DXM

Dextromethorphan is sold as a cough suppressant, filed by this wiki among the dissociatives, and its highest-affinity target is neither of those. It is the serotonin transporter, which it inhibits as potently as fluoxetine, at concentrations the brain actually reaches at the doses sought for their effects. The only controlled trial ever run at those doses further finds that subjects, blinded, name a classical hallucinogen and almost never a dissociative. A molecule whose class name, whose box and whose affinity say three different things is a molecule whose risks are hard to read, and that is exactly what makes it dangerous in combination.

Neurochemical action

DXM is a morphinan derivative, the dextrorotatory enantiomer of a skeleton whose laevorotatory enantiomer, levorphanol, is a potent opioid. Chirality does everything: DXM lacks the phenanthrene structure of the classical opioids and acts directly at neither mu, nor kappa, nor delta.

It is O-demethylated by CYP2D6 into dextrorphan, which binds the NMDA receptor with a higher affinity than the parent molecule. Two psychoactive compounds therefore coexist in the same user, in a proportion set by their genotype and their other medications. A secondary pathway, carried by CYP3A4, gives 3-methoxymorphinan.

On the targets, one value has been settled, and it is the foundation of everything else. Four affinity measurements for SERT are in circulation: three fall between 23 and 68 nM across two species and two assay types, the fourth, at n equals 2, is three to eight times lower and must be set aside. The reference measurement is that of Rickli et al. on human SERT, IC50 68 nM, 95% confidence interval 47 to 100, against 92 nM for fluoxetine measured in the same assay, with a SERT to NET ratio of 85. NMDA sits one to two orders of magnitude lower in affinity.

The same source gives the human concentrations, which forbids treating this point as theoretical: plasma Cmax of 0.01 to 1.4 µM at doses of 75 to 100 mg, up to 5.6 µM after doses of up to 400 mg, and a maximum concentration measured in cerebrospinal fluid of 0.44 µM, that is several times the SERT IC50. At a recreational dose, serotonin reuptake inhibition is not a laboratory possibility, it is massive and it is measured on the human transporter. The rest of the profile, sigma-1 agonism, alpha3beta4 nicotinic antagonism, calcium channel blockade, is cited constantly and has never been isolated.

What a dose produces, and what has never been measured

There exists a single controlled clinical trial of recreational doses, Reissig et al. 2012: twelve healthy volunteers, single oral double-blind doses of 100 to 800 mg per 70 kg, triazolam as active comparator, placebo. That is little, and it is good.

At 400 mg per 70 kg, eleven subjects out of twelve indicate on an eighteen-class questionnaire that they believe they received a classical hallucinogen of the LSD, psilocybin, DMT or mescaline type; mean similarity score 92.7 out of 100. Between 400 and 500 mg, all of them do so, and DXM is rarely identified as a dissociative of the ketamine or PCP type. The authors report that their subjects, all experienced with hallucinogens, were familiar with the laboratory's work, which may explain the classification; they note in counterpoint that triazolam, in the same setting and with the same expectations, produced nothing of the sort. The class declared by this wiki therefore deserves to be discussed rather than asserted.

The phenomenological threshold is high and abrupt: seven subjects out of eleven meet the criteria for a so-called complete mystical experience at the maximum dose, two out of eleven at the preceding dose, none below 400 mg per 70 kg. The price can be read in the same study. Six subjects out of twelve vomit at 400 mg. Blood pressure and heart rate rise dose-dependently, up to +20.8 mmHg systolic and +26 beats per minute, where triazolam slightly lowers the systolic. Seven of the twelve stopped the escalation for behavioural impairment, and the individual stopping points range from 400 to 800 mg per 70 kg: the maximum tolerated dose varies by a factor of two between healthy adults, screened, prepared and supervised.

What does not exist deserves to be written out in full, because the ketamine page, for its part, can lean on it. Three things are missing, and they are the ones that would count: human brain imaging at a psychoactive dose, the human tolerance study, the controlled study of DXM with cannabis, which is nonetheless the most frequent combination in practice. The overall mechanism has never been described in humans other than by inference from the kinship with ketamine and PCP, and that kinship is an analogy, not a measurement.

