VKORC1 and CYP2C9 Genetic Variant-Associated Warfarin Treatment Outcomes and Artificial Intelligence-Based Dosing Optimization Among Chinese Patients

Q: What are the key findings of this study?

Carrying >1 VKORC1/CYP2C9 variant allele was associated with a higher 30-day time in therapeutic range (33% vs 17%, p<0.001) and faster time to therapeutic INR (8.7 vs 13.7 days, p=0.018).

Q: What is the clinical evidence from this study?

Study design: Cohort. Population: Heart valve replacement surgery requiring warfarin anticoagulation (n=146). Intervention: >1 VKORC1/CYP2C9 variant allele during warfarin therapy vs. ≤1 VKORC1/CYP2C9 variant allele. Primary outcome: Time in the therapeutic range (TTR) at 30 days (p=<0.001).

Q: What does this research mean for the field?

Carrying more than one VKORC1 or CYP2C9 variant allele is associated with requiring lower warfarin doses, achieving therapeutic INR more quickly, and having a higher 30-day time in therapeutic range in Chinese patients. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Q: What question did this study set out to answer?

To evaluate warfarin response across VKORC1 and CYP2C9 genotype subgroups and develop an artificial intelligence-assisted dosing regimen for Chinese patients undergoing heart valve replacement.

Fangli Liu; Dongyun Nan; Shaoke Li; Yi Wang; Yiqun Cai; Bo Liu; Jialing Peng; Jinyun Ye; Zilong Pang; Jiexin Deng

doi:10.31083/pharmazie58987

October 23, 2026Die PharmazieOpen Access

VKORC1 and CYP2C9 Genetic Variant-Associated Warfarin Treatment Outcomes and Artificial Intelligence-Based Dosing Optimization Among Chinese Patients

View Full Paper

Ask AI

Bookmark

Share

Key result

Carrying multiple VKORC1/CYP2C9 variant alleles is linked to ~16-point higher 30-day TTR during warfarin initiation.

P<0.001
n=146

Why the study?

Warfarin therapy outcomes in Chinese patients can differ from other populations due to differences in genotype frequencies and demographic characteristics, motivating an investigation of warfarin treatment patterns and dose optimization in this population.

Does VKORC1 and CYP2C9 genetic variation influence warfarin anticoagulation outcomes, and can generative AI-assisted pharmacometric modeling optimize warfarin dosing to achieve target INR faster in Chinese patients undergoing heart valve replacement surgery?

Key Points

To evaluate warfarin response across VKORC1 and CYP2C9 genotype subgroups and develop an artificial intelligence-assisted dosing regimen for Chinese patients undergoing heart valve replacement.
Enrolled 146 adult patients undergoing heart valve replacement surgery receiving warfarin therapy.
Assessed anticoagulation parameters, including international normalized ratio (INR) and time in therapeutic range (%TTR), across VKORC1 (1639G>A) and CYP2C9*3 (1075A>C) genotype subgroups.
Constructed a population-based pharmacometric model and simulated initiation dosing schedules using generative artificial intelligence (GAI).
Patients carrying VKORC1 AA and CYP2C9*1/*1 (73.3%) achieved 39% time in the therapeutic range (%TTR) within week 1 and stabilized at approximately 50% within four weeks.
Patients with >1 variant allele required significantly lower daily hospital doses (2.91 vs. 3.69 mg/day) and achieved a therapeutic INR faster (8.7 vs. 13.7 days) than those with ≤1 variant allele.
A simulated GAI-based dosing protocol with higher loading doses on Days 1–5 and lower maintenance doses thereafter enabled achieving target INR within one week.

Study Design

Type

Cohort (n=146)

Multicenter

Yes

Structured PICO

Does VKORC1 and CYP2C9 genetic variation influence warfarin anticoagulation outcomes, and can generative AI-assisted pharmacometric modeling optimize warfarin dosing to achieve target INR faster in Chinese patients undergoing heart valve replacement surgery?

P

Population

146 Chinese adults (mean age 58 years, 42% female) undergoing cardiac valve replacement surgery and receiving warfarin with a target INR of 2.0-3.0, followed for up to 90 days.

E

Exposure

Warfarin oral anticoagulation prescribed by clinicians based on regular INR monitoring (target INR 2.0-3.0) for up to approximately 90 days post-surgery, followed by construction of a population-based kinetic/pharmacodynamic (K/PD) model and generative artificial intelligence (GAI)-assisted simulation (using Qwen2.5-72B-Instruct and mrgsolve/Python optimization) to propose optimized 21-day warfarin dosing regimens (5-day loading phase and 16-day maintenance phase) for patients with >1 VKORC1/CYP2C9 variant alleles.

