Objectives:

This comparative effectiveness review evaluated the analytic validity, prognostic value, and comparative effectiveness of two types of medical tests (genetic testing for CYP2C19 variants and phenotypic testing to measure platelet reactivity) to identify patients who are most likely to benefit from clopidogrel-based antiplatelet therapy and to guide antiplatelet therapy in patient populations who are eligible to receive or are already receiving clopidogrel treatment.

Data sources:

We searched MEDLINE^®, the Cochrane Central Trials Registry, the Cochrane Database of Systematic Reviews, the Human Genome Epidemiology Network database, and the National Institutes of Health Genetic Association Database, from inception to July 27, 2012. We also searched the Food and Drug Administration Web site and ClinicalTrials.gov, and contacted test manufacturers and authors of primary studies.

Review methods:

We used established systematic review methods to identify English-language articles describing studies performed in all relevant care settings on the basis of predetermined eligibility criteria: adult patients with cardiovascular, cerebrovascular, or peripheral arterial disease who were candidates for or were receiving clopidogrel; use of genetic testing (for CYP2C19 variants) or phenotypic testing (for platelet reactivity). Studies had to report information on the analytic validity, prognostic ability for intermediate (platelet reactivity) or clinical outcomes, use of tests to guide antiplatelet therapy, or adverse events from testing itself or from test-directed treatment.

Results:

The literature search yielded 10,475 unique citations, 1,419 of which were obtained in full text and reviewed. A total of 326 publications were judged to have met the inclusion criteria. (Some publications contributed data to multiple analyses.)

Eleven studies provided information for the analytic validity of genotyping assays and 105 for the analytic validity of platelet reactivity assays. Test-retest reliability and interassay agreement for genotyping assays appeared adequate; however, few studies were available. Agreement between assays for measuring platelet reactivity was poor to moderate. Generally, agreement was higher between measurements obtained from the same assay using different agonist concentrations than between different assay types. Only 12 studies provided information on analytic test performance, and they reported variable results.

One hundred six studies provided information on the ability of genetic testing for CYP2C19 variants to predict clinical outcomes or platelet reactivity during followup. The majority of studies were conducted in populations with ischemic heart disease. Under a dominant model, loss-of-function alleles were associated with higher on-clopidogrel platelet reactivity and were statistically significantly associated with stent thrombosis (relative risk [RR] = 1.52; 95% confidence interval [CI], 1.17 to 1.97); major adverse cardiovascular events (RR=1.20; 95% CI, 1.04 to 1.39); and cardiovascular mortality (RR=1.98; 95% CI, 1.13 to 3.46). Associations were nonstatistically significant in meta-analyses of loss-of-function alleles and all-cause mortality, acute coronary syndromes, stroke, and bleeding events. Sensitivity analyses under recessive and additive models for the two outcomes with adequate data (major adverse cardiovascular events and stent thrombosis) produced results in the same direction but with larger effect sizes than the dominant model. Under a dominant model, gain-of-function alleles were significantly associated with reduced risk of major adverse cardiovascular events (RR=0.82; 95% CI, 0.74 to 0.92).

One hundred twenty-eight studies provided information on the ability of baseline on-clopidogrel platelet reactivity to predict clinical outcomes or platelet reactivity during followup. The majority of studies were conducted in populations with ischemic heart disease. Patients with high platelet reactivity at baseline were more likely to be clopidogrel nonresponders during followup. The ability to predict clinical outcomes was reported for various assays; the most commonly assessed were light-transmission aggregometry (55 studies); VerifyNow P2Y12 (38 studies); vasodilator-stimulated phosphoprotein assay (19 studies); Multiplate analyzer (18 studies); and Platelet Function Analyzer-100 (11 studies). Overall, studies suggested that increased on-clopidogrel reactivity was associated with increased rates of adverse cardiovascular outcomes.

Fourteen studies provided information on the use of genetic testing for CYP2C19 variants for guiding treatment choice: 1 randomized trial of testing versus no testing, 12 randomized treatment trials assessing effect modification by CYP2C19 status, 1 randomized treatment trial selecting patients on the basis of CYP2C19 genotype. Twenty-four studies provided information on the use of platelet reactivity measurements to guide antiplatelet therapy: 7 comparative studies of alternative testing strategies (6 with random allocation), 3 randomized treatment trials assessing effect modification by reactivity status, 14 randomized treatment trials selecting patients on the basis of platelet reactivity levels. Studies had heterogeneous designs and compared different treatment strategies. Overall, there was insufficient evidence regarding the use of genotyping CYP2C19 variants or phenotypic testing for platelet reactivity to guide treatment selection.

Conclusions:

We found evidence to support an association between loss-of-function CYP2C19 variants and increased risk of adverse cardiovascular outcomes. Similarly, we found evidence that high on-clopidogrel platelet reactivity is associated with an increased risk of adverse cardiovascular outcomes, at least for some of the available assays. The strength of evidence regarding these prognostic effects was judged to be low or moderate because of concerns regarding selective outcome reporting and publication bias, and the relatively small number of studies reporting data on most clinical outcomes. The strength of evidence regarding the use of genetic or platelet reactivity testing to guide antiplatelet treatment selection was judged to be insufficient because studies reporting on clinical outcomes were few, had diverse designs, and included heterogeneous populations. Comparative data on alternative test strategies (genetic versus phenotypic) are lacking.