The plateau system, in mg/kg, comes from the corpus of user reports and not from the clinic. The fifth plateau described as neurotoxic and the so-called fifty-trip limit beyond which tolerance would become irreversible have no support in the literature read. The first is even actively contradicted: in the rat, in a single oral dose of 120 mg/kg as well as in daily administration for thirty days up to 400 mg/kg per day, Carliss et al. find no neuropathological change, at any dose, whereas their MK-801 positive control does produce the expected vacuolation. A negative result whose measurement could see the lesion sought is worth something. It bears on one species, one route and one specific lesion, not on all neurotoxicity nor on humans.

Duration, and why one enzyme decides everything

In the CYP2D6 extensive metaboliser, the half-life of DXM is 2 to 3 hours. In the poor metaboliser, it is 29.5 hours according to the only direct measurement read, 15 to 20 hours according to a tertiary review, and a median of 10.1 hours circulates with no identified primary source: the real uncertainty is a factor of three on how long someone stays intoxicated. Under quinidine, 16.4 hours. Under bupropion, about 22 hours, three times as long, with an exposure increased twenty to thirty times. A user on bupropion, fluoxetine or paroxetine remains exposed well after the end of the effects they feel.

The direction of this effect is counter-intuitive and it has to be given the right way round. Blocking CYP2D6, by genetics or by medication, does not make the experience stronger in the sense the user means. Under quinidine pretreatment, the measures of euphoria and of drug liking fall and displeasure rises. Poor metabolisers, about 8 to 10% of populations, barely tolerate 3 mg/kg where extensive metabolisers tolerate 3 to 6, with more psychomotor impairment, more sedation, more dysphoria; their authors conclude that they are less likely to abuse DXM. The opposite reputation circulates because it confuses plasma concentration with quality of experience. A nuance to keep: the single poor metaboliser in the Reissig trial received 600 mg per 70 kg with no obvious difference in that subject's data. That is n equals 1, it cancels nothing, it forbids presenting the effect of genotype as massive and systematic.

No human study of tolerance to DXM exists: neither its existence, nor its kinetics, nor its reversibility. There is an isolated clinical observation of dependence, physical withdrawal and psychosis, and one case is not a series.

On felt durations, the user corpus gives a total of 8 to 12 hours and reports that the extended-release polistirex forms would carry the effects to 18 or 24 hours, which exposes the user to overdose by redosing: consistent with the formulation, without clinical confirmation in the sources read. The controlled trial, for its part, measures significant effects from 2 hours, maximal between 2 and 4 hours, declining at 6 hours, with balance and coordination deficits still present at 6 hours at the high doses.

Harm reduction

Salt and freebase, the dosing trap peculiar to this molecule. The commercial product is almost always the hydrobromide monohydrate, mass 370.33, against 271.4 for the freebase: a ratio of 1.36. A box labelled 30 mg therefore contains about 22 mg of freebase. Now the only controlled trial expresses its doses as freebase, and says so in as many words, while all the regulatory labels are written in hydrobromide. The direction of the error matters and it is reassuring: taking the salt in the belief that one is following a freebase figure underdoses by 27%. It is the reverse operation, reading a quantity of salt as though it were base, that overestimates by 36% the active quantity supposedly received. About 5% uncertainty remains depending on whether the label denotes the monohydrate or the anhydrous form. No clinical source read gives the user a conversion rule, and the ranges from the user corpus do not specify their unit: the ambiguity is worth about a third of the dose.

MAOIs are the only absolute contraindication, and it has killed. Gillman's reference review establishes that severe serotonin toxicity, the kind that produces hyperthermia and the risk of death, is reached only by the combination of two serotonergic agents acting by different mechanisms, the most frequent being an MAOI combined with any reuptake inhibitor. Dextromethorphan is named there among the weakly reuptake-inhibiting opioids implicated in reactions with MAOIs, fatal outcomes included. Two index cases circulate, Rivers and Horner 1970 and Sovner and Wolfe 1988: neither one could be read, and the details that circulate about the first come from secondary summaries and must not be taken up as established.

