Intravenous Ketamine for Adolescents with Treatment-Resistant Depression: An Open-Label Study

Depression often begins in adolescence and can cause substantial disability and suicide risk. Standard pharmacological and psychotherapeutic interventions frequently fail: about 40% of adolescents do not respond to a first treatment and only half of those nonresponders improve with a second trial. Ketamine, an N-methyl-D-aspartate (NMDA) receptor antagonist used in anaesthesia, has shown rapid antidepressant effects in adults with treatment-resistant depression (TRD), including after single low-dose (0.5 mg/kg) intravenous infusions and in studies of repeated administrations. However, little is known about ketamine's efficacy or tolerability in adolescents, a developmental period with heightened neuroplasticity that might influence both response and duration of benefit. Cullen and colleagues set out to evaluate six serial open-label intravenous ketamine infusions (0.5 mg/kg over 40 minutes) administered across 2 weeks in adolescents with TRD. The study had three aims: to measure the extent and short-term duration of clinical response and remission (primary outcome: percent change on the Children's Depression Rating Scale–Revised, CDRS-R, with response defined as ≥50% reduction and remission as CDRS-R ≤28), to assess tolerability including dissociative and haemodynamic effects, and to explore clinical and demographic predictors of response. The investigators also prespecified a naturalistic follow-up of responders with weekly visits for 6 weeks and a 6-month visit to capture relapse timing and persistence of benefit.

This was a single-arm, open-label clinical trial conducted at a single centre and approved by the local Institutional Review Board. Adolescents aged 12–18 years with a current diagnosis of Major Depressive Disorder and treatment resistance (failure to respond to at least two adequate antidepressant trials, as rated on the Antidepressant Treatment History Form) were eligible. Required baseline severity was a CDRS-R raw score >40. Current psychotropic medications had to be stable for 2 months; pre-study washout periods were specified (2 weeks for mood stabilisers/antipsychotics, 4 weeks for antidepressants, 1 week for stimulants). Exclusion criteria included current substance use disorder, primary psychotic disorder, bipolar disorder, autism spectrum disorder, intellectual disability, nonpsychiatric neurological disorder, or significant medical illness. After informed consent/assent and diagnostic assessment using the K-SADS-PL (adolescents and parents interviewed separately), baseline symptom measures included the CDRS-R (primary outcome), MADRS, BDI-II, Snaith–Hamilton Pleasure Scale (SHAPS), and Temporal Experience of Pleasure Scale (TEPS). Intelligence was estimated with the Wechsler Abbreviated Scale of Intelligence. Participants who met inclusion criteria received six open-label intravenous ketamine infusions administered on alternate weekdays over 2 weeks, each infusion given at 0.5 mg/kg over 40 minutes followed by 2 hours of monitoring. For the first five participants, dose was calculated using ideal body weight; after observing no responders in that subgroup, dosing was switched to actual body weight for subsequent participants. Safety monitoring during and after each infusion included vital signs (heart rate, blood pressure, respiratory rate, oxygen saturation) measured at baseline and every 15 minutes, and dissociation assessments with the Clinician-Administered Dissociative States Scale (CADSS) immediately postinfusion and at 1 and 2 hours. Depression and clinical global measures (MADRS, CGI) and self-report measures (BDI-II, SHAPS, TEPS) were collected before each infusion and at 2 hours postinfusion. The primary endpoint was percent change in CDRS-R from baseline to the day following the final infusion; response was defined as ≥50% reduction and remission as CDRS-R ≤28. Responders were invited for six weekly follow-up visits and a 6-month visit during which clinical care was naturalistic. Statistical analysis comprised descriptive statistics, paired t-tests for pre/post changes on primary and secondary measures, and linear regression analyses using CDRS-R percent change as the continuous outcome to explore potential predictors (age, sex, age at depression onset, illness duration, baseline severity measures, number of past medication trials, trauma history, BMI and dosing strategy). The investigators documented adverse events. The extracted text does not state use of an intention-to-treat framework; sample size was small (13 completers) and no formal power calculation is reported in the extracted material.

