Lancet Psychiatry’s CYP2D6 Venlafaxine Cohort Data Signals a Pre-Emptive Genotyping Inflection Point for CNS Trial Design

On October 6, 2026, Lancet Psychiatry published a retrospective analysis across two real-world cohorts, one from Norway and one from the UK, demonstrating that CYP2D6 poor metabolizer (PM) status is consistently associated with increased risk of venlafaxine treatment failure. The study provides what smaller candidate-gene analyses have not: population-scale replication across independent healthcare systems, covering a drug that ranked as the 51st most commonly prescribed medication in the United States in 2023 and remains a first-line SNRI three decades after its December 1993 FDA approval.

The finding does not arrive in a vacuum. The Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines, updated July 2023, already carry an optional recommendation to avoid venlafaxine in CYP2D6 PM individuals, citing increased probability of side effects, and to consider a clinically appropriate alternative antidepressant not predominantly metabolized by CYP2D6. What the Lancet Psychiatry data adds is the treatment-failure signal at scale: two independent real-life cohorts, not curated trial populations, bearing out the PM-failure association with the consistency that moves a recommendation from “optional” toward a standard-of-care conversation.

What the Data Actually Establish

The study’s design matters for how sponsors should read the finding. Both cohorts draw on routine clinical records, which means the PM-failure association is not an artifact of the controlled exposure conditions that define a randomized trial. Participants were prescribed venlafaxine in standard psychiatric practice, then followed for treatment outcomes. CYP2D6 PM status, rather than being a theoretical pharmacokinetic concern, tracked with failure events across both geographic and healthcare-system contexts.

The prevalence arithmetic makes this operationally significant. Among 26,957 individuals of European ancestry in the UK Biobank, 7.29% carried CYP2D6 PM phenotype, a figure drawn from the same UK population contributing to one of the study’s cohorts. Applied to a Phase 3 depression trial enrolling 400 participants, that prevalence implies roughly 29 PM individuals in an ungenotyped arm, each carrying elevated treatment-failure probability. Their inclusion dilutes observed efficacy and inflates apparent nonresponse in the active arm, producing estimates that do not reflect the drug’s performance in the metabolically normal population.

CYP2C19 genotype also featured in the analysis, consistent with venlafaxine’s partial CYP2C19 metabolism, though the primary signal in both cohorts centered on CYP2D6 PM status. Sponsors designing CNS protocols should treat CYP2C19 as a secondary stratification variable rather than a primary exclusion criterion, pending additional evidence on the magnitude of its independent contribution to failure risk.

The IND Submission Pressure Point

FDA’s pharmacogenomics regulatory framework creates a specific decision node for sponsors reading this data. FDA’s guidance on pharmacogenomics in IND applications requires submission of pharmacogenomic data in a full report when a sponsor uses pre-emptive genotyping as a clinical decision-making criterion for entry, specifically when that criterion is intended to enhance the efficacy profile of the drug under investigation. A sponsor that excludes CYP2D6 PM participants from a venlafaxine-based comparator arm using pre-emptive genotyping as an entry criterion triggers that reporting obligation.

That obligation is not a deterrent; it is the mechanism by which FDA accumulates the cross-program PM-failure dataset that will eventually inform labeling language. FDA’s pharmacogenetic biomarker table, published February 25, 2020, lists established gene-drug interactions present in approved labeling and identifies additional interactions supported by sufficient scientific evidence, with the explicit intent of helping practitioners determine appropriate testing. The Lancet Psychiatry dual-cohort replication positions the CYP2D6-venlafaxine interaction for movement within that table’s evidentiary tiers, which in turn affects what reviewers expect to see in submissions involving venlafaxine as either the investigational agent or an active comparator.

Sponsors with active venlafaxine INDs filed in the next six months face a specific protocol question: whether to collect CYP2D6 genotype data prospectively, even where exclusion is not planned, to enable post-hoc stratified efficacy analysis. Collecting and reporting that data without using it as an entry criterion may avoid triggering the full pharmacogenomics report requirement while building the evidentiary record that supports a future PM-stratified label claim; sponsors should confirm the applicable reporting tier with FDA. The Lancet Psychiatry paper provides the precedent population on which to anchor sample-size calculations for that stratified analysis.

Operational Consequences by Stakeholder

For site networks operating in the UK and Norway, the study’s real-world provenance is both the finding’s strength and an immediate clinical-operations signal. Norway has already implemented a nationwide web-based genotyping guidance tool for psychotropic prescribing, including venlafaxine, via cypinfo.no and the Centre for Psychopharmacology at Diakonhjemmet Hospital in Oslo. UK sites contributing to the Biobank-linked cohort operate under a healthcare system that has not yet issued equivalent national guidance, creating asymmetric genotyping practice between the two study populations. That asymmetry implies that any sponsor conducting a multi-national CNS trial spanning these two markets should expect site-level heterogeneity in baseline genotyping data availability and should build a prospective genotyping protocol rather than relying on existing records.

