Primum Non Nocere: The Onus to Characterize the Potential Harms of Psychedelic Treatment

Hallucinogen persisting perception disorder (HPPD) refers to reexperiencing perceptual disturbances experienced while intoxicated with a hallucinogen, i.e., psychedelic, when no longer under the influence, resulting in impairment (7). These can include geometric hallucinations, false perception(s) of movement in peripheral vision, light flashes, color intensification, visual trails from moving objects, halos around objects, and distortions to object size. HPPD can present as brief “flashbacks” or persist in a chronic form (8). Distress with the former often presents as anxiety and panic attacks, as well as depressive features, while the latter is associated with suicidal ideation and suicide (9). Unfortunately, the current data do not allow us to estimate the prevalence of HPPD (8, 10) as it consists largely of case reports following the use of lysergic acid diethylamide (LSD), psilocybin, 3,4-methylenedioxymethamphetamine (MDMA), 5-methoxy-N, N-Dismethylryptamine (5-MeO-DMT), and ibogaine (9). As a result, there is little understanding of the risk factors that might predispose to HPPD or what effect dosing or number of uses have on developing HPPD (9). Notably, despite the numbers of people who have tried psychedelics, the fact that large case series of HPPD have not emerged provides some reassurance that it is likely rare, though in the absence of systematic surveillance, the true incidence and prevalence remain unknown (10).

Major clinical trials of psychedelics, namely MDMA-assisted psychotherapy for posttraumatic stress disorder (PTSD) (11, 12) and psilocybin for treatment resistant depression (TRD) (13), major depression (14), and alcohol use disorder (15), have excluded patients with a history of mania or psychosis, and even a family history of mania or psychosis. This was prudent in early development, given the accounts of manic and psychotic reactions to psychedelics in the first wave of research on psychedelics (16, 17). More recently, a cross-sectional study examining the association of manic and psychotic symptoms with psychedelic use in a sample of twin adolescents (the Swedish Twin Registry) found that in those with a genetic vulnerability to schizophrenia or bipolar I disorder there was an association between psychedelic use and self-reported manic symptoms, but not self-reported psychotic symptoms (18). Interestingly, in those with a genetic vulnerability to schizophrenia or bipolar I disorder, psychedelic use was associated with parent-reported psychotic symptoms. In the event that studies do show a higher risk of manic and psychotic symptoms in those with a family history of bipolar disorder or psychotic disorders, research will also be needed on whether there are ways to mitigate personal and family history as risk factors for those whom psychedelic compounds might provide benefit over other treatment options. Notably, there has been a small open-label pilot study of psilocybin in 15 patients with treatment resistant type II bipolar depression, with 12 of 15 meeting criteria for treatment response at week 3, with no increase in manic or psychotic symptoms up to 12 weeks out (19).

Finally, there have been recent case reports of mania and persistent psychosis following the use of psychedelics recreationally (19–21) after just a single dose of MDMA and two doses of psilocybin (22) in individuals with no history of hypomania/mania or psychotic symptoms, suggesting there is a need to further characterize the risk factors for these harms, even in the absence of a personal or family history of bipolar disorder or psychosis. Arguably, although the recent case reports cited were of individuals using psychedelic compounds outside of clinical trial settings, they nonetheless signal potential harms that may be more of an issue with widespread use following FDA approval. It is not uncommon for rare, but significant, adverse events caused by drugs to become apparent only with wider use in the post marketing phase (23).

Early animal studies of self-administration of MDMA suggest that, while not as potent as some drugs of abuse, such as cocaine, MDMA does have the potential for abuse and dependence (24–26). In line with that, there are case reports of MDMA dependence (27). Studies of recreational MDMA users have shown that between 17% and 70% of users, depending on the study population, meet criteria for dependence (28). Notably, there was little evidence of physical symptoms or withdrawal; behavioral symptoms such as financial, relationship, and social problems were more characteristic of MDMA dependence (28, 29).