Thirteen adolescents completed the six-infusion protocol and the posttreatment assessment; the consort diagram indicated initial contact with parents of 36 adolescents, 14 completed consent and diagnostic interview, and one dropped out before receiving infusions. Participant characteristics reported elsewhere in the extracted text include a mean age of 16.9 years (range 14.5–18.8) and eight biologically male, although the Methods section itself did not re-state the full demographic table. On the primary outcome, mean CDRS-R percent change from baseline to the day after the final infusion was 42.5% (SD = 31%). Five participants (38%) met the predefined response criterion (≥50% reduction); three of these responders were in remission at the posttreatment assessment (CDRS-R ≤28). During follow-up, two responders relapsed within 1–2 weeks, two responders remained in remission at 6 weeks, and a third participant who was just below remission at posttreatment reached remission by week 6. Four of the five responders attended the 6-month visit and two remained both in response and remission at that time. Secondary clinician- and self-rated measures (MADRS, BDI-II, CGI) showed significant pre/post improvements; measures of anhedonia (SHAPS, TEPS) did not change significantly. Adverse events were generally transient. Blood pressure increases were modest and peaked during infusions (mean systolic increases up to 6.8 mmHg at 30 minutes and 7.3 mmHg at 45 minutes; maximum mean diastolic increase 5.1 mmHg at 45 minutes); no elevations required intervention. Heart rate, respiratory rate and oxygen saturation did not show clinically significant changes. CADSS dissociation scores were elevated immediately after infusions, with larger effects during earlier infusions, but returned to baseline by the 1-hour assessment. Two participants experienced dysphoria during an infusion; three reported nausea (one vomited, subsequently treated with ondansetron); one reported post-procedural hand pain from IV placement. No respiratory adverse events were observed. One participant with high baseline suicidal thinking reported subjective worsening but structured assessments did not document a change. Exploratory predictor analyses identified actual ketamine dose (F = 8.84, p = 0.01) and BMI (F = 6.28, p = 0.03) as significant predictors of greater CDRS-R percent change. Dosing strategy (actual versus ideal body weight) showed a trend in the regression (F = 4.43, p = 0.06); in a subgroup comparison those dosed by actual body weight had larger mean percent CDRS-R change (55%, SD 31%) than those dosed by ideal body weight (22%, SD 21%; t = 2.3, p = 0.04). None of the five participants dosed by ideal body weight were responders, whereas five of eight participants dosed by actual body weight were responders (chi-square = 2.79, p = 0.10). Trauma history did not emerge as a significant predictor in the regression (trend-level p = 0.15), but a follow-up t-test reported greater improvement among those without a trauma history (mean percent change 60.2%, SD 26.1%) versus those with trauma history (mean 21.9%, SD 24.2%; t = 2.74, p = 0.02). No significant predictive effects were observed for age, sex, estimated IQ, number of past medication trials, age at depression onset, duration of current episode, or baseline severity scores (CDRS-R, BDI-II, SHAPS, TEPS).

Cullen and colleagues interpret these findings as preliminary evidence that serial low-dose intravenous ketamine can produce rapid reductions in depressive symptoms in some adolescents with TRD. They note this is, to their knowledge, the first report of ketamine treatment in an adolescent TRD sample and that six open-label infusions were associated with significant symptom reductions and a 38% responder rate, with several responders maintaining remission through the 6-week follow-up and two remaining in response/remission at 6 months. The authors contrast the observed responder rate with higher rates reported in adult TRD studies and suggest several possible explanations. Developmental factors and earlier illness onset in adolescents may be associated with a more severe or distinct illness course less amenable to brief interventions. A procedural factor likely contributing to lower response is dosing: the initial use of ideal body weight for the first five (all nonresponding, four overweight) resulted in lower mg/kg exposures (mean ~0.35 mg/kg), after which dosing was changed to actual body weight and responders were more frequent. From this they infer a possible dose–response relationship in adolescents, consistent with adult data showing greater efficacy at higher doses in some studies. Regarding safety, the investigators report a low rate of adverse events limited chiefly to transient dissociative symptoms and modest haemodynamic changes, and no respiratory complications. Nonetheless, they emphasise important unknowns: long-term safety of repeated or maintenance low-dose ketamine is inadequately characterised, particularly in a developing brain where animal and some human data suggest potential neurotoxic risk with higher or chronic exposures. Concerns about potential induction of substance use disorder are acknowledged; the trial excluded participants with recent substance use disorder and the authors note that recreational ketamine doses are far larger than those used here, but they call for research on long-term risk. The authors explicitly acknowledge limitations that constrain inference: the open-label, single-arm design cannot rule out placebo or regression-to-the-mean effects; small sample size limits generalisability and statistical power for predictor analyses; the study was not designed to determine optimal dosing or maintenance strategies; and follow-up was limited, so questions about sustained benefits remain. They conclude that larger, double-blind, placebo-controlled trials are required to confirm efficacy, to establish optimal dosing and administration schedules, to characterise predictors of response, and to evaluate long-term safety and strategies to sustain remission in adolescent TRD.

Ketamine produced rapid antidepressant effects in a subset of adolescents with TRD in this open-label series, and the findings provide preliminary evidence for a dose–response relationship favouring dosing by actual body weight. However, the authors caution that definitive conclusions require large-scale, double-blind, randomised trials to determine safety and efficacy in this population, to optimise dosing and patient selection, and to address strategies for maintaining remission and long-term safety after acute treatment.