Key result
Adding VKORC1 testing to clinical factors cuts acenocoumarol dose prediction error by ~41%.
Why the study?
Acenocoumarol maintenance dose varies with clinical and pharmacogenetic factors, but evidence for predicting stable weekly dosing in older Southeast Asian patients with atrial fibrillation is limited.
Does an integrated pharmacogenomic-clinical model incorporating VKORC1 -1639G>A genotype improve the prediction of stable weekly acenocoumarol maintenance dose compared to a clinical-only model in older Vietnamese patients with atrial fibrillation?
Observational (n=161)
Yes
Does an integrated pharmacogenomic-clinical model incorporating VKORC1 -1639G>A genotype improve the prediction of stable weekly acenocoumarol maintenance dose compared to a clinical-only model in older Vietnamese patients with atrial fibrillation?
Adding VKORC1 -1639G>A genotype to age and Charlson Comorbidity Index substantially improves the prediction of stable weekly acenocoumarol maintenance dose in older Vietnamese patients with atrial fibrillation, though external validation is needed.
Answers are grounded in this paper and cite their sources.
May refine acenocoumarol dosing in older Vietnamese AF patients; extends prior models but remains hypothesis-generating pending external validation.
Background: Acenocoumarol maintenance dose varies with clinical and pharmacogenetic factors, but evidence in older Southeast Asian patients is limited. Objective: To develop and internally validate a pharmacogenomic–clinical model predicting stable weekly acenocoumarol dose in older Vietnamese patients with atrial fibrillation. Methods: We studied 161 patients aged ≥ 60 years receiving stable acenocoumarol therapy. Clinical factors, Charlson Comorbidity Index (CCI), CYP2C9 *3 and VKORC1 − 1639G>A were assessed. A clinical model (age and CCI) was compared with an integrated model adding VKORC1 . Apparent R2 , mean absolute error (MAE) and root mean squared error (RMSE) were optimism-corrected using 2,000 bootstrap resamples. Calibration, residual diagnostics and HC3 robust inference were assessed. Results: Mean age was 75.64 ± 7.74 years; 56.52% were women. The VKORC1 A-allele frequency was 82.61%, and CYP2C9 *3 frequency was 1.55%. Median stable doses were 5.0, 7.0 and 14.0 mg/week in AA, GA and GG carriers, respectively (P< 0.001). The integrated model improved apparent R2 from 0.180 to 0.788; optimism-corrected R2 was 0.158 versus 0.783. Optimism-corrected MAE was 1.187 versus 0.698 mg/week and RMSE was 1.667 versus 0.849 mg/week. Heteroscedasticity was detected (Breusch–Pagan P=0.005), and HC3 inference retained significant associations for age, CCI and VKORC1 . Excluding three GG carriers reduced adjusted R2 to 0.638. Conclusion: Incorporating VKORC1 improved the internally assessed performance of an exploratory model for estimating stable acenocoumarol maintenance dose. However, the high explained variance was partly influenced by three GG carriers, all receiving 14.0 mg/week; this small subgroup limits the precision and generalizability of genotype-specific predictions. Independent external validation is required before clinical implementation.
No takes yet. Share an insight, caveat, or question.
Lâm et al. (2026) conducted an observational in Atrial fibrillation requiring acenocoumarol anticoagulation (n=161). Pharmacogenomic–clinical model incorporating VKORC1 −1639G>A genotype, age, and Charlson Comorbidity Index vs. Clinical model (age and Charlson Comorbidity Index alone) / VKORC1 −1639AA reference genotype was evaluated on Stable weekly acenocoumarol maintenance dose prediction (mg/week; model explained variance and prediction error) (Optimism-corrected R2 0.783 vs 0.158 (apparent R2 0.788 vs 0.180; optimism-corrected MAE 0.698 vs 1.187 mg/week), p=<0.001). Adding VKORC1 −1639G>A genotype to age and CCI improved optimism-corrected R2 for stable weekly acenocoumarol dose from 0.158 to 0.783 and lowered MAE from 1.187 to /week.
Synapse has enriched 10 related papers. Consider them for comparative context: