Impact of Implementing CYP2C19 Genotype-Guided Antiplatelet Therapy on P2Y12 Inhibitor Selection and Clinical Outcomes in Acute Coronary Syndrome Patients After Percutaneous Coronary Intervention: A Real-World Study in China

Background: CYP2C19 loss-of-function (LOF) alleles reduce the effectiveness of clopidogrel in patients undergoing percutaneous coronary intervention for acute coronary syndrome. However, the clinical impact of implementing CYP2C19 gene-guided pharmacotherapy is unclear, especially among the Chinese population. The purpose of this study was to evaluate P2Y12 receptor inhibitor selection and clinical outcomes upon implementation of CYP2C19 genotype-guided pharmacotherapy in current clinical practice. Methods: This was a single-center observational cohort study. Adult percutaneous coronary intervention patients who received CYP2C19 genetic testing (*2, *3, *17 alleles) were included. Ticagrelor was recommended for patients with a LOF allele. Factors related to P2Y12 inhibitor selection were determined by logistic regression. The primary endpoint was major cardiac or cerebrovascular adverse events (MACCE) within 12 months. MACCE and clinically significant bleeding events (BARC ≥2) in the LOF-clopidogrel group, non-LOF-clopidogrel group, and non-LOF-ticagrelor group were compared with those in the LOF-ticagrelor group. The inverse probability of treatment weighting (IPTW) was adjusted in a Cox regression analysis to eliminate confounding factors. Results: Among 1,361 patients, 826 (60.7%) had a LOF allele. Patients with a LOF allele were more likely to be prescribed ticagrelor (multivariate-adjusted OR 1.349; 95% CI 1.040 to 1.751; p = 0.024). The MACCE rate was higher in the LOF-clopidogrel group than in the LOF-ticagrelor group (7.8 vs. 4.0%; log-rank p = 0.029; IPTW-adjusted HR 2.138; 95% CI 1.300–3.515). Compared with the LOF-ticagrelor group, the non-LOF-clopidogrel group showed no significant difference in MACCE rate (5.8 vs. 4.0%; log-rank p = 0.272; IPTW-adjusted HR 1.531; 95% CI 0.864–2.714). Among the patients treated with ticagrelor, there was no significant difference in the MACCE rate between the LOF group and non-LOF group (4.3 vs. 4.0%; log-rank p = 0.846; IPTW-adjusted HR 1.184; 95% CI 0.582–2.410). There was no significant difference in the incidence of clinically significant bleeding events among the four groups. Conclusion: This study confirms that efficiently returned CYP2C19 genotype results did partially guide cardiologists to prescribe ticagrelor for patients with a LOF allele, and that clopidogrel had a higher risk of MACCE than ticagrelor in these patients, which provides support for the implementation of CYP2C19 gene-guided antiplatelet therapy in clinical practice.


INTRODUCTION
Coronary artery disease (CAD), specifically acute coronary syndrome (ACS), is still the leading cause of disability and mortality worldwide (Fox et al., 2010;Jernberg et al., 2015). Oral dual antiplatelet therapy with aspirin and a P2Y 12 receptor inhibitor is the standard postoperative maintenance strategy for patients undergoing percutaneous coronary intervention (PCI) for ACS (Levine et al., 2011;Levine et al., 2016). Clopidogrel, the most widely used P2Y 12 inhibitor, was previously shown to improve the prognosis of ACS patients. However, evidence from pharmacogenomics has gradually raised concerns about the clinical efficacy of clopidogrel. As a prodrug, clopidogrel needs to be converted into active substances by cytochrome (CYP2C19) enzymes. Patients with a CYP2C19 loss-of-function (LOF) allele were shown to have lower conversion rates of clopidogrel by their CYP2C19 enzymes (Mega et al., 2009;Shuldiner et al., 2009). The population frequency of a LOF allele is approximately 60% in East Asian countries and approximately 30% in the rest of the world (Klein et al., 2019). Our previous study identified possession of a CYP2C19 LOF allele is an independent risk factor for clopidogrel-related high platelet responsiveness (Peng et al., 2019), which reduces the effectiveness of the drug (Mega et al., 2009;Shuldiner et al., 2009;Mao et al., 2013;Xi et al., 2019).
