Table 3.
Univariate analysis of AUC of dose-adjusted tacrolimus trough levels for CYP3A4 and CYP3A5 variants.
| Gene | Variant | Genetic Model | AUC (ng/ml/mg/day * days) of dose-adjusted tacrolimus trough levels | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Day 1-7 | Day 8-14 | Day 1-16 | |||||||||
| Median ± IQR | Effect size | P value | Median ± IQR | Effect size | P value | Median ± IQR | Effect size | P value | |||
| CYP3A4 | *1B, T>C rs2740574 |
dominant: TC vs TT |
3.8 ± 3.5 vs 5.8 ± 4.2 |
-1.6 (-3.4, -0.02) |
0.046§ | 3.8 ± 3.9 vs 4.6 ± 3.4 |
-0.77(-2.3, 0.69) | 0.249 | 9.3 ± 10 vs 12 ± 8.6 |
-3.2 (-6.8, 0.33) |
0.071 |
| log additive: 0, 1, 2 |
5.5 ± 3.9 | -1 (-2,0) | 0.048§ | 4.5 ± 3.5 | 0 (-1, 1) | 0.244 | 12 ± 8.9 | -1 (-4, 1) | 0.072 | ||
| CYP3A4 | *22, G>A rs35599367 |
dominant: GA vs GG |
6.7 ± 3.4 vs 5.3 ± 4.0 |
1 (-0.65,2.9) | 0.173 | 6.9 ± 4.1 vs 4.2 ± 3.5 |
2.4 (1.0, 4.1) |
0.002 | 17 ± 10 vs 12 ± 8.9 | 4.3 (0.6, 9.1) | 0.027§ |
| log additive: 0, 1, 2 |
5.5 ± 3.9 | 1 (-1,2) |
0.17 | 4.5 ± 3.5 | 2 (1, 5) | 0.003 | 12 ± 8.9 | 4 (0, 8) | 0.029§ | ||
| CYP3A5 | *3, T>C rs776746 |
dominant TT + TC vs CC |
2.9 ± 2 vs 6.0 ± 4.1 |
-2.3 (-3.6,-1.2) |
6 × 10-5 | 2.5 ± 1.8 vs 4.9 ± 3.3 |
-1.9 (-3, -0.9) |
6.1 × 10-5 | 6.7 ± 5.9 vs 13 ± 8.7 |
-5.2 (-7.9, -3) |
1.5 × 10-5 |
| log additive: TT = 0, TC = 1, CC = 2 | 5.5 ± 3.9 | -3 (-3,-2) | 1.2 × 10-4 | 4.5 ± 3.5 | -2 (-3, -2) |
1.2 × 10-4 | 12 ± 8.9 | -6 (-8, -4) | 4.8 × 10-5 | ||
| CYP3A | CYP3A4/5 combined genotypes † | dominant: IM + PM vs EM |
6 ± 4 vs 2.8 ± 1.7 |
2.4 (1.2,3.7) | 5.8 × 10-5 | 4.9 ± 3.2 vs 2.4 ± 1.4 |
2 (1, 3.1) |
1.8 × 10-5 | 13 ± 8.5 vs 6.7 ± 3.9 |
5.3 (3.1, 8.1) | 1.1 × 10-5 |
| log additive: EM = 0, IM = 1, PM = 2 | 5.5 ± 3.9 | 2 (1.4,3.1) | 2.9 × 10-4 | 4.5 ± 3.5 | 2.5 (1.7, 3.0) |
4.8 × 10-6 | 12 ± 8.9 | 5.6 (4.3, 7.0) | 1.9 × 10-5 | ||
95% CI, 95% confidence interval; IQR, interquartile range; AUC, area under the curve; P-values are unadjusted P-values. P-values in bold remain significant after adjustment for multiple comparison. Wilcoxon–Mann–Whitney test was used for single variants; the Kruskal–Wallis test or linear median regression analysis was used for CYP3A4/5 combined genotypes (†according to Elens et al., 2013, Lloberas et al., 2017).