Stanford, California (TFC) – Stanford University’s most popular researcher in biology and neurology, Dr. Robert Sapolsky, wrote a column about a recent plane crash, and a lot of people didn’t like it. There was too much about pilot Andreas Lubitz’s mood disorder, and not enough about his personal responsibility. I didn’t like it either, but for a different reason.
There is more to the story than Lubitz’s major depression and personal problems. Dr. Sapolsky assigns blame to the pilot’s brain, and I do too. But his brain did not act alone. Lubitz had been using, for depression and perhaps some other ailments, psychotropic drugs.
Though we do not know what his ailments were, we know that Lubitz had sought psychiatric care. He might have been using psychotropic drugs since 2009, possibly accumulating neurologic and other side-effects for 6 years. At the time of his last flight there were a large number of pill bottles in his home, containing lorazepam, an anti-anxiety agent, and Valdoxan, a new type of antidepressant. He had received an injection of olanzapine, an antipsychotic used to quell mania or psychosis, in 2010. Why didn’t these drugs make him well?
Doctors prescribe drugs that have been approved by a government agency. Approvals are based on the results of clinical trials. Those trials must demonstrate that the health benefits of a new drug outweigh its known risks in order to gain approval. In addition, the proposed labeling is submitted, and must be approved. The goal is to approve only those drugs whose benefits outweigh their risks. 150 people may have died because Andreas Lubitz’s drugs were neither beneficial or safe.
Beneficial? Clinical trials and meta-analyses typically show unremarkable differences between the therapeutic value of antidepressants and placebo. The findings of a 2008 meta-analysis of clinical trial data for four drugs (one of which was discontinued because of liver toxicity) were grim, and this is not an unusual finding:
…there was virtually no difference in the improvement scores for drug and placebo in patients with moderate depression and only a small and clinically insignificant difference among patients with very severe depression. The difference in improvement between the antidepressant and placebo reached clinical significance, however, in […] the most severely depressed patients.
Valdoxan’s efficacy is in question, too:
We found evidence suggesting that a clinically important difference between agomelatine and placebo in patients with unipolar major depression is unlikely. There was evidence of substantial publication bias.
Safe? The FDA provides guidelines for the handling of adverse events that occur during clinical trials. Clinical trials are meant to find out what these are, and whether they are acceptable given the severity of the conditions the drug is meant to treat. If the researchers believe there is a causal relationship between the drug and adverse event, it must be recorded and reported along with information including the frequency of the event, and any known ways to predict, avoid, and manage it.
Clinical trials are meant to keep patients safe, but drugs have gone to market that are unacceptably dangerous, and this possibility is built into the FDA’s guidelines, which allow the researchers discretion in some aspects of testing and recording, discretion that can be abused. The guidelines allow for under-reporting of adverse events, because causality of adverse events can be a matter of opinion. Adverse events can be wrongly ascribed to the condition being treated, and this is especially so with psychiatric drugs tested on psychiatric patients, whose self-reports are not always believed.
In trials of drugs intended for long term use, a six-month trial is considered adequate. After a drug trial ends, the FDA recommends, but does not require, following drug-exposed participants for twelve months. There is no recommendation to follow placebo-exposed participants. Thus, the health of psychiatric patients who were on the drug, and discontinued it at the end of the trial, cannot be compared to the health of psychiatric patients who were not drug-exposed. This means that any discontinuation symptoms in the drug-exposed group will be labeled as relapse or recurrence; psychiatric patients on placebo, who will not experience discontinuation symptoms, are not available as a comparison.
Drugs prescribed to Lubitz have had troubling psychiatric and long-term side-effects, including these:
Lorazepam: blurred vision, seeing, hearing, or feeling things that are not there, aggressive, angry, attack, assault, force, false or unusual sense of well-being, excitation.
Olanzapine: blurred vision, changes in vision, delusions, manic reaction, somnolence, delirium, dementia, depersonalization, hallucinations, abnormal dreams, agitation, nervousness, and hostility.
Valdoxan (agomelatine): blurred vision, suicidal thoughts or behavior, irritability, restlessness, aggressive behavior, nightmares, abnormal dreams, mania, hallucinations.
These three excerpts from online support groups, all pertaining to Valdoxan, which was touted as aiding sleep with few side effects, bring the lists to life:
In addition to immediate adverse events, psychiatric drugs can also cause disabling physical and mental symptoms, distinct from relapse, after they are discontinued. Unlike withdrawal from heroin or methamphetamine, discontinuation is not accompanied by cravings. Nonetheless, the brain and the body react to discontinuation.
