Cannabinoid Drug Interactions and Polypharmacy

By Robert Seik

Last Updated

This article ranks cannabis-drug interactions by evidence tiers—from confirmed clinical PK/PD effects to in-vitro predictions only—summarizes by drug class, and guides monitoring.

Cannabinoid Drug Interactions and Polypharmacy

Overview

  • Why In Vitro Inhibition Does Not Predict Clinical Effect
  • Tier 1a: Confirmed Pharmacokinetic Interactions
  • Tier 1b: Confirmed Pharmacodynamic Interactions
  • Tier 2: Interactions Supported by Case Reports and Observational Cohorts
  • Tier 3: Interactions Predicted From In Vitro Data Alone
  • Tier 4: Where the Evidence Argues Against a Meaningful Interaction
  • Interaction Summary by Drug Class
  • What the Evidence Supports for Monitoring
  • Future Research Directions
  • Challenges and Ethical Considerations
  • Frequently Asked Questions
  • The Takeaway
  • References

Key Findings

  • Cannabidiol raises N-desmethylclobazam exposure roughly three- to fivefold, making it the best-characterized cannabinoid drug interaction in the literature, though the two largest datasets come from open-label epilepsy cohorts in which clobazam doses were actively reduced during observation, so the untreated magnitude is inferred and not directly measured [9, 10, 11].
  • Formulation and food alter cannabinoid exposure more than most co-prescribed drugs do. A high-fat meal increased cannabidiol area under the curve 3.8-fold and peak concentration 5.2-fold compared with fasting in a randomized phase 1 crossover, which exceeds the magnitude of most interactions clinicians are warned about [5].
  • Predicted interaction magnitude separates products by an order of magnitude. Static modeling built on human liver microsome data predicts a 14.8-fold midazolam exposure ratio for oral cannabidiol at 700 mg against 1.8-fold for oral tetrahydrocannabinol at 130 mg, but these are model outputs awaiting in vivo verification and not observed clinical results [8].
  • A transplant recipient receiving 2,000 to 2,900 mg of cannabidiol daily showed an approximately threefold rise in dose-normalized tacrolimus concentrations, the highest-consequence interaction described to date, though the evidence base is one case report plus a seven-patient open-label series with no control arm [13, 14].
  • Cannabis and alcohol effects on simulated driving were additive and not synergistic in a within-subject controlled trial, with blood tetrahydrocannabinol near 8 µg/L producing lane-weave comparable to a 0.05 g/210 L breath alcohol concentration; participants were young occasional users, not the older polypharmacy population the interaction literature is usually invoked to protect [16].
  • A randomized crossover in 18 healthy adults found that a cannabidiol-dominant extract produced greater anxiety, sedation, and cognitive impairment than a tetrahydrocannabinol-dominant extract at an identical tetrahydrocannabinol dose, contradicting the widely repeated claim that cannabidiol buffers the adverse effects of tetrahydrocannabinol; the result comes from a single small trial using oral administration [17].
  • Concurrent cannabis use during nivolumab therapy was associated with a lower tumor response rate in a retrospective cohort of 140 patients, odds ratio 3.13 (95% CI, 1.24 to 8.1), with no difference in progression-free or overall survival and no adjustment for why patients sought cannabis, a variable that plausibly tracks symptom burden and therefore disease severity [22].

Introduction

Cannabis use among older adults in the United States has been tracked as a distinct trend in national survey data [1]. That surveillance matters clinically because of a question that arrives in exam rooms far more often than it arrives in the literature: a patient on warfarin, or tacrolimus, or four psychoactive medications, discloses that they have started taking cannabidiol, and the clinician has to decide whether anything needs to change.

The available answers are unsatisfying in a specific way. Cannabinoids inhibit a long list of cytochrome P450 enzymes in vitro, and that list has been reproduced across reviews, product inserts, and interaction checkers until it reads like established clinical fact [2, 3]. It is not. The concentrations that produce enzyme inhibition in a microsomal preparation are not always reached in a person, and whether they are reached depends heavily on which cannabinoid, at what dose, by which route, in whom. Meanwhile a much smaller set of interactions has been documented in controlled human studies with effect sizes large enough to change management, and those tend to get the same visual weight in a warning list as the theoretical ones.

