TABLE 4.
Cox proportional hazards regression with time-varying analysis assessing the association of additive PK-related genetic variant models with DOACs bleeding risk.
| CYP3A4 (rs35599367) | HR (95%CI; p-value) |
| AA vs. AG vs. GG genotypes | |
| Model 1 | 0.876 (0.691–1.110); 0.274 |
| Model 2 | 0.891 (0.708–1.122); 0.327 |
| CYP3A5 (rs776746) | HR (95%CI; p-value) |
| GG vs. GA vs. AA genotypes | |
| Model 1 | 0.960 (0.685–1.347); 0.814 |
| Model 2 | 0.943 (0.687–1.294); 0.716 |
| CYP2J2 (rs890293) | HR (95%CI; p-value) |
| GG vs. GT vs. TT genotypes | |
| Model 1 | 1.133 (0.873–1.471); 0.349 |
| Model 2 | 1.131 (0.871–1.468); 0.357 |
| ABCG2 (rs2231142) | HR (95%CI; p-value) |
| CC vs. CA vs. AA genotypes | |
| Model 1 | 1.076 (0.882–1.314); 0.469 |
| Model 2 | 1.055 (0.863–1.289); 0.602 |
| ABCB1 (rs4148732) | HR (95%CI; p-value) |
| GG vs. GA vs. AA genotypes | |
| Model 1 | 1.113 (0.969–1.277); 0.129 |
| Model 2 | 1.096 (0.956–1.256); 0.188 |
| ABCB1 C-G-C diplotypes | HR (95%CI; p-value) |
| Homozygous vs. hetero vs. other | |
| Model 1 | 0.999 (0.868–1.148); 0.983 |
| Model 2 | 1.027 (0.895–1.179); 0.707 |
Model 1: Unadjusted model. Model 2: Fully adjusted for age, previous bleeding, Elixhauser comorbidities score, previous thromboembolism, smoking, normalized dose, and DOAC. The underlined genotypes were encoded as the risk genotype in the additive genetic model, and thus the HR for each variant was hypothesized to be > 1.