Enzyme:CYP2C9

CYP2C9 (cytochrome P450 2C9) is one of the most clinically consequential of the human cytochrome P450 enzymes, second only to CYP3A4 in the breadth of its everyday relevance to prescribing. It is encoded by the CYP2C9 gene on chromosome 10q23.33, sits within a cluster of four CYP2C genes (CYP2C8, CYP2C9, CYP2C18, CYP2C19), and is expressed almost exclusively in the liver. Its substrate list is shorter than CYP3A4's but contains medicines whose narrow therapeutic windows have made the enzyme one of the central concerns of clinical pharmacogenomics: the oral anticoagulant warfarin, the antiepileptic phenytoin, many of the NSAIDs, several sulfonylurea oral hypoglycaemics, and the angiotensin-receptor blocker losartan (whose conversion to its active metabolite E-3174 is a CYP2C9 reaction).

The foundational mechanistic insight was published in 1992 by Allan Rettie and colleagues at the University of Washington in Seattle, who showed using cDNA-expressed human cytochromes P450 that the more potent S-enantiomer of warfarin, the one that does most of the anticoagulant work, is hydroxylated predominantly by CYP2C9. That observation, together with the recognition during the same period that CYP2C9 was polymorphic and that some patients carried reduced-function alleles, became the mechanistic foundation for everything that followed in warfarin pharmacogenomics.

Tissue distribution

CYP2C9 is overwhelmingly a hepatic enzyme. It accounts for roughly 20% of total cytochrome P450 protein in human liver, placing it second only to CYP3A4 in hepatic abundance. Its expression in the small-intestinal wall is modest by comparison to CYP3A4, and in clinical practice the intestinal contribution is rarely the dominant factor for any single CYP2C9 substrate. This distribution matters for understanding why CYP2C9 interactions behave differently from CYP3A4 interactions: for CYP3A4 substrates, the gut wall and the liver are both at risk from interactions; for CYP2C9 substrates, the liver is essentially the whole story.

Function and substrate spectrum

CYP2C9 catalyzes hydroxylation, epoxidation, and N- and O-dealkylation across a substrate spectrum that, while narrower than CYP3A4's, contains a striking concentration of medicines with narrow therapeutic windows and well-defined dose-response toxicities. The combination of warfarin and phenytoin alone, both of which are major CYP2C9 substrates with narrow therapeutic windows, accounts for a substantial share of the inpatient pharmacogenomic-testing case that has been built around this enzyme over the last two decades.

The table below collects the clinically important CYP2C9 substrates by therapeutic class, with each entry tagged by the contribution CYP2C9 makes to overall clearance: major (CYP2C9 is the predominant route; genotype-modulated and inhibitor-modulated interactions are clinically expected), moderate (CYP2C9 contributes meaningfully but other routes carry comparable load), and minor (CYP2C9 contributes but other pathways dominate). The list is curated for clinical relevance and is not exhaustive; see Comprehensive substrate and interaction tables below for the authoritative maintained resources. Click Expand at right to view the table.

near-complete CYP2C9 substrate table (click to expand)

