Sanjay J Mathew, MD
Associate Professor of Psychiatry & Behavioral Sciences, Baylor College of Medicine, Staff Physician, Michael E. Debakey VA Medical Center, Houston, TX
Dr. Mathew has disclosed that he has worked as a paid consultant for AstraZeneca, Naurex, Bristol-Myers Squibb, and Roche/Genentech. Dr. Balt has reviewed this interview and found no evidence of bias in this educational activity.
TCPR: Were any other medications given during the infusion in your study?
Dr. Mathew: Participants in our study were washed off of any other psychotropic drug, including antidepressants, for at least a seven-day period prior to infusion. And the reason for that was that we want to look at other biological measures such as neurotrophic factors and brain-imaging markers, so we wanted to have as “clean” a sample as possible.
TCPR: People may argue that simply the psychoactive effects of ketamine—for instance, the transient dissociative symptoms and other unusual phenomena that people experience—might have antidepressant effects in and of themselves. Please explain how you took this into consideration in your study design.
Dr. Mathew: Instead of using a saline or an inert substance as a control, we randomized patients to ketamine or to midazolam (Versed), a commonly used anesthetic benzodiazepine. This was to make sure that patients were truly adequately blinded. Midazolam doesn’t have the dissociative side effects of ketamine, but it does have the anxiolytic, sedative, and some of the transient cognitive side effects. We found that midazolam was effective as an antidepressant in only about 20% of patients at 24 hours, and it persisted at 72 hours. So the drug-versus-placebo difference in our study was significant. However, it’s notable that when a treatment-resistant group of patients is given an intravenous medication that is not known to have any antidepressant properties, we still found 20% mounting a robust response. So we can’t discount the power of expectations for these types of studies and for clinical practice as well.
TCPR: So you attribute midazolam’s response rate primarily to the expectation that the medication will work. Is it possible that a rapidly acting psychotropic drug may disrupt the system so that the patient is no longer experiencing symptoms of depression?
Dr. Mathew: The patients who responded to midazolam said they felt lighter and more relaxed. That early anxiolytic effect might have influenced how they felt overall in terms of their mood. Nevertheless, in our study, the fact that the response rate was still significantly lower than with ketamine affirmed that there was something real about the ketamine antidepressant effect beyond its anxiolytic properties, and beyond its sedative or amnestic properties.
TCPR: Are there any predictors of the type of patient who might respond to ketamine?
Dr. Mathew: Researcher Carlos Zarate has found that a family history of alcohol dependence seems to be associated with early response in both unipolar and bipolar depression (Phelps LE, Biol Psychiatry 2009;65(2):181–184; Luckenbaugh D et al, Bipolar Disord 2012;14(8):880–887). His data also support that a previous history of suicide attempt is a predictor of good outcome. In addition, we have found that a history of childhood sexual abuse seems to be a predictor of good response, which is sort of at odds with much of the pharmacotherapy literature, in which early childhood trauma is a negative prognostic factor. We are looking at a number of other things as well, including a previous history of suicidal ideation and comorbid anxiety.
TCPR: How is this treatment being used in real-life clinical practice?
Dr. Mathew: There are a number private clinics now offering this treatment. Many of them use the 0.5 mg/kg protocol that we have studied, although there seems to be a lot of variability. Some clinics are doing bolus infusions over a couple of minutes, while others are doing different dosing, like 0.25 mg/kg over a longer period of time. Some practices are doing this under anesthesiologist monitoring with continuous EKG and pulse oximetry and nasal cannula, with the psychiatrist’s role much like in ECT. There are other clinics being run completely by psychiatry without the same degree of monitoring.
TCPR: Would you advise the average office-based psychiatrist to do this? If so, what are some of the practical issues?