This review grades the evidence by strength instead of grouping it by enzyme or by drug class, because the central problem in this literature is not that clinicians lack a list of possible interactions. It is that the list does not distinguish between what has been observed in patients and what has been calculated from a test tube. Six drug classes receive detailed treatment: antiepileptics, anticoagulants and antiplatelets, transplant immunosuppressants, central nervous system depressants, oncology agents including immunotherapy, and cardiovascular medications. A summary table organized by drug class appears after the tiered sections for readers who need a lookup and not an argument.

Why In Vitro Inhibition Does Not Predict Clinical Effect

Why does the same cannabinoid produce a clinically significant interaction in one patient and nothing measurable in another? Three variables account for most of the spread, and none appear on a standard interaction warning.

Route and First-Pass Exposure

Cannabidiol bioavailability after smoking is approximately 31 percent, and no published human study has established absolute oral bioavailability despite intravenous formulations being available [4]. Elimination half-life ranges from 1.4 to 10.9 hours after oromucosal spray to 2 to 5 days after chronic oral dosing [4]. Oral dosing also routes the compound through gut and hepatic first-pass metabolism at concentrations inhaled cannabinoids never reach in the portal circulation, where enzyme inhibition matters most for a co-administered oral drug.

The consequence shows up in modeling. Static prediction from human liver microsome data forecasts a 14.8-fold midazolam exposure ratio for 700 mg oral cannabidiol against 1.8-fold for 130 mg oral tetrahydrocannabinol. For diclofenac, the one substrate the same analysis splits by route, the predicted ratio falls from 6.6 for oral tetrahydrocannabinol at 130 mg to 2.3 for inhaled at 75 mg [8]. These remain predictions requiring in vivo confirmation, and the authors say so. Even so, they establish that a warning applied uniformly to "cannabis" describes several different pharmacologic situations at once.

Formulation, Food, and Dose

In a randomized phase 1 trial of 750 mg cannabidiol, a high-fat meal increased area under the curve 3.8-fold and peak concentration 5.2-fold relative to the fasted state, with a low-fat meal producing 2.7-fold and 3.8-fold increases and whole milk 2.4-fold and 3.1-fold [5]. Alcohol produced more modest elevations of 1.6-fold and 1.9-fold. Taking the same product with breakfast instead of before it shifts exposure more than most interactions catalogued in this article.

Cytochrome P450 Genotype

Cannabidiol is metabolized to its active 7-hydroxy metabolite by CYP2C19 and CYP2C9, with CYP3A handling much of the remaining clearance at other oxidation sites [6]. Formation of the active metabolite tracked measured CYP2C19 activity in genotyped human liver microsomes but not CYP2C19 genotype, which complicates predicting exposure from a genetic panel [6].

The tetrahydrocannabinol signal is cleaner. Among 43 healthy volunteers given oral tetrahydrocannabinol, median area under the curve was threefold higher and the carboxy metabolite 70 percent lower in CYP2C9*3 homozygotes compared with wild-type, with a trend toward greater sedation in *3 carriers [7]. That threefold figure is the primary observation. Secondary sources have circulated one two orders of magnitude larger, and clinicians meeting that number should return to the original report.

Tier 1a: Confirmed Pharmacokinetic Interactions

These are exposure changes measured in prospective human studies, fewer than the warning lists imply and clustered in two therapeutic areas.

Antiepileptics

The clobazam interaction is the anchor. In 13 children on cannabidiol under an expanded access protocol, mean clobazam levels rose 60 percent and N-desmethylclobazam levels rose 500 percent at four weeks [9]. Side effects appeared in 10 of the 13 and resolved on dose reduction. Because doses were cut in those same 10 during observation, the reported increase understates the untreated magnitude.

A larger open-label safety study of 39 adults and 42 children escalating cannabidiol from 5 to 50 mg/kg/day found rising serum topiramate, rufinamide, and N-desmethylclobazam alongside falling clobazam, with zonisamide and eslicarbazepine rising in adults only. That divergence, parent drug down while active metabolite climbs, is the mechanism the clobazam story turns on. All mean changes except clobazam and its metabolite stayed within the accepted therapeutic range [10]. Sedation frequency in adults tracked N-desmethylclobazam levels, and aminotransferases were significantly higher in participants on concomitant valproate.

A phase 1 fixed-sequence trial in healthy volunteers isolated it. Cannabidiol produced only a 1.2-fold change in clobazam exposure while raising N-desmethylclobazam 3.4-fold, raised stiripentol exposure 1.3- to 1.6-fold, and had no clinically relevant effect on valproate [11]. That last result is the instructive one. The valproate signal is real but hepatic, not pharmacokinetic. The pivotal randomized Dravet syndrome trial recorded abnormal liver function results more often with cannabidiol than placebo without separating patients on concomitant valproate, which is where the cohort study localizes the effect [10, 12].