Substrate	Therapeutic class	CYP2C9 contribution	Clinical notes
Acenocoumarol	Anticoagulant	major	Coumarin anticoagulant; reduced-function CYP2C9 alleles reduce dose requirement and raise bleeding risk, as with warfarin.
Bosentan	Endothelin-receptor antagonist	major	CYP2C9 and CYP3A4 dual metabolism; auto-induces CYP2C9 and CYP3A4 over the first weeks of therapy.
Candesartan	Angiotensin-receptor blocker	minor	Mostly cleared unchanged renally; CYP2C9 contribution is small.
Celecoxib	NSAID (COX-2 selective)	major	CPIC recommends dose reduction in CYP2C9 poor and intermediate metabolizers.
Cyclophosphamide	Antineoplastic (alkylating)	partial	Bioactivation to 4-hydroxycyclophosphamide is multi-CYP (CYP2B6 dominant, CYP2C9 and CYP3A4 contributing).
Diclofenac	NSAID	major	A canonical CYP2C9 probe substrate in research pharmacology.
Dronabinol	Cannabinoid (synthetic Δ9-THC)	moderate	Mixed CYP2C9 and CYP3A4 metabolism; CYP2C9 inhibition raises plasma THC.
Fluoxetine	SSRI antidepressant	partial	S-enantiomer is partly CYP2C9-cleared; fluoxetine is itself a moderate CYP2C9 inhibitor (see Inhibitors).
Flurbiprofen	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM.
Fluvastatin	Statin (HMG-CoA reductase inhibitor)	major	The only statin metabolized predominantly by CYP2C9 rather than CYP3A4; CYP2C9 poor metabolizers have higher plasma exposures and higher rhabdomyolysis risk.
Glimepiride	Sulfonylurea (oral hypoglycaemic)	major	Hypoglycaemia risk raised in CYP2C9 PM/IM.
Glipizide	Sulfonylurea	major	Same pattern as glimepiride.
Glyburide	Sulfonylurea (glibenclamide)	major	Same pattern as glimepiride.
Ibuprofen	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM.
Indomethacin	NSAID	moderate	Both CYP2C9 and conjugation routes; less affected by CYP2C9 genotype than ibuprofen or celecoxib.
Irbesartan	Angiotensin-receptor blocker	major	CYP2C9 is the predominant oxidative route.
Lornoxicam	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM.
Losartan	Angiotensin-receptor blocker (prodrug)	major	Activation reaction, not inactivation: CYP2C9 converts losartan to E-3174, the active metabolite responsible for most of the antihypertensive effect. CYP2C9 poor metabolizers may have reduced antihypertensive response.
Meloxicam	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM; long half-life magnifies the accumulation problem in poor metabolizers.
Naproxen	NSAID	moderate	Less CYP2C9-dependent than diclofenac or ibuprofen; conjugation contributes substantially.
Nateglinide	Meglitinide (oral hypoglycaemic)	moderate	Mixed CYP2C9 and CYP3A4.
Phenprocoumon	Anticoagulant (coumarin)	major	Same anticoagulant-class CYP2C9 dependence as warfarin and acenocoumarol; more relevant in European practice than US.
Phenytoin	Antiepileptic	major	CPIC dose-reduction guidance in CYP2C9 PM/IM; nonlinear pharmacokinetics make small clearance changes produce large concentration changes. Minor CYP2C19 contribution.
Piroxicam	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM; long half-life.
Rosiglitazone	Thiazolidinedione (oral hypoglycaemic)	moderate	Mixed CYP2C8 and CYP2C9.
Sildenafil	PDE5 inhibitor	minor	CYP3A4 is the dominant clearance route; CYP2C9 contribution is real but small.
Sulfamethoxazole	Sulfonamide antibiotic	moderate	Self-metabolised partly by CYP2C9; sulfamethoxazole is also a moderate CYP2C9 inhibitor (the source of trimethoprim-sulfamethoxazole interactions with warfarin).
Tamoxifen	Anti-estrogen (breast cancer)	minor	4-hydroxylation pathway is partly CYP2C9; the major activation pathway (to endoxifen) is CYP2D6, dominantly.
Tenoxicam	NSAID	major	CPIC dose-reduction guidance in CYP2C9 PM/IM; long half-life.
Tetrahydrocannabinol (Δ9-THC)	Cannabinoid	moderate	See cannabinoid-pharmacology page; mixed CYP2C9 + CYP3A4 metabolism.
Tolbutamide	Sulfonylurea (historical)	major	The original probe substrate for CYP2C9 in research pharmacology; rarely prescribed clinically today but central to the historical literature.
Torsemide	Loop diuretic	major	CYP2C9 PM status raises plasma torsemide and prolongs diuretic effect.
Valproic acid	Antiepileptic, mood stabilizer	minor	Predominantly cleared by glucuronidation and beta-oxidation; CYP2C9 contribution is small.
Valsartan	Angiotensin-receptor blocker	minor	Cleared mostly unchanged in bile; CYP2C9 contribution is small.
Voriconazole	Triazole antifungal	moderate	Mixed CYP2C19 (major), CYP2C9, and CYP3A4; CYP2C19 genotype dominates the clinical pharmacogenomics story for voriconazole, but CYP2C9 contributes.
Warfarin	Anticoagulant	major	Canonical CYP2C9 substrate. CYP2C9 clears the more potent S-enantiomer. CPIC warfarin guideline integrates CYP2C9 with VKORC1 for dose-prediction.
Zafirlukast	Leukotriene-receptor antagonist	moderate	Also a CYP2C9 inhibitor (significant warfarin interaction).