This bears directly on this wiki, and without extrapolation on the reader's part: Syrian rue and ayahuasca are reversible MAO-A inhibitors, and Gillman himself names moclobemide, the only reversible MAO-A inhibitor in clinical use, among the MAOIs at risk. The Auvelity label plainly contraindicates, with no waiting period, starting DXM in a patient on a reversible MAOI. The MAOIs one does not recognise as such count too: linezolid, intravenous methylene blue. And something real is missing here: no safety interval after a reversible MAOI is established anywhere. The fourteen days of the labels are written for MAOIs without partition by reversibility, and nothing dates the lifting of an enzyme blockade by the beta-carbolines of a seed.

The pivotal case comes with its reservations. A 46-year-old man, carrier of factor V Leiden, after three consecutive days of ayahuasca on retreat, took a DXM cough syrup and diphenhydramine for respiratory symptoms. Presentation: respiratory rate 27 per minute, blood pressure 166/98, saturation 93%, temperature 38.9 °C, creatine kinase 1,432 IU/L. Management: saline, oral cyproheptadine 2 mg every 2 hours, intravenous midazolam 1 mg per hour, emergency intubation, extubation on day 2, discharge on day 3. Three reservations inseparable from this case and which must travel with it: three products are present together, the urine toxicology was additionally positive for benzodiazepines without it being known whether that precedes or follows the midazolam, and only the conference abstract was read, the full article having proved impossible to open. This case does not on its own demonstrate a causality. It is exactly what Gillman's mechanism predicts, and the DXM side of that mechanism is now quantified in humans. The practical message holds independently of the case: DXM and ayahuasca, or DXM and any harmala-based preparation, do not combine.

Opioids are not to be treated as a block, and the distinction is actionable for someone on substitution. Implicated in serotonin toxicities with MAOIs, deaths included: pethidine, tramadol, methadone, dextromethorphan, propoxyphene. Not reuptake inhibitors and not taking part in this mechanism: morphine, codeine, oxycodone, buprenorphine, to which modern measurement adds heroin, hydrocodone, hydromorphone, oxymorphone. This second half is not an absolution, and the same study shows it: oxycodone figures among the opioids most frequently reported in serotonin syndrome in the WHO database, by mechanisms independent of the transporter.

The combination syrup is the real object consumed, and it carries two dangers that DXM does not have. The first is paracetamol: the quantity of combination product needed for a dissociative effect brings with it a dose of paracetamol that massively exceeds the hepatotoxic threshold, and the danger is delayed, liver failure declaring itself over 24 to 72 hours, after the effects have ended, when the person believes they are in the clear. There is an antidote, N-acetylcysteine, whose efficacy depends on the delay. Of 337 deaths from paracetamol combination products reported to the American poison centres over ten years, paracetamol is the presumed cause in 60.8% of cases. That figure bears on a category of products, not on the population that consumes DXM, whose fraction is not isolable.

The second is chlorphenamine, and it is the most counter-intuitive point on this page. It is not an inert antihistamine: a systematic review concludes that it is a serotonin reuptake inhibitor of a potency comparable to that of dextromethorphan, and in three of the six cases of serotonin toxicity it retains under the Hunter criteria, DXM and chlorphenamine were the only drugs taken in overdose. The product most used for recreational DXM in North America therefore contains two reuptake inhibitors at once, without any SSRI being needed. A reservation to keep: this review speaks of overdose. To this are added a metabolic competition that prolongs both, and an anticholinergic side. One documented case, a 31-year-old woman on about 360 mg of DXM hydrobromide and 48 mg of chlorphenamine a day for several weeks, was intubated for acute psychosis, hyperthermia and tachycardia at 166; its authors attribute the picture to anticholinergic and serotonergic synergy, and no neurological examination could be carried out there, the diagnosis being presumptive.

The brand name does not determine the composition. Coricidin HBP Cough and Cold contains chlorphenamine 4 mg and DXM 30 mg with no paracetamol, the initials HBP denoting a decongestant-free formula. Other products in the same commercial range do contain it. One has to read the list of active ingredients on the box, at every purchase, including for a product already used.