Transplant Immunosuppressants

A participant in a cannabidiol epilepsy trial who was also on tacrolimus showed an approximately threefold rise in dose-normalized tacrolimus concentrations at 2,000 to 2,900 mg of cannabidiol daily [13]. A subsequent open-label series of seven kidney transplant recipients on 100 to 300 mg daily found tacrolimus variable enough to require dose reduction in two, while those on cyclosporine stayed stable [14]. Both are small and uncontrolled, but the mechanism is well supported and the downside of a missed interaction here is graft loss [15].

Drugs Acting on Cannabinoids

The reverse direction is usually omitted. In 36 healthy male volunteers split into three parallel groups of 12, rifampicin reduced peak 11-hydroxy-tetrahydrocannabinol concentration by 87 percent while ketoconazole raised it by 204 percent, which the authors read as evidence that these analytes are CYP3A4 substrates [18]. An enzyme inducer can quietly cost a patient the cannabinoid effect.

Tier 1b: Confirmed Pharmacodynamic Interactions

Exposure is not the only axis. The best-documented human evidence in this literature is behavioral, and it sits where clinicians are least likely to look for it.

Additive Central Nervous System Depression

In a within-subject controlled trial, 18 occasional cannabis users drove a simulator after vaporized cannabis with or without low-dose alcohol. Blood tetrahydrocannabinol concentrations of 8.2 and 13.1 µg/L increased lane weave comparably to breath alcohol concentrations of 0.05 and 0.08 g/210 L, and the combined effects were additive and not synergistic [16]. The population was young and healthy, which limits extrapolation to older adults on sedating medications, though the direction of that bias is toward underestimating risk in the frailer group.

Cannabinoid on Cannabinoid

A randomized double-blind crossover in 18 healthy adults administered brownies containing either a tetrahydrocannabinol-dominant extract or a cannabidiol-dominant extract, matched at 20 mg tetrahydrocannabinol, alongside a cytochrome P450 probe drug cocktail [17]. The cannabidiol-dominant preparation produced higher peak concentration and area under the curve for tetrahydrocannabinol and both major metabolites, and produced greater self-reported anxiety, sedation, and memory difficulty, higher heart rate, and more pronounced cognitive and psychomotor impairment.

This contradicts the common claim that cannabidiol attenuates the adverse effects of tetrahydrocannabinol. The trial is small and used oral dosing, so the magnitude in other routes is unknown, but the direction of effect is the opposite of what the claim predicts. Where evidence positively refutes a claim, the honest report is that the claim is false, and this claim is false as stated for oral products at these ratios.

Tier 2: Interactions Supported by Case Reports and Observational Cohorts

This tier carries genuine signal without controlled confirmation. The interactions here are plausible, sometimes serious, and consistently under-quantified.

Anticoagulants and Antiplatelets

Two case reports anchor the warfarin concern. A 44-year-old man with Marfan syndrome, a mechanical mitral valve, and post-stroke epilepsy entered an open-label cannabidiol program on warfarin 7.5 mg daily, with an international normalized ratio that had held between 2.0 and 2.6 for at least six months beforehand. As cannabidiol was titrated upward in 5 mg/kg/day increments every two weeks, his ratio rose non-linearly and his warfarin dose was eventually cut by approximately 30 percent, with no bleeding complications [19]. The stable six-month baseline is what gives a single case weight, since it removes the obvious alternative explanations for drift. Separately, a 56-year-old man on warfarin for 11 years following mechanical valve replacement presented with an international normalized ratio of 10.41 and gastrointestinal bleeding, then again at 11.55 with epistaxis, during a period of increased cannabis smoking; over nine months of abstinence his values ranged from 1.08 to 4.40 [20]. The authors graded the interaction probable on the Horn scale. Neither establishes incidence, and confounding by adherence, diet, and intercurrent illness cannot be excluded.

A review of cannabis interactions with anticoagulant and antiplatelet agents screened 665 articles across PubMed and EMBASE and found four case reports, one in vitro study, and one pharmacokinetic paper meeting criteria [21]. That yield is the finding. The review also raises the opposite concern: cannabidiol inhibition of CYP2C19 could reduce conversion of clopidogrel to its active thiol metabolite, producing underanticoagulation rather than bleeding. No clinical study has tested this.

