LAUSR

We thank the authors for their insightful comments on our recent article on the association between the prothrombin time international normalized ratio and concomitant use of antibiotics in warfarin users.1 We agree that for our study to be a clue to personalized precision medicine in the field of pharmacomicrobiomics and pharmacogenomics, we need to evaluate using larger cohorts and investigate associations with Single Nucleotide Polymorphism (SNPs) of the metabolizing enzymes and microbiomics. Also, as noted in our article, our results did not adjust the effect of CYP2C9 and VKORC1 genetic variants and did not cover information on some potential confounding and risk factors for the outcomes (eg, overthe-counter medications, diet, herbal medicines, microbiology data). Our study design was chosen to minimize the influence of these confounding factors; however, further investigation into consideration of pharmacomicrobiomics and pharmacogenomics will contribute to predicting precision warfarin dosing in a patient with antimicrobial therapy. The reports for the prediction of precision warfarin dosing have been published in several races and settings. Nguyen et al2 showed the prediction model for warfarin dosing using multiple linear regression in a Korean population. This model included demographic information, indications, comorbidities, comedications, habits, and genetic factors (VKORC1 genotype, CYP2C9 genotype), and they suggested that VKORC1 genotype, CYP2C9 genotype, age, and weight were the highest contributors in this multiple linear model. Wright and Duffull3 developed a Bayesian dose individualization tool for warfarin. This Bayesian model also used some parameters, including VKORC1 and CYP2C9 genotypes. To develop an optimal design to predict precision warfarin dosing, we have to consider the influence of VKORC1 and CYP2C9 genotypes. Although we did not consider the impact of VKORC1 and CYP2C9 genotypes in our previous study, we plan to investigate the association between warfarin anticoagulation and concomitant use of antibiotics with information on VKORC1 and CYP2C9 genotypes in the Japanese population using a prospective study. Gut microbiota composition is associated with chronic diseases (eg, gastrointestinal inflammatory and metabolic conditions to neurological, cardiovascular, and respiratory illnesses), race, and environmental conditions.4 The composition of Bacteroides spp varies with obesity and age among individuals.5,6 However, our study lacked microbiology data. Thus, we are not certain that susceptibility to Bacteroides fragilis causes a higher risk of excessive anticoagulation. Unfortunately, it is difficult to evaluate the microbiology data with our knowledge; thus, we would like to collaborate with other research groups specializing in pharmacomicrobiomics. Furthermore, we will design the retrospective cohort study using the Swedish resister system and the Japanese medical database system to adjust for some confounders, such as chronic diseases, age, race, and so on. Thanks to their essential comment, we could find the necessity of further investigations to evaluate the association between warfarin anticoagulation and concomitant use of antibiotics using a larger cohort and prospective study. With these further investigations, we may provide novel recommendations for warfarin dose in the clinical setting when starting antibiotic therapy. I hope that our research will contribute to personalized precision medicine in pharmacomicrobiomics and pharmacogenomics.