Efficacy and safety of ketamine in bipolar depression: A systematic review

The investigators conducted an electronic search of PubMed and Embase for prospective studies, case series or clinical trials in humans published from January 2012 to October 2015, limited to English and Spanish. Search terms included "ketamine", "bipolar disorder" and "bipolar depression". A manual search and reverse citation searching were also performed. Studies were included if they enrolled patients diagnosed with bipolar disorder and evaluated ketamine's efficacy for depression; papers with mixed diagnostic samples or that did not assess depressive outcomes were excluded. Two reviewers performed study selection and data extraction, resolving disagreements by consensus. Included papers were appraised using the online critical-reading tool "Osteba Fichas de lectura crítica" to assess methodological quality. The authors report following a PRISMA-style flow for study selection; the initial yield was 30 records, of which 10 met the inclusion criteria.

Ten studies met inclusion criteria: one double-blind randomised crossover clinical trial, five cohort studies and four case series. Intravenous (i.v.) infusion at a dose of 0.5 mg/kg was the most commonly used regimen. Other routes reported were intramuscular (i.m.; 50–100 mg), intranasal (i.n.; 10 mg per administration) and sublingual (s.l.; 10 mg). All studies prespecified a response criterion, defined as a ≥50% reduction on standard depression scales (HDRS or MADRS). The double-blind crossover trial (Zarate et al.) found rapid improvements in anxiety and depressive symptoms within 40 minutes of i.v. ketamine versus placebo (HDRS: F = 3.08, P = .001; VAS-Anxiety: F = 2.12, P = .030). The 40-minute response rate (MADRS ≥50%) was 64% after ketamine; it fell to 43% at 24 hours and roughly 30% were in remission (MADRS <10). Antidepressant effects measured by HDRS and MADRS lasted about 3 days, whereas BDI scores showed a longer effect (up to 14 days). The study also reported a rapid reduction in suicidal ideation in the first 40 minutes. Luckenbaugh et al. performed a pooled post hoc analysis of two earlier double-blind crossover studies and reported that participants with a family history of alcohol dependence (FHAP) experienced greater and longer-lasting antidepressant responses to ketamine than those without FHAP (MADRS: P = .030, d = 0.9; HDRS: P = .010, d = 0.25). Those with a personal history of alcoholism displayed lower scores on dissociation and psychotomimetic scales after ketamine. Among cohort and case reports, results were heterogeneous but generally indicated rapid onset of antidepressant effect with variable durability. Rybakowski et al. (n = 25) reported 24% response at 24 hours and 52% response at 7 days after a single i.v. 0.5 mg/kg dose, with 12 patients in remission by day 14. Permoda-Osip et al. found 10 of 20 patients responded at 7 days and identified higher vitamin B12 levels in responders (P = .047), noting the sample was predominantly female. Kantrowitz et al. combined a single ketamine infusion with an 8-week d-cycloserine regimen in treatment-resistant bipolar depression; of seven completers, four achieved remission (HDRS <7) and large effect sizes were reported early and at 8 weeks (Cohen's d = 2.0 at day 1; d = 1.1 at week 8). Case reports described durable remissions in some individuals, including patients who had previously responded to electroconvulsive therapy. Route-specific reports indicated that i.m., i.n. and s.l. administration can produce clinically meaningful benefits. Papolos et al. reported i.n. ketamine in 12 children and adolescents produced improvements lasting 36–60 hours and remission maintained for 3–7 days in many cases. Lara et al. observed that very low-dose s.l. ketamine (10 mg) induced remission in 4 of 14 bipolar patients and maintained responses in 5 others; tolerability was generally favourable. Adverse effects were typically transient and occurred around the time of administration. Studies documented dizziness, somnolence, perceptual disturbances, nausea, headache and mild dissociative symptoms that usually resolved within 1–2 hours. No consistent changes in vital signs or laboratory values were reported in the clinical trial. Tolerance to acute perceptual effects was noted with repeated administrations in some series.

Across the assembled studies, ketamine produced a rapid antidepressant effect in bipolar depression, often within minutes to hours, but the duration of benefit was generally short, typically ranging from a few days up to two weeks depending on the assessment instrument. The authors interpret these findings as consistent with evidence from unipolar depression and suggest ketamine may serve as an adjunctive or second-line option for patients who require rapid symptom relief or who are treatment-resistant, provided mood stabilisers are co-administered. The review highlights putative moderators and mechanistic leads identified in the literature. Family history of alcohol dependence emerged as a potential predictor of greater ketamine responsiveness, possibly reflecting NMDA receptor genetic variation. Vitamin status (notably B12) and inflammatory markers such as IL-6 were raised as candidate biomarkers that merit further exploration. Adjunctive agents such as d-cycloserine and pyridoxine were discussed as possible strategies to augment or prolong effects, but evidence is preliminary. Safety considerations receive detailed attention. Short-term adverse effects are usually transient and mild to moderate, but concerns remain about long-term safety with repeated dosing. The authors note theoretical risks related to mTOR pathway upregulation (with a speculative link to tumour growth), cardiovascular stimulation, and known associations of ketamine abuse with urological and hepatobiliary problems. They caution about use in patients with current or prior neoplasia, cardiovascular disease, or substance-use disorders and emphasise the potential for abuse with oral formulations. The variable bioavailability across routes is discussed: i.v. and i.m. have high bioavailability, i.n. is intermediate, and oral is low but yields higher relative norketamine exposure and longer half-life. Alternative routes (i.m., i.n., s.l.) may be attractive for maintenance dosing due to practicality and tolerability, but robust safety and efficacy data are lacking. The authors acknowledge the limited and heterogeneous evidence base: only one randomised controlled trial was identified alongside cohort studies and case series, and many findings derive from small samples or post hoc analyses. They therefore call for more studies specifically designed for bipolar depression to define optimal dosing, maintenance strategies, predictors of response and long-term safety.

The review concludes that ketamine, as an NMDA receptor antagonist, demonstrates rapid antidepressant and anti-suicidal effects in some patients with bipolar depression, but these effects are typically short-lived. On the basis of available evidence, ketamine may be considered as an adjunct to mood stabilisers or as a second-line option for selected subgroups (for example, treatment-resistant patients or those with a family history of alcohol dependence), while emphasising the need for maintenance strategies to prevent relapse. The authors recommend further controlled studies to evaluate different administration routes, dosing schedules, predictors of response and long-term safety, and they advise caution when treating patients with neoplastic disease, cardiovascular comorbidity or a history of ketamine abuse.