Ketamine’s dose-related effects on anxiety symptoms in patients with treatment refractory anxiety disorders

Previous research has shown that low-dose ketamine produces rapidly acting antidepressant effects in treatment-resistant depression (TRD), and preliminary studies have reported rapid improvement in obsessive–compulsive disorder and post-traumatic stress disorder. However, at the time of this study there were no published data on ketamine’s effects in treatment-resistant generalized anxiety disorder (GAD) or social anxiety disorder (SAD), despite shared alterations in brain network activity, grey matter and glutamate signalling across depression and anxiety disorders. Glue and colleagues set out to assess whether ketamine reduces anxiety symptoms in patients with treatment-resistant GAD or SAD. The study aimed to evaluate the feasibility, safety and dose-related effects of ascending single subcutaneous ketamine doses on standard anxiety rating scales in this patient group.

This was an open-label, ascending single-dose feasibility study approved by a regional ethics committee and prospectively registered with the Australian New Zealand Clinical Trial Registry; the extracted text does not clearly report the full registration identifier. Adults with DSM-IV GAD and/or SAD who met severity thresholds (Hamilton Anxiety Rating Scale (HAM-A) score ≥20 and/or Liebowitz Social Anxiety Scale (LSAS) score ≥60 at screening) were eligible. Patients with severe medical disorders, pregnancy or lactation, current use of monoamine oxidase inhibitors, thyroxine or stimulants, active suicidal ideation, or a Montgomery–Åsberg Depression Rating Scale (MADRS) score ≥20 at screening were excluded. All participants provided informed consent and underwent screening laboratory and vital-sign checks. Participants were allowed to continue stable ongoing medications and psychotherapy but could not start new treatments or change doses during the study. Three ascending subcutaneous ketamine dose levels were given in fixed order (0.25, 0.5 and 1 mg/kg), each separated by one week. The investigators selected subcutaneous administration for perceived similar efficacy to intravenous dosing with better tolerability; dose selection was informed by prior work in TRD that suggested a threshold near 0.25–0.3 mg/kg and durability at higher doses. Safety and efficacy monitoring included vital signs predose and at 15, 30, 45, 60, 90 and 120 minutes post-dose. Anxiety was assessed using the Fear Questionnaire (FQ) and HAM-A at predose and at 1, 2, 24, 72 and 168 hours post-dose. Dissociative effects were measured with the Clinician Administered Dissociative States Scale (CADSS) predose and at 30 and 60 minutes. Adverse events were recorded throughout. For analysis, summary statistics described demographics, vital signs and scales. To account for repeated measures within participants and across the three doses, the investigators used linear mixed models with random effects for participant and participant-by-dose and fixed effects for time, dose and the time-by-dose interaction; residual maximum likelihood estimation was used. Conditional residuals were inspected and log transformations considered when appropriate. Statistical testing used Stata 14.1 with two-sided p<0.05 considered significant; as an exploratory study no adjustment was made for multiple comparisons. A responder analysis defined response as a ≥50% reduction in HAM-A or FQ total score after any dose.

Seventeen patients were screened and 12 were enrolled; all 12 completed the protocol. The sample comprised four females (33%) and eight males (67%), mean age 32 years (range 23–55), with a mean duration of anxiety disorders of 16.5 years (range 5–30). Ten participants (83%) met criteria for GAD and nine (75%) for SAD; five (42%) also had panic disorder and two had post-traumatic stress disorder. All patients were taking antidepressants and had not responded to prior antidepressant trials and psychotherapy. Eleven subjects (92%) had a lifetime history of major depressive disorder, but none appeared to be currently depressed (mean MADRS at screening 8.0, range 5–12). Baseline severity scores were consistent with marked illness: mean HAM-A at screening was 26 and mean LSAS was 89. On the FQ, mixed model analysis showed a significant dose-by-time interaction (p=0.033), with dose-related differences in the magnitude and duration of initial reductions. HAM-A profiles showed dose-related patterns for initial decrease and duration, but the time-by-dose interaction did not reach conventional significance (p=0.068), described by the investigators as a non-significant tendency. Overall, 10 of 12 patients (83%) achieved a ≥50% reduction on HAM-A and/or FQ after the 0.5 or 1 mg/kg doses. All participants reported dissociative symptoms beginning around 5 minutes post-injection, peaking at about 20–30 minutes and largely returning towards baseline by 60 minutes. CADSS scores demonstrated dose-dependent increases at 30 minutes with a significant dose-by-time interaction (p=0.008); after 1 mg/kg two participants rated the dissociative experience as very intense and feeling out of control. Two subjects experienced transient nausea at 30 minutes. Cardiovascular effects were dose related: after 1 mg/kg mean systolic and diastolic blood pressure rose by 9.4 mm Hg and 7.5 mm Hg respectively at 30 minutes, and heart rate increased by 4.5 beats per minute at 15 minutes. No participants withdrew for adverse events.

Glue and colleagues interpret these results as the first demonstration that ketamine can produce rapid anxiolytic effects in patients with severe, treatment-refractory GAD and/or SAD who are not currently depressed. The temporal pattern resembled that reported in TRD: rapid onset of effect with waning benefit over 3–7 days. The investigators report a dose–response relationship for both anxiolytic effects and dissociative side effects; 0.25 mg/kg appeared to be a threshold producing only brief reductions at 1 hour, whereas 0.5–1 mg/kg produced larger reductions and the longest duration of benefit appeared at 1 mg/kg. The authors acknowledge several important limitations. The open-label ascending-dose design means both participants and investigators knew ketamine was administered, which may introduce bias in reported outcomes. Sample size was small and diagnostic comorbidity common: seven patients met criteria for both GAD and SAD, six had current or prior panic disorder and almost all had prior major depressive disorder. Concomitant antidepressant medications were continued, which could confound attribution of effects to ketamine. The investigators note that this cohort reflects a ‘‘real-world’’ treatment-resistant population but that prior MDD could represent a phenotype particularly responsive to ketamine; this possibility and the suggestion that high baseline anxiety in TRD predicts ketamine response warrant testing in future randomised trials. In their concluding remarks the study team state that ketamine was generally safe and well tolerated in this small sample and suggest ketamine may offer a therapeutic alternative for patients with few options. They also raise the broader hypothesis that ketamine could have efficacy across disorders characterised by negative emotional states, but indicate that controlled studies are needed to confirm efficacy and define optimal dosing.