Ticagrelor is a P2Y 12 inhibitor that acts directly on platelets, and its efficacy is hardly affected by CYP2C19 gene polymorphisms (Wallentin et al., 2010;Li et al., 2017). The PLATO trial confirmed that ticagrelor is superior to clopidogrel in reducing adverse cardiovascular events in patients with ACS (Wallentin et al., 2009). Furthermore, current guidelines recommend the use of ticagrelor or prasugrel in preference to clopidogrel in ACS patients (Levine et al., 2011;Levine et al., 2016). However, the high incidences of bleeding and discontinuation due to dyspnea, as well as the high price, have limited the widespread clinical use of ticagrelor (Wallentin et al., 2009;Bonaca et al., 2015).
Emerging evidence from observational studies and randomized controlled trials has confirmed the feasibility of implementing CYP2C19 gene-guided pharmacotherapy (Cavallari et al., 2018b;Lee et al., 2018;Notarangelo et al., 2018;Claassens et al., 2019;Tuteja et al., 2020). In addition, increasing numbers of medical institutions are beginning to incorporate CYP2C19 genotyping into clinical practice. However, the PCI guidelines state that CYP2C19 pharmacogenetic testing should only be considered for highrisk patients (Class IIb; Level of Evidence: C) (Levine et al., 2011;Levine et al., 2016). In the absence of strong support from current guidelines, there are limited real-world data in East Asian populations to validate the implementation of genotype-guided antiplatelet therapy in clinical practice. Furthermore, the impact of genetic testing results on P2Y 12 inhibitor selection by doctors in China remains unclear.
Therefore, we performed a single-center retrospective observational study in Chinese patients with ACS after PCI, with the following objectives: 1) to determine the effects of returned genotype results on the prescription behavior of cardiologists and 2) to understand the relationships of the identified genotypes with P2Y 12 receptor inhibitor selection and clinical outcomes.

Patients and Study Design
PHARM-ACS is an ambispective single-center ongoing observational registry study on PHARMacotherapy and longterm clinical outcomes in patients with ACS after PCI, initiated by the Department of Pharmacy at Beijing Anzhen Hospital (NCT04184583). In this registry, consecutive adult patients in our hospital with successfully indexed PCI with indication of STsegment elevation myocardial infarction (STEMI), non-STEMI, or unstable angina were eligible between April 2018 and December 2021. Patients were recruited retrospectively from April 2018 to September 2019 and prospectively from October 2019 to December 2021. Follow-up will be conducted until December 2023. All available clinical features, detailed medical history, medication information, genetics information, and clinical outcomes were retrospectively/prospectively documented in an electronic data capture system (EDCs) and regularly monitored for data quality. The Ethics Committee of the Clinical Research Center at Beijing Anzhen Hospital approved the research protocol. All registered patients signed informed consent.
Data for the present study were extracted for consecutive patients registered between April 2018 and December 2018 as part of the PHARM-ACS registry. The inclusion criteria were: 1) age ≥18 years and 2) detection of CYP2C19 genotype (*2, *3, *17 alleles) during hospitalization. The exclusion criteria were: 1) patients who were not prescribed a P2Y 12 inhibitor at discharge, 2) patients who were prescribed cilostazol at discharge, 3) patients who could not cooperate with the research protocol or withdraw informed consent, and 4) patients with in-hospital death that was not due to stent thrombosis.
After obtaining informed consent from the patient, and within 48 h after the order from their doctor came into effect, a genotype test report with drug selection recommendations was reviewed by the pharmacist and returned to the clinician via the electronic medical record system (EMRs). Although ticagrelor was recommended for patients with PM/IM, the clinical characteristics (such as ischemia or bleeding risk factors) and economic conditions of patients were also considered by cardiologists when selecting P2Y 12 inhibitors. The P2Y 12 inhibitor prescribed on discharge was at the discretion of the cardiologist. Because prasugrel is not licensed in China, the maintenance of antiplatelet therapy for ACS patients after standard PCI procedures was recommended as aspirin (100 mg/day) combined with clopidogrel (75 mg/day) or ticagrelor (90 mg/twice a day) for at least 12 months. Prescription of ticagrelor for IM/PM patients at discharge was considered to comply with the genotype-guided therapy.