A distressing restlessness called akathisia is among the most difficult discontinuation effects to endure. The patient cannot stop moving. Pacing, stamping the feet, and rocking, characterize the condition; the agitation is internal and external. Dizziness, headache, nausea, fatigue, irritability, mania, and even psychosis can occur. The suffering is so great that suicide is sometimes chosen rather than enduring it. Because clinical trials do not reliably record and report discontinuation symptoms, drugs can be approved and sold without any mention of them. (The following account links to the longer story from which it was excerpted.)
Dr. Sapolsky wrote that untreated major depression is one of the most life-threatening diseases on Earth, perhaps hoping to ease misgivings about antidepressants and encourage depressed people to get treated, but the statement is misleading, because it implies that treatment for depression reduces suicide risk. If the treatment includes antidepressants, it may not. A study of suicide and attempted suicides recorded in the FDA database showed that rates did not differ significantly among placebo- and drug-treated individuals. A meta-analysis reported twice the rate of suicidality (thoughts and behavior, including attempts) in the drug group compared to the placebo group. In any event, 1999 study found that the lifetime risk of suicide in major depression may be as low as 3.4 percent. Another study found that antidepressant use does not reduce it.
The meta-analysis that reported double the rate of suicidality in the drug group compared to the placebo group might have missed a few cases. The researchers searched the database of patient information for the following terms, but not for “kill.” They searched for “gun,” but not for “weapon,” “razor-,” “knife,” or “machete.”
“accident-”, “attempt”, “burn”, “cut”, “drown”, “gas”, “gun”, “hang”, “hung”, “immolat”, “injur-”, “jump”, “monoxide”, “mutilat-”, “overdos-”, “self damag-”, “self harm”, “self inflict”, “self injur-”, “shoot”, “slash”, “suic-”, “poison”, “asphyxiation”, “suffocation”, “firearm”
Newer data would be valuable, but not likely to be different, because newer antidepressants are not more effective than older ones. Given side-effects and discontinuation symptoms, and the availability of non-drug therapies, it is arguably a disservice to urge depressed people to seek treatment with antidepressants.
Treatment aimed at preventing mass-murder may also be ill-advised. The behavioral side effects of psychotropic drugs include include violence. Further, there is reason to believe that since the advent of psychotropic drugs, mass-murder is almost always committed by people who are on, or just off, them, although the claim has only been supported by compilations of anecdotes, thus far.
Those who have not seen or experienced extreme adverse events can be forgiven for not, at first, believing that drugs that can do so much harm or that increase suicidality would still be in wide use. News articles are not always accurate. The week after the crash, Dr. Richard Friedman answered readers’ questions about antidepressants for the New York Times, misstating the findings of an FDA-sponsored meta-analysis meant to discover the suicide risks of the drugs. He acknowledged that suicidal thinking and behavior in the drug groups occurred at twice the rate of that in the placebo groups, and added “none of the suicide attempts documented in the clinical trials were fatal.” In fact, eight suicides had occurred (five in drug groups, three in placebo groups).
Media reports of drug effectiveness can be overly positive, if they cover results from studies that compare one drug to another. Self-reported response-to-medication data from clinical trials that compare one or more antidepressants, without any placebo group, may be inflated. Participants in those studies report improvement at nearly twice the rate of those in drug/placebo comparison trials, presumably because they are certain they are getting a drug, not a placebo, and thus more likely to perceive and report improvement.
With limited information, each of us can only form an opinion on why Lubitz committed mass-murder and killed himself. Some say it was evil. Dr. Sapolsky says it was incapacitating depression, while noting correctly that it is “immensely rare for depression to result in violence to others.”
I say it was the drugs.
There have already been enough lives ruined by unpredictable interactions between psychotropic drugs and individual attributes like age and genetics. That adverse events are said to be rare does not make them acceptable. To keep another Lubitz from the cockpit, the indications for psychotropic drugs should be limited, and we must learn to predict which patients will have a serious adverse reaction to a given drug. Genetic research has associated psychotropic-drug-induced akathisia and violence, including murder, with genes in the CYP450 family. Despite the cost and inconvenience, testing for these genes should be required before psychotropic drugs can be prescribed, and more tests of susceptibility to serious adverse events like akathisia should be developed quickly. We have to raise awareness of, and develop treatments for, psychotropic-induced akathisia, which occurs as often as 65% of the time with some psychotropic drugs. All studies of psychotropic drugs should require long-term follow-up of drug-exposed and placebo-exposed psychiatric patients, in order to document disabling long-term discontinuation symptoms that are currently under-acknowledged.
A few years ago, Dr. Sapolsky was making waves for his work on a stress vaccine that would use modified herpes simplex virus to bypass the blood-brain barrier and deliver ‘neuroprotective’ genes deep into the brain. Is another unnatural foray into natural brains something we need, or does it just sound cool?
Author: Caroline Collins, PhD