In the case of classic psychedelics, including LSD, psilocybin, and mescaline, there is evidence of rapid tolerance or tachyphylaxis (30), making physiological dependence unlikely. Consistent with tachyphylaxis, self-administration studies in animals with classic psychedelics have generally been unsuccessful even with extensive training, suggesting low abuse potential (31). However, given the reversal of tachyphylaxis on the order of days (5 days for LSD), abuse of classic psychedelics remains possible (30). Hallucinogen use disorder (HUD) requires hallucinogen use meeting at least two of the 11 criteria laid out in the DSM for substance use disorders: taking the substance in larger amounts or for longer than planned; wanting to cut down or stop using the substance but not managing to; spending a lot of time getting, using, or recovering from use of the substance; cravings and urges to use the substance; not managing to do what one should at work, home, or school because of substance use; continuing to use, even when it causes problems in relationships; giving up important social, occupational, or recreational activities because of substance use; using substances again and again, even when it puts one in danger; continuing to use, even when one knows one has a physical or psychological problem that could have been caused or made worse by the substance; needing more of the substance to get the same effect (tolerance); development of withdrawal symptoms, which can be relieved by taking more of the substance (7). Data from the National Epidemiologic Study on Alcohol and Related Conditions-III (N=36,309), which examined 12 months and lifetime use of hallucinogens (LSD, peyote, mescaline, psilocybin, anticholinergics, N, N-Dimethyltryptamine [DMT], 2,5-Dimethoxy-4-methylamphetamine [DOM], Dimethoxybromoamphetamine [DOB], Salvia divinorum, dextromethorphan, and phencyclidine), found twelve-month and lifetime prevalence in the general population of HUD to be 0.05% and 0.6% respectively (32). Among hallucinogen users, 6.4% and 8.1% developed lifetime and 12-month HUD, respectively (32).

While descriptions of positive and profound meaningful experiences with psychedelic treatments abound and indeed dominate the popular narrative, the actual experience of psychedelic-assisted psychotherapy is not always so beatific. As one might expect during any course of therapy, there are difficult moments with accompanying unease as well. Coupled with the unique effects of psychedelics, ranging from perceptual disturbances to a loss of sense of self, difficult moments can rise to the level of harmful events when there is marked distress, especially if it leads to distress that persists as anxiety, derealization, or at an extreme, existential distress, accompanied by dread or even nihilism. In the current lexicon, both expected transient episodes of difficulty or unease, as well as marked and persisting distress during and after a psychedelic experience, have been grouped together under the term “challenging experiences,” previously known as bad trips” (33). Avoidance of all discomfort is undesirable (34, 35) since discomfort can signal difficult issues or unpleasant content coming to the forefront that would be beneficial to address and process and difficult moments can precede psychological breakthroughs and growth. In line with this, in an online survey of 1,993 individuals who have had a challenging experience with psilocybin in settings outside of clinical trials, 84% endorsed benefit from the experience (36). However, persistent and marked distress is undesirable and indeed harmful.

Unfortunately, there are concerning reports of such marked and persistent distress, a negative corollary to the more often discussed “afterglow,” in the days to weeks following psychedelic use. Whereas the afterglow experience is characterized by positive mood states, increased wellbeing, and potentially beneficial changes in perceptions of self, others, and the environment (37), persistent distress lasting days to weeks afterward ranges from new onset anxiety to unsettling nihilism (38). In a recent survey of 613 individuals with a lifetime history of using classic psychedelics, 8.9% of respondents reported functional impairment lasting longer than a day, and 2.6% reported seeking medical, psychiatric, or psychological assistance in the days or weeks after a distressing psychedelic experience (39). This raises the question of what factors predispose to distressing and harmful experiences with psychedelics. In the aforementioned survey of individuals with a lifetime history of using classic psychedelics, respondents reported co-use of at least one psychiatric medication (39). Co-use of lithium and other mood stabilizers was found to be associated with the degree of difficulty during respondents’ most challenging classic psychedelic experience and higher odds of overall risk of harm.

