Safety of ibogaine administration in detoxification of opioid dependent individuals: a descriptive open-label observational study

Knuijver and colleagues situate ibogaine as the principal indole alkaloid of Tabernanthe iboga rootbark, traditionally used as an entheogen and increasingly used outside medical settings to treat substance use disorders. Earlier clinical case series and small observational studies suggest ibogaine can reduce withdrawal, craving and drug self-administration in humans and animals, but safety concerns have been repeatedly raised. In vitro and clinical reports implicate inhibition of hERG potassium channels, QT interval prolongation on the electrocardiogram and risk of torsades de pointes (TdP). Additional concerns include transient cerebellar ataxia observed in animals and limited human reports, as well as anecdotal accounts of seizures, mania and persistent perceptual disturbances. The study aimed to evaluate the clinical safety of a single administration of ibogaine-HCl in patients with opioid use disorder (OUD) who were receiving opioid substitution therapy (OST) and seeking detoxification. Specifically, the investigators tested the hypotheses that ibogaine produces reversible QTc prolongation, induces cerebellar ataxia, and causes psychomimetic behavioural effects. The team also collected limited observations of opioid withdrawal to assess potential short-term benefit and durability.

This was a descriptive, open-label observational study conducted between October 2015 and November 2017 at two outpatient addiction clinics. Patients were selected from roughly 500 clinical files; 130 files were reviewed for psychosocial stability and OST status, 36 patients met preliminary eligibility and were approached, 29 were screened and 14 were ultimately included. Inclusion criteria included age 20–60 years, desire for opioid detoxification and prior treatment failure; key exclusions were clinically significant cardiac disease, long QT syndrome or baseline QTc >450ms (men) / >470ms (women), abnormal serum potassium, severe liver or renal dysfunction, and pregnancy. A post hoc power calculation is reported indicating 80% power with 10 patients to detect a 10% QTc increase from baseline under specified assumptions. Participants were admitted and converted from OST to oral morphine sulphate for 8 days to avoid methadone-related QT effects and to standardise baseline pharmacotherapy; the last morphine dose was given four hours before ibogaine administration. Each subject received a single dose of ibogaine-HCl 10 mg/kg. To reduce nausea and ensure ingestion, 20 mg metoclopramide was given before dosing. Electrolytes (K+, Ca2+, Mg2+) were checked and had to be within normal ranges prior to administration. Cardiac monitoring was intensive: 12-lead ECGs were taken every 30 minutes for the first 12 hours, then hourly if QTc remained prolonged (>450ms men, >470ms women) or every four hours if QTc shortened and remained below 500ms; monitoring continued for 24 hours and was extended if a cardiologist advised. The QT interval was corrected for heart rate using Bazett's formula (QTc). QT intervals were measured manually by two independent researchers (averaging leads V5 and II) and automatic QTc values from a Philips TC50 device were also recorded for clinical monitoring. If QTc exceeded 500ms the protocol specified magnesium supplementation: a 2 g bolus over 10 minutes followed by 2 g infused over the next 10 hours, and transfer to a cardiac care unit if necessary. Secondary safety assessments included heart rate and blood pressure (bradycardia defined <60 bpm, systolic hypotension <90 mmHg), cerebellar ataxia measured with the Scale for the Assessment and Rating of Ataxia (SARA), and psychomimetic/delirium signs assessed hourly with the Delirium Observation Screening (DOS) scale (score ≥3 indicates delirium). Opioid withdrawal was monitored with the Clinical Opioid Withdrawal Scale (COWS). Adverse events were recorded. Statistical analysis was descriptive, focusing on the change from baseline to maximum QTc per subject (ΔQTcMax), the proportion of subjects with QTc >500ms at any time, and other safety counts; psychomimetic effects were reported qualitatively because scores were low. Analyses used IBM SPSS 25 and Excel.

