Ketamine Treatment for Depression in Patients With a History of Psychosis or Current Psychotic Symptoms: A Systematic Review

Ketamine is an N-methyl-d-aspartate (NMDA) receptor antagonist developed as a short-acting dissociative anaesthetic that has been repurposed over recent decades because of rapid antidepressant and antisuicidal effects observed in unipolar and bipolar depression. Short-term psychotomimetic and dissociative effects are among ketamine's dose-related adverse effects, and because of its transient schizophrenia-like effects ketamine has commonly been used as a human model of psychosis in research. As a result, most clinical trials of subanaesthetic ketamine for depression have excluded patients with current or past psychotic symptoms. This exclusion has important clinical consequences because psychotic features are relatively common in severe depression and treatment-resistant depression (TRD), and negative symptoms in schizophrenia overlap phenomenologically with depressive syndromes. Veraart and colleagues therefore set out to review the published literature on ketamine or esketamine treatment in patients with a history of psychosis or current psychotic symptoms, including studies addressing depressive and negative symptoms, to determine whether ketamine administration in this vulnerable group is associated with exacerbation of psychosis or with therapeutic benefit.

The investigators performed a systematic search of PubMed/MEDLINE from database inception to March 2020 with no restrictions on publication year or language. The search terms combined ketamine with psychosis, psychotic or schizo*, and a human-studies filter was applied. Articles were eligible if they reported effects of ketamine (or esketamine) administered as a treatment for depressive symptoms or for negative symptomatology in patients who had a history of psychosis or current psychotic symptoms. The authors screened 482 abstracts and retained studies that presented original data on this specific patient population. The extraction and synthesis emphasised study design, population characteristics, ketamine dosing and route, outcomes on depressive, negative and psychotic symptom scales, and reported adverse events.

Nine reports met the inclusion criteria, comprising pilot studies and case reports with a combined total of 41 patients. All included studies were uncontrolled; study types ranged from single-case reports and small case series to a post hoc pooled analysis of randomized crossover trials and two small pilot series. Treatment targets were predominantly depressive symptoms; one small study specifically targeted persistent negative symptoms in schizophrenia. A post hoc analysis by Pennybaker and colleagues combined data from three randomized, placebo-controlled crossover trials of a single 0.5 mg/kg IV ketamine infusion over 40 minutes in patients with a current depressive episode. For the 69 patients for whom history of psychosis information was available, 12 had a prior psychotic episode (2 major depressive disorder with psychotic features and 10 bipolar disorder with psychotic features). Patients with a history of psychosis showed significant improvement on the Montgomery–Åsberg Depression Rating Scale (MADRS) from baseline up to 3 days post-infusion (P < .01). The antidepressant effect versus placebo was significant in both groups but had a smaller effect size in those with a history of psychosis (Cohen d = 0.35) than in those without (Cohen d = 1.17, P < .001 for both). Dissociation scores on the Clinician-Administered Dissociative States Scale (CADSS) were higher in the history-of-psychosis group at 40 minutes post-infusion (Cohen d = 0.23, P < .01), but Brief Psychiatric Rating Scale–Positive Symptoms (BPRS-P) scores did not differ between groups (P = .47). Multiple single-case and small-series reports described clinical improvements without sustained psychotic exacerbation. Medeiros da Frota Ribeiro and colleagues reported two cases (one long-standing unipolar depression with auditory hallucinations and paranoia; one schizoaffective disorder with depression, suicidality and catatonia) who both showed rapid and marked reductions in Hamilton Depression Rating Scale scores and disappearance of psychotic symptoms after 0.5 mg/kg IV ketamine; both patients remained stable on monthly maintenance infusions for up to one year. Ajub and Lacerda described four patients treated with esketamine (0.5 mg/kg IV in one case, subcutaneously in three): three patients had marked improvement or remission of depressive and psychotic symptoms at 24 hours and at 2 and 4 weeks, while one patient showed no benefit. Reported adverse events were transient dissociation, nausea, vomiting and light-headedness, resolving within about two hours. Zarrinnegar et al. documented a 15-year-old with severe TRD and psychotic features who received six 0.5 mg/kg infusions over three weeks; depressive, psychotic and suicidal symptoms were absent one month after the final infusion and the patient remained well on follow-up of several months. Bartova and colleagues reported a 30-year-old woman with chronic schizophrenia and residual depressive symptoms who responded to IV esketamine (initial 0.22 mg/kg then increased to 0.33 mg/kg thrice weekly) with a robust antisuicidal and antidepressant response lasting several days and sustained remission with the higher dose; only transient dissociation occurred and psychotic symptoms did not re-emerge. A pilot study in 15 patients with treatment-resistant schizophrenia and concurrent treatment-resistant depressive symptoms administered 0.5 mg/kg IV ketamine every 3 days for 3.5 weeks. Mean Calgary Depression Scale for Schizophrenia (CDSS) scores fell by 64% (baseline 16.50 [SD 3.94]) and PANSS general psychopathology scores decreased by about 30% between day 7 and day 14; PANSS positive symptom scores remained essentially unchanged (baseline 25.60 [SD 9.23]). One patient experienced transient visual hallucinations within 30 minutes of the first infusion. When the same protocol was repeated after 58 days, no significant differences in CDSS or PANSS were observed at later assessments, suggesting the initial benefit was not sustained. Nunes and colleagues investigated subcutaneous esketamine for persistent negative symptoms in six patients with schizophrenia, using weekly escalating doses from 0.5 mg/kg to 1 mg/kg over four weeks. Five of six patients showed clinically significant improvement in negative symptoms: mean Brief Negative Symptom Scale scores decreased from 59.5 to 37.3 (-37.3%), and mean BPRS scores decreased from 21.2 to 17.3 (-18.1%). Psychotic symptoms did not increase in any patient. Reported adverse events across studies were mostly mild and transient (dissociation, derealisation, nausea, vomiting, transient sedation); no sustained exacerbations of psychosis were reported. Several studies noted most patients continued standard psychopharmacotherapy, including antipsychotics, which complicates causal attribution of effects.

Veraart and colleagues interpret the available literature as not supporting the prevailing assumption that subanaesthetic ketamine or esketamine will exacerbate psychotic symptoms in patients predisposed to psychosis. Across the small case series, pilot studies and pooled post hoc data, ketamine produced good short-term antidepressant effects in many patients and in several reports psychotic symptoms improved or resolved. Dissociative and psychotomimetic side effects were observed more frequently in patients with a history of psychosis in one pooled analysis, but these phenomena were transient, peaking shortly after infusion and often diminishing with repeated dosing. The authors caution that effect sizes may be smaller in patients with a history of psychosis (Pennybaker et al. reported Cohen d = 0.35 versus 1.17), and that some longer-term or repeated-dosing strategies yielded limited durability of benefit, raising the possibility of tolerance. Interpretation is further complicated because most patients continued concurrent psychotropic treatment, notably antipsychotics, and pharmacodynamic interactions between ketamine and agents such as haloperidol, risperidone and clozapine have been demonstrated, so additive or interactive effects cannot be excluded. Key limitations acknowledged include the small number of studies and participants, the uncontrolled nature of all included reports, short follow-up periods, and likely publication bias favouring positive cases. On balance, the authors suggest these observations provide preliminary evidence that ketamine or esketamine can be administered to some patients with a history of psychosis or current psychotic symptoms without short-term exacerbation of psychotic illness, and they propose that future randomised controlled trials specifically include patients with vulnerability to psychosis to assess efficacy, tolerability and longer-term outcomes more robustly.