Special K: Ketamine for Treatment Resistant Depression

A few weeks ago, the final report on the death of Jim Irsay was released. For those who may not know, Mr. Irsay was the owner of the Indianapolis Colts. It was well known that he battled alcohol and drug addiction throughout his life. Toward the end, he was being treated regularly for depression with ketamine.

This marks the second high-profile case in which ketamine contributed to a celebrity death. Actor Matthew Perry drowned in his hot tub after receiving a physician-administered ketamine infusion. Although he had other sedating substances in his system, ketamine was likely a major factor.

Humans and ketamine have a complicated relationship. Ketamine was originally developed as an anesthetic, meant to be an improvement over phencyclidine (PCP). Yes—that PCP: angel dust, sherms, rocket fuel. It was once given to patients as a pre-operative hypnotic, leading to some wild recoveries in the PACU. I can’t imagine waking up from surgery while effectively tripping on PCP.

Ketamine was somewhat better than PCP. It was an effective pain reliever, and at higher doses, a functional hypnotic. But it still caused hallucinations and delusions—less than PCP, but still present. It also produced a nasty “hangover” phase and had various gastrointestinal and cardiac side effects, though these were considered tolerable compared to earlier alternatives.

As pharmaceuticals improved, ketamine was phased out of mainstream anesthesia. Benzodiazepines, Propofol, Precedex, and halogenated volatile agents rendered it largely obsolete.

But ketamine didn’t fade away quietly. In the ’80s and ’90s it became a popular recreational drug. Known as “Special K,” it infiltrated rave and festival culture and saw widespread abuse among young adults.

Like Viagra—which was originally designed to treat angina—ketamine has now been repurposed. Any objective scientist or logistician might look at its history with humans and think: This is probably a bad idea. But in science, enthusiasm for new tools is often blinding.

A 2023 study showed that patients with treatment-refractory depression who received ketamine infusions twice weekly for six weeks had outcomes non-inferior to electroconvulsive therapy (ECT). More than 50% of participants experienced at least a 50% reduction in symptoms. However, the study was relatively small (around 200 participants per arm), short in duration (six weeks), and had limited long-term safety data.

For those unfamiliar, non-inferiority trials are used when the goal is to reduce harm or patient burden. ECT is not particularly harmful, but it is burdensome and intimidating.

This study appeared to solve a problem in an area of enormous need—and ketamine clinics rapidly sprang up, often charging cash for treatments.

Notably, the above mentioned study followed a 2022 meta-analysis that favored ECT over ketamine for reducing depressive symptoms. However, this was an amalgamation of small, low-quality studies. Hence, the controversy.

Meanwhile, studies dating back to the 1980s showed that frequent recreational ketamine users developed significant problems: hallucinations, delusions, social withdrawal, and a variety of other psychiatric complications.

Still, the need for effective depression treatment is massive, and ketamine appears to offer a possible strategy—so naturally, it’s gaining traction.

Approximately 70 million people in the U.S. have been diagnosed with major depressive disorder. Roughly 30% of them experience a protracted course—treatment-resistant or refractory depression, which is a staggering number.

Treatment-resistant depression refers to depressive episodes unresponsive to at least two medications or failure of cognitive behavioral therapy. Chronic depression refers to symptoms lasting more than two years.

In the trial, patients received ketamine twice per week. Patients tend to be depressed most, if not all, days. So one must ask themselves, what are we to do with the remaining days of the week? This is an area where long-term risk emerges. If some is good, many will assume more is better. Conveniently—or problematically—the study administered follow-up surveys three days after infusion, making the results more susceptible to recency bias, since ketamine has a known euphoric effect.

In my clinical experience, very few patients with chronic depression respond fully within six weeks and then require no further treatment. Long-term, this is concerning. Although some were better, how many need to continue the treatment to keep symptoms at bay? How many will become tolerant, needing more frequent doses? Here in lies a problem in the same vein as above.

If the dots are beginning to connect, you can see where this train is headed—and it’s picking up momentum.

Another concerning issue is what happens when released into the wild. It’s difficult to predict behavior in uncontrolled environments. Clinical studies are controlled: patients behave according to protocol, and physicians follow strict guidelines. But in the real world, practices vary. Lines blur—and often cross—in the name of altruism or intellectual openness. If a patient is suffering and I can alleviate that suffering, why shouldn’t I? That mindset quickly moves from twice weekly, to daily, then multiple times per day, which is exactly what happened to Matthew Perry and Jim Irsay.

Depression and pain are similar syndromes, and often overlap. The psychiatric elements of both tend to fuel abuse. Also, patients with mental illness are more susceptible to addiction. Similar to the study, opioids were designed for short-term use, but pain often persists indefinitely. Stopping treatment becomes difficult due to mental and chemical dependency. Monthly refills become routine. The incentives—economic and emotional and medical—encourage continuation. One can see this can be a great business model, all in the name of relieving suffering (with plenty of nefarious actors along the way), yet it resulted in more deaths and destroyed lives than the Vietnam War.

Which brings us to the lesson of the day: when you, as a patient, engage with the healthcare system, the goal is to avoid certain ruin—in other words, to survive. There is the risk of your illness weighed against the known and unknown risks of the intention to treat the illness. Example: If your aorta ruptures, and death is minutes away, you accept any surgical risk, because death is at your doorstep. But if you sprain your knee and someone offers you a knee replacement, you’re likely to try Advil and physical therapy first.

Mental health is trickier. Knowing ketamine’s risk profile, if I were actively suicidal and had access to a firearm, or some rope and a banister, I might choose ketamine to break the crisis. But outside that extreme, What are the alternatives? I don’t say that to make light of suicide. I say it because ketamine carries real danger. It has always carried danger. Not for everyone, but for someone. And you never know if that someone is you.

Who that “someone” is emerges randomly. It’s noisy. The cause and effect aren’t always obvious because the harm isn’t immediate.

The best way to understand this is through ergodicity. Ergodicity is a mathematical concept describing how an entity in a dynamic, random environment will eventually visit all possible states. If some states guarantee harm or elimination, the process is non-ergodic. If the states are mostly beneficial with negligible downside, it’s ergodic.

To illustrate: Imagine entering a room filled with chairs. You must sit in one. One-third of the chairs award you a million dollars. One chair is rigged to explode and kill you. After you sit, if nothing happens, you must leave, and the chairs are rearranged. Would you enter the room? What about for $10 million? Or $100 million?

Even if you wouldn’t, someone will. That's the beauty of being in the wild. We don't share the same value structure. Variety is inherent when there is room for decision. Winners tend to return, thinking they’ve found a path to riches—until they choose the chair that blows them up.

This is a non-ergodic system. Even though many chairs offer rewards, the one that guarantees ruin makes the entire game un-winnable over time. If you never enter the room, you guarantee survival and can pursue wealth some other way.

So the question becomes: Do we, as physicians, really understand the downside? Would patients like Matthew Perry and Jim Irsay take the risk knowing the complexity of this treatment? One question can't be measured, the other, we'll never know.