Immunotherapy

Two Israeli cohorts examined cannabis use during checkpoint inhibitor therapy. A retrospective analysis of 140 patients receiving nivolumab found a response rate of 15.9 percent among cannabis users against 37.5 percent among non-users, odds ratio 3.13 (95% CI, 1.24 to 8.1), with no difference in progression-free or overall survival [22]. A prospective observational study of 102 patients starting immunotherapy found cannabis use correlated with shorter time to tumor progression and shorter overall survival, alongside fewer immune-related adverse events [23].

Both are observational, and the confounding runs one way: patients who seek cannabis during cancer treatment have more symptoms, and symptom burden tracks disease burden. Neither study adjusted for indication. The mechanism proposed is immunomodulatory, not metabolic, which places it outside the cytochrome P450 framework entirely and means the usual reasoning about dose and route does not apply.

Tier 3: Interactions Predicted From In Vitro Data Alone

Most of the cannabis interaction literature lives here. These findings are real measurements of enzyme behavior; what they lack is any demonstration that the effect occurs in a patient.

Cannabinoid metabolites, not only the parent compounds, inhibit major hepatic enzymes. The abundant tetrahydrocannabinol metabolites 11-hydroxy-tetrahydrocannabinol and the carboxy glucuronide competitively inhibit CYP2B6, CYP2C9, and CYP2D6, while cannabidiol inhibits CYP3A4, CYP2B6, CYP2C9, CYP2D6, and CYP2E1 [24]. Since metabolites circulate at higher concentrations and for longer than parent cannabinoids, restricting attention to tetrahydrocannabinol and cannabidiol underestimates the theoretical exposure.

Binding-corrected inhibition constants sharpen the picture. After accounting for aqueous solubility, microsomal protein binding, and nonspecific binding to labware, cannabidiol showed half-maximal inhibitory concentrations between 0.17 and 0.95 µM across CYP1A2, 2C9, 2C19, 2D6, and 3A, with time-dependent inactivation of CYP1A2, CYP2C19, and CYP3A [25]. Tetrahydrocannabinol showed a distinct profile, most potent against CYP2C9 at 0.012 µM.

A systematic review comparing in vitro inhibition parameters against physiologically achievable cannabinoid concentrations concluded that CYP2C9, CYP1A1/2, and CYP1B1 are likely inhibited by all three major cannabinoids, with CYP2D6, CYP2C19, CYP2B6, and CYP2J2 inhibited by tetrahydrocannabinol and cannabidiol. The same review places UDP-glucuronosyltransferase and carboxylesterase 1 inhibition on thinner evidence [26]. A separate review surveying efflux transporters, including P-glycoprotein and breast cancer resistance protein, finds the same pattern of in vitro signal without clinical confirmation [27].

For cardiovascular agents specifically, the entire concern sits in this tier. Statins, calcium channel blockers, and antiarrhythmics that depend on CYP3A4 are theoretically susceptible to high-dose oral cannabidiol on the basis of the modeling above, and no clinical interaction study has been performed in any of these classes. The same holds for CYP2D6-dependent prodrugs, where reduced conversion to active metabolite is a coherent hypothesis with no clinical outcome data behind it.

Tier 4: Where the Evidence Argues Against a Meaningful Interaction

This tier is small, and its contents deserve more attention than they get.

The clearest negative result comes from oncology, the setting where theoretical concern runs highest. Twenty-four cancer patients received intravenous irinotecan or docetaxel, then repeated the same treatment after 12 days of daily medicinal cannabis as herbal tea [28]. Neither exposure nor clearance changed meaningfully: the irinotecan exposure ratio was 1.04 (95% CI, 0.96 to 1.11) and the docetaxel ratio 1.11 (95% CI, 0.94 to 1.28). The confidence intervals are tight enough to exclude a clinically relevant effect. The important caveat is the preparation, since 200 mL of tea brewed at 1 g/L delivers far lower cannabinoid exposure than a concentrated oral product, so this null applies to the exposure tested and not to cannabis generally.

Omeprazole, a CYP2C19 inhibitor, produced no significant change in peak concentration or area under the curve for tetrahydrocannabinol, cannabidiol, or 11-hydroxy-tetrahydrocannabinol when co-administered with oromucosal spray, which argues against CYP2C19 as a governing pathway for that formulation [18].