There has also been an emphasis on the potential role of an individual’s mindset going into a psychedelic treatment session and the importance of preparation, as well as an emphasis on the potential role of setting, or the environment and people present during a psychedelic administration (33, 36, 40). Indeed, respondents in the survey study of individuals reporting lifetime use of classic psychedelic use also reported six set and setting variables associated with their most challenging, difficult, or distressing classic psychedelic experience: no preparation; negative mindset; no psychological support; disagreeable social environment; disagreeable physical environment; and major life event prior to experience (39). Negative mindset, no psychological support, and major life events prior to experience were associated with the degree of difficulty during the respondents’ most challenging psychedelic experience and higher odds of overall risk of harm. However, notably, even in carefully controlled clinical trials, with careful attention to set and setting, there are case reports of harmful, challenging experiences (38), suggesting the need for attention to risk factors beyond “set and setting” (34). Recent research suggests trait neuroticism may be a potential risk factor for challenging experiences (41, 42), and as is well known, trait neuroticism is prevalent in the same psychiatric populations (43) for which psychedelic treatments are being developed. Research also suggests that challenging experiences are more likely with high doses of psychedelics (36, 42), which are the very doses that have been used in promising clinical trials to date (13, 44). While we learn more about challenging experiences and the potential for harm, we need to maintain a more nuanced portrayal of psychedelic experiences, one that is inclusive of difficult moments during the experience and the potential for an altogether distressing and harmful experience.

Preclinical research in animals shows accumulating evidence for an increased capacity for structural and functional neuronal plasticity (45, 46) and even the reopening of critical periods of development (47, 48) with psychedelics. Provided the results are found to be applicable in humans, this speaks to the potential for great benefit as well as harm. Increased capacity for plasticity, or change, does not direct the actual course of the change itself but instead is highly context-dependent and increases the influence of the environment to shape outcomes (49). Given the challenging environments in which our patients sometimes find themselves, whether at home, in their immediate partnerships and families, at work, or in other environments such as the battlefield, plasticity can translate to greater impact or influence of the stressful and harmful elements of patients’ environments, which might lead to further harm. Consistent with this, we should be mindful of the environment patients return to following psychedelic treatments, as these periods of increased capacity for plasticity could potentially confer vulnerability as well as opportunity.

When used at standard doses in appropriately selected patients under medical supervision and for a limited number of exposures, classic psychedelics are generally safe and well-tolerated from a cardiovascular standpoint (50–53). Available evidence from clinical studies, as well as decades of use outside the medical environment, have not identified serious cardiac safety signals or concerns when the drugs are used at standard doses (50–53). However, for regulatory and practical reasons, rigorously collected and adjudicated large-scale clinical trial-based safety data of the type that is typically obtained during new drug development are lacking for both classic and non-classic psychedelics. Additionally, there is a paucity of data in patients with clinically significant cardiovascular disease (e.g., ischemic heart disease with angina pectoris or prior myocardial infarction, valvular heart disease, arrhythmias, hypertension, or congestive heart failure) on which to base safety assessments.

The use of classic psychedelics at standard doses is commonly associated with adrenergically mediated increases in heart rate, blood pressure, and respiratory rate, which are generally mild, transient, and rarely require intervention (54, 55). While the overall cardiac safety of a limited number of exposures to classic psychedelics at standard doses in healthy individuals is well-established, rare serious adverse cardiovascular events have been reported when classic psychedelics, most commonly magic mushrooms (psilocybin) or LSD, are used recreationally at high doses or in combination with other drugs (56–63). These cardiovascular toxicities have included rare cases of myocardial infarction, cardiomyopathy, and cardiac arrest (57–60). In addition, rare cases of intense vasoconstriction leading to limb ischemia and stroke have been reported, most commonly in association with high doses of LSD (61). While LSD is the most potent vasoconstrictor among the classic psychedelics, psilocybin and mescaline have also been shown to exert vasoconstrictive effects.

