Studies of ketamine in the treatment of depression
Numerous studies, literature reviews and meta-analyses have been published on intravenous ketamine for major depressive disorder. Following a broad review, we selected 22 controlled studies for closer evaluation (see Appendix). The patients mostly had treatment-resistant depression, although a couple of studies did not explicitly use this term. In most studies, doses of 0.5 mg/kg were administered as an infusion. Many ketamine studies have considerable shortcomings in terms of methodology and reporting. Nearly half of the studies (10/22) were single-dose studies focusing on immediate effects. On average, the studies included 30 patients receiving ketamine, with a follow-up period of 14 days. Several were effectively pilot studies, exploring different doses in a small patient group.
When administered intravenously, ketamine produced a rapid effect in most studies, as reflected by improvements on depression rating scales such as the Montgomery–Åsberg Depression Rating Scale (MADRS) (5–7). In several studies, the effect waned over the course of days. It is not surprising that administration of a substance such as ketamine produces observable effects. This is a substance with marked psychoactive properties. The question is whether this immediate effect can truly be used to assess ketamine's benefit or role in the treatment of a chronic condition such as depression. The acute antidepressant effect of ketamine was substantially reduced when the opioid antagonist naltrexone was administered prior to ketamine infusion in patients with depression (4, 8). Many psychoactive substances are used as self-medication in psychiatric disorders, with well-known harmful effects over time.
There appears to be a general consensus that ketamine treatment must be repeated over time, but few studies had multiple dosing and follow-up beyond a couple of weeks after treatment cessation. We identified four studies that used multiple dosing (2–3 doses per week for 1–4 weeks), included at least 50 patients, and had a minimum follow-up of four weeks (9–12).
In one study, repeated ketamine infusions over two weeks were compared with midazolam infusions followed by a single ketamine infusion (n = 54) in outpatient veterans (9). The authors found no difference in response between the groups when measured 24 hours after infusion of the final dose. During a 12-week follow-up phase among responders, there was a substantial relapse rate within a few weeks in both groups. In the recently published KARMA-Dep 2 study, 65 hospitalised patients were randomised to up to eight ketamine or midazolam infusions, with six months of follow-up (12). Repeated ketamine infusions were not more effective than midazolam in reducing depressive symptoms during the treatment phase. Results from the follow-up phase are difficult to interpret because of attrition.
Ketamine studies are characterised by significant bias. Many are designed as randomised, controlled double-blind trials, but blinding is very difficult due to the acute hallucinogenic effects of ketamine
Two large-scale studies compared repeated ketamine infusions with electroconvulsive therapy (ECT) (10, 11). ELEKT-D was a pragmatic three-week study that included outpatients with no psychotic symptoms (10). The study found that ketamine (n = 195) was non-inferior to ECT (n = 170). Ketamine was 'the new treatment', and the drop-out rate in the ECT group was high (approximately 20 %) before treatment started, primarily because patients wanted ketamine. The study included a six-month follow-up of responders, but the results are difficult to interpret due to the high drop-out rate and because many participants received ketamine during the follow-up period.
The KetECT study included inpatients referred for ECT (n = 186) (11). The authors found a lower remission rate with ketamine compared with ECT (46 % versus 63 %). Little information is given about the one-year follow-up phase, but relapse was observed in 64–70 % of patients, with a median remission duration of just less than a couple of months. Differences in results between KetECT and ELEKT-D have been discussed in terms of patient population differences, expectation bias, and the unusually low remission rate with ECT in ELEKT-D (22 %) compared with what was expected (13).
Based on both studies, ketamine does not appear to be more effective than ECT in treatment-resistant depression. The Norwegian health technology assessment only included the ELEKT-D study (1), not KetECT, and therefore does not provide a representative picture of the evidence base.
In the studies with long-term follow-up, there was substantial relapse within weeks to a few months (9–11), and the development of tolerance has been discussed (4).