C

Comparator

Patients with <=1 VKORC1/CYP2C9 variant allele (and comparisons across major genotype subgroups: VKORC1 AA & CYP2C9*1/*1, VKORC1 AA & CYP2C9*1/*3, and VKORC1 GA & CYP2C9*1/*1) receiving clinician-prescribed warfarin.

O

Outcome

Percentage of time in the therapeutic INR range (2.0-3.0) (%TTR) and time to achieve initial therapeutic INR (2.0-3.0) over up to 90 days post-surgery and in 21-day in silico simulations.surrogate

Main Result

Absolute Event Rate: 33% vs 17%

p-value: p=<0.001

Chinese heart valve surgery patients carrying >1 VKORC1/CYP2C9 variant allele (representing >70% of the cohort) require lower warfarin doses and reach therapeutic INR faster than those with <=1 variant allele, and a generative AI-assisted K/PD modeling approach can optimize loading and maintenance dosing to achieve target INR within one week.

Limitations

Enrolled only patients who underwent heart valve replacement surgeries, limiting confirmed generalizability to other warfarin-treated populations.
Accuracy of AI-assisted dose recommendations is bounded by the structural assumptions of the underlying K/PD model and may not extrapolate to extreme BMI, severe hepatic/renal impairment, or rare genetic polymorphisms.
Lack of prospective real-world clinical validation; in silico simulations cannot fully capture real-world complexities like dietary vitamin K fluctuations, drug interactions, and compliance.
Practical implementation challenges, including lack of insurance coverage for genetic testing and need for a convenient clinical software application.
Formal statistical power calculation was not performed
Plasma pharmacokinetic (PK) data were not collected in the study, requiring PK parameters to be fixed to a previously published model

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

May inform genotype-guided warfarin dosing in Chinese valve patients; extends observational data but leaves prospective AI validation open.

Abstract

BACKGROUND: As warfarin therapy outcomes in Chinese patients can differ from those in other populations owing to differences in genotype frequencies and demographic characteristics (e.g., weight and height), this study was developed to investigate warfarin treatment patterns among Chinese patients and to propose dose optimization. METHODS: This study recruited adults undergoing heart valve replacement surgery who received warfarin treatment. Several anticoagulation response-related parameters were measured in these patients and compared among major genotype subgroups. These data were used to construct a population-based pharmacometric model, and warfarin initiation dosing regimens were proposed using generative artificial intelligence (GAI)-assisted model simulation. RESULTS: We found that those patients carrying the Vitamin K epoxide reductase complex subunit 1 (VKORC1) AA & Cytochrome P450 family 2 subfamily C member 9 (CYP2C9)*1/*1 genotype, which accounted for the majority of patients (73.3%), achieved 39% time in the therapeutic range (%TTR) within the first week of treatment, stabilizing to approximately 50% within four weeks after initiating warfarin therapy. In contrast, those patients carrying the VKORC1 AA & CYP2C9*1/*3 genotype (8.9%) exhibited more pronounced fluctuations in %TTR, while those with the VKORC1 GA & CYP2C9*1/*1 genotype (15.1%) tended to exhibit subtherapeutic %TTR. The GAI-based dosing regimen proposed in this study, consisting of higher loading doses on Days 1-5 and lower maintenance doses thereafter, was found to enable patients to achieve target international normalized ratio (INR) range within one week. CONCLUSION: As over 70% of Chinese patients have the VKORC1 AA/CYP2C9*1/*1 genotype, this means that most carry more than one VKORC1/CYP2C9 variant allele. On average, those patients with >1 VKORC1/CYP2C9 variant allele required lower warfarin doses and reached the therapeutic INR range more quickly than those with ≤1 variant allele. GAI-based approaches can provide value as a means of facilitating dose optimization based on modeling and simulation.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Liu et al. (2026) conducted a cohort in Heart valve replacement surgery requiring warfarin anticoagulation (n=146). >1 VKORC1/CYP2C9 variant allele during warfarin therapy vs. ≤1 VKORC1/CYP2C9 variant allele was evaluated on Time in the therapeutic range (TTR) at 30 days (p=<0.001). Carrying >1 VKORC1/CYP2C9 variant allele was associated with a higher 30-day time in therapeutic range (33% vs 17%, p<0.001) and faster time to therapeutic INR (8.7 vs 13.7 days, p=0.018).

synapsesocial.com/papers/6aee6bc771a7594806220b54 https://doi.org/10.31083/pharmazie58987

View Full Paper

Ask AI

Bookmark

Share

VKORC1 and CYP2C9 Genetic Variant-Associated Warfarin Treatment Outcomes and Artificial Intelligence-Based Dosing Optimization Among Chinese Patients

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

VKORC1 and CYP2C9 Genetic Variant-Associated Warfarin Treatment Outcomes and Artificial Intelligence-Based Dosing Optimization Among Chinese Patients

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading