CYP2D6 inhibitors can convert a patient's genotype-predicted metabolic function into a different functional phenotype entirely, a process called phenoconversion, and strong inhibitors can push enzyme activity toward near-null levels regardless of what the genotype report says. The FDA classifies inhibitors as strong, moderate, or weak based on measured AUC fold-changes, but a scoping review found only 62% concordance between those classifications and the underlying data, so moderate and weak designations deserve case-by-case scrutiny.
TL;DR:
- Strong CYP2D6 inhibitors like paroxetine and fluoxetine reliably cause near-complete enzyme silencing, significantly affecting drugs like codeine or tamoxifen.
- The clinical impact of inhibitors varies depending on dosing pattern and substrate reliance, with steady-state use often causing more inhibition than single-dose studies suggest.
- Phenoconversion can shift patients from predicted genotype-based metabolizer status to phenotypically poor metabolizers in 20% to 70% of cases with CYP2D6 drugs.
- Incorporating inhibitor classification and phenoconversion logic into automated lab reports improves prescribing safety and reduces reliance on manual, static interaction lists.
Table of Contents
- How regulators and the literature define strong, moderate, and weak CYP2D6 inhibitors
- What the evidence shows: scoping-review concordance with FDA classifications and suggested reclassifications
- Mechanisms of CYP2D6 inhibition and how mechanism shapes clinical effect
- Clinically relevant inhibitors: a categorized reference for medication review
- Concrete substrate cases: tamoxifen, tramadol, metoprolol, and antiarrhythmics
- Phenoconversion and integrating inhibitors into genotype-guided prescribing
- Where evidence is weak and next steps for research and practice
- Operationalizing inhibitor-aware PGx reporting in laboratory workflows
- What clinicians should prioritize when an inhibitor shows up on a patient's list
- Build inhibitor-aware reporting into your lab's pharmacogenomics program
- FAQ
- Sources
How regulators and the literature define strong, moderate, and weak CYP2D6 inhibitors
The FDA's Table of Substrates, Inhibitors and Inducers sets the reference framework that most clinical labs and prescribers rely on today. A drug earns its classification based on how much it raises the area under the curve (AUC) of a sensitive CYP2D6 substrate when the two are coadministered in a controlled study. Strong inhibitors raise AUC at least fivefold, moderate inhibitors raise it between two- and fivefold, and weak inhibitors produce an increase of 1.25- to twofold. These thresholds are not arbitrary: they reflect the magnitude of exposure change a prescriber should expect when adding the inhibitor to a patient already taking a CYP2D6 substrate.
The classification depends heavily on which substrate was used to measure it. An index substrate, typically a drug metabolized almost exclusively by CYP2D6, such as dextromethorphan or desipramine, gives a cleaner read on inhibitory potency than a substrate metabolized through multiple pathways. When a study uses a substrate with secondary metabolic routes, the observed AUC change can understate or overstate the true CYP2D6-specific effect, because other enzymes partially compensate for the inhibition.
Several caveats shape how these numbers should be applied in practice.
- Substrate sensitivity varies: a sensitive substrate shows a larger AUC change than a substrate with multiple clearance pathways, even with the same inhibitor.
- Multi-pathway inhibition confounds results: some drugs inhibit CYP2D6 alongside other enzymes (CYP3A4, CYP2C19), making it hard to isolate the CYP2D6-specific contribution.
- Study design matters: single-dose interaction studies can underestimate the effect of an inhibitor that accumulates to steady state, especially with long half-life agents.
- Dose and formulation differ by indication: the inhibitor dose used in a labeling study may not match the dose a patient receives in a different clinical context.
For clinicians interpreting a drug label or an interaction checker, the practical takeaway is that the strong/moderate/weak label is a starting point, not a guarantee. A drug labeled moderate based on one substrate might behave closer to strong when paired with a different, more sensitive substrate, and the index substrate used in the pivotal study rarely matches the actual substrate a given patient is taking.
What the evidence shows: scoping-review concordance with FDA classifications and suggested reclassifications
A 2020 scoping review set out to test how well the FDA's strong, moderate, and weak categories actually track the primary pharmacokinetic literature. The authors compiled observed AUC fold-changes across a large set of inhibitor-substrate pairs and compared each one against the FDA's stated classification for that inhibitor.
