CYP2C19-guided antiplatelet therapy after stroke and TIA: From genotype result to clinical decision

Clinical bottom line: CYP2C19 testing is most useful when the result is available early enough to influence antiplatelet selection. Point-of-care testing (POCT) can bring the result to the bedside during the high-risk early period. In carriers of loss-of-function alleles, clopidogrel activation is reduced; an alternative such as ticagrelor may be considered when clinically appropriate. Testing should support, not delay urgent secondary prevention.

Why genotype matters

CYP2C19 loss-of-function (LOF) alleles are particularly common in Asian populations, with carrier frequencies of approximately 50-60% reported in many East Asian and selected Southeast Asian cohorts. A Singapore cost-effectiveness model used an overall LOF prevalence of 61%: 65% in Chinese, 60% in Indians and 53% in Malays. A multi-ethnic pharmacogenetic study detected CYP2C19*2, *3 and *17 in Malays and recommended including all three variants when investigating clopidogrel response in this population. Population background informs prevalence, but individual genotype determines the predicted phenotype.

The genotype is translated into a predicted metaboliser phenotype. Normal metabolisers generally activate clopidogrel as expected. Intermediate metabolisers carry one LOF allele, while poor metabolisers carry two; both groups may have reduced response. CYP2C19*17 is an increased-function allele that contributes to rapid or ultrarapid metaboliser phenotypes, although reduced- and no-function phenotypes are generally more relevant to clopidogrel decisions after stroke.

What does the evidence show?
The clearest stroke-specific randomised evidence comes from CHANCE-2. The trial enrolled 6,412 Chinese patients with minor ischaemic stroke or high-risk transient ischaemic attack (TIA) who carried a CYP2C19 LOF allele. Ticagrelor–aspirin reduced 90-day stroke compared with clopidogrel–aspirin (6.0% versus 7.6%; hazard ratio 0.77). Severe or moderate bleeding was similar (0.3% in each group), but any bleeding was more frequent with ticagrelor (5.3% versus 2.5%). The benefit was therefore real but modest, and accompanied by more minor bleeding.

This evidence should be applied to the population studied: early-presenting Chinese patients with minor stroke or high-risk TIA suitable for short-course dual antiplatelet therapy (DAPT). CHANCE-2 should not be interpreted to mean CYP2C19-guided prescribing is relevant only to Chinese patients; LOF variants also occur in Malays and Indians. Nor does it mean every LOF carrier with any stroke syndrome requires ticagrelor, or that prolonged ticagrelor–aspirin is appropriate.

The 2022 CPIC guideline recommends considering an alternative P2Y12 inhibitor not significantly affected by CYP2C19 in intermediate and poor metabolisers when clinically appropriate. NICE HealthTech guidance HTG724 recommends testing after ischaemic stroke or TIA, regardless of ethnicity, using laboratory testing or appropriately quality-assured POCT when laboratory testing is unavailable. A 2026 expert guideline developed by the UK CERSI-PGx recommends testing patients about to receive clopidogrel where available; for cerebrovascular disease, consider an alternative in intermediate metabolisers and avoid clopidogrel in poor metabolisers. Antiplatelet treatment should not be delayed while results are pending.

Why point-of-care testing matters
The clinical value of pharmacogenomics depends on turnaround time. Recurrent risk is front-loaded after minor stroke or TIA, and antiplatelet decisions are often made in the emergency department, stroke unit or TIA clinic. A result returned after discharge may miss the decisive window. POCT can move directly from sample to actionable phenotype, reduce dependence on central laboratories and support same-visit prescribing, especially in spoke hospitals, weekend admissions and services with limited molecular capacity.

Speed must not compromise accuracy. POCT requires trained staff, internal quality control, external quality assurance, defined handling of failed or indeterminate results, and coverage of locally relevant alleles. Reports should translate genotype into phenotype and a concise prescribing recommendation. POCT is therefore a clinical pathway linking testing, interpretation and treatment while the decision remains actionable.

Case in practice. A 62-year-old Malay man presents within 12 hours with a minor non-cardioembolic ischaemic stroke (NIHSS 2). Short-term DAPT is being considered. CYP2C19 POCT detects *1/*2, predicting an intermediate-metaboliser phenotype and reduced clopidogrel activation. Antiplatelet treatment is started without delay. After reviewing bleeding risk, drug interactions, tolerability, cost and access, the team considers a ticagrelor-based alternative appropriate to the stroke syndrome and treatment duration. The result is recorded for future prescribing, illustrating how rapid testing can influence a decision while it remains actionable.

A practical bedside approach
1.  Confirm the indication. Establish that the event is non-cardioembolic and requires antiplatelet rather than anticoagulant therapy. Genotyping does not replace evaluation for atrial fibrillation, large-artery disease or other mechanisms.
2.  Test early when clopidogrel is being considered. The greatest value is early after the event, when recurrent risk is highest and the result can change treatment. Use POCT or rapid laboratory testing where possible, but do not delay guideline-directed antiplatelet therapy.
3.  Check what the assay detects. A locally appropriate panel should reliably identify common actionable LOF alleles, particularly *2 and *3; including *17 supports fuller phenotype assignment. Prefer phenotype reporting over raw diplotypes. A limited negative panel does not exclude rare variants, and adherence, interactions, absorption and competing mechanisms also influence response.
4.  Act on the phenotype in context. Clopidogrel can generally be used in normal, rapid or ultrarapid metabolisers when otherwise indicated. For intermediate metabolisers, consider an alternative not dependent on CYP2C19; for poor metabolisers, avoid clopidogrel where feasible. Weigh syndrome, timing, bleeding risk, tolerability, cost and access. Prasugrel is contraindicated after stroke or TIA.
5.  Match treatment duration to the evidence. In eligible minor stroke or high-risk TIA, DAPT is short term and should not be extended solely because a LOF allele is present. Thereafter, choose single antiplatelet therapy according to stroke mechanism, guidelines and patient factors.