For IRBs and ethics committees, the PM-failure association in a second published cohort changes the ethical calculus on pre-emptive testing disclosure. A protocol that does not collect or disclose CYP2D6 PM status while enrolling patients in a venlafaxine trial can now be evaluated against a published body of evidence showing that PM participants face meaningfully higher failure probability. IRBs at UK institutions should expect sponsor teams to address this in the protocol rationale section beginning immediately.

Investor-side analysis of CNS programs using venlafaxine as a comparator carries an analogous implication. A Phase 3 trial that reports a superiority outcome against an ungenotyped venlafaxine arm has a confounded comparator. Post-hoc FDA reviewer scrutiny of comparator-arm composition is a documented source of complete response letters in CNS submissions. The Lancet Psychiatry data gives reviewers the evidentiary anchor to question whether an ungenotyped comparator arm reflects venlafaxine’s true efficacy potential or an artificially diluted estimate inflated by PM nonresponders.

Cost-effectiveness framing will also become a component of payer submissions in this space. A 2022 cost-utility analysis published in Clinical Drug Investigation modeled pre-emptive CYP2C19 and CYP2D6 genotyping for major depressive disorder from an Italian societal perspective, identifying testing scenarios in which genotype-guided prescribing reached cost-effectiveness thresholds. The dual-cohort replication in Lancet Psychiatry strengthens the population-level denominator for those models, particularly in European health-economic contexts where the Norwegian and UK healthcare systems set the comparator cost reference.

The Strategic Pattern

This publication fits a broader trajectory in which real-world evidence generated outside industry-sponsored trials begins to drive protocol design requirements. The FDA’s pharmacogenomics biomarker table is a living document; agency reviewers consult it when evaluating CNS submissions, and its contents shift as the external literature accumulates. The Lancet Psychiatry paper is the kind of dual-cohort replication that moves an interaction from “supported by sufficient scientific evidence” toward the “actionable” tier, the designation that most directly influences what IND reviewers ask about in Type B meetings.

That movement is not hypothetical. Venlafaxine’s efficacy advantage over SSRIs as a class, established at a number needed to treat of 17 in intent-to-treat analyses, is modest enough that PM-driven failure dilution is materially relevant to any superiority claim. A 7.29% PM prevalence in an ungenotyped trial arm is not a rounding error; it is a systematic bias source that FDA reviewers now have published evidence to cite.

Sponsors with CNS programs in active development should, in this quarter, review their IND pharmacogenomics data collection plan against the CPIC July 2023 CYP2D6-venlafaxine guidance and the FDA February 2020 biomarker table. Where venlafaxine functions as an active comparator, the protocol rationale should address PM-status distribution in the comparator arm or document the prospective genotyping plan that will enable post-hoc stratification. Where venlafaxine is the investigational agent, pre-emptive CYP2D6 genotyping as a stratification variable, rather than an exclusion criterion, preserves statistical power and may avoid triggering mandatory pharmacogenomics full-report submission while positioning the program for a future PM-stratified label.

The next regulatory signal to track is FDA’s handling of any CNS new drug application that incorporates CYP2D6 stratification data submitted in the 12 months following this publication, specifically whether advisory committee briefing documents begin citing the Lancet Psychiatry dual-cohort data as a benchmark for what constitutes adequate genotypic characterization of a depression trial population.

References

1. Lancet Psychiatry, “Association between CYP2D6 and CYP2C19 genotypes and venlafaxine treatment failure: a retrospective study on two cohorts from Norway and the UK” (2026)
2. CPIC, “Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2D6 and Venlafaxine,” updated July 2023
3. MDPI Pharmaceutics, “CYP2D6 Metabolizer Phenotype Is Associated with Early Antidepressant Discontinuation in UK Biobank,” 7.29% PM prevalence in 26,957 European-ancestry participants
4. FDA, Guidance on Pharmacogenomics and IND Applications, full-report requirement for pre-emptive genotyping used as entry criterion
5. FDA, Pharmacogenetic Biomarker Table, published February 25, 2020
6. Synapse/PatSnap, Venlafaxine FDA approval history (December 28, 1993) and 2023 prescribing rank
7. The Nordic Psychiatrist, Norway nationwide pharmacogenomics guidance tool (cypinfo.no) and Centre for Psychopharmacology, Diakonhjemmet Hospital
8. ClinPGx / Clinical Drug Investigation 2022, Cost-utility analysis of pre-emptive CYP2C19 and CYP2D6 genotyping in major depressive disorder (Squassina et al.)
9. University of Miami, Venlafaxine vs. SSRI meta-analysis, NNT 17, 5.9% ITT remission advantage