Follow-Up and Endpoints
All patients had scheduled visits at 6 and 12 months, and annually thereafter. During the examinations, the patients were asked to fill out questionnaires by trained followers. The questionnaire was designed by the researcher team and involved the following content: demographic information, medication information, medication compliance, adverse drug reactions, clinical events and quality of life. Medication compliance was evaluated with the 8-item Morisky medication adherence scale (MMAS-8) score (Krousel-Wood et al., 2009). Information on any outcome event in patients was obtained by the investigator via electronic medical records, telephone, or WeChat, and uploaded to the EDCs.
The primary endpoint was a composite of major adverse cardiovascular or cerebrovascular events (MACCE), including all-cause death, stent thrombosis, stroke, myocardial infarction, and any urgent coronary revascularization within 1 year after the indexed PCI. The secondary endpoint was a composite of MACCE plus unstable angina within 12 months of follow-up. The safety endpoint was clinically significant bleeding events, defined as Bleeding Academic Research Consortium (BARC) class 2 or higher bleeding events within 12 months. All bleeding events were defined according to the BARC criteria (Ndrepepa et al., 2012). BARC class 2 was considered a moderate bleeding event, and BARC class 3 or higher was considered a severe bleeding event. Identification of a MACCE was based on diagnostic records derived from the EMRs or diagnostic reports provided by patients. Self-reported information from patients was adopted to assist in the identification of bleeding events and unstable angina. All clinical events were verified by at least two cardiologists.

STATISTICAL ANALYSIS
Baseline demographic and clinical factors were assessed using descriptive statistics. Frequency (percentage) was used to report discrete variables. Mean ± standard deviation was used to report continuous variables. Unadjusted comparisons between groups were performed using Fisher's exact test, chi-square test, or Student's t-test, as appropriate. Demographic and clinical factors related to P2Y 12 inhibitor selection were determined by univariate and multivariate logistic regression. Patients with no events within 1 year after PCI were censored at the time of the most recent follow-up. The 1-year cumulative event rates of MACCE, MACCE plus unstable angina, and clinically significant bleeding events were plotted using the Kaplan-Meier method and evaluated by the log-rank test. Survival analysis was implemented using Cox proportional hazards models. Three pairs of comparisons were constructed separately, comprising LOF-clopidogrel vs. LOF-ticagrelor, non-LOF-clopidogrel vs. LOF-ticagrelor, and non-LOFticagrelor vs. LOF-ticagrelor. Inverse probability treatment weighting (IPTW) was performed to avoid bias in comparisons. Differences between groups were examined with standardized differences using a 10% threshold to indicate significant meaningful imbalances in covariates (Austin, 2015). Three sensitivity analyses and secondary analyses were conducted.
The baseline demographic and clinical factors in the LOFclopidogrel group, non-LOF-clopidogrel group, and non-LOFticagrelor group were compared with those in the LOF-ticagrelor group ( Table 1). As shown in Table 2, after the values were wellbalanced by IPTW adjustment, the differences between the LOFclopidogrel group and LOF-ticagrelor group were anticoagulant agent use and end-stage kidney disease. The imbalances between the non-LOF-clopidogrel group and LOF-ticagrelor group were ticagrelor use before admission, anticoagulant agent use, and endstage kidney disease. End-stage kidney disease differed between the non-LOF-ticagrelor group and LOF-ticagrelor group.

