Last verified: April 2026

CYP2C9 — The Primary THC Metabolizer

The cytochrome P450 enzyme CYP2C9 is responsible for approximately 70% of delta-9-THC's phase I metabolism, catalyzing the hydroxylation of THC to its primary active metabolite 11-hydroxy-THC (11-OH-THC) and subsequently to the inactive metabolite 11-nor-9-carboxy-THC (THC-COOH). This single enzyme's activity is the dominant determinant of THC clearance rate, peak plasma concentrations, and duration of effect — and its activity varies dramatically across individuals based on genetic polymorphisms.

CYP2C9 belongs to the CYP2C subfamily, which collectively metabolizes approximately 20% of clinically used drugs, including warfarin, phenytoin, NSAIDs (celecoxib, ibuprofen), and several sulfonylurea hypoglycemics. The enzyme's role in THC metabolism means that the same genetic variants known to affect warfarin dosing also affect cannabis pharmacokinetics — a connection that is clinically underappreciated.

The CYP2C9*3 Polymorphism

The CYP2C9 gene has several well-characterized polymorphisms, but the *3 allele (Ile359Leu, rs1057910) has the most dramatic impact on THC metabolism. This single amino acid substitution in the enzyme's active site substantially reduces catalytic efficiency for THC hydroxylation.

Individuals homozygous for the *3 allele (CYP2C9*3/*3) exhibit approximately 300% higher THC plasma concentrations compared to wild-type (*1/*1) individuals after an equivalent dose. Simultaneously, they produce approximately 70% less of the inactive metabolite THC-COOH, confirming that the elevated THC levels result from impaired metabolic clearance rather than altered absorption or distribution. The net effect is profoundly stronger and longer-lasting psychoactive effects from the same dose.

CYP2C9*3 homozygotes showed approximately 3-fold higher THC exposure with significantly reduced formation of the inactive metabolite, resulting in prolonged psychoactive effects.

Sachse-Seeboth et al., Clinical Pharmacology & Therapeutics 2009

Heterozygous carriers (CYP2C9*1/*3) show intermediate effects — approximately 50–80% higher THC levels compared to wild-type, with proportionally reduced clearance. The gene-dose effect is consistent with a co-dominant inheritance pattern for metabolizer phenotype.

Population Frequencies — Who Carries *3?

CYP2C9*3 allele frequencies vary substantially across ethnic groups. In populations of European descent, approximately 15–20% carry at least one *3 allele (heterozygous), while 1–2% are homozygous *3/*3. The allele is less common in African (approximately 1–3%) and East Asian (approximately 2–4%) populations, though other CYP2C9 variants (e.g., *2, *5, *6, *8, *11) contribute to reduced-function phenotypes in these groups.

The population math is striking: in a room of 100 European-descent cannabis users, 1–2 people will experience roughly triple the pharmacological effect of the same dose, and another 15–20 will experience substantially elevated effects. These individuals are not "lightweights" or "inexperienced" — they are genetically programmed to metabolize THC slowly. They are also the most likely to report adverse experiences from standard edible doses (10 mg) that other users tolerate comfortably.

Clinical Implications — Beyond Body Weight

The conventional wisdom that cannabis dosing should be adjusted for body weight has minimal pharmacokinetic support. Unlike alcohol, which distributes into total body water (making body weight relevant), THC is highly lipophilic and distributes primarily into adipose tissue and lipid-rich membranes. Acute psychoactive effect is determined more by the rate of THC delivery to CB1-rich brain regions than by total body distribution.

CYP2C9 genotype explains far more of the inter-individual variability in THC response than body weight. A 200-pound CYP2C9*3/*3 individual will experience substantially stronger effects from a standard edible dose than a 130-pound wild-type metabolizer — the opposite of what body-weight-based dosing would predict. This disconnect is one reason why universal edible dosing (e.g., "10 mg is a standard dose") fails so many patients.

The FDA-approved dronabinol (Marinol) prescribing information now includes language acknowledging CYP2C9 polymorphisms as a factor in variable patient response, making it one of the earliest cannabinoid label warnings related to pharmacogenomics. As cannabis enters mainstream medicine, CYP2C9 genotyping may become a standard pre-treatment consideration — analogous to CYP2C19 testing before clopidogrel (Plavix) therapy or CYP2D6 testing before codeine prescribing.

The *2 Allele and Other Variants

The CYP2C9*2 allele (Arg144Cys) produces a more modest reduction in enzyme activity — approximately 30–40% lower catalytic efficiency compared to wild-type. It is more common than *3 in European populations (approximately 10–15% allele frequency). Compound heterozygotes (*2/*3) show reduced-function phenotypes intermediate between *3/*3 and wild-type but closer to *3 homozygotes in clinical impact.

Beyond CYP2C9, secondary metabolic pathways involving CYP3A4 and CYP2C19 contribute to THC metabolism, particularly at higher doses when CYP2C9 approaches saturation. Polymorphisms in these enzymes add additional variability, though their contribution is less than CYP2C9's dominant role. The UGT (UDP-glucuronosyltransferase) family — particularly UGT1A9 and UGT2B7 — catalyzes phase II conjugation of THC metabolites, and polymorphisms in these enzymes affect the duration of detectable THC-COOH in urine, relevant to drug testing timelines.

Pharmacogenomics and Edible Dosing Disasters

The emergency department visits attributed to cannabis edibles disproportionately involve first-time or infrequent users who consumed "standard" doses. CYP2C9 pharmacogenomics provides a mechanistic explanation for many of these incidents. A CYP2C9*3/*3 individual who consumes a 10 mg THC edible will achieve approximately 30 mg equivalent plasma exposure — well above the MacCallum/Russo adverse-effect threshold of 20–30 mg — from a dose that most users tolerate without difficulty.

Compounding the problem, oral administration maximizes the CYP2C9 contribution through first-pass hepatic metabolism. While inhalation delivers THC directly to systemic circulation (partially bypassing CYP2C9-mediated first-pass clearance), oral THC is completely subject to hepatic CYP2C9 metabolism. Poor metabolizers thus experience the greatest amplification of effect specifically from the route most likely to produce delayed onset and difficulty titrating.

The convergence of genetic poor metabolism, oral first-pass dynamics, delayed edible onset (30–90 minutes), and the common error of re-dosing creates a pharmacokinetic perfect storm. A *3/*3 individual who takes 10 mg, waits 45 minutes, concludes "it's not working," and takes another 10 mg may ultimately reach plasma levels consistent with a 60 mg dose in a normal metabolizer — a genuinely dangerous exposure that can produce severe dysphoria, psychotic-like symptoms, and tachycardia.

Pharmacogenomic testing is increasingly accessible and affordable (often under $200, sometimes covered by insurance when ordered for other indications). As cannabis medicine matures, the case for pre-treatment CYP2C9 genotyping — particularly before oral cannabinoid therapy in treatment-naive patients — becomes difficult to argue against.