Implementing a Pharmacogenomic-driven Algorithm to Guide Antiplatelet Therapy among Caribbean Hispanics: A non-randomized prospective cohort study

Abstract

Background: After percutaneous coronary intervention (PCI), clopidogrel resistant patients are at an increased risk of major adverse cardiovascular and cerebrovascular events (MACCEs). We aimed to assess whether genotype-guided selection of oral antiplatelet drugs using a clinical decision support (CDS) algorithm reduces the occurrence of these ischemic events and improves outcomes among Caribbean Hispanic patients from Puerto Rico, who are underrepresented in clinical pharmacogenomic (PGx)-guided implementation studies.

Methods: Individual platelet function testing (PRU) measures, CYP2C19*2 and PON1 rs662 genotypes, clinical and demographic data from 8 medical facilities were included. Patients were separated into standard of care (SoC) and genotype-guided groups (150 each). Risk scores were calculated based on a previously developed CDS risk prediction algorithm designed to make actionable treatment recommendations for each patient. Alternative therapy with ticagrelor was recommended for patients with a high risk score ≥2. Statistical associations between patient time free of MACCEs and predictor variables (i.e., treatment groups, risk scores) were tested in this population using Kaplan-Meier survival analyses and Cox proportional-hazards regression models.

Results: Median age of participants is 67 years; BMI: 27.8; 48% women; 14% smokers; 59% with type-2 diabetes mellitus (T2DM). Among patients with high-risk scores who were free from MACCE events 6 months after coronary stenting, genotype-driven guidance of antiplatelet therapy showed superiority over SoC in terms of reducing the incidence rate of atherothrombotic events.

Conclusions: The clinical utility of our PGx-driven CDS algorithm to reduce the incidence rate of MACCEs among post-PCI Caribbean Hispanic patients on clopidogrel was externally demonstrated.

Clinical trial registration unique identifier: NCT03419325.

Keywords: Adverse Cardiovascular and Cerebrovascular Events; Antiplatelet Therapy; Cardiovascular Diseases; Caribbean Hispanics; Clinical Decision Support Tool; Clopidogrel; Pharmacogenomics; Polygenic Risk Factor.

Figure 1.

Experimental design. Matched non-concurrent sub-cohort of patients used as SoC controls. Data censored for 6 months. Further details can be found elsewhere.

Figure 2:

Kaplan-Meier plot of time free of MACCEs plus major bleeding episodes combined at 6 months of follow-up, grouped by treatment interventions (i.e., SoC vs. Genotype-guided groups). The adjusted hazard ratio (HR) and the corresponding log-rank test p-value are presented. Data censored for 6 months.

Figure 3.

The PGx-guided CDS algorithm reduced the incidence rate of MACCEs and major/minor bleedings in Caribbean Hispanic patients. The incidence of MACCEs and major/minor bleeding events in the SoC group was 40% while in the genotype-guided group was 23%. The odds ratio (OR, 95%CI) and p-value of the association were computed using a one-sided chi-square test. Data are presented as frequency of events in patients.

Figure 4:

Kaplan-Meier plots of time free of MACCEs grouped by risk scores (6 months of follow-up). Panel A depicts the comparison between SoC and genotype-guided treatment groups for low-risk patients (risk score <2 whereas Panel B depicts the comparison between SoC and genotype-guided treatment groups for high-risk patients (risk score ≥2). Hazard ratios (HRs) and corresponding log-rank test p-values are shown, and data censored for 6 months.

Figure 5.

Forest plot of unadjusted hazard ratios (±95%CI) of MACCEs and bleeding for comparisons between treatment groups (i.e., PGx CDS vs SoC) and after stratifying by risk subgroups (high-risk) and relevant covariate (T2DM).