CYP2C19 Genotypes and P2Y 12 Inhibitor Selection
An overview of the genotype results is provided in Supplementary Figure S1. All genotype results were available before the patients were discharged. Among the total included patients, 826 (60.7%) carried at least one CYP2C19 LOF allele [IM: 660 (48.5%); PM: 166 (12.2%)], and no UM patients were detected. Clopidogrel (67.7%) was the most frequently used P2Y 12 inhibitor, even in patients with a LOF allele (63.4%). Compared with non-LOF patients, ticagrelor was prescribed more frequently in IM/PM patients (36.6 vs. 25.8%) ( Figure 1). Furthermore, 44.0 and 34.7% of PMs and IMs were prescribed ticagrelor for maintenance therapy (Supplementary Figure S2). The clinical factors related to the P2Y 12 inhibitor selection are shown in Supplementary

Clinical Outcomes
The median time from indexed PCI to MACCE or last follow-up was 15.6 (14.5-17.4) months. During the 12-months follow-up period after PCI, a total of 82 (6.0%) patients experienced MACCE, defined as a composite of death, myocardial infarction, stroke, urgent need for revascularization, and stent thrombosis. Clinically significant bleeding events occurred in 25 (1.8%) patients. The distributions of specific clinical events are shown in Table 3. The MACCE rate was higher in the LOFclopidogrel group compared with the LOF-ticagrelor group (7.8   Values are mean ± SD or n (%). If the Standardized Difference ≥0.10, it indicates that the variable is not well balanced between groups after adjusting IPTW. LOF, loss-of-function; PCI, percutaneous coronary intervention; CABG, coronary-artery bypass grafting; STEMI, ST-segment elevation myocardial infarction; NSTEMI, non-ST-segment elevation myocardial infarction; EF, ejection fraction; eGFR, estimate glomerular filtration rate; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker.
Compared with the LOF-ticagrelor group, the incidence of MACCE plus unstable angina was significantly higher in the LOF-clopidogrel group (10.1 vs. 4.6%; log-rank p 0.006; IPTWadjusted HR 2.442; 95% CI 1.566-3.809). There was no significant difference in the event rate of MACCE plus unstable angina between the non-LOF-clopidogrel group and LOF-ticagrelor group or between the non-LOF-ticagrelor group and LOFticagrelor group ( Figure 2B and Table 4). We also compared clopidogrel vs. ticagrelor for the incidence of MACCE plus unstable angina in IMs and PMs. Among IMs, patients who were prescribed clopidogrel had a higher risk of developing MACCE than those who were prescribed ticagrelor (9.5 vs. 4.8%; log-rank p 0.033) (Supplementary Figure S3). In PMs, clopidogrel was associated with a significantly higher incidence of MACCE than ticagrelor (event rate: 12.9 vs. 4.1%; log-rank p 0.050) (Supplementary Figure S4).
During the 1-year follow-up, 25 patients suffered clinically significant bleeding events (BARC ≥2). There was no significant difference in the incidence of clinically significant bleeding events across the four groups, even after distinguishing between severe and moderate bleeding events ( Figure 2C and Tables 3, 4).
After further adjustment of unbalanced covariates between the IPTW-adjusted groups, the HRs of the endpoints remained stable (Supplementary Table S5).

DISCUSSION
In this large single-center observational cohort study, we investigated the effect of CYP2C19 genotype-guided antiplatelet therapy on P2Y 12 inhibitor selection and clinical outcomes. The significant findings were as follows: 1) efficient returned CYP2C19 genotype results did, to a certain extent, guide cardiologists to prescribe ticagrelor for ACS in IM/PM patients; 2) for IM/PM patients, use of clopidogrel was associated with a higher risk of developing MACCE than ticagrelor; and 3) drug selection based on CYP2C19 genotype guidance did not increase the risk of significantly clinical bleeding events within 1 year in patients.
The frequency of CYP2C19 LOF alleles in the present study was 60.7%, which was similar to previous data in other East Asian populations (Xie et al., 2013;Shen et al., 2016;Wang et al., 2016) and higher than data in other populations (approximately 30% in African American and Caucasian populations) (Klein et al., 2019). Because clopidogrel remains the most widely used P2Y 12 inhibitor after PCI, it is vital to verify the effects of CYP2C19 genotype-guided antiplatelet therapy, especially for patients in East Asia. However, there are limited data in East Asia that have focused on CYP2C19 genotype-guided antiplatelet therapy in patients with coronary heart disease. A randomized controlled trial (Xie et al., 2013) and a prospective study (Shen et al., 2016) from China confirmed that personalized antiplatelet therapy based on CYP2C19 genotype after PCI could decrease the rates of cardiovascular adverse events with no difference in bleeding in Chinese populations. A small sample-size prospective study in Japan reported that fewer MACCE and hemorrhagic events were observed compared with conventional treatments after genotype-guided antiplatelet therapy was implemented in patients with ACS (Ozawa et al., 2018). Our real-world data support the above-mentioned studies, MACCE: Major Cardiac or Cerebrovascular Adverse Events, defined as a composite of death, myocardial infarction, stroke, urgent need for revascularization, and stent thrombosis. Bleeding events were defined according to the Bleeding Academic Research Consortium (BARC) criteria. Clinically significant bleeding events was defined as a BARC score greater than or equal to 2. BARC class 2 was considered a moderate bleeding event, and BARC class 3 or higher was considered a severe bleeding event.