Fourteen patients were included. The extracted text reports that participants were psychosocially stable with ASI domain scores mostly 0–1 except for drug use, and that all had histories of polysubstance use. Baseline heart rate, blood pressure and QTc were within normal ranges. Primary cardiac outcomes showed large and variable QTc prolongation after ibogaine. The median maximum change in QTc (ΔQTcMax) was 102 ms. Half of the participants reached a QTc >500 ms at some point during monitoring; at any given measurement the proportion with QTc >500 ms ranged between 7% and 28% across timepoints. After 24 hours, QTc remained >450 ms in 50% of participants. No episodes of torsades de pointes were observed on the recorded ECGs and no clinical signs of TdP were documented. Eight participants received magnesium infusion according to protocol because automatic ECG monitoring flagged QTc >500 ms. The manual QT measurements by the two researchers showed a moderate correlation (r = 0.64) and a mean inter-rater difference of 0.89 ms; the average assessor QTc correlated strongly with the automatic TC50 QTc (r = 0.71) with a mean automatic-minus-manual difference of +7 ms (SD 26 ms). Heart rate and blood pressure decreased mildly during the first 12 hours: transient bradycardia around 50 bpm and reductions in blood pressure were recorded. The only adverse event explicitly noted in the extracted text was vomiting in two patients occurring more than two hours after ingestion; no seizures were observed. Cerebellar effects were consistent: all subjects developed signs of ataxia on the SARA, with scores rising between two and six hours after dosing and returning to full remission within 24–48 hours (five patients had minimal residual scores at 24 hours but were normal 24 hours later). Affected SARA items were primarily gait, stance and heel–shin testing; some participants required nurse support for ambulation. Psychomimetic phenomena were generally mild. DOS scores were zero at baseline; ten participants showed no signs of delirium and four had scores of one to two during treatment. Clinicians observed wakeful dreaming, reliving of memories and closed-eye visuals lasting approximately 3–7 hours; one participant exhibited a brief visual hallucination and another transient misperception of nonexistent objects. Withdrawal scores (COWS) remained low over 24 hours, with only five measurements scoring 1 and most subjects showing little objective withdrawal. Three participants requested resumption of morphine substitution stating subjective treatment failure.

Knuijver and colleagues interpret their findings as evidence that a single 10 mg/kg dose of ibogaine-HCl produces clinically relevant, reversible QTc prolongation in patients with OUD even when administered in a medically supervised setting to individuals without known cardiac disease. The investigators emphasise that half of their small sample exceeded a QTc of 500 ms, a threshold associated with markedly increased risk of adverse cardiac events in overdose populations. They note that bradycardia and hypotension observed here further compound proarrhythmic risk. The authors link the clinical QTc findings to in vitro data implicating ibogaine and its metabolite noribogaine in hERG potassium channel inhibition, leading to delayed cardiac repolarisation. Inter-individual variability in the magnitude and timing of QTc changes is discussed in the context of pharmacokinetic differences, particularly CYP2D6-mediated metabolism to noribogaine; noribogaine's longer half-life (reported here as 28–49 hours) may account for persistent QTc prolongation in some participants. Transient cerebellar ataxia observed in every subject is consistent with prior animal and limited human reports and may reflect Purkinje cell vulnerability; the authors recommend further study, including neuroimaging, to characterise possible neurotoxic effects. Psychomimetic "oneirogenic" experiences were typically short-lived and tolerable in this setting, but the investigators caution that more persistent or severe behavioural disturbances have been reported elsewhere and clinical settings should be prepared to manage such reactions. Several limitations are acknowledged. The sample was small (n = 14) and highly selected: participants were stable, motivated for abstinence and excluded for hepatic or cardiac disease, which limits generalisability to broader populations of people with OUD where comorbidities and concomitant medications (including CYP2D6 inhibitors) are common. ECG monitoring was not continuous so brief arrhythmias could have been missed, and the study was underpowered to detect rare but severe adverse events such as TdP or prolonged psychosis. All participants received metoclopramide, which may have a minor QTc effect, and there was no placebo control for withdrawal assessments. Given the observed cardiac risks and the limited evidence for durable effectiveness, the authors conclude that ibogaine administration should be limited to well-controlled settings with strict cardiac monitoring and that the risk–benefit ratio for routine clinical use is unfavourable based on these findings. They suggest that alternative dosing regimens, such as repeated low dosing, might warrant investigation if dose-related toxicity can be reduced, and call for further research into cerebellar effects and mechanisms underlying any anti-addictive properties.