Only 62% of inhibitor-substrate pairs matched their FDA classification, based on an analysis of 89 pairs drawn from the published pharmacokinetic literature. That means more than a third of the comparisons the review examined showed a measured AUC change that did not line up with the category the drug was assigned. For a clinician relying on a static drug interaction list to triage risk, that gap is large enough to change a prescribing decision.
The review flagged specific drugs where the primary data suggested a different category than the one commonly cited.
- Cimetidine: observed effects were inconsistent across studies, likely reflecting its inhibition of multiple cytochrome pathways beyond CYP2D6.
- Desvenlafaxine: classified data did not consistently match predicted AUC changes, raising questions about the strength assigned to it in some reference lists.
- Fluvoxamine: multi-pathway inhibition (affecting CYP1A2 and CYP2C19 alongside CYP2D6) complicated attempts to isolate its CYP2D6-specific potency.
- Celecoxib: study-specific factors, including dosing and substrate choice, produced AUC changes that diverged from its assigned category.
The common thread across these examples is confounding: a drug that inhibits more than one enzyme, or that has been studied with substrates of differing sensitivity, is harder to pin to a single clean category. The review's authors suggested that some of these drugs warrant reclassification, or at minimum a cautionary note that the assigned category rests on thinner evidence than the strong inhibitors do.
The practical interpretation for clinical use splits along a clear line. Strong inhibitors, the agents that push AUC fivefold or more, tend to show consistent effects across different substrates and studies, because an effect that large is hard to miss regardless of study design. Moderate and weak inhibitors sit on much shakier ground: the review's findings suggest these categories should be treated as directional guidance rather than precise, substrate-independent numbers. When a patient's regimen includes a moderate or weak inhibitor alongside a narrow therapeutic index substrate, it is worth checking whether the classification was built on a sensitive index substrate or on the specific drug the patient is actually taking.
Mechanisms of CYP2D6 inhibition and how mechanism shapes clinical effect
Not all CYP2D6 inhibition happens the same way, and the mechanism behind a given inhibitor influences how much clinical effect to expect and how long that effect persists. Most CYP2D6 inhibitors act through competitive inhibition: the inhibitor molecule competes with the substrate for the same active site on the enzyme, and the degree of inhibition tracks closely with the inhibitor's plasma concentration. As the inhibitor is cleared, competitive inhibition fades roughly in step with its own elimination half-life.
Quinidine behaves differently. It binds CYP2D6 at a site distinct from the substrate pocket, an allosteric or noncompetitive mechanism, and its inhibitory effect does not scale cleanly with plasma concentration the way competitive inhibitors do. This distinction matters clinically because an allosteric inhibitor like quinidine can suppress CYP2D6 activity at doses too low to produce meaningful antiarrhythmic or other pharmacologic effects of its own, yet still functionally silence the enzyme for coadministered substrates.
The dosing pattern used to study an inhibitor also changes what gets observed. A single-dose interaction study captures the acute effect of one exposure, but many CYP2D6 inhibitors are prescribed chronically, and steady-state concentrations can produce a larger and more sustained degree of inhibition than a single dose predicts. This is particularly relevant for antidepressants like paroxetine and fluoxetine, which patients typically take daily for months, not once.
- Competitive inhibitors fade in step with the inhibitor's own clearance, so a short washout often restores baseline activity.
- Allosteric inhibitors like quinidine can suppress enzyme activity disproportionately to their own plasma levels.
- Steady-state dosing of a chronic inhibitor tends to produce more complete and sustained inhibition than single-dose data suggest.
- Substrate identity changes the outcome: the same inhibitor can produce a small AUC change for one CYP2D6 substrate and a large one for another, depending on each substrate's reliance on the CYP2D6 pathway specifically.
Pro Tip: When reviewing a medication list for CYP2D6 interaction risk, check whether the inhibitor is dosed chronically or as a single agent: chronic dosing of even a "moderate" inhibitor can approach the clinical impact of a strong inhibitor at steady state.
Substrate-dependent variability is the mechanistic thread that ties classification and clinical effect together. A drug that relies on CYP2D6 for 80% of its clearance will show a much larger AUC shift under inhibition than a drug where CYP2D6 handles only a minor fraction of total metabolism, even when the inhibitor itself is identical in both cases. Clinicians should resist the instinct to treat an inhibitor's classification as a fixed property independent of what it is being combined with.
Clinically relevant inhibitors: a categorized reference for medication review
A working reference list helps during medication reconciliation, particularly when a patient's genotype report needs to be interpreted alongside their current prescriptions. The categories below follow the FDA framework, with notes on where substrate-specific judgment still applies.
