Botanical QT Risk Starts In The Extract

By Odette Hauke, Odette Alina LLC

Ask a development team what their compound's proarrhythmic liability is and you get one number. For a plant-derived candidate that answer is wrong before anyone runs a study, because the liability belongs to the preparation rather than to the molecule named on the label.

The iboga alkaloids demonstrate this more plainly than anything else I have found in the published literature, and the numbers deserve to travel outside that field. Measuring human ether-à-go-go-related gene (hERG) blockade by whole-cell patch clamp in human embryonic kidney (HEK) 293 cells, one group reported half-maximal inhibitory concentrations (IC50) of 4.09 µM for ibogaine made by semisynthesis via voacangine, 3.53 µM for ibogaine extracted from Tabernanthe iboga, 2.86 µM for the principal metabolite noribogaine, and 2.25 µM for voacangine itself.¹

Read that list with a chemistry, manufacturing, and controls (CMC) eye. The same named active carries a different hERG potency depending on how someone made it. The metabolite exceeds the parent. And voacangine, both a plant congener and the semisynthetic precursor, tops the set, which turns residual precursor into a cardiac specification rather than an impurity one.

What The Data Show

The same study measured 18-methoxycoronaridine (18-MC), an iboga analogue from rational synthesis, above 50 µM.¹ That is a more than tenfold drop in hERG blockade within a single structural family. The mechanism is what makes it instructive: binding affinities across the whole set, 18-MC included, clustered in a narrow 0.71 to 3.89 µM band.¹ So 18-MC occupies the channel about as readily as the others and simply blocks it far less. A screening cascade built on affinity would have ranked these compounds as equivalent.

The clinic bears the in vitro signal out, at magnitudes that are difficult to overstate. In an open-label observational study, 14 patients with opioid use disorder received a single 10 mg/kg dose of ibogaine hydrochloride and underwent monitoring at regular intervals for at least 24 hours.² Maximum QTc prolongation by Fridericia averaged 95 ms, range 29 to 146 ms.² Half the subjects passed a QTc of 500 ms, and in six of 14, prolongation above 450 ms outlasted 24 hours.² ICH E14 governs how sponsors evaluate QT prolongation for non-antiarrhythmic drugs, and a mean effect approaching 100 ms is not a borderline finding under it. It is the central safety problem of a program.³ Case reports link ibogaine to torsades de pointes and fatal arrhythmia at doses given with therapeutic intent, in people with no pre-existing cardiac disease.⁴ No regulator has approved ibogaine for any indication, so there is no labeled dose. The 10 mg/kg above is the dose used in the study that produced the 95 ms mean.²

Eight Things To Build Into The Package

1. Screen the congeners, not only the labeled active. Identify every constituent your extract carries above a meaningful threshold and run hERG on each. The iboga data show a precursor out potentiating the active it precedes, which no assay of the labeled compound alone would reveal. Include the metabolites in that panel, and judge each result against the exposure your clinical dose will produce rather than against the parent's IC50. A congener present at a low percentage can still dominate the liability if it is potent enough and your extraction concentrates it. Budget for a panel rather than a single curve.

2. Fix the manufacturing route early and own its safety consequence. Extraction and semisynthesis produced measurably different hERG potencies for the same named compound.¹ Treat route selection as a safety-relevant CMC decision made in early development, not a cost decision deferred to scale-up. If you change route later, assume you owe a bridging cardiac assessment, and plan the comparison before you need it.

3. Write the drug-extract ratio and the solvent into your safety rationale. European product definitions already carry both. Filsuvez, the birch bark triterpene gel that holds marketing authorizations in both the European Union and the United States, specifies a drug-extract ratio of 5-10:1 with n-heptane 95% w/w as extraction solvent.⁵ Those two parameters determine which congeners you have and in what proportion, which determines your cardiac profile. Teams that treat them as quality-only attributes discover the connection during review.

4. Let the metabolite set your monitoring window. Investigators attribute QT prolongation observed days after ibogaine ingestion to noribogaine's extended half-life.¹ Characterize metabolites for channel activity, then build the observation period around the longest-lived active species rather than the parent. A standard 24-hour Holter sized to parent pharmacokinetics will miss the tail, and a missed tail becomes a post-marketing commitment.

5. Genotype, or document why you did not. Population pharmacokinetic modeling in opioid use disorder patients found ibogaine clearance strongly dependent on CYP2D6 genotype: 0.82 L/h at an activity score of zero, rising 30.7 L/h per additional point, with a sigmoid concentration-QTc relationship.⁶ A fixed milligram-per-kilogram dose therefore delivers very different exposures, and a nonlinear exposure-response converts that spread into a safety problem rather than a potency one. Where a polymorphic enzyme dominates clearance, stratify prospectively. Retrofitting genotype onto a finished dataset rarely rescues it.

