Journal Club: Trial of Ketamine Masked by Surgical Anesthesia in Depressed Patients
Over the weekend, a new preprint was released from Stanford University focused on better understanding potential drug placebo effects in clinical trials assessing ketamine as a treatment for treatment resistant depression (TRD). While this paper has not been officially peer-reviewed, it provides the public and science communities early access to results that authors believe to be important.
This paper comes at a time where ketamine has been popularized by the media and believed to be the hot new cure to TRD. In 2019, ketamine nasal spray was approved for at-home use for those with TRD. Following this, multiple companies started providing at-home and in-clinic ketamine treatments in the form of nasal sprays, injectibles and infusions (and not for cheap). In most clinical trials of ketamine for depression, patients are able to correctly identify their treatments, which is a major issue in result interpretation.
Despite the promising results for TRD, it's important to continue to research these new indications of ketamine and whether the drug is really producing all that we expect. Which is exactly what this paper is trying to do.
The goal of this study is to figure out whether ketamine given at a low dose during general surgical anesthesia produces greater antidepressant effects than a saline placebo. They hypothesize that ketamine is superior to saline in reducing depression symptoms within 3 days post-infusion.
In contrast to many other clinical trials for ketamine and psychedelics, this trial was university-sponsored and investigator-initiated. This means that a drug company did not pay for this trail.
Adults undergoing elective surgeries (no heart or brain surgery) were recruited from the Stanford Medical Center and surgical clinics. Exclusion criteria included pregnancy/breastfeeding,
moderate to severe SUD, high risk of suicidal behavior via suicide-severity scale, history of schizophrenia/schizoaffective disorder, dementia/amnesia cognitive disorders, history of brain surgeries, degenerative nervous system disorder, thyroid dysfunction in the last 6 months, and chronic use of greater than or equal to morphine mg equivalents per day.
Only one participant out of the 40 was discontinued due to death via witnessed cardiac arrest that occurred 2 days after discharge, unrelated to the ketamine trial. Recruitment of participants and study design/information can be seen the right. ITT means intention-to-treat.
This study is triple masked, which means that participants/experimenters/anesthesiologists/data analysts were all blinded to treatment groups, randomized and placebo-controlled.
Ketamine or saline was given IV infused over 40 minutes, initiated after anesthetic induction and surgical incision to ensure participant masking.
EEG was used to confirm anesthetic depth, and anesthesia was not altered in response to any EEG changes unless patient was at risk of intraoperative awareness.
Anesthetic agents used included: propofol and inhaled sevoflurane or isoflurance. Nitrous oxide use was minimized due to it's reported anti-depressant effects.
The primary outcome measure was the MADRS, which we've discussed extensively on the podcast and in other journal clubs. It is used for indicating depression severity. The secondary outcomme was clinical response.
MADRS given 1, 2 and 3 days post-infusion (primary outcome)
Additional assessments at 5, 7 and 14 dats post-infusion (exploratory outcome)
Clinical response defined as greater than or equal to 50% reduction in MADRS scores from screening baseline (secondary outcome)
Remission defined as MADRS scores less than or equal to 12 in this study (exploratory outcome)
Other exploratory outcomes includes the HADS self-reported questionnaire to assess anxiety/depression in hospital patients.
Post-operative pain was assessed using Pain Inventory Short Form
In-patient post-operative opioid use was also assessed.
14 days post-treatment, participants were asked to guess their treatment.
The moment we've all been waiting for...
No differences between MADRS scores for the placebo and ketamine groups on 1-3 days post-infusion were noted. Differences were not seen in the exploratory post-infusion days either. This pre-infusion baseline MADRS scores did not differ between groups.
Day 0 to Day 1 (first day post-infusion) there was an average change in MADRS scores of around -12 points in the ketamine group and around -15 in the placebo group. This produces around a 50% decrease in both groups.
Antidepressant response rates similar in both groups at >50%, while remission rates were different, 50% in ketamine group and 35% in placebo group. This is comparable with other studies assessing ketamine.
To determine patient unblinding, they asked participants at the end of the follow up to guess their treatment group. Only around 37% of patients were able to guess correctly and the distribution of guesses was pretty even between the groups (see figure 2B on the right). One caveat of this analysis is that the authors did not ask about expectancy of outcomes prior to treatment, which would allow us to determine if expectancy of ketamine plays a large role in the therapeutic outcome. It also begs the question, should we start assessing the unblinding and expectancy of experimenters/medical staff? Do their expectations of the trial confound results?
Interestingly, the participants in this study were mostly female (70%!), which is very different from other studies. The mean age of participants was 51 years old, and most were non-hispanic white and majority were employed. How would the results would look in a sample with equal distribution of males? Would there be any sex differences in response rates or in the ability to identify the correct group?
Thoughts & Caveats
The increased response in the control group allows us to explore this question: is it really high placebo or is there something else at play?
The first thing that comes to mind is the question of whether anesthetics also produce antidepressant like effects. It was mentioned that nitrous oxide, an inhaled anesthetic, does produce some potential antidepressant effects and it's a large question of whether it also produces psychedelic-like effects. Other anesthetics have been found to produce some antidepressant like effects, but this is mostly in doses that suppress EEG activity over multiple administrations which differs GREATLY from the methods here.
To get at this question, the researchers tried to exclude nitrous from the mix. They also did a sensitivity analysis where nitrous was present to determine if it was confounding results. They find that when adjusting for nitrous, they still had no between-group differences in MADRS. They even go a step further to exclude these folks from the analysis and found that the null-hypothesis still stands: No difference between placebo and ketamine groups.
Another potential explanation for this high placebo response is that the effects of ketamine are being blocked by this general anesthesia. Propofol, which is used throughout this study, is a sedative-hypnotic that acts primarily on GABA receptors in the brain, which are inhibitory and reduce excitation. Ketamine's mechanism of action is mixed-pharmacology, but there is little activity of ketamine at GABA-receptors specifically.
The hypothesized primary mechanism of action for ketamine to produce antidepressant effects is through inhibition of NMDA receptors on GABA inhibitory neurons, allowing the excitatory neurons to produce glutamate and increase cell-firing and synaptic plasticity (see figure below Credit: Yang H. Ku/C&EN). This is interesting because the interaction between direct GABA-activation of anesthetics and indirect GABA neuron inhibition via ketamine has not been explored in detail. Perhaps there is a cancelling out - or forced equilibrium - between these two mechanisms, causing this high placebo/low ketamine effect. Even with this potential explanation, the ketamine responses (60%) are quite similar to other clinical trials.
There are a lot of questions raised from this study and we are excited to see what changes/commentary arises once the paper is officially peer-reviewed and published. A few thoughts to leave our readers with:
If ketamine/psychedelic therapy proves to be just placebo, is it appropriate for companies to profit from vulnerable groups seeking treatment?
What drug-drug interactions should be assessed to optimize treatment outcomes?
If there is a possible therapeutic benefit from anesthetics, is it appropriate to put people under anesthesia for an hour and call it good?
Do these results allow for some interpretation of the importance of being in the experience of the ketamine, the ritual of going to the doctor or therapist office, the routine of taking medication to your better yourself?
Does this work in other surgical populations? (see study with delirium here, and a follow-up analysis from UMich)