Frontiers in Pharmacology | www.frontiersin.org January 2021 | Volume 11 | Article 582929 and demonstrate clinical benefits of genotype-guided antiplatelet therapy.
Our conclusions are also consistent with previous studies conducted in other populations. A multicenter study conducted by the IGNITE network in the United States examined the outcomes of genotype-guided P2Y 12 -receptor inhibitor therapy after PCI, and prasugrel/ticagrelor were recommended for IM/ PM patients. These retrospective data proved that clopidogrel, as a maintenance treatment, had a higher MACCE risk for IM/ PM patients than prasugrel or ticagrelor, especially for those with ACS indications (Cavallari et al., 2018b). Meanwhile, a single-center observational study from the United States, including 1,193 post-PCI patients demonstrated that IM/PM patients receiving prasugrel/ticagrelor had a lower risk of developing cardiovascular adverse events than those receiving clopidogrel, and no increased risk of clinically significant bleeding events was observed (Lee et al., 2018). Recently, the Patient Outcome after Primary PCI (POPular Genetics) trial in the The Netherlands involving 2,751 patients with STEMI who underwent PCI showed that genotype-guided therapy was not inferior to standard treatment with prasugrel or ticagrelor in terms of major cardiovascular adverse events and had a lower incidence of bleeding risk events (Claassens et al., 2019). The Pharmacogenetics of Clopidogrel in Acute Coronary Syndromes (PHARMCLO) trial in Italy also demonstrated that composite endpoints of ischemic events and major bleeding events were reduced in the genotype-guided group compared with the conventional treatment group (Notarangelo et al., 2018).
In this non-intervention study, 36.6% of patients with a LOF allele were prescribed ticagrelor, which was lower that the  corresponding proportions of 60.5% in the IGNITE study (Cavallari et al., 2018b) and 53% in another study (Tuteja et al., 2020). Although important, the CYP2C19 genotype is not the only factor considered when prescribing P2Y 12 inhibitors. Various clinical factors, especially risk factors for bleeding, are related to the use of clopidogrel. Some of these potential factors also make cardiologists very cautious about prescribing ticagrelor. First, previous reports indicated that Asian ACS patients had a higher risk of drug-related bleeding and a lower risk of ischemia than Caucasian ACS patients (Mak et al., 2009;Kumar et al., 2013;Levine et al., 2014). Second, ticagrelor had higher discontinuation rates and costs than clopidogrel (Wallentin et al., 2009;Bonaca et al., 2015). These concerns may stop clinicians from prescribing ticagrelor even after receiving an unoptimistic pharmacogenomics result. However, given the higher risk of ischemia associated with clopidogrel use in patients with a LOF allele, and the clear cardiovascular benefits of using ticagrelor, it is recommended that clinicians should increase the weight of genetics when choosing P2Y 12 inhibitors.
We observed an imbalance between IMs (44.0%) and PMs (34.7%) when physicians implemented genotype-guided drug selection. Such inequalities also existed in other previous studies (Cavallari et al., 2018b;Tuteja et al., 2020). Controversy about the value of IM status in genotype-guided antiplatelet therapy has always existed, despite sufficient evidence that IMs are associated with reduced clinical efficacy of clopidogrel (Xi et al., 2019). Previously, the CPIC recommended an alternative therapy for IM patients (Scott et al., 2013). However, in the clopidogrel drug manual, the black box warning from the FDA only recommends that PMs should receive P2Y 12 receptor inhibitors other than clopidogrel, and does not mention the risk for IMs (Holmes et al., 2010). Our real-world data prove that an increased risk of cardiovascular adverse events exists in both PMs and IMs. This conclusion not only validates the results of the IGNITE study (Cavallari et al., 2018b) and earlier studies (Xie et al., 2013;Shen et al., 2016;Gross et al., 2018;Lee et al., 2018;Notarangelo et al., 2018;Claassens et al., 2019;Khan et al., 2019), but also supports recommendations for prescribing alternative therapies to IMs and PMs from the CPIC (Scott et al., 2013). Therefore, we recommend that cardiologists should pay attention to the risk of cardiovascular adverse events in IMs, and increase the weight of IMs when making drug selection decisions, similar to the case for PMs.