6. Treat a proposed mitigation as a hypothesis until a controlled trial closes. Investigators have proposed magnesium coadministration to reduce arrhythmia risk. The clinical dataset behind that proposal is an open-label observational study of 30 veterans at Stanford University, which reported no unexpected or serious adverse events and whose authors called for controlled trials to validate their open-label findings.^7,8 That is a reasonable thing to study and an unreasonable thing to assume. Do not let a mitigation that rests on an uncontrolled cohort shrink the monitoring you plan.

7. Do not ask ethnobotanical history to carry an electrophysiological claim. A delayed, dose-dependent effect on cardiac repolarization is invisible without instrumentation, so centuries of use establish nothing about it either way. This is not a reason to discount traditional use. It is a reason to be exact about which hazards it speaks to, which are the ones a community could see, and to generate the cardiac data separately.

8. File what you learned before Nov. 3. The FDA has a docket open on botanical drug development that asks which study designs are scientifically credible for complex botanical mixtures, and what makes minimally purified products harder than highly purified ones.⁹ A worked example of preparation-dependent channel liability answers both questions better than any general statement about heterogeneity. Submit the specifics: which congeners you screened, which assay you used, and how large of a difference you measured between preparations. If the detail is commercially sensitive, file the shape of the problem publicly and the numbers through the confidential route rather than withholding both.

What This Does And Does Not Establish

One alkaloid family cannot prove a rule for all botanicals, and I am not claiming it does. What the iboga data establish is narrower and still consequential: where a botanical substance contains structurally related congeners, the named active does not determine the channel liability of the mixture, and the manufacturing route can change it measurably.

Whether that pattern recurs elsewhere is an empirical question nobody has answered systematically, which is itself worth telling the FDA. The mechanism follows from what a botanical substance is: composition depends on species, plant part, extraction solvent and ratio, and manufacturing route. The labeled active contributes one part of the pharmacology among several. The consequence for a development plan is uncomfortable and simple. Your safety package attaches to the preparation you will actually market, so a change in sourcing or route changes your cardiac dataset. Teams that establish this early build one package. Teams that find out later build two.

References

1. hERG Blockade by Iboga Alkaloids. Cardiovasc Toxicol, 2016. doi:10.1007/s12012-015-9311-5
2. Safety of ibogaine administration in detoxification of opioid-dependent individuals: a descriptive open-label observational study. Addiction, 2021. doi:10.1111/add.15448
3. Food and Drug Administration. E14 Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs, Questions and Answers (R3), guidance for industry, June 2017. Docket No. FDA-2004-D-0241.
4. Rare but relevant: Ibogaine and cardiovascular complications, prolonged QT interval and ventricular arrhythmias. Addiction, 2026. doi:10.1111/add.70319. See also Anti-addiction drug ibogaine inhibits hERG channels: a cardiac arrhythmia risk. Addict Biol, 2012. doi:10.1111/j.1369-1600.2012.00447.x
5. European Medicines Agency, Filsuvez European public assessment report. https://www.ema.europa.eu/en/medicines/human/EPAR/filsuvez
6. The pharmacokinetics and pharmacodynamics of ibogaine in opioid use disorder patients. J Psychopharmacol, 2024. doi:10.1177/02698811241237873
7. Magnesium-ibogaine therapy in veterans with traumatic brain injuries. Nat Med, 2024. doi:10.1038/s41591-023-02705-w
8. ClinicalTrials.gov NCT04313712, Pre-post Evaluation of the Safety and Efficacy of Ibogaine-Magnesium Therapy in Veterans With Repeated Blast Exposure, Stanford University, actual enrollment 30, completed Sept. 13, 2023; registry record updated Oct. 27, 2026. Reference 7 is the published report of this study.
9. Food and Drug Administration. "Advancing Development of Botanical Drug Products; Request for Information." Docket No. FDA-2026-N-9550. 91 FR 56887, Sept. 4, 2026. Comments close Nov. 3, 2026. https://www.federalregister.gov/documents/2026/09/04/2026-18133/advancing-development-of-botanical-drug-products-request-for-information

Note: Doses and clinical findings cited here come from published studies and appear as development evidence. Nothing in this article constitutes clinical guidance.

About The Author

Odette Alina Hauke is a regulatory strategist and epidemiologist and the founder of Odette Alina LLC, where she advises sponsors on global regulatory strategy and clinical development. She has served as regulatory lead for three first-in-class Schedule I programs, among them ibogaine, and researches ibogaine's cardiac safety, including starting-dose justification, CYP2D6 pharmacogenomics, and exposure-response. She testified at FDA's September 2026 public hearing on the therapeutic use of psychedelic drugs, cardiac risk stratification, and pre-treatment screening. She holds an MS in regulatory affairs from Northeastern University, a BS in epidemiology from the CUNY Baccalaureate for Unique and Interdisciplinary Studies at Hunter College.