Dr. Mathew: From our experience, even among carefully selected patients, there will be a small number who experience significant changes in blood pressure and pulse, and some patients will have relatively dramatic elevations in those parameters. In our studies we did have a portion who needed intervention by the anesthesiologist with antihypertensive medications. So I urge caution for the clinical use of this. If we see those type of side effects in a small proportion of highly selected patients who don’t have histories of hypertension or obstructive sleep apnea, and who have normal EKGs and are on no other psychotropic medications, it would suggest that in the general clinical practice you run into much higher risk.
TCPR: So there is really no accepted protocol for how often the infusion should be given, how many infusions should be given, when they stop, whether there should be maintenance infusions, and/or whether to use concurrent medications?
Dr. Mathew: That is correct. We don’t know the best way to conduct repeated administration dosing strategies—for instance, should the regimen be three times a week over two weeks, or three times a week over three weeks? At what point do you taper? And how do you manage a patient who has responded beautifully to ketamine? There is some preclinical literature suggesting lithium may have some common mechanisms with ketamine. And so there are studies ongoing looking at this lithium/ketamine combination as a way to prolong the benefit of ketamine. There really isn’t much literature guiding clinicians on this point. We decided initially three times a week because we were trying to follow an ECT model, and we felt three times a week over two weeks would be acceptable to patients and represent a minimal ECT course. But ultimately the treatment needs to be individualized.
TCPR: Let’s discuss how ketamine works as an antidepressant. Can you explain the history behind the use of ketamine and other NMDA-based antidepressant approaches?
Dr. Mathew: Back in the ‘80s it was discovered by Skolnick and others that most standard antidepressants, including the tricyclics, down-regulated NMDA receptor activity. They also found that NMDA antagonists may have acute antidepressant properties in laboratory experiments. Then, a number of years later, a pilot study done at Yale (Berman RM et al, Biological Psychiatry 2000;47(4):351–354) found that low-dose ketamine could have very rapid antidepressant properties. Since that study was published in 2000, there have been a number of replications supporting the idea that NMDA receptor blockade can result in a very rapid effect whose time course is beyond the half-life of the drug. The elimination half-life of ketamine is approximately two to four hours, and most of these studies have supported the notion that there is a persistence of benefit certainly beyond the half-life and in some patients several weeks, and these were by and large treatment-resistant patients.
TCPR: There are other NMDA antagonists, however, like memantine (Namenda), which don’t seem have any antidepressant effects. So what is it about ketamine?
Dr. Mathew: What is unique about ketamine? Ketamine binds within the channel of the NMDA receptor, specifically at same site as phencyclidine (PCP, “angel dust”), but with less affinity. This is important to its therapeutic effect. On the other hand, memantine is a relatively weak NMDA blocker, and other drugs that are available, like amantadine (Symmetrel), have weak NMDA receptor activity. By blocking NMDA, you may be preferentially engaging AMPA receptor activity, and AMPA receptor activity may be important for many of the downstream effects, including those underlying the long duration of action. So while the initial pharmacologic target is NMDA, one of the theories is that AMPA receptor potentiation is actually critical to the mechanism of action. That may also be more unique to the ketamine story than some of the other NMDA-like substances. And the way we know this preclinically is that when you block AMPA receptors you can undo some of the behavioral effects of ketamine in the animals.
TCPR: Given what we know about how ketamine works, what other molecules or other agents are on the horizon that might take advantage of this NMDA mechanism?
Dr. Mathew: One study is looking at the isomer of ketamine—S-ketamine—in an intranasal form for treatment-resistant depression. S-ketamine has a higher affinity at the PCP site, so the idea would be that can you use lower doses, achieve equal efficacy with fewer side effects and less dissociation. There are also a number of drugs that modulate NMDA-receptor functioning, but are more selective for specific subunits of the NMDA receptor (for example, NR2B receptor antagonists), which are now in clinical trials and being investigated. An NMDA channel blocker known as AZD6765, developed by AstraZeneca, is now in late phase 2 studies. Finally, there is a glycine site agonist called GLYX-13 (developed by Naurex) that is a partial NMDA agonist that is being studied in Phase 2 trials.