Improving the adoption of genetic testing results is very urgent and challenging. In our study, although all of the genotype results were available before the patients were discharged, our data and those in previous studies (Bagai et al., 2014;Lee et al., 2018;Tuteja et al., 2020) suggested that clinicians were reluctant to switch existing treatments regardless of genetic outcomes. It is necessary to provide feedback on the results of pharmacogenomics tests as quickly as possible, because this can help physicians to adopt CYP2C19 genotype results. We agree that delayed return of genetic test reports may result in lower acceptance of the genetic results (Cavallari et al., 2018a;Tuteja et al., 2020). Regular conduct of clinician education is also essential, because it allows cardiologists to continue to pay attention to pharmacogenomics reports (Bell et al., 2014). Besides, the participation of clinical pharmacists can support the clinical decision-making of cardiologists, which is very important (Owusu-Obeng et al., 2014).
Based on our real-world data, clopidogrel was similar to ticagrelor in reducing cardiovascular clinical events in patients without a LOF allele. Because prescription of clopidogrel has better safety and economic profiles, it is worth encouraging the prescription of clopidogrel for NM/RM/UM patients. For patients prescribed ticagrelor after PCI, early de-escalation to clopidogrel may be a strategy worth considering. The results of The TROPICAL-ACS genotyping substudy (Gross et al., 2018) and a recent observational study (Martin et al., 2020) confirmed that de-escalation to clopidogrel in UM/RM/NM patients did not significantly increase the risk of MACCE compared with continuation of ticagrelor/prasugrel. However, there is very little clinical evidence from the East Asian population to verify this strategy, and this will be the focus of our future studies.
Our study has several limitations. First, due to the inherent limitations of observational studies, the implementation of genotype-guided drug therapy in this study was nonrandomized. Therefore, whether to detect genotypes and whether to obey genotype guidance were dependent on the decisions of individual doctors. Although we tried to balance the covariates at baseline by IPTW adjustment, we were still unable to rule out residual confounding. Second, the proportions of STEMI and non-STEMI were relatively low in real-world ACS patients who received genotype testing, which is inconsistent with epidemiological data (Wang et al., 2020). The reason may be that some patients with myocardial infarction were prescribed ticagrelor upon admission, and their genetic testing was unnecessary. However, in the Supplementary Material, we have demonstrated the clinical benefits of geneguided drug selection in both STEMI/non-STEMI and UA. Therefore, the conclusions of this study are robust. Third, patients who did not have CYP2C19 genotype results were not included in the study. Therefore, we did not explore the factors that influence clinicians' indications and frequencies of CYP2C19 genotype testing. Fourth, since prasugrel is not licensed in China, ticagrelor was the only P2Y 12 receptor inhibitor used for alternative therapy. However, the results of a previous meta-analysis suggested that ticagrelor and prasugrel showed no difference in preventing adverse cardiovascular events and bleeding events in patients with ACS (Khan et al., 2019). Fifth, the conclusions of the study were based on a registration database from a single medical center in China, and thus the results may not be widely applicable to other settings or populations.
In conclusion, this observational study in a real-world setting confirmed that CYP2C19 genomics test results could guide cardiologists to prescribe ticagrelor for IM/PM patients, but this is not universal. In patients with a LOF allele, the incidence of MACCE with clopidogrel was significantly higher than that with ticagrelor. The findings of the present study support the implementation of CYP2C19 gene-guided antiplatelet therapy in clinical practice. However, our conclusions still need to be proven by large multicenter randomized controlled trials.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any qualified researcher.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committee of the Clinical Research Center at Beijing Anzhen Hospital. The patients/participants provided their written informed consent to participate in this study.