Strong inhibitors are the agents most likely to functionally silence CYP2D6 regardless of the patient's underlying genotype. Paroxetine and fluoxetine, both selective serotonin reuptake inhibitors, fall in this category, as does bupropion, whose metabolites contribute to the inhibitory effect documented in its prescribing label. Quinidine, used far less often today as an antiarrhythmic but still cited in pharmacokinetic studies, is the classic allosteric strong inhibitor. When any of these is added to a regimen, expect a normal metabolizer to behave pharmacokinetically like a poor metabolizer for any coadministered CYP2D6 substrate, which matters most for opioids like codeine or tramadol, tamoxifen, and several antiarrhythmics.
Moderate inhibitors carry more evidence variability, consistent with the scoping review's concordance findings. Duloxetine and terbinafine are commonly cited examples, and both require substrate-specific judgment rather than a blanket assumption of moderate risk across every coadministered drug. Mirabegron, used for overactive bladder, also falls into this group and is a useful reminder that CYP2D6 interaction risk is not limited to psychiatric medications.
Weak inhibitors rarely matter in isolation but can become clinically relevant in combination. Sertraline is frequently cited as a weak-to-moderate inhibitor depending on dose, and its effect can compound when a patient is also taking another weak or moderate inhibitor, or when the coadministered substrate has a narrow therapeutic index.
| Category | Example inhibitors | Expected phenotype shift | Watch for |
|---|---|---|---|
| Strong | Paroxetine, fluoxetine, bupropion, quinidine | Normal metabolizer functions as poor metabolizer | Codeine, tramadol, tamoxifen, antiarrhythmics |
| Moderate | Duloxetine, terbinafine, mirabegron | Variable, substrate-dependent reduction | Substrate-specific AUC data over category label |
| Weak | Sertraline (dose-dependent) | Minimal alone, additive in combination | Co-prescribed inhibitors, narrow therapeutic index drugs |
Combination scenarios deserve particular attention during medication review. When a patient takes two weak inhibitors, or a weak inhibitor alongside a moderate one, the combined effect on CYP2D6 activity can approach what a single moderate or even strong inhibitor would produce, especially for a substrate that is nearly exclusively cleared through CYP2D6. This additive risk is easy to miss when each drug is checked individually against a static interaction list rather than evaluated as part of the full regimen.
Concrete substrate cases: tamoxifen, tramadol, metoprolol, and antiarrhythmics
Tamoxifen depends on CYP2D6 to form endoxifen, its most pharmacologically active metabolite, and coadministration with a strong CYP2D6 inhibitor can blunt that conversion in patients who would otherwise metabolize tamoxifen normally. Paroxetine is the inhibitor most frequently raised in this context because it is also commonly prescribed for hot flashes in the same patient population taking tamoxifen for breast cancer. When a strong SSRI cannot be avoided, consulting with endocrine oncology about an alternative antidepressant, or selecting one with weaker CYP2D6 inhibition, is a reasonable step before assuming tamoxifen efficacy is unaffected.
Tramadol presents one of the clearest pharmacokinetic demonstrations of inhibitor impact. PBPK modeling shows that coadministering quinidine with a single dose of tramadol increases tramadol's AUC by around half while decreasing the AUC of its active metabolite, O-desmethyltramadol, by about half, with the model predicting the inhibitory effect can persist for nearly two days. Because O-desmethyltramadol carries most of tramadol's opioid analgesic activity, this shift can mean a patient receives less pain relief than expected while accumulating more of the parent compound. The same logic extends to codeine, which depends on CYP2D6 to form morphine: coadministration with a strong inhibitor can leave a patient with inadequate analgesia despite a seemingly appropriate dose.
Metoprolol and several antiarrhythmics are CYP2D6 substrates with narrow therapeutic margins, meaning small increases in exposure can produce clinically meaningful bradycardia or hypotension. Adding a strong or moderate inhibitor to a patient stabilized on metoprolol warrants a dose review and closer monitoring of heart rate and blood pressure, rather than assuming the existing dose remains appropriate. The same caution applies to antiarrhythmic agents cleared through CYP2D6, where the margin between therapeutic and toxic exposure is often narrow enough that a twofold AUC increase has real consequences.
