Joint relevance of CYP2D6 and H558R in the efficacy and toxicity of flecainide in patients with atrial fibrillation: a cohort study

Abstract Background A cohort study conducted by our group has demonstrated that the heterozygous H558R polymorphism of the SCN5A gene (H558R+/−) is associated with lower toxicity and fewer recurrences at six months in patients with atrial fibrillation (AF) treated with flecainide (1). Moreover, a subgroup analysis of this cohort, currently pending publication, revealed that patients with a poor metaboliser (PM) phenotype of CYP2D6 exhibited a trend towards lower recurrence and toxicity rates at six months, which reached statistical significance at one year of follow-up. Purpose This study aims to determine whether an interaction exists between these two pharmacogenetic predictors. Methods We conducted a cohort study using data collected through indirect methods, ensuring blinding for all parties involved except the statistician. Patients with flecainide prescriptions (100 mg/12 h) and follow-up within our healthcare area between 2017 and 2021 were identified. A pre-specified primary composite endpoint was defined, consisting of six-month recurrence or toxicity, along with a secondary objective extending the follow-up to one year. Multivariable analysis will be performed, including PMs, H558R+/−, and the relevant variables selected using the backward stepwise method (p0.1). Results Table 1 presents the baseline characteristics of the patients. Multivariate analysis of the incidence of the combined outcome of toxicity or AF recurrence at 6 and 12 months was conducted using logistic regression. The variables selected for multivariate analysis were the same at both 6 and 12 months (Figure 1). The PMs phenotype was associated with a non-significant trend towards a lower incidence of the combined outcome at 6 months (p=0.13; OR 0.17; 95% CI: 0.00–1.54), which became significant at 12 months (p=0.03; OR 0.09; 95% CI: 0.00–0.79) (Figure). The H558R+/- polymorphism was significantly associated with a lower incidence of the outcome at 6 months (p=0.02; OR 0.30; 95% CI: 0.10–0.81), but not at 12 months (p=0.09; OR 0.44; 95% CI: 0.16–1.14) (Figure 1). It should be noted that the 12-month analysis included only 88 of the 104 patients, as the remaining patients had discontinued follow-up before 12 months for reasons unrelated to toxicity or recurrence. Conclusions The H558R+/- polymorphism and the CYP2D6 PMs phenotype are independent predictors of a better response to flecainide at 6 and 12 months, respectively. Further studies are needed to confirm the clinical relevance of these findings in real-world practice.Table 1 Figure 1