Electrocardiographic abnormalities attributed to high dose psilocybin, including QT-prolongation and ST-segment elevation, have also been reported (60, 64). Individuals with preexisting comorbidities are likely to be more susceptible to these rare cardiovascular adverse events. As is often the case with toxicities related to substance use, the presence of confounding factors can make it difficult to definitively establish causal links. Nevertheless, given the number of well-documented case reports, it is important to be aware of the potential of these drugs to cause rare but serious adverse cardiovascular events when used at excessively high doses alone or in combination with other drugs, particularly in susceptible individuals.

With a few exceptions, there is very limited data available on the effects of both classic and non-classic psychedelic drugs on cardiac ion channels and the electrocardiogram. Psilocybin has been shown to be a relatively weak hERG blocker at standard therapeutic exposures (65, 66). In a study of 12 healthy adult subjects, psilocybin at a standard therapeutic dose of 25 mg caused a mild degree of QTc prolongation (mean 2.1 ms; 90% upper confidence limit 6.6 ms) that was well below the 10 ms level of regulatory concern (65–68). However, the drug does display a shallow concentration effect relationship, and at higher exposures than those seen with standard clinical doses, psilocybin could cause clinically significant QTc prolongation (66). A study in 23 healthy adult subjects pre-treated with escitalopram prior to psilocybin exposure at a standard dose of 25 mg was not associated with significant QTc prolongation either after escitalopram or following the addition of psilocybin (68). However, since the QTc effects were not assessed at Cmax, the investigators cautioned that significant QTc prolongation may occur in susceptible individuals during peak effects, as evidenced by one subject who exhibited a prolonged QTc of 489 ms after psilocybin and 509 ms after escitalopram and psilocybin (68). These findings suggest that absent additional safety data on specific drug-drug interactions, psilocybin should be used with caution when combined with other drugs known to prolong cardiac repolarization, as well as in patients with preexisting QT prolongation.

Ibogaine, a non-classic psychedelic, in contrast to psilocybin, causes significant QTc interval prolongation at standard therapeutic doses. Preclinical studies on ibogaine have demonstrated significant inhibition of hERG and prolongation of cardiac repolarization at typical therapeutic exposures in the 1–4 µM range (69). Clinical observations have demonstrated significant dose-dependent prolongation of the QTc interval by noribogaine, the active metabolite of ibogaine, with mean effect sizes of 16 ms, 20 ms, and 42 ms at relatively low doses of 60 mg, 120 mg, and 180 mg, respectively (70). Cases of cardiac arrest and sudden death have been reported with the use of ibogaine (71, 72). DMT has been shown in preclinical studies to inhibit cardiac voltage-gated sodium channels via sigma-1 receptor binding (73). To date, there have been no clinical studies published on the ECG effects of DMT in humans; hence, the potential clinical significance of this ion channel effect is unknown.

A recent study of ibogaine in 30 special operations forces veterans with traumatic brain injury treated patients with 1 g of intravenous magnesium sulfate prior to and 12 hours after administration of ibogaine (74). In addition to close monitoring of vital signs, QTc intervals were monitored visually via continuous 5-lead ECG and trained medical staff onsite at all times. While quantitative ECG data were not provided, the investigators reported that there were no instances of bradycardia, tachycardia, or “clinically meaningful, qualitatively detectable” QT prolongation and no significant disturbances in blood pressure during the study (74). Prophylactic treatment with magnesium has been shown to prevent or mitigate QTc interval prolongation associated with ibutilide in patients with atrial fibrillation and normal baseline QTc intervals (75). While the rationale for this risk mitigation strategy is sound, quantitative ECG/QTc data and additional clinical experience with the magnesium-ibogaine combination are needed before this approach can be widely recommended. In addition, pre-treatment with magnesium, if used, should be viewed as complimentary to and not a replacement for careful patient selection with exclusion of high-risk individuals, ECG assessments before and after drug administration, and avoidance of pharmacokinetic and pharmacodynamic interactions.

