Approximately 50% of the inter-individual variations in warfarin doses could be attributed to genetic polymorphisms of the CYP2C9 and VKORC1 genes, together with the patient's clinical and demographic factors [19]. Here we report a case of warfarin hyper-responsiveness, manifested by above the therapeutic range INR (ie, a supratherapeutic INR) in response to standard warfarin doses in a young man in his twenties with a history of CD managed by sulfasalazine and azathioprine.
Patients with inflammatory bowel disease (IBD), including CD, are at two to three times higher risk for developing thrombosis. Noticeably, morbidity and mortality linked to thromboembolism are significantly higher in young IBD patients and those who are hospitalized due to a flare [20]. Moreover, patients with Crohn's disease may also have reduced small intestine absorption of drugs, including warfarin, as a result of losing the effective surface area secondary to chronic inflammation, ulcerative lesions, or resection [7, 20].
On the other hand, azathioprine, indicated in a variety of autoimmune disorders such as CD, was reported to induce warfarin resistance and therefore a dose-dependent increase of at least 2.5-fold in warfarin dose requirement with the initiation of azathioprine at 75–200 mg daily [21]. Taken together, CD per se and concomitant azathioprine are expected to increase warfarin dose requirements in our case patient. Intriguingly, our patient attained high INR at a standard dose of warfarin, which could have predisposed him to serious bleeding. Due to the failure of non-genetic factors guided warfarin dose prediction, it was rational to investigate the underlying genetic factors that may elucidate this patient's hyper-responsiveness to a standard dose of warfarin.
Based on the FDA’s label and dosing algorithms tailored to individual genetic factors, the VKORC1 − 1639G > A polymorphism (the AA genotype) demands lower dose requirements of warfarin [22]. Moreover, patients with CYP2C9 *1/*3, *2/*2, *2/*3, and *3/*3 genotypes are associated with lower warfarin dose requirements accompanied by a greater tendency to experience bleeding complications with standard warfarin dosing [22]. However, genotyping analysis of our patient proved the absence of the most clinically relevant CYP2C9 reduced function alleles (namely the *2,*3, *4, *5, *8, and *14) and heterozygosity of VKORC1 − 1639G > A.
Accordingly, the calculated daily warfarin dose for our patient was estimated to be 6.6 mg/day based on the Gage algorithm that takes into account the patient's environmental factors and his genotype [23, 24]. Almost identically, a daily dose of 6.7 mg was conceived based on the International Warfarin Pharmacogenetics Consortium (IWPC) algorithm advocated by the Clinical Pharmacogenetics Implementation Consortium (CPIC) [24, 25]. However, such algorithm-guided dose was found to be three times higher than the real-world dose (2.5 mg/day) needed to reach and maintain INR values within the target range.
Puzzled by the failure of the different algorithm-guided warfarin dose estimations, we sought to take advantage of the lengths of the PCR products and their inclusiveness of other alleles with unknown impact or with proven in vitro reduced activity but not established clinical phenotype.
We explored the presence of unknown impact alleles (i.e., *20, *41, *47 *62, *63, *66, *68) and those with proven in vitro reduced activity (i.e., *24, *26, *33, *42, *43, *45, *46, and *55) [26]. Of these investigated allelic variants, one copy of the CYP2C9*46 (3623G > A, Ala149Thr) was detected, which makes our patient heterozygous for this allele.
Our search for the predicted phenotype of the CYP2C9*46 allele revealed paucity of relevant data. A single study by Dai and colleagues adapted an in vitro setting approach to investigate the impact of the CYP2C9*46 allele, among others, on the functionality of the encoded enzyme [18]. Based on the differential clearance rate of diclofenac as a substrate, Dai's findings proved a relatively reduced enzymatic activity of the CYP2C9*46 allele in comparison with the wild type as well as the well-established reduced function CYP2C9*3 allele [18]. Nevertheless, no in vivo investigation nor clinical evidence has supported this in vitro finding, partially due to the rarity of this allele and/or its detection in individuals not on any of the CYP2C9 substrate drugs. Building on the findings of these in vitro comparisons, we speculate a reduced enzymatic activity of the CYP2C9*46 allele by more than 90%.
The calculated daily dose for a carrier of the CYP2C9*1*3 and VKORC1 -1639GA genotype is suggested to be 4.5 mg based on the Gage algorithm [24]. Hence, warfarin dose requirement is expected to be lower than 4.5 mg daily for patient(s) with the CYP2C9*1*46 and VKORC1 GA genotype.
This is the first report of a patient on warfarin with the CYP2C9*1*46 genotype. The CYP2C9*46 allele has not been previously reported in populations other than Han Chinese [27]. The genetic makeup of the Middle East populations, including Syrians, is thought to be similar to that of the Caucasians' rather than East Asians' [28]. However, this case of a Syrian carrier of the CYP2C9*46 allele sheds the light on the diversity of the Middle Eastern populations due to human migrations, wars, and trade, which resulted in a remarkable ethnic, cultural, and genetic diversity. The currently implemented dosing algorithms (i.e., Gage, IWPC, and CPIC) take into consideration only the well-characterized and relatively frequent allele variants that result in altered activity or inactive proteins. Nevertheless, our data suggest that more attention should be paid to subjects carrying the corresponding infrequent unknown function or in vitro predicted reduced enzymatic function variants CYP2C9 alleles when prescribing warfarin. Incorporating this and other less well-characterized alleles, and probably genes, in warfarin dosing algorithms may contribute to the enhancement of their performance in the prediction of warfarin optimal dose.