Reading a drug label for DDI magnitude means looking past the summary classification to the actual fold-change reported in the clinical pharmacology section, and checking which substrate was used to generate that number. A label that reports a 2.5-fold AUC increase with a sensitive index substrate carries different weight than one reporting the same number with a substrate that has multiple clearance pathways, and matching the patient's actual medication to the closest studied substrate gives a more reliable read than the category label alone.
Phenoconversion and integrating inhibitors into genotype-guided prescribing
Phenoconversion describes the mismatch that occurs when a patient's genotype predicts one metabolic phenotype, say a normal or intermediate metabolizer, but concurrent medication exposure shifts their functional phenotype toward poor metabolizer status. A 2023 review estimates that 20% to 70% of patients taking CYP2D6-metabolized drugs may be at risk of phenoconversion, a range wide enough to reflect how much it depends on the specific population and medication combination studied, but large enough to make genotype-only interpretation unreliable on its own.
The Clinical Pharmacogenetics Implementation Consortium (CPIC) approach, echoed by the 2023 phenoconversion review, offers a structured way to adjust a genotype-based activity score when inhibitors are present: multiply the baseline activity score by 0 for a strong inhibitor, and by 0.5 for a moderate inhibitor, then reassign the phenotype category based on the adjusted score.
A practical workflow follows from this adjustment logic.
- Reconcile the full medication list, not just the drug being evaluated for pharmacogenomic risk, since any concurrent CYP2D6 inhibitor is relevant.
- Identify inhibitor strength for each relevant medication using the FDA classification, cross-checked against substrate-specific data when the inhibitor falls in the moderate or weak category.
- Adjust the activity score using the strong (×0) or moderate (×0.5) multiplier, then reassign the functional phenotype.
- Document the adjusted phenotype and the reasoning in the chart, so the next prescriber understands why the functional phenotype differs from the genotype-only prediction.
- Monitor and reassess after any washout period, since discontinuing the inhibitor should eventually restore genotype-predicted function.
Pro Tip: Build inhibitor washout into the monitoring plan, not just the initial dose adjustment: a patient's functional phenotype can shift twice, once when the inhibitor is added and again when it is stopped.
Washout timing varies by drug. Many competitive inhibitors clear within 5 to 7 days, but long half-life agents break that rule. Fluoxetine's label notes that its inhibitory effect, driven by both the parent drug and its active metabolite norfluoxetine, can persist for weeks after discontinuation, so reassessing phenotype on a standard 5 to 7-day timeline would be premature for that agent specifically.
Where evidence is weak and next steps for research and practice
The concordance gap identified in the scoping review points to a broader evidence problem: for many inhibitor-substrate pairs, especially those involving moderate or weak inhibitors, high-quality human AUC data simply does not exist in sufficient volume to support confident classification. Much of what clinicians rely on traces back to a small number of interaction studies, often conducted with a single index substrate that may not represent the drug actually being prescribed.
Several priorities would close this gap over time.
- Standardize DDI study designs so that future interaction studies use sensitive, well-characterized index substrates rather than convenience substrates with multiple clearance pathways.
- Report participant genotypes alongside inhibitor dosing in published studies, since an interaction study conducted only in normal metabolizers tells little about how the same inhibitor behaves in intermediate metabolizers.
- Expand use of physiologically based pharmacokinetic (PBPK) modeling, following the approach used for the tramadol-quinidine interaction, to predict effects for inhibitor-substrate pairs that lack direct clinical trial data.
- Connect pharmacokinetic changes to real-world clinical outcomes, since an AUC fold-change is a surrogate marker, and the clinical consequences (treatment failure, toxicity, hospitalization) are what ultimately matter for prescribing decisions.
Until that evidence base matures, the most defensible clinical posture treats strong inhibitor classifications as reliable and moderate or weak classifications as provisional, subject to revision as substrate-specific data accumulates.
Operationalizing inhibitor-aware PGx reporting in laboratory workflows
Translating inhibitor classification and phenoconversion logic into a working clinical report is where most pharmacogenomics programs struggle, because it requires reconciling a patient's genotype, their full medication list, and current guidance in a format a prescriber can act on quickly. Pharmacogenomics reporting platforms can build this reconciliation into the reporting layer itself rather than leaving it to manual review. Medication intelligence simulation can evaluate a patient's active medication list against genotype data to flag relevant inhibitor exposure, while evidence fusion can draw on multiple authoritative sources to keep inhibitor classifications current.