Phenotype categories

Unlike CYP3A4, which lacks a well-defined poor-metabolizer phenotype, CYP2C9 has a clean tiered phenotype classification grounded in star-allele genotype. The CPIC convention recognises four metabolizer phenotypes:

• Normal metabolizer (NM): two normal-function alleles, typically *1/*1.
• Intermediate metabolizer (IM): one normal-function and one decreased- or no-function allele.
• Poor metabolizer (PM): two decreased- or no-function alleles.
• Indeterminate: genotype not interpretable under the current allele-function table.

There is no ultra-rapid CYP2C9 category in current CPIC usage; gene duplications and gain-of-function variants are not established for this enzyme in the way they are for CYP2D6.

The phenotype matters most for warfarin and phenytoin: poor metabolizers require substantially lower doses than normal metabolizers, and intermediate metabolizers sit in between with intermediate dose requirements. CPIC has issued specific dosing guidance for both indications (see Clinical implications below).

Major variants

The clinically actionable star alleles of CYP2C9 cluster into two ancestry groups, and the European-ancestry literature considerably overrepresents the field. Note that under current CPIC convention, all listed reduced-function alleles below contribute to the IM and PM phenotypes; ancestry-aware testing panels are the practical solution.

European-ancestry-prevalent reduced-function alleles:

• *2 (rs1799853, Arg144Cys), a decreased-function allele present in roughly 10 to 20% of European-ancestry populations. About 70% of wild-type activity for most substrates.
• *3 (rs1057910, Ile359Leu), a more severely decreased-function allele present in roughly 5 to 10% of European-ancestry populations. About 10 to 20% of wild-type activity. The *3 allele was characterised in the tolbutamide-polymorphism work of Sullivan-Klose and colleagues in 1996 and has been the single most heavily studied CYP2C9 variant.

African-ancestry-prevalent reduced-function alleles (frequently missing from older testing panels, a recognised health-equity problem):

• *5 (Asp360Glu), decreased function.
• *6 (rs9332131, frameshift), no function.
• *8 (Arg150His), decreased function. Population frequencies of *8 in African-ancestry groups can reach 5 to 10%, making it as important as *2 in European populations, yet many pre-2015 testing panels did not capture it.
• *11 (Arg335Trp), decreased function.

A more complete PharmVar haplotype table sits at the PharmVar page; this list covers the alleles with established CPIC actionability.

Inhibitors

CYP2C9 inhibition is the mechanism behind some of the most dangerous warfarin interactions in clinical practice. Major inhibitors include:

Strong inhibitors:

• Fluconazole (the prototype strong CYP2C9 inhibitor; routinely doubles or triples warfarin INR within days of co-prescription).
• Miconazole (including topical oral gels, which are systemically absorbed enough to matter for warfarin INR).
• Amiodarone (a mixed CYP inhibitor that hits CYP2C9 hard; expect roughly a one-third warfarin dose reduction after a few weeks of amiodarone exposure).

Moderate inhibitors:

• Trimethoprim-sulfamethoxazole (sulfamethoxazole is the CYP2C9-inhibiting component; a major warfarin interaction).
• Fluoxetine and fluvoxamine.
• Capecitabine and fluorouracil (a severe and sometimes catastrophic warfarin interaction in oncology practice).
• Tigecycline.

The clinical pattern is consistent: any patient on a stable warfarin dose who is started on a CYP2C9 inhibitor needs prompt INR monitoring within days, not weeks.

Inducers

CYP2C9 is inducible, though not as strongly or as broadly as CYP3A4. Induction is mediated chiefly by the constitutive androstane receptor (CAR) and the pregnane X receptor (PXR), as for CYP3A4. Major inducers include:

• Rifampin, again the clinical archetype, with substantial induction over a two-to-three-week window.
• Carbamazepine, phenobarbital.
• Phenytoin itself, which auto-induces its own metabolism. This auto-induction is part of why phenytoin's pharmacokinetics are so difficult to manage clinically: the first dose of a new regimen, the steady-state on chronic therapy, and the period after a missed week look pharmacokinetically distinct from each other.
• St John's Wort.