In the larger antiepileptic dataset, all mean serum changes other than clobazam and its metabolite remained within accepted therapeutic ranges despite statistical significance [10]. Statistical detection and clinical relevance parted company there, and the distinction is the point.

Drug transporters are the cleanest negative. A systematic review identified no clinical studies suggesting cannabinoid influence on transporter activity, and concluded from in vitro data that a clinically meaningful transporter interaction is unlikely [26].

Interaction Summary by Drug Class

Drug or class Principal mechanism Strongest available evidence Monitoring parameter
Clobazam CYP2C19 inhibition, active metabolite accumulation Controlled human, multiple cohorts Clobazam and N-desmethylclobazam levels; sedation
Valproate Hepatic, mechanism unresolved Randomized trial and open-label cohort Aminotransferases at baseline and during titration
Tacrolimus, mTOR inhibitors CYP3A4 inhibition Case report plus small open series Trough levels within 1 to 2 weeks of any change
Warfarin CYP2C9 inhibition Case reports only International normalized ratio within 1 to 2 weeks
Clopidogrel CYP2C19 inhibition reducing activation Mechanistic inference only No validated parameter
CNS depressants, alcohol Additive pharmacodynamic depression Controlled human trial Sedation, gait, falls, driving counseling
Checkpoint inhibitors Immunomodulatory, not metabolic Two observational cohorts Response assessment; no laboratory parameter
Irinotecan, docetaxel CYP3A4 substrate Controlled crossover, null result None indicated at tested exposure
CYP2D6 prodrugs Reduced activation In vitro only No validated parameter
Statins, CYP3A4 substrates CYP3A4 inhibition In vitro and modeling only No validated parameter

What the Evidence Supports for Monitoring

The evidence does not support dose adjustment rules, and this article does not offer any. It does support a small number of monitoring actions tied to specific pairs.

Where a patient on clobazam begins or changes cannabidiol, serum clobazam and N-desmethylclobazam levels give a direct readout, and sedation is the clinical correlate that tracked metabolite concentration in the adult cohort [9, 10]. Where cannabidiol is added to valproate, aminotransferases at baseline and through titration are the parameter with trial support behind them [10, 12]. For transplant recipients, trough immunosuppressant levels drawn within one to two weeks of any change in cannabis or cannabidiol use address a threefold exposure shift that no clinical examination would detect [13, 14].

For warfarin the case evidence supports checking international normalized ratio on a similar interval after any change in cannabis use, understanding that the underlying evidence is two case reports and that most patients will show nothing [19, 20].

Everything else in this article is either pharmacodynamic, where the monitoring is clinical and not laboratory, or theoretical, where there is no validated parameter to monitor. Ordering a level because an interaction checker flagged an in vitro finding generates cost and anxiety without generating information.

Future Research Directions

The most useful study in this field would be an unglamorous one. Run a probe drug cocktail trial in older adults on established polypharmacy, using commercially available cannabis products at doses patients actually take, with genotyping for CYP2C9 and CYP2C19. The probe cocktail methodology already exists and has been applied to cannabinoids in a healthy young cohort [17]. Extending it to the population where the clinical question actually arises would convert a large share of Tier 3 into either Tier 1a or Tier 4.

Second, the modeling predictions now published deserve direct verification [8, 25]. The predicted 14.8-fold midazolam ratio for high-dose oral cannabidiol is either approximately correct, in which case it is among the most significant interactions in ambulatory medicine, or substantially wrong, in which case the field should know that. A single-arm crossover with midazolam as probe would settle it.

Third, the immunotherapy signal requires a design that can separate cannabis use from the symptom burden that motivates it [22, 23]. An active comparator study, in which patients seeking symptom relief are randomized to cannabis or a non-cannabinoid regimen, would address the confounding by indication that neither existing cohort can.

Fourth, product characterization needs to enter interaction research as a variable and not an assumption. Route, cannabinoid ratio, and food state each moved exposure by multiples in the studies reviewed here [4, 5, 8]. An interaction trial that does not specify and verify these is measuring something other than what it reports.

Finally, the warfarin question is answerable with existing data. Anticoagulation clinics maintain longitudinal international normalized ratio records; linking those to documented cannabis use would produce an incidence estimate where the field currently has two case reports [19, 20, 21].