There remains what can be said about the other substances in this set, and what must not be invented.

• Syrian rue and ayahuasca: contraindication, see above. It is the only combination in this set to carry a serious human case.
• AMT, the other established MAO inhibitor in this set, and the only one that carries its inhibitor with it. AMT releases serotonin, inhibits its reuptake and inhibits MAO-A, reversibly and competitively. DXM therefore falls exactly on the axis identified by Wagmann, a monoamine reuptake inhibitor combined with an alpha-methylated tryptamine that blocks MAO: this is Gillman's mechanism set out above, with the two serotonergic agents of different mechanisms that this mechanism requires. No AMT plus DXM case has been published. Two reservations that bound the inference without lifting it: the MAOI potency of AMT is not quantified, the only range actually read bearing on a panel of fourteen compounds and no value specific to AMT having come out of it; and the inhibition here is endogenous to the molecule taken, not a medication at a controlled dose. The inference is therefore weaker than the one aimed at Syrian rue, and it aims at the same mechanism.
• Diphenhydramine: co-present in the pivotal case, and present in three of the five adolescents who died in a series of deaths from powdered DXM bought from the same internet supplier, where four of the five had a co-ingestion. Presence is not attribution. To this is added an inference that no one has measured but whose direction is established: diphenhydramine is a CYP2D6 inhibitor measured in humans, DXM is a substrate of it, and that kind of inhibition, measured under bupropion and on that pair only, increases exposure twenty to thirty times. That pair has never been studied as such, despite its frequency. The direction matters: diphenhydramine increases exposure to CYP2D6 substrates, the reverse is neither demonstrated nor expected. And a reservation about strength travels with that direction: diphenhydramine is a real inhibitor but weak in intrinsic potency, the weakest of the six classical antihistamines tested in vitro, where it comes last. What makes it important is not its strength, it is its over-the-counter availability. The bupropion figure therefore gives the direction of the effect, not its magnitude for that pair. It also figures among the drugs producing false positives on urine screening, including for phencyclidine, and one can ingest it unknowingly, since it has been identified as an undeclared adulterant in street powders.
• MDMA, which is not in this set but belongs to the terrain: a CYP2D6 inhibitor by suicide mechanism, near-irreversible in vitro, at potencies reached from 50 to 100 mg, and recovery of the enzyme goes through its re-synthesis. Taking DXM the day after MDMA is therefore not equivalent to taking it cold. Status: a prediction derived from in vitro data and from modelling, no clinical study of this combination exists.
• Nutmeg, morning glory, Hawaiian baby woodrose: cough syrup figures among the co-ingestants recorded in a series of intentional nutmeg exposures, and dextromethorphan among the co-uses encountered in the user corpus for lysergamide seeds. No clinical outcome is attached to them: these are records of co-presence, not interactions.
• Nitrous oxide: two NMDA antagonists, and it would be easy to manufacture a danger out of that. No source proposes this combination as a risk, and the only datum available runs in a counter-intuitive direction, the good safety record of nitrous oxide being explained by its usual association with GABAergic anaesthetics that counter its neurotoxic potential. To be posed as an unexamined question, never as a danger.
• The seizure threshold adds up, and DXM contributes the least supported piece of it. It lowers the threshold in overdose, but in its case this is an unquantified clinical mention, with no case counted, whereas the diphenhydramine that accompanies it in the pivotal case rests on counted human cases. This imbalance is the reason to count: ten of the fourteen substances in this corpus carry a seizure signal and four do not, which makes it likely that a combination will put DXM up against another carrier, often better documented than it. Two named supports from the set give the measure of the gap, that of DPT, of which 90% of mice convulse at 20 mg/kg, and that of 2C-T-7, of which two of the three published deaths carry convulsions. Tramadol and lithium, which are not in this set, lower this threshold too.
• Vomiting under a sedative is the most commonplace and the most lethal chain. DXM is one of the sedatives in this set and an emetic in its own right, and Syrian rue, nutmeg, morning glory and Hawaiian baby woodrose induce vomiting almost without fail. To these are added the synthetic emetics: profuse and repeated vomiting figures in the picture of the three 2C-T-7 deaths, and it is DPT that supplies the only published case establishing the chain. A 20-year-old man, after insufflation, collapses, then has a seizure, then vomits, in someone already unconscious; the autopsy concludes there was deep aspiration of gastric contents and cerebral oedema from oxygen deprivation. This death is the anchor of the chain, and it names DXM: it is one of the three sedatives in this set, along with diphenhydramine and nitrous oxide, that is to say one of the three products that blunt the airway protective reflexes and therefore aggravate precisely the mechanics that killed. No case documents the combination of DPT with any of the three, and the reasoning bears on the mechanics of aspiration, not on a pharmacology. There is nothing to expect from a product: DXM being itself on both sides of this chain, the one who vomits may already be the one whose reflexes no longer respond, and what keeps them alive is someone who has stayed sober and knows to lay them on their side. A useful detail: in the controlled trial, vomiting typically occurred 90 minutes or more after ingestion, so vomiting does not protect against the dose already taken.

