Serum brain-derived neurotrophic factor following oral esketamine in treatment-resistant depression: Results from a randomized placebo-controlled trial

Major depressive disorder is common and burdensome, and treatment-resistant depression (TRD) remains difficult to treat after failure of multiple antidepressants. The paper notes that ketamine and esketamine have shown antidepressant effects in some patients, but the biological mechanisms behind this response are still not fully established. One proposed mechanism involves brain-derived neurotrophic factor (BDNF), a protein linked to synaptic plasticity, with earlier studies suggesting that peripheral BDNF may rise after acute intravenous ketamine in people who respond to treatment. However, the authors emphasise that evidence is inconsistent and largely based on single-dose intravenous studies, leaving it unclear whether other dosing schedules or routes of administration produce similar changes. Massetti and colleagues set out to examine whether six weeks of daily, low-dose oral esketamine changes serum BDNF levels compared with placebo in patients with TRD. They also aimed to assess whether any BDNF changes were related to changes in depressive symptoms or to circulating ketamine metabolites. The study uses biomaterial from a multicentre, randomised, double-blind, placebo-controlled trial, and is presented as a way to better understand whether oral esketamine engages the same proposed neurobiological pathway as other ketamine regimens.

The researchers analysed data and blood samples from a multicentre, double-blind, randomised placebo-controlled trial conducted in the Netherlands between February 2017 and February 2021. The trial took place at three psychiatric centres, although blood samples were collected at two of them. Adults aged 18 to 80 years with TRD were recruited through psychiatric departments, patient and family associations, and media advertising. Eligible participants had a DSM-5 diagnosis of current major depressive disorder confirmed by MINI interview, a Hamilton Depression Rating Scale 17-item version (HDRS-17) score above 18, and insufficient lifetime response to three or more antidepressant classes given at adequate dose and duration. Key exclusions included psychosis, bipolar depression, substance use disorder, personality disorder, recent substantial benzodiazepine use, ongoing electroconvulsive therapy, active suicidal intent, pregnancy or lactation, and use of agents known to interact with ketamine. Participants were randomised by computer to oral esketamine or placebo. The intervention lasted six weeks, followed by a four-week washout period. Esketamine was given three times daily as fixed low-dose oral capsules, starting at 30 mg/day, increasing to 90 mg/day over the first three days, maintained at 90 mg/day, and then tapered back to 30 mg/day during the final three days. Placebo capsules matched esketamine in appearance and scent. Usual antidepressant medication was continued, and participants were hospitalised for the first five days. Compliance was checked by asking about each intended dose. Depression severity and serum BDNF were measured at baseline, end of treatment, and end of washout. Ketamine and related metabolites were measured at week 1 and week 6. BDNF was assayed in serum using ELISA, with samples collected at 12:00 pm, processed promptly, and stored at -80°C. Ketamine metabolites were quantified using liquid chromatography tandem mass spectrometry. The primary analyses used repeated measures ANOVA with time as the within-subject factor and treatment group as the between-subject factor. Pearson’s correlations examined relationships between changes in BDNF and changes in HDRS-17 scores, as well as between BDNF and metabolite levels; Spearman’s rho was used when data were non-normally distributed.

Of the 111 participants who started the original trial, 97 completed the six-week treatment phase, and 54 had sufficient blood samples for the present biomarker analysis: 27 received esketamine and 27 placebo. Baseline and follow-up HDRS-17 data were available for all 54, while washout HDRS-17 data were available for 51. BDNF analyses used baseline and follow-up serum samples from 52 participants, with 43 providing samples at follow-up and washout. The included sample was 55.6% male, with ages ranging from 22 to 73 years, and had long-standing depressive episodes and substantial prior antidepressant exposure. Clinical outcomes did not differ significantly between groups. At the end of treatment, 11 participants (20.4%) showed a partial response, 3 (5.6%) responded, and 2 (3.7%) remitted overall. In the esketamine group, 5 participants (18.5%) showed partial response, but none responded or remitted. In the placebo group, 6 (22.2%) had partial response, 3 (11.1%) responded, and 2 (7.4%) remitted. Differences between groups were not significant for response, partial response, or remission. Repeated measures ANOVA showed no significant time-by-treatment interaction for HDRS-17 scores during treatment (F(1,52)=0.514, p=0.477), although there was a significant main effect of time, with scores falling from baseline to follow-up across both groups. During washout, neither the interaction nor the main effect of time was significant. For serum BDNF, there was no significant time-by-treatment interaction during the treatment phase (F(1,50)=0.915, p=0.343). There was, however, a significant overall increase over time irrespective of treatment, from baseline mean 37,948.5 pg/ml to follow-up mean 43,409.2 pg/ml. During washout, the direction of change differed between groups: BDNF decreased in the esketamine group and increased in the placebo group, but the interaction was not significant. Changes in BDNF were not correlated with changes in depressive symptoms in the full sample or within either treatment group. Likewise, serum BDNF levels did not correlate with ketamine, norketamine, dehydroxynorketamine, or hydroxynorketamine levels at week 1 or week 6. Overall, the data did not show a treatment-specific rise in BDNF with oral esketamine.

The authors interpret the findings as showing that six weeks of daily low-dose oral esketamine did not increase serum BDNF relative to placebo in patients with TRD, matching the lack of clinical superiority of this regimen in the parent trial. They argue that the absence of a BDNF effect may mean that the oral dose and schedule did not generate sufficient circulating esketamine levels to trigger the synaptic plasticity processes thought to underlie antidepressant action. They connect this possibility to oral bioavailability and first-pass metabolism, and suggest that the low, divided dosing may have produced only modest NMDA receptor blockade rather than the transient, substantial blockade associated with acute intravenous ketamine. The authors place their findings in the context of earlier work showing BDNF increases after acute high-dose intravenous ketamine, particularly among clinical responders. They note that their results differ from that literature and from meta-analytic evidence, but point out that prior studies often used different routes, doses, and sampling schedules, and many included antidepressant washout. They also note an open-label follow-up study in the same patient group where higher individualised oral doses were associated with antidepressant benefit, which they cite as consistent with the idea that higher exposure may be necessary. The paper also discusses the unexpected overall increase in BDNF across both groups. The authors suggest this could reflect ongoing antidepressant co-medication, which all participants continued during the trial, or possibly placebo effects. They state that concomitant psychotherapy is unlikely to explain the finding. During washout, BDNF trended down in the esketamine group and up in the placebo group, but this was not statistically significant. The authors speculate that stopping oral esketamine might reduce glutamatergic signalling and thereby lower BDNF, while acknowledging that previous intravenous ketamine studies do not clearly support this pattern. Several limitations are acknowledged. The authors emphasise the difficulty of inferring central nervous system biology from peripheral serum BDNF, and they note multiple confounders such as platelet release, circadian rhythms, acute stress, and physical activity. They also point to selection bias and reduced power due to missing blood samples. Another limitation is the absence of BDNF Val66Met genotyping, which could have masked effects in genetically susceptible subgroups. Because the oral esketamine regimen was not clinically effective, the study could not stratify analyses by response status. Finally, they note that sampling at a single time of day may have reduced but not eliminated variation, and that acute changes may have been missed. In terms of implications, the authors suggest that future studies should examine higher oral doses of esketamine, better control for antidepressant co-medication, and consider genetic stratification and more precise biomarker sampling. They conclude that achieving higher acute blood concentrations may be important for engaging the biological mechanisms thought to support antidepressant effects.