brain-derived neurotrophic factor (BDNF) might be involved in its mechanism of rapid antidepressant action. We aimed to evaluate the differential impact of ketamine and esketamine on serum BDNF levels and its association with response patterns in treatment-resistant depression (TRD). Methods Participants (n = 53) are from a randomized, double-blind clinical trial comparing the efficacy of single-dose ketamine (0.5mg/kg, n = 27) and esketamine (0.25mg/kg, n = 26) in TRD. Depression severity was assessed before and 24 hours, 72 hours, and 7 days after the intervention, using the Montgomery-Åsberg Depression Rating Scale (MADRS). Blood samples were collected before infusion, 24 hours, and 7 days afterwards. Results There were no significant changes in BDNF levels at post-infusion evaluation points, and no difference in BDNF levels comparing ketamine and esketamine. Both drugs exhibited similar therapeutic effect. There was no association between BDNF levels and response to treatment or severity of depressive symptoms. Conclusion There was no significant treatment impact on BDNF serum levels – neither with ketamine nor esketamine – despite therapeutic response. These results suggest that ketamine or esketamine intervention for TRD has no impact on BDNF levels measured at 24 hours and 7 days after the infusion. This clinical trial is registered on the Japan Primary Registries Network: UMIN000032355.


Introduction
Major depressive disorder (MDD) is a major social and economic burden, being the leading cause of incapacity worldwide. 1 Traditional antidepressants require several weeks to exert therapeutic effect, and only one third of patients with MDD achieve remission after multiple treatments. 2 The N-methyl-D-aspartate (NMDA) receptor antagonist, ketamine, has emerged as an alternative with rapid antidepressant action for patients who did not respond to monoaminergic antidepressants. [3][4][5] Ketamine is a racemic mixture of two enantiomers: S(+)-ketamine (or esketamine) and R(-)-ketamine (or arketamine), with the S(+) stereoisomer presenting a greater affinity for the NMDA receptor than R(-)-ketamine. 6 Intranasal esketamine was the first FDA-approved rapid antidepressant for treatment-resistant depression (TRD), 7 and intravenous racemic ketamine was considered to have Level 1 evidence for efficacy in the latest update of the Canadian Network for Mood and Anxiety Treatment (CANMAT) guidelines for MDD. 8 Evidence links the antidepressant action of ketamine to increases in brain-derived neurotrophic factor (BDNF) levels. [9][10][11] BDNF is a growth factor from the family of neurotrophins, which plays a key role in development, differentiation, and survival of neurons, as well as in neuroplasticity. 12,13 There is mounting evidence indicating involvement of BDNF in stress response and in the pathophysiology of mood disorders. 14,15 In depressive patients, BDNF levels are reduced in some brain regions and in peripheral blood [16][17][18][19] and seem to correlate with improvement of depression. 20 Animal studies even suggest that BDNF is necessary for the efficacy of antidepressants, including ketamine. 21 Having a valid biomarker to guide treatment of mood disorders has always been longed-for in Psychiatry, with BDNF as one of the most promising candidates. Assessing peripheral BDNF in the context of treatment with different forms of ketamine can not only contribute to the elucidation of its role as a biomarker, but also to clarification of the mechanism of action of ketamine, since the racemate and esketamine may have different pharmacological effects over BDNF.
The main objective of the present study was to explore the differential impact of a single dose of ketamine or esketamine on BDNF serum levels in TRD subjects. We also aimed to evaluate the association of BDNF levels with response patterns in TRD, including the utility of baseline BDNF for predicting response with ketamine and esketamine treatment.

Study design
This study was conducted as a secondary analysis of data from a randomized, double-blind clinical trial designed to compare the efficacy of ketamine and esketamine for TRD, registered on the Japan Primary
Both drugs were diluted in saline and administered intravenously for 40 minutes. Participants were randomized into esketamine and ketamine groups at a 1:1 ratio, using electronic randomization software.

Outcomes
Depression severity was assessed with the

Participant flow
Participants were recruited between March 2017 and June 2018. A total of 63 patients were enrolled on the present trial; a participants' flowchart can be seen in Figure 1. Baseline characteristics were well balanced between groups and can be consulted in Table 1.

Effect of treatment on BDNF levels
Mean values of BDNF for the intervention groups at each timepoint can be consulted in Table 2 Figure 2B).

Treatment response and BDNF levels
With the GLM repeated measures of BDNF over time, but with both intervention group and 24h response status as between-subjects factors, we did not find a

Discussion
We found no significant alteration in BDNF levels 24 hours and 7 days after the infusion of ketamine or esketamine, compared to baseline. We also did not find any association between BDNF levels and response to treatment. As far as we know, this is the first study directly comparing the effects of ketamine and esketamine on BDNF levels in TRD patients. Liu et al. 25 31 Liu et al. 25 found elevated BDNF levels from three days to one month after ketamine and esketamine administration in depressed subjects, but, as stated before, the infusions occurred before surgery, and the effects of the procedure cannot be discarded. Therefore, the optimal period for assessment of peripheral BDNF is still unclear.
In our sample, BDNF levels did not correlate with depression severity. Similar findings were obtained in a study with 962 depressed patients, 19 but they contrasted with other studies that found a negative correlation. 10,18,32 In our study, lower levels of BDNF at baseline were not associated with a lack of therapeutic response, a finding that is in line with one previous study. 33 We chose to measure serum BDNF instead of plasma levels because it is more stable, reproductible, 34 and considered more reliable. 35 The BDNF concentration in serum is about 100 times higher than in plasma, which is probably due to release of BDNF by platelets during the coagulation process. 36,37 Therefore, measuring peripheral BDNF in serum, compared to plasma, might influence the assessment sensibility, since the significant platelet pool of BDNF -released after platelet activation -could mask possible discrete acute variations in this neurotrophin.
There is evidence that alterations in BDNF serum levels in humans and rats primarily reflect processes occurring in megakaryocytes and platelets, not necessarily reflecting neuronal levels. 38 Furthermore, a recent study in animal models has called into question the correlation between central and peripheral BDNF levels and their use as a biomarker for ketamine's effects. 39 Regarding a possible differential effect of the enantiomers of ketamine, previous studies in murine models highlighted R(-)-ketamine as having a longerlasting antidepressant action when compared to S(+)ketamine, 40 suggesting an impact on BDNF levels and BDNF-TrkB signaling as a mediator of its longer-lasting effect. 41 In the present study, however, there was no differential impact of racemic ketamine or esketamine on BDNF levels at the timepoints evaluated.
Our study has some limitations: not being specifically designed to evaluate BDNF levels; the absence of a control group receiving placebo; and non-standardized concomitant medication use, which, on the other hand, may better reflect the naturalistic scenario of clinical practice.

Conclusion
In conclusion, serum measurements of BDNF levels 24 hours and 7 days after treatment were not different between ketamine and esketamine groups, and BDNF levels were not associated with treatment response to either drug. Possibly, this negative result, rather than discrediting the role of BDNF in the antidepressant mechanism of action of ketamine, suggests that these post-infusion times may not be the most appropriate for peripheral BDNF measurement. Further studies with larger samples will be needed to clarify this possible differential effect of ketamine forms. Additionally, we suggest measuring BDNF within a shorter period after ketamine infusion since it might be a more suitable window for detection of variations in this neurotrophin.