Plasma BDNF concentrations and the antidepressant effects of six ketamine infusions in unipolar and bipolar depression

Accumulating evidence links glutamatergic dysfunction to the pathophysiology of mood disorders, and subanesthetic doses of the NMDA (N-methyl-D-aspartate) receptor antagonist ketamine have been reported to produce rapid antidepressant and antisuicidal effects in unipolar and bipolar depression. Neurotrophic mechanisms, particularly brain-derived neurotrophic factor (BDNF), have been implicated in depression and in antidepressant treatment responses: lower peripheral BDNF levels have been observed in depressed patients versus healthy controls and have risen after successful treatments. Prior studies examining the relationship between peripheral BDNF and ketamine's antidepressant efficacy have produced inconsistent results, and most have considered single infusions rather than repeated administrations; no published work had examined this question in a Chinese sample receiving serial ketamine infusions. Zheng and colleagues designed the present study to test whether six subanesthetic intravenous ketamine infusions (0.5 mg/kg, administered three times weekly over two weeks) change plasma BDNF (pBDNF) concentrations, and whether baseline pBDNF is associated with subsequent antidepressant response. The investigators hypothesised that serial ketamine would increase pBDNF and that higher baseline pBDNF would be associated with better antidepressant outcomes in individuals with unipolar or bipolar depression.

This was an open-label clinical study conducted between November 2016 and December 2017 (registration: ChicCTR-OOC-17012239) with ethical approval from the Affiliated Brain Hospital of Guangzhou Medical University. Participants provided written informed consent. Eligibility criteria required age 18–65 years, a DSM-5 diagnosis of unipolar or bipolar depression established by SCID-5, HAMD-17 score ≥17, absence of psychotic symptoms, negative urine toxicology, no neurological disease or substance abuse, not pregnant or breastfeeding, and no unstable medical illness. Treatment-resistant depression (TRD) was defined as nonresponse to two or more antidepressant treatments; suicidal ideation also qualified some participants under the study inclusion criteria. All enrolled patients received six intravenous infusions of ketamine at 0.5 mg/kg delivered over 40 minutes, administered three times per week for two weeks, with a subsequent two-week follow-up. Vital signs and clinical status were monitored during infusions. Participants continued their existing psychotropic medications during the study (no washout). Depressive symptoms were assessed with the Montgomery–Åsberg Depression Rating Scale (MADRS) at baseline, 1 day after the sixth infusion (day 13), and two weeks after the last infusion (day 26). Response was defined as a ≥50% reduction in MADRS score and remission as a MADRS score <10. Plasma samples for pBDNF were collected at the same three time points, stored at −80 °C, and analysed by a commercial ELISA kit (EMD Millipore). Statistical analyses included Mann–Whitney U tests for non-normally distributed continuous data, independent t tests for normally distributed continuous data, and Fisher’s exact or Chi-squared tests for categorical comparisons. Linear mixed models tested changes over time and subgroup (responders versus nonresponders; remitters versus nonremitters). Bivariate correlations examined associations between baseline pBDNF and MADRS at days 13 and 26; multiple linear regression models tested whether baseline pBDNF predicted MADRS while adjusting for covariates (age, sex, weight, BMI, psychiatric family history, previous hospitalisation, psychiatric comorbidity, and age at onset). An additional analysis focused on the TRD subsample. Significance was set at p<0.05 and analyses were performed in IBM SPSS v23.

Ninety-four participants with unipolar or bipolar depression who provided baseline blood samples were enrolled; ages ranged from 18 to 62 years. Most participants met criteria for treatment-resistant depression: 81.9% (77/94). Baseline plasma BDNF concentrations averaged 10.1 ng/ml (range 0.9–27.2 ng/ml). Clinical outcomes after the six ketamine infusions showed a response rate of 68.1% (64/94) and a remission rate of 51.1% (48/94) at the post-treatment assessment. Within the TRD subgroup, response and remission rates were 68.8% (53/77) and 51.9% (40/77), respectively. Linear mixed-model analyses indicated significant time effects for both MADRS scores and pBDNF concentrations: ketamine produced significant reductions in MADRS and increases in pBDNF at day 13 and day 26 compared with baseline. Comparing clinical subgroups, baseline pBDNF concentrations were higher in responders than nonresponders (11.0 ± 6.2 ng/ml versus 8.0 ± 5.5 ng/ml) and higher in remitters than nonremitters (10.1 ± 5.8 ng/ml versus 9.2 ± 6.4 ng/ml); these differences reached statistical significance (p<0.05). The time-by-group analyses showed significant main effects between responders and nonresponders and between remitters and nonremitters for both MADRS and pBDNF. Correlation analyses demonstrated that baseline pBDNF was significantly associated with MADRS scores at day 13 and day 26 (all p<0.05). In multiple linear regression models that adjusted for demographic and clinical covariates, baseline pBDNF remained significantly associated with MADRS at day 13 (t = −2.011, p = 0.047) and day 26 (t = −2.398, p = 0.019). Subgroup analyses restricted to the TRD sample produced similar findings.

The investigators report this as the first study examining plasma BDNF concentrations following six subanesthetic intravenous ketamine infusions in a Chinese sample with unipolar and bipolar depression and the first to study the relationship between baseline pBDNF and serial-ketamine antidepressant outcome in this population. Key findings highlighted were that pBDNF increased at 13 and 26 days post-baseline, responders and remitters had higher baseline pBDNF than nonresponders/nonremitters, MADRS scores improved significantly at both follow-up time points, and baseline pBDNF was associated with subsequent MADRS scores; these associations were preserved in the TRD subsample. The authors place their results in the context of prior animal and human work: an animal study showed increased BDNF expression after NMDA receptor blockade, and some human studies have reported increased peripheral BDNF following ketamine, although results have been inconsistent across studies of single versus repeated infusions. Zheng and colleagues note that while baseline pBDNF differed between outcome groups, repeated ketamine infusions did not produce a significantly greater pBDNF increase in responders/remitters versus nonresponders/nonremitters in their sample, echoing mixed findings in the literature. Mechanistically, the paper reiterates the role of BDNF in synaptic plasticity and its putative place in ketamine's effects, including links to AMPA receptor trafficking and mTOR pathway activation reported in preclinical studies. The authors acknowledge several limitations: participants continued concomitant psychotropic medications without a washout, which may have influenced pBDNF; the sample size was modest; the open-label design and lack of a control group introduced the possibility of subjective bias; mediation and moderation analyses were not performed; and central measures (brain BDNF or mTOR) were not obtained, requiring reliance on the assumption that peripheral BDNF reflects central levels. Given these caveats, the authors present the results as preliminary and call for confirmation in randomised controlled trials.