Clinical implications, summary

CYP2C9 is one of the few enzymes for which pharmacogenomic-guided prescribing has entered routine practice in some health systems, and the clinical case is built on three CPIC guidelines:

Warfarin (CPIC 2017 update): integrates CYP2C9 genotype with VKORC1 genotype, CYP4F2 genotype, age, body size, and concomitant medicines into a dosing algorithm. The clinical evidence for genotype-guided initiation versus standard-of-care initiation has been mixed across trials (GIFT, COAG, EU-PACT), but CPIC's position is that when genotype is already available before initiation, it should be used. For patients with reduced-function CYP2C9 alleles and reduced-function VKORC1 alleles in combination, initial dose requirements can be a fraction of the standard dose, and starting at the standard dose risks bleeding.

Phenytoin (CPIC 2020 update): integrates CYP2C9 genotype with HLA-B*15:02 status. CYP2C9 poor metabolizers should receive about 50% of the standard maintenance dose; intermediate metabolizers should receive about 75%. HLA-B*15:02-positive patients (chiefly people of Han Chinese, Thai, Vietnamese, and other Southeast Asian ancestry) should not receive phenytoin at all because of severe-cutaneous-adverse-reaction risk, regardless of CYP2C9 genotype.

NSAIDs (CPIC 2020): recommends CYP2C9-genotype-guided dose reduction for celecoxib, flurbiprofen, ibuprofen, lornoxicam, piroxicam, meloxicam, and tenoxicam in poor and intermediate metabolizers, or selection of an alternative non-CYP2C9 NSAID. The clinical concern is that the longer-half-life NSAIDs (piroxicam, meloxicam, tenoxicam) accumulate substantially in poor metabolizers, with corresponding increases in gastrointestinal and renal toxicity.

The original clinical-genetic observation linking CYP2C9 polymorphism to warfarin dose requirements was the 1999 Lancet paper by Aithal and colleagues, who showed in a Newcastle cohort that patients with reduced-function CYP2C9 alleles required lower warfarin doses and had a higher risk of bleeding complications than wild-type patients. That paper opened the modern era of CYP2C9 pharmacogenomics; the CPIC guidelines above are its lineal descendants.

Comprehensive substrate and interaction tables

The substrate and interaction tables on this page are curated for clinical relevance, not for completeness. Three authoritative external resources maintain comprehensive lists of CYP2C9 substrates, inhibitors, and inducers, and the wiki recommends them to any reader who needs an exhaustive look-up:

• Flockhart Cytochrome P450 Drug Interaction Table, maintained by the Department of Medicine at Indiana University School of Medicine. The most widely cited clinical-reference cytochrome P450 table; substrate-, inhibitor-, and inducer-tiered, updated regularly. Available at https://drug-interactions.medicine.iu.edu/.
• U.S. Food and Drug Administration Drug Development and Drug Interactions Table, the regulatory-grade list FDA uses for labeling and clinical-trial design decisions. Smaller than Flockhart but every entry is FDA-vetted. Available via the FDA Center for Drug Evaluation and Research clinical drug interaction page.
• PharmGKB, the pharmacogenomics knowledge base hosted at Stanford University; the CYP2C9 gene page indexes substrate-, inhibitor-, and inducer-relationships with their underlying primary literature, and links each gene-drug pair to the CPIC dosing guideline where one exists. Available at https://www.pharmgkb.org/.

For a comprehensive review of CYP2C9 (and the rest of the human cytochrome P450 family) covering regulation, polymorphism, and substrate spectrum in detail, the Zanger and Schwab 2013 review in Pharmacology and Therapeutics remains the standard reference.

See also

• Enzyme:CYP2C19, the closely related CYP2C-cluster enzyme with its own large pharmacogenomic story (clopidogrel, PPIs, several antidepressants).
• Enzyme:CYP3A4, Enzyme:CYP2D6, Enzyme:CYP1A2, Enzyme:CYP2B6.
• Enzyme:VKORC1, the warfarin-target gene whose polymorphisms are dosed together with CYP2C9 in the CPIC warfarin algorithm.
• Warfarin, Phenytoin, Celecoxib (canonical CYP2C9 substrate examples with CPIC guidelines).
• Fluconazole, Amiodarone (canonical CYP2C9 inhibitor examples).
• Category:Drug-metabolizing enzymes (when this category is built out).