Rifampicin-warfarin interaction represents a clinically important challenge because rifampicin can substantially reduce anticoagulant efficacy through enzyme induction. Warfarin consists of two enantiomers, with S-warfarin primarily metabolized by CYP2C9 and R-warfarin metabolized mainly by CYP1A2 and CYP3A4. Rifampicin induces these metabolic pathways, resulting in increased clearance of warfarin and reduced anticoagulant activity6,7,8. The onset of enzyme induction generally occurs within 1–2 weeks after rifampicin initiation and gradually resolves after discontinuation as enzyme turnover returns to baseline9,13.
In this case, the patient's warfarin dose increased from 1.5 mg/day before rifampicin initiation to 4.5 mg/day during concomitant therapy, representing an approximately 200% increase. The timing of dose escalation and subsequent INR stabilization was consistent with the expected pharmacokinetic characteristics of rifampicin-mediated enzyme induction rather than a population-specific effect.
However, it should be noted that INR stabilization in this case required approximately 2 months, which is longer than the typically reported 2–4 weeks for rifampicin enzyme induction to reach steady state9,13. This prolonged course may be partly explained by an intervening clinical event: pyrazinamide was discontinued on day 19 of antituberculosis therapy because of drug-induced hepatotoxicity (alkaline phosphatase increased from 117 IU/L to 193 IU/L). Hepatic dysfunction and subsequent recovery during this period could have introduced additional variability in warfarin metabolism, potentially confounding the time course of INR stabilization. Therefore, the 2-month timeline observed in this case should not be interpreted as a typical or expected duration of rifampicin-warfarin interaction, but rather as a reflection of the complex clinical context involving concurrent hepatotoxicity and medication adjustment.
Previous management approaches for patients requiring rifampicin therapy often involve switching from oral anticoagulation to parenteral anticoagulation because of concerns regarding reduced anticoagulant efficacy. However, long-term low-molecular-weight heparin therapy may be associated with practical limitations, including daily injections, treatment burden, and reduced feasibility for prolonged outpatient management.
Compared with DOACs, warfarin offers a unique advantage in this setting because anticoagulant intensity can be directly assessed via INR monitoring, enabling individualized dose adjustment. Although DOACs are effective alternatives for many indications, their exposure may be significantly reduced during rifampicin therapy due to CYP3A4 and P-glycoprotein induction, and routine monitoring equivalent to INR is generally not available8,9. Therefore, in selected patients requiring prolonged rifampicin treatment, carefully monitored warfarin therapy may represent a practical alternative when frequent INR assessment and clinical follow-up are feasible.
The literature on the increase in warfarin dosage during co-administration with rifampicin shows considerable heterogeneity in the magnitude of the required increase. The degree of increase in warfarin dosage ranged from approximately 50% to more than 600%. The time required to regain therapeutic INR after rifampicin discontinuation also varied, often taking between one and five months10,11,12,14,15,16,17,18,19,20 (Table 1, Figure 2). A clinical practice summary on the practical management of warfarin during co-administration with rifampicin is shown in Table 2.
Pharmacogenetic factors contribute substantially to baseline warfarin dose requirements. CYP2C9 and VKORC1 polymorphisms account for a considerable proportion of interindividual variability in warfarin dosing, with some studies suggesting that these genetic factors may explain up to approximately 50% of dose variation19,20. Reduced-function CYP2C9 alleles decrease warfarin clearance and are associated with lower dose requirements21,22. The CYP2C9*3 allele has been reported to occur in approximately 3%–5% of individuals of Chinese ancestry, although the frequency varies among different Chinese cohorts and geographic regions3,21,22,23. In contrast, CYP2C9*2 is uncommon among Chinese populations, whereas VKORC1 sensitivity-associated variants, particularly VKORC1 −1639G>A and VKORC1 1173T, are highly prevalent in East Asian populations17,18.
These pharmacogenetic characteristics, especially the higher prevalence of VKORC1 sensitivity-associated variants, may partially explain why many East Asian patients require lower baseline warfarin maintenance doses compared with individuals of European ancestry23,24. However, these genetic differences primarily influence baseline warfarin sensitivity and dose requirements rather than indicating a different susceptibility to rifampicin-mediated enzyme induction. Therefore, pharmacogenetic and ethnic characteristics should not be used to infer the magnitude of the rifampicin-warfarin interaction, which remains highly individualized and requires patient-specific monitoring and dose adjustment25,26.
This case provides practical evidence supporting a structured pharmacist-supported approach for managing warfarin therapy during rifampicin treatment. The main contribution of this report is not to establish population-level estimates of rifampicin-warfarin interaction but to illustrate how intensive INR monitoring and individualized dose adjustment can maintain safe anticoagulation in a clinically challenging situation.
Future prospective studies involving larger cohorts are needed to better characterize predictors of warfarin dose escalation during rifampicin therapy, including pharmacogenetic markers, inflammatory status, concomitant medications, and clinical characteristics. Development and validation of evidence-based dosing prediction models may further improve individualized anticoagulation management.
This report has several limitations. First, it represents a single patient experience and cannot determine interindividual or population variability. Second, pharmacogenetic testing was not performed; therefore, the contribution of CYP2C9 and VKORC1 variants could not be directly assessed. Third, the independent effects of isoniazid could not be completely separated from those of rifampicin. Fourth, INR variability during hospitalization may have been partially influenced by bridging anticoagulation, as warfarin interruption and enoxaparin overlap may affect early INR interpretation. Therefore, this case should be considered a practical clinical example rather than evidence for population-specific differences in rifampicin–warfarin interaction.
This report highlights the need for frequent monitoring of INR levels and prompt dose adjustments (increase) when co-administering rifampicin with warfarin. An increase in the patient's maintenance warfarin dose was observed (approximately 200%), and a therapeutic INR was achieved after approximately 2 months, which was longer than the typically reported 2–4 weeks, likely due to confounding factors such as concurrent hepatotoxicity during the enzyme induction period. The most important elements of this report are the importance of structured INR monitoring, recognition of the timing of enzyme induction and de-induction, and pharmacist involvement to ensure safe management of this type of drug-drug interaction. Although pharmacogenetic and ethnic status may affect the base warfarin dose requirement(s), this report provides no support for making population-based inferences regarding the strength or magnitude of the rifampicin and warfarin interaction. This case is an illustration of clinical management rather than a comparative ethnic study. Additional studies with larger patient populations are needed to characterize variability in warfarin dose requirements when used concurrently with rifampicin and to develop improved evidence-based monitoring guidelines.