There remain the depressants, and the question of naloxone. Concomitant use of alcohol, benzodiazepines, opioids or gabapentinoids produces additive central nervous and respiratory depression. And while DXM is not a mu agonist, an opioid-type presentation exists, drowsiness, lowered respiratory rate, bilateral miosis, which its own authors describe as rare and which responded to naloxone. The two statements do not exclude each other, one bearing on a receptor affinity and the other on a clinical presentation: faced with respiratory depression, naloxone keeps its place. Agitation, psychosis and seizures are treated with benzodiazepines. Rhabdomyolysis with renal failure requiring dialysis has been reported in DXM serotonin syndrome, in a case we have not read.

A behavioural signal must not be passed over in silence. A forensic series of five cases describes mania, psychosis and hallucinations at high dose, and concludes that users are exposed to a risk of violent and self-destructive acts. It is selected on the judicial outcome, therefore massively biased towards the worst, and it gives no denominator. It does not state a frequency, it states that a high dose of DXM belongs on the list of things to consider in the face of an unexplained mania or psychosis.

Recognising serotonin toxicity. The standard is the Hunter criteria, derived from 2,222 patients: inducible, spontaneous or ocular clonus, agitation, sweating, tremor, hyperreflexia, plus hypertonia and a temperature above 38 °C, these last two universal in the life-threatening forms. The cardinal sign to remember is clonus, ocular in particular, with hyperreflexia predominating in the lower limbs. It is neither confusion, commonplace under DXM alone, nor tachycardia, nor hypertension, which DXM alone produces in an expected and dose-dependent way. Fever together with rigidity or muscle jerks is an emergency, not a bad trip. The treatment of the serotonergic side is named, it is benzodiazepines and cyproheptadine, and it has been used in the published cases.

When to call. The consensus recommendation of the American poison centres, for the ingestion of DXM alone, directs towards follow-up every 2 hours between 5 and 7.5 mg/kg, and towards the emergency department above 7.5 mg/kg. If the person is also taking a serotonergic agent, MAOI or SSRI, follow-up becomes every 2 hours for 8 hours. Any suicidal intent or any intentional abuse sends the person to the emergency department. Do not induce vomiting, no activated charcoal at home. Ask actively whether other substances are involved, paracetamol in particular. These are medical triage thresholds, not effect thresholds: they do not say that nothing happens below them.

What to tell the hospital, and why it is not the instruction on the other pages of this wiki. The tryptamine and phenethylamine pages all carry the same warning, no routine test looks for these molecules and a negative screen rules nothing out. That is not the situation with DXM, which is a pharmaceutical product perfectly amenable to assay, and pinning that formula on it would be false. The side that concerns it lies elsewhere, and it comes down to four points. Do not wait for a blood concentration to settle the matter, it does not sort the survivors from the dead, see just below. Say what was taken, in what quantity and how long ago. Bring the box or photograph it, because the brand name does not determine the composition and one may have ingested DXM unknowingly in a badly read fixed combination. And name paracetamol, which must be assayed actively, whose damage is delayed by 24 to 72 hours and whose antidote loses its efficacy with delay. Report any possible intake of diphenhydramine too, which gives false positives on urine screening including for phencyclidine and can mislead the reading.