Challenges and Ethical Considerations

Disclosure is the practical constraint on all of this. Patients underreport cannabis use to clinicians, and the interactions that matter most cluster in populations with reason to withhold, including transplant candidates for whom documented use can affect listing [15]. A monitoring recommendation that depends on disclosure will fail exactly where it is needed. Framing the question as a medication safety matter instead of a compliance matter is a reasonable response, though no study has tested whether it improves disclosure rates.

Research access remains restricted in ways that shape what is known. The review that screened 665 articles and found six eligible papers on anticoagulant interactions was not searching poorly [21]. The papers are not there, and regulatory barriers to conducting controlled interaction studies with a Schedule I substance are part of why.

There is also an asymmetry in how this literature gets used commercially. Interaction warnings are cheap to publish and rarely wrong in a way anyone can demonstrate, so they proliferate. Negative findings, like the irinotecan and docetaxel result, receive far less circulation despite being the more actionable result for a patient on chemotherapy who wants to know whether to stop [28]. Any organization publishing in this space, including this one, should be conscious that overwarning has costs that fall on patients who forgo useful symptom relief.

Frequently Asked Questions

Which cannabinoid drug interaction has the strongest evidence? Cannabidiol with clobazam. Three independent human datasets, including a phase 1 trial in healthy volunteers, document accumulation of the active metabolite N-desmethylclobazam at three- to fivefold [9, 10, 11].

Does cannabidiol interact with more drugs than tetrahydrocannabinol? In vitro and in modeling, yes, particularly at the high oral doses used therapeutically [8, 25]. The comparison is confounded by dose, since pharmaceutical cannabidiol is given in hundreds of milligrams while tetrahydrocannabinol exposure from typical use is far lower.

Is smoked or vaporized cannabis less likely to cause a metabolic interaction than an edible? Modeling predicts a substantially smaller effect for inhaled tetrahydrocannabinol, driven by the absence of first-pass exposure [8]. No clinical study has compared routes directly for interaction outcomes.

Does cannabidiol reduce the adverse effects of tetrahydrocannabinol? No. A randomized crossover found a cannabidiol-dominant extract produced more anxiety, sedation, and cognitive impairment than a tetrahydrocannabinol-dominant extract at the same tetrahydrocannabinol dose [17]. The finding is from one small trial using oral dosing, but it contradicts the claim outright and does not merely fail to support it.

Should patients on chemotherapy stop using cannabis? For irinotecan and docetaxel, a controlled crossover found no pharmacokinetic interaction at the exposure tested [28]. The separate question of checkpoint inhibitors involves observational data with unresolved confounding [22, 23].

Does cannabis affect drug transporters? No clinical study has demonstrated an effect, and in vitro data suggest a clinically meaningful transporter interaction is unlikely [26].

The Takeaway

The cannabinoid interaction literature is inverted relative to how it is usually presented. The interactions that appear most often in warnings, involving CYP2D6 substrates, statins, and a long tail of CYP3A4-dependent drugs, rest almost entirely on in vitro inhibition data at concentrations that oral and inhaled products may or may not achieve. The interactions with real human evidence behind them are few: clobazam, valproate hepatic tolerability, tacrolimus, and additive sedation with central nervous system depressants.

Two findings deserve to travel further than they have. Food and formulation shift cannabinoid exposure by multiples larger than most catalogued drug interactions, which means the product and the meal are variables of the same order as the co-prescribed drug. And a cannabidiol-dominant preparation intensified the adverse effects of tetrahydrocannabinol in controlled conditions instead of softening them, which contradicts a claim that circulates widely in consumer-facing material.

For clinicians the operational summary is short. Monitor levels and liver enzymes for the four pairs where human data justify it, counsel on additive sedation, and treat the rest as unresolved, not as established risk. For researchers the gap is specific and closable. Probe drug cocktail studies run in the polypharmacy population itself, using the products patients actually buy, would close it.