Many of the classic psychedelics, in addition to their primary effect on the 5-HT_2A receptor, also interact with the 5-HT_2B receptor, which is known to mediate a proliferative effect on heart valves that can lead to an overgrowth valvulopathy arising from mitogenic stimulation of quiescent cardiac fibroblasts resulting in deposition of connective tissue plaques on valve leaflets and chordae (76). Despite this effect, there is no evidence to date that any of the classic psychedelics are causally linked to the development of valvular heart disease. Nevertheless, these agents are not well studied, and while short-term use appears to be safe, the possible risks of repetitive long-term use at standard or non-standard doses are unknown. The non-classic psychedelic MDMA, an amphetamine derivative, and its N-demethylated metabolite 3,4-methylenedioxyamphetamine (MDA) both elicit prolonged mitogenic responses in human valvular interstitial cells via activation of 5-HT_2B receptors (77). In a small clinical study of 29 chronic MDMA users, typical valve lesions were detected in 28% of MDMA users compared to 0% of gender and age-matched control subjects (78). Additional clinical studies are needed to better define the extent and significance of potential valvular heart disease liability with the chronic use of MDMA. While the absence of data linking any of the classic psychedelics to the occurrence of valvular heart disease is reassuring, cardiac safety data on the chronic use of these agents both at standard and non-standard doses (e.g., prolonged microdosing) are lacking, and further studies are warranted as the use of these agents expands.

Notably, while the use of classic psychedelics for a limited number of exposures at standard doses in healthy individuals is safe from a cardiovascular standpoint, safety data are lacking in several important areas. For patients with underlying cardiovascular disease, among whom anxiety and depression are widely prevalent and who might benefit from psychedelic therapy, clinical experience and safety data are essentially absent. Once initial psychedelic drug approvals are achieved and clinical trials expand to include patients with underlying cardiovascular comorbidities, rigorous heart rate, blood pressure, and ECG monitoring should be carried out as a routine part of these studies. In addition to standard vital signs and appropriately timed ECG interval and PK measurements, baseline and follow up echocardiograms to assess cardiac valve structure and function and pulmonary pressures should be incorporated for drugs that display functional activity at the 5-HT_2B receptor (79). Another area in which echocardiographic data are needed as a matter of public safety is the repetitive, long-term use of psychedelics, including microdosing. In addition, data on clinically important drug-drug interactions, particularly with commonly used psychoactive and cardiac drugs, many of which carry potential cardiac liability in terms of QTc or QRS interval prolongation, is needed. Finally, as second and third generation psychedelics enter preclinical and early phase clinical testing, development programs should include cardiac ion channel profiling as well as standard ECG safety studies and, where indicated based on affinity for the 5-HT_2B receptor, assessments in relevant preclinical models and clinical echocardiographic studies to evaluate potential liability for cardiac valve disease.

In highlighting the known and potential harms of psychedelic treatments, our aim is not to detract from the potential or promise of these treatments but rather to urge maintaining the dialectic that these treatments, like nearly all therapeutic interventions, hold potential for both healing and harm. This awareness serves to temper the hype around psychedelics that can lead to unrealistic expectations of a panacea or cure. Indeed, those who expect psychedelics to be their last hope may be at risk of demoralization and despair if they do not respond to a psychedelic treatment. This caution also addresses the false dichotomy that current treatments cause harm, whereas psychedelics are natural and harmless. At its essence, the dictum primum non nocere serves to remind physicians that all therapeutic interventions carry risk and reinforces the need for humility about our work and our treatments. We are reminded to avoid being blinded by our desire to do good or the promise of healing with novel treatments. Just as the early demonstrated potential of psychedelic treatments places a duty on us to explore their potential in the service of our patients, their potential harms place a similar responsibility on us to conduct rigorous research on the risks of psychedelic treatments and continue to monitor adverse outcomes as they gain FDA approval and become more widely available.

The authors acknowledge Sarah G. Richter and Charlotte Schiewe for assistance with manuscript preparation. Dr. Ghaznavi acknowledges support from the Arielle Soussan Memorial Fund for Psychedelic Research and the Goldsmith and Malievskaia Family Fund for Clinical Neuroscience Research.