Adapting warfarin to Asian genetics


A genetics-based approach for tailoring patients’ initial dose of the blood thinner warfarin could make the dosing process more efficient—even in Asia’s genetically diverse population.


Despite its humble beginnings as a rat poison, warfarin has risen to become a widely used drug for preventing blood clots in patients at risk of strokes or heart attacks. Doctors have been prescribing warfarin as an anticoagulant for more than 60 years, but determining the right dose for each individual is still notoriously tricky, especially when initiating therapy. A key reason for this is that the maintenance dose—the amount of the drug that needs to be administered to maintain an effective concentration in the blood—is influenced by a multitude of factors, with genetics playing a large part. In fact, variations in two genes called VKORC1 and CYP2C9 account for up to 40 to 45 percent of the variability in maintenance dose requirements among individual patients.

In light of this, genotype-guided dosing which aims to determine the most effective the warfarin dose according to a person’s genetics, could hold the answer. However, findings from a small number of clinical trials on this approach in Europe and North America have been mixed. Meanwhile, studies in Asia have been comparatively overlooked—that is until two recent reports from researchers in Singapore began filling in this important gap in our knowledge.

Speedier dosing

Determining a patient’s maintenance dose can be a long and tedious process that sometimes extends over several weeks. Patients are typically given an initial concentration of warfarin and monitored using a standardised value called the international normalised ratio (INR). The dose is then adjusted until the INR is stable within a narrow therapeutic range. The number of adjustments, known as titrations, needed to achieve this can be highly variable.

Promisingly, studies in Europe suggest that genotype-guided dosing can speed up the time to reach a stable dose. Professor Goh Boon-Cher, a Clinician Scientist at the Cancer Science Institute of Singapore and Deputy Director of Research at the National University Cancer Institute, wanted to know whether this held true in Asia, where genetic variations in VKORC1 and CYP2C9 differ greatly among people of Chinese, Malay and Indian descent—three of the continent’s major ethnic groups.

To this end, Goh and his team randomly assigned 322 patients who were representative of the ethnic diversity of the Asian population into either a traditional dosing group or a genotype-guided dosing group. The first group received a standardised initial dose of warfarin that was based on age. In the genotype-guided group, the dose was calculated using an algorithm that considered each patient’s age, weight, and the presence of specific genetic variations in VKORC1 and CYP2C9.

After monitoring patients’ INR, performing drug titrations over 14 days and following up with patients up to 90 days later, the team found that patients in the genotype-guided dosing group required fewer titrations to reach a stable dose over both time periods. These findings, published in BMC Medicine 1, have significant practical implications for patients. “Lower frequency of dose titrations means fewer clinic visits for patients,” explained Goh, adding that this would enhance the convenience of blood tests for patients who must travel long distances to a hospital, saving them both time and resources.

Tailoring benefits to Singapore

A key difference between genotype-guided and traditional dosing approaches lies in the way the initial warfarin dose is selected. As well as the clinical variables that some traditional methods account for, genotype-guided algorithms factor in genetic variations in VKORC1 and CYP2C9 that are known to be present in the population of interest.

Selecting the right dosing algorithm can therefore be pertinent for precise dosing. In a study2 published in European Cardiology Review, Dr Grace Shu-wen Chang from Khoo Teck Puat Hospital and NUS Department of Pharmacy Associate Professor Doreen Su-Yin Tan propose that an inappropriate choice of algorithm could explain why the Clarification of Optimal Anticoagulation Through Genetics (COAG) trial3, a large clinical study on genotype-guided warfarin dosing in North America, found that genotype-guided dosing did not improve anticoagulation control in their tested cohort of over 500 patients compared to the control group.

Chang and Tan point out that despite 30 percent of the trial participants being of African-American ancestry, the trial and dosing strategy did not account for the rare genetic variations in CYP2C9 found in this population.

To determine which algorithm would be best for predicting patients’ maintenance dose in Singapore’s genetically diverse population, Chang and Tan compared the algorithm used in the COAG trial with the one delivered by Goh’s team, which was developed and validated locally. Surprisingly, both algorithms predicted patients’ actual maintenance doses with similarly high accuracy, suggesting that either one would be effective for genotype-guided dosing in Singapore.

According to Tan, who pioneered the iRight4Me Yishun Health strategic plan to deploy warfarin and CYP2C9 genotyping in clinical tests, genotype-guided dosing may be of most benefit to certain patient populations, such as newly-initiated high risk patients who need to be pushed into therapeutic INR range quickly or those concurrently on interacting drugs like the antibiotic rifampicin.

“Patients who require rifampicin cannot be given other direct-acting oral anticoagulants because the anticoagulation effects cannot be measured. Thus, we have to use warfarin for anticoagulation and try to maintain INR within a therapeutic range as much as possible” explained Tan. “We used warfarin genotyping in these patients and managed to maintain them in excellent time-in-therapeutic range while starting and stopping rifampicin.”

Despite the potential benefits of such an approach, Goh anticipates that a significant hurdle in bringing a tailored dosing strategy to the clinic may lie in convincing physicians to adopt it in the first place. “Implementing genotype-guided dosing clinically would need a mindset shift, increased awareness, education, and development of clinical decision support infrastructure,” he said.

Tan echoed Goh’s sentiments, adding that the long turnaround time, limited access and high cost of genotyping also further deterred the routine ordering of genotype testing. “The key factor is having pharmacists who are trained to manage warfarin and deploy genotyping calculations into clinical assessment and titrations,” Tan concluded.


1 Syn, N.L., Wong, A.L-A., Lee, S-C., Teoh, H-L., Yip, J.W.L., et al. Genotype-guided versus traditional clinical dosing of warfarin in patients of Asian ancestry: a randomized controlled trial. BMC Medicine 16, 104 (2018).
2 Chang, G.S., Tan, D.S. Using pharmacogenetic testing to tailor warfarin therapy: The Singapore experience and what the future holds. European Cardiology Review 15, e53
3 Kimmel, S.E., French, B., Kasner, S.E., Johnson, J.A, Anderson, J.L., et al. A pharmacogenetic versus a clinical algorithm for warfarin dosing. N Engl J Med 24. (2013)