References

  • [1]: Han BH, Palamar JJ. Trends in Cannabis Use Among Older Adults in the United States, 2015-2018. JAMA Intern Med. 2020;180(4):609-611. https://pubmed.ncbi.nlm.nih.gov/32091531/
  • [2]: Cox EJ, Maharao N, Patilea-Vrana G, Unadkat JD, Rettie AE, McCune JS, Paine MF. A marijuana-drug interaction primer: Precipitants, pharmacology, and pharmacokinetics. Pharmacol Ther. 2019;201:25-38. https://pubmed.ncbi.nlm.nih.gov/31071346/
  • [3]: Antoniou T, Bodkin J, Ho JM. Drug interactions with cannabinoids. CMAJ. 2020;192(9):E206. https://pubmed.ncbi.nlm.nih.gov/32122975/
  • [4]: Millar SA, Stone NL, Yates AS, O'Sullivan SE. A Systematic Review on the Pharmacokinetics of Cannabidiol in Humans. Front Pharmacol. 2018;9:1365. https://pubmed.ncbi.nlm.nih.gov/30534073/
  • [5]: Crockett J, Critchley D, Tayo B, Berwaerts J, Morrison G. A phase 1, randomized, pharmacokinetic trial of the effect of different meal compositions, whole milk, and alcohol on cannabidiol exposure and safety in healthy subjects. Epilepsia. 2020;61(2):267-277. https://pubmed.ncbi.nlm.nih.gov/32012251/
  • [6]: Beers JL, Fu D, Jackson KD. Cytochrome P450-Catalyzed Metabolism of Cannabidiol to the Active Metabolite 7-Hydroxy-Cannabidiol. Drug Metab Dispos. 2021;49(10):882-891. https://pubmed.ncbi.nlm.nih.gov/34330718/
  • [7]: Sachse-Seeboth C, Pfeil J, Sehrt D, Meineke I, Tzvetkov M, Bruns E, Poser W, Vormfelde SV, Brockmöller J. Interindividual variation in the pharmacokinetics of Delta9-tetrahydrocannabinol as related to genetic polymorphisms in CYP2C9. Clin Pharmacol Ther. 2009;85(3):273-276. https://pubmed.ncbi.nlm.nih.gov/19005461/
  • [8]: Bansal S, Paine MF, Unadkat JD. Comprehensive Predictions of Cytochrome P450 (P450)-Mediated In Vivo Cannabinoid-Drug Interactions Based on Reversible and Time-Dependent P450 Inhibition in Human Liver Microsomes. Drug Metab Dispos. 2022;50(4):351-360. https://pubmed.ncbi.nlm.nih.gov/35115300/
  • [9]: Geffrey AL, Pollack SF, Bruno PL, Thiele EA. Drug-drug interaction between clobazam and cannabidiol in children with refractory epilepsy. Epilepsia. 2015;56(8):1246-1251. https://pubmed.ncbi.nlm.nih.gov/26114620/
  • [10]: Gaston TE, Bebin EM, Cutter GR, Liu Y, Szaflarski JP. Interactions between cannabidiol and commonly used antiepileptic drugs. Epilepsia. 2017;58(9):1586-1592. https://pubmed.ncbi.nlm.nih.gov/28782097/
  • [11]: Morrison G, Crockett J, Blakey G, Sommerville K. A Phase 1, Open-Label, Pharmacokinetic Trial to Investigate Possible Drug-Drug Interactions Between Clobazam, Stiripentol, or Valproate and Cannabidiol in Healthy Subjects. Clin Pharmacol Drug Dev. 2019;8(8):1009-1031. https://pubmed.ncbi.nlm.nih.gov/30791225/
  • [12]: Devinsky O, Cross JH, Laux L, Marsh E, Miller I, Nabbout R, Scheffer IE, Thiele EA, Wright S. Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome. N Engl J Med. 2017;376(21):2011-2020. https://pubmed.ncbi.nlm.nih.gov/28538134/
  • [13]: Leino AD, Emoto C, Fukuda T, Privitera M, Vinks AA, Alloway RR. Evidence of a clinically significant drug-drug interaction between cannabidiol and tacrolimus. Am J Transplant. 2019;19(10):2944-2948. https://pubmed.ncbi.nlm.nih.gov/31012522/
  • [14]: Cuñetti L, Manzo L, Peyraube R, Arnaiz J, Curi L, Orihuela S. Chronic Pain Treatment With Cannabidiol in Kidney Transplant Patients in Uruguay. Transplant Proc. 2018;50(2):461-464. https://pubmed.ncbi.nlm.nih.gov/29579828/