A blood concentration does not say what is asked of it. The ranges overlap between survivors and the dead: 950 and 2,820 ng/mL in two survivors of a serotonin syndrome, 950 to 3,230 ng/mL in five who died. The numbers are tiny on both sides, post mortem samples undergo redistribution, four of the five who died had co-ingestions, and what mostly separates these two series is the management. No lethality threshold should be published. A frequently cited figure, 400 ng/mL, is read the wrong way round: the source says that among the patients presenting with adverse effects, 87.5% were above that threshold, which is not the inverse probability.

Bromism is a laboratory trap peculiar to this molecule. The commercial salt is a hydrobromide, and chronic high-dose use loads the body with bromide, which interferes with the hospital chloride assay and produces a factitious hyperchloraemia with a zero or negative anion gap. In a patient seen on several occasions over several months, chloride at 125 mmol/L and anion gap at 1, and she was otherwise asymptomatic. The message is stronger than the list of symptoms one often reads: chronic bromism under DXM may have no clinical expression at all and be visible only on the electrolyte panel, and nothing in the sources read says at what dose or after what duration it appears.

Populations at risk. CYP2D6 poor metabolisers, and anyone on a CYP2D6 inhibitor, fluoxetine, paroxetine, duloxetine, bupropion, terbinafine, ritonavir, quinidine, in whom dissociation, hallucinations and toxicity can occur even at normal antitussive doses. A history of seizures. A personal or first-degree family history of psychotic or bipolar disorder: the only trial that has given high-dose DXM to humans explicitly excluded this population, and nothing allows one to say what happens to it. Children, not recommended before the age of 2. Pregnancy: a controlled study of 184 women exposed at therapeutic doses finds no increase in major malformations, 2.3% against 2.8% in the controls, which says nothing about the recreational dose. Driving or operating machinery: the balance deficits persist at 6 hours.

Two aggravating factors are less well known. Grapefruit juice and Seville orange juice increase the bioavailability of DXM and return only to half the baseline value after three days; the effect bears on the secondary CYP3A4 pathway, hence a plausible aggravating factor and not a major potentiator. And DXM as a pure powder is a distinct risk factor, because it removes the natural limitation imposed by the volume of syrup and the other ingredients, and turns a weighing error into an overdose: it is the context of the five adolescent deaths cited above.

One last point frames all the others. The Cochrane review of over-the-counter cough medicines, 29 placebo-controlled trials and 4,835 participants, finds no good evidence for or against their efficacy in acute cough, and notes that the most numerous adverse effects concerned preparations containing antihistamines and dextromethorphan. Being sold over the counter proves neither safety nor even efficacy.

Sources

• A. Rickli, E. Liakoni, M.C. Hoener, M.E. Liechti (2018), opioid inhibition of the human serotonin and noradrenaline transporters, Br J Pharmacol (full text read; reference source for the SERT affinity, the human concentrations and the WHO pharmacovigilance data).
• C.J. Reissig et al. (2012), high doses of dextromethorphan and effects resembling those of the classical hallucinogens, Psychopharmacology (full text read; the only controlled trial of recreational doses, doses expressed as freebase).
• P.K. Gillman (2005), MAOIs, opioid analgesics and serotonin toxicity, Br J Anaesth (opioid-by-opioid distinction, moclobemide named).
• E.J.C. Dunkley et al. (2003), the Hunter serotonin toxicity criteria, QJM.
• A.A. Monte, R. Chuang, M. Bodmer (2010), dextromethorphan, chlorphenamine and serotonin toxicity, a systematic review, Br J Clin Pharmacol.
• P.A. Chyka et al. (2007), dextromethorphan poisoning, consensus guideline for out-of-hospital management, Clin Toxicol (triage thresholds, valid for DXM alone).
• M. Schadel et al. (1995), pharmacokinetics of dextromethorphan by CYP2D6 phenotype, J Clin Psychopharmacol (abstract read; half-lives).
• L.A. Zawertailo et al. (1998), psychotropic effects modified by the CYP2D6 polymorphism, J Clin Psychopharmacol and (2010), effect of metabolic blockade, Hum Psychopharmacol.
• S. Doyon et al. (2013), deaths involving paracetamol combination products, Clin Toxicol (337 deaths; the figure of 79.2% for non-medical use bears on the whole category, it does not transpose to DXM).
• B.K. Logan et al. (2009), five deaths from abuse of dextromethorphan sold over the internet, J Anal Toxicol and (2012), DXM abuse leading to assault, suicide or homicide, J Forensic Sci (series selected on the outcome, with no denominator).
• A.R. Schwartz, A.F. Pizon, D.E. Brooks (2008), DXM-induced serotonin syndrome, Clin Toxicol (two cases with serum concentrations).
• M. Bachar et al. (2021), acute psychosis and serotonin syndrome on Triple-C overdose, J Med Case Rep (full text read; no neurological examination could be carried out in this patient).
• S. Monks, M. Yen, J. Myers (2020), bromism, an unrecognised toxidrome of chronic dextromethorphan abuse, Am J Emerg Med.
• R.D. Carliss et al. (2007), oral administration of dextromethorphan does not produce neuronal vacuolation in the rat, Neurotoxicology (MK-801 positive control present and positive).
• B.A. Hamelin et al. (1998), in vitro inhibition of CYP2D6 by the classical antihistamines, Drug Metab Dispos (Ki of about 11 µM; diphenhydramine is the weakest inhibitor of the six, and the authors' conclusion is conditional, clinically relevant interactions could occur in specific circumstances).
• A. Heydari et al. (2004), inactivation of CYP2D6 by MDMA, Drug Metab Dispos and L.M. Van et al. (2006), J Psychopharmacol (duration of the inhibition, recovery by enzyme re-synthesis).
• A. Einarson, D. Lyszkiewicz, G. Koren (2001), the safety of dextromethorphan in pregnancy, Chest (therapeutic doses only).
• S.M. Smith, K. Schroeder, T. Fahey (2014), over-the-counter medicines for acute cough, Cochrane review.
• M.P. Di Marco et al. (2002), the effect of grapefruit juice and Seville orange juice on the pharmacokinetics of DXM, Life Sci.
• StatPearls, Dextromethorphan, NCBI Bookshelf (tertiary clinical review); regulatory labels Coricidin HBP Cough and Cold, Auvelity and Nuedexta on DailyMed.
• PsychonautWiki, DXM (collaborative corpus of user reports, source of the plateaus, the felt durations and the tolerance claims; to be read as such, never as clinical material).
• Erowid, DXM Vault (reference corpus of experience reports, to be read on site; page returned HTTP 403 during this search, and its licence forbids reproduction).
• S. Pack, T. Ellett, J. Pham, Y. Ahmad (2026), severe serotonin syndrome with acute respiratory failure after ayahuasca and dextromethorphan, Am J Respir Crit Care Med 212(Suppl 1), aamag162.4811. Conference abstract read; full article in Cureus 500155 NOT READ. This patient's urine toxicology was moreover positive for benzodiazepines.
• N. Rivers and B. Horner (1970), Can Med Assoc J 103(1):85, and R. Sovner and J. Wolfe (1988), N Engl J Med 319(25):1671. Both unread: the first with no abstract and the PMC full text returned empty, the second a letter with no abstract. The clinical details that circulate about the 1970 case come from secondary summaries and are not verified.
• E.E. Codd et al. (1995), J Pharmacol Exp Ther 274(3):1263. Not read: the 23 nM value at SERT is known by citation. And C.P. Taylor et al. (2016), Pharmacol Ther 164:170. Abstract read, affinity tables not read, full text in PDF only.
• Case report of rhabdomyolysis and acute renal failure requiring dialysis in DXM serotonin syndrome, Renal Replacement Therapy 2025, DOI. Not read.