Because guidance in this area shifts as new interaction data emerges, living reanalysis features can re-evaluate existing reports as classifications or guidelines change, rather than leaving a report static from the day it was generated. For labs managing narrow therapeutic index substrates such as tamoxifen or antiarrhythmics, that reanalysis capability can alert a prescriber when new evidence affects a patient already on file, not only at the time of initial testing. Reports may carry physician review and an audit trail before reaching a provider, and integration with electronic health records through HL7/FHIR standards can help inhibitor-adjusted phenotype information reach the prescribing workflow rather than sitting in a separate portal.
For laboratories building or expanding a pharmacogenomics program, such capabilities may turn inhibitor-aware prescribing from a manual chart review exercise into a structured, repeatable part of report generation.
What clinicians should prioritize when an inhibitor shows up on a patient's list
The evidence here splits cleanly into what is certain and what is not. Strong inhibitors are well characterized, consistent across studies, and should trigger an immediate phenotype adjustment: treat the patient as functionally poor metabolizer for any coadministered CYP2D6 substrate, full stop. That is the highest-value, lowest-ambiguity action available, and it should happen before worrying about anything else on the medication list.
Moderate and weak inhibitors are where clinical judgment earns its keep. The concordance data makes clear that a category label alone is not a reliable guide for these agents, so substrate-specific evidence, and a willingness to monitor rather than guess, matters more than memorizing a static list.
The bigger shift I would argue for is structural: inhibitor-aware phenotype adjustment should not depend on an individual clinician remembering to check a table. Labs and health systems that build this logic into automated reporting will catch interactions that busy prescribers miss during a ten-minute visit.
— Tarek
Build inhibitor-aware reporting into your lab's pharmacogenomics program
SignalPGx was built for laboratories that need to turn genotype and medication data into reports a prescriber can trust without second-guessing. The platform's white-label infrastructure lets a lab deploy its own branded pharmacogenomic reporting service, with evidence fusion, medication intelligence simulation, and living reanalysis built into the system rather than bolted on afterward.
For labs evaluating how inhibitor classification and phenoconversion logic would fit into an existing CLIA workflow, the Discovery & Strategy service is the starting point, followed by System Design and Deployment & Integration once requirements are scoped. Full platform details, including the Pilot, Standard, Scale, and Enterprise plans, are available on the White-Label PGx Reporting page. Reach out to discuss how your lab's medication review workflow could incorporate automated, inhibitor-aware phenotype adjustment.
This article is general information, not a substitute for advice from a qualified doctor. Consult a qualified healthcare professional about your own circumstances before acting on anything here.
FAQ
What are some natural inhibitors of CYP2D6?
Evidence for herbal or dietary CYP2D6 inhibition in humans is limited compared to the pharmaceutical data behind the FDA's classification system, and most claims about botanical inhibitors come from in vitro or preliminary work rather than controlled human AUC studies. Clinicians evaluating a patient's full intake, including supplements, should weigh this evidence cautiously and prioritize substrate-specific pharmaceutical interaction data, discussing supplement use openly since patients frequently omit it from medication histories; resources like this overview of psychotropic and supplement interactions illustrate how uneven the evidence base can be outside prescription drugs.
Does Wellbutrin block CYP2D6?
Yes, bupropion (marketed as Wellbutrin) and its active metabolites inhibit CYP2D6, and the drug's prescribing label documents increased exposure to coadministered CYP2D6 substrates. The FDA classifies bupropion as a strong CYP2D6 inhibitor, meaning it can functionally phenoconvert a normal metabolizer toward poor metabolizer status for affected substrates.
Is Prozac a strong CYP2D6 inhibitor?
Yes, fluoxetine (Prozac) is classified by the FDA as a strong CYP2D6 inhibitor, capable of raising a sensitive substrate's AUC at least fivefold. Its active metabolite, norfluoxetine, has a long half-life, so inhibitory effects can persist for weeks after the drug is stopped, which matters when planning washout before reassessing a patient's functional phenotype.
What drugs should a CYP2D6 poor metabolizer avoid?
A genotype-predicted or phenoconverted CYP2D6 poor metabolizer generally should avoid or use with caution prodrugs that require CYP2D6 activation for efficacy, such as codeine and, to a lesser extent, tramadol, since reduced conversion to active metabolites can mean inadequate pain relief. Certain antiarrhythmics and standard-dose tamoxifen also warrant extra scrutiny, since reduced endoxifen formation has raised clinical concern, and any decision should weigh substrate-specific evidence alongside CPIC guidance rather than a single rule applied across all substrates.