  • [15]: Rein JL. The nephrologist's guide to cannabis and cannabinoids. Curr Opin Nephrol Hypertens. 2020;29(2):248-257. https://pubmed.ncbi.nlm.nih.gov/31972598/
  • [16]: Hartman RL, Brown TL, Milavetz G, Spurgin A, Pierce RS, Gorelick DA, Gaffney G, Huestis MA. Cannabis effects on driving lateral control with and without alcohol. Drug Alcohol Depend. 2015;154:25-37. https://pubmed.ncbi.nlm.nih.gov/26144593/
  • [17]: Zamarripa CA, Spindle TR, Surujunarain R, Weerts EM, Bansal S, Unadkat JD, Paine MF, Vandrey R. Assessment of Orally Administered Δ9-Tetrahydrocannabinol When Coadministered With Cannabidiol on Δ9-Tetrahydrocannabinol Pharmacokinetics and Pharmacodynamics in Healthy Adults: A Randomized Clinical Trial. JAMA Netw Open. 2023;6(2):e2254752. https://pubmed.ncbi.nlm.nih.gov/36780161/
  • [18]: Stott C, White L, Wright S, Wilbraham D, Guy G. A Phase I, open-label, randomized, crossover study in three parallel groups to evaluate the effect of Rifampicin, Ketoconazole, and Omeprazole on the pharmacokinetics of THC/CBD oromucosal spray in healthy volunteers. Springerplus. 2013;2(1):236. https://pubmed.ncbi.nlm.nih.gov/23750331/
  • [19]: Grayson L, Vines B, Nichol K, Szaflarski JP. An interaction between warfarin and cannabidiol, a case report. Epilepsy Behav Case Rep. 2017;9:10-11. https://pubmed.ncbi.nlm.nih.gov/29387536/
  • [20]: Yamreudeewong W, Wong HK, Brausch LM, Pulley KR. Probable interaction between warfarin and marijuana smoking. Ann Pharmacother. 2009;43(7):1347-1353. https://pubmed.ncbi.nlm.nih.gov/19531696/
  • [21]: Greger J, Bates V, Mechtler L, Gengo F. A Review of Cannabis and Interactions With Anticoagulant and Antiplatelet Agents. J Clin Pharmacol. 2020;60(4):432-438. https://pubmed.ncbi.nlm.nih.gov/31724188/
  • [22]: Taha T, Meiri D, Talhamy S, Wollner M, Peer A, Bar-Sela G. Cannabis Impacts Tumor Response Rate to Nivolumab in Patients with Advanced Malignancies. Oncologist. 2019;24(4):549-554. https://pubmed.ncbi.nlm.nih.gov/30670598/
  • [23]: Bar-Sela G, Cohen I, Campisi-Pinto S, Lewitus GM, Oz-Ari L, Jehassi A, Peer A, Turgeman I, Vernicova O, Berman P, Wollner M, Moskovitz M, Meiri D. Cannabis Consumption Used by Cancer Patients during Immunotherapy Correlates with Poor Clinical Outcome. Cancers (Basel). 2020;12(9):2447. https://pubmed.ncbi.nlm.nih.gov/32872248/
  • [24]: Nasrin S, Watson CJW, Perez-Paramo YX, Lazarus P. Cannabinoid Metabolites as Inhibitors of Major Hepatic CYP450 Enzymes, with Implications for Cannabis-Drug Interactions. Drug Metab Dispos. 2021;49(12):1070-1080. https://pubmed.ncbi.nlm.nih.gov/34493602/
  • [25]: Bansal S, Maharao N, Paine MF, Unadkat JD. Predicting the Potential for Cannabinoids to Precipitate Pharmacokinetic Drug Interactions via Reversible Inhibition or Inactivation of Major Cytochromes P450. Drug Metab Dispos. 2020;48(10):1008-1017. https://pubmed.ncbi.nlm.nih.gov/32587099/
  • [26]: Qian Y, Gurley BJ, Markowitz JS. The Potential for Pharmacokinetic Interactions Between Cannabis Products and Conventional Medications. J Clin Psychopharmacol. 2019;39(5):462-471. https://pubmed.ncbi.nlm.nih.gov/31433338/
  • [27]: Alsherbiny MA, Li CG. Medicinal Cannabis-Potential Drug Interactions. Medicines (Basel). 2018;6(1):3. https://pubmed.ncbi.nlm.nih.gov/30583596/
  • [28]: Engels FK, de Jong FA, Sparreboom A, Mathot RA, Loos WJ, Kitzen JJ, de Bruijn P, Verweij J, Mathijssen RH. Medicinal cannabis does not influence the clinical pharmacokinetics of irinotecan and docetaxel. Oncologist. 2007;12(3):291-300. https://pubmed.ncbi.nlm.nih.gov/17405893/