Does prophylactic antibiotic administration for tooth extraction affect PT-INR in patients taking warfarin?

Various antibiotics and analgesics have been reported to interact with warfarin [1‐5]. When tooth extraction is performed, patients are usually prescribed antibiotics to prevent surgical site infection (SSI) or infective endocarditis (IE) and analgesics to decrease pain. These drugs may also interact with warfarin and produce a clinically significant alteration in anticoagulation status. However, most previous reports were on long-term treatment cases and cases where oral intake was not possible [1‐5]. To our knowledge, there is only one report that investigated the effects of medication that was administered for just a few days for tooth extraction on the prothrombin time-international normalized ratio (PT-INR) [6]. This study investigated the effect of azithromycin (AZM) on INR values in patients taking warfarin [6].

In the present study, we retrospectively investigated the effects of various antibiotics administered during tooth extraction on the INR values in patients who were on stable warfarin therapy.

In the present study, we investigated the effects that various antibiotics and analgesics taken during tooth extraction had on the INR values of patients taking warfarin. Initially, we classified the patients into two groups (INR value increasing group and decreasing group after tooth extraction) and compared these two groups. However, there were no significantly different factors (e.g., warfarin dose, number of extracted teeth and type or dose of antibiotics) between the two groups (Additional file 1: Table S1). In addition, the increase/decrease values were mostly within 20% of the INR before extraction, which is within the range of error. Therefore, we focused on the fact that the medication before and after tooth extraction hardly raised the INR value sharply and we presented three patients in which the value increased beyond the therapeutic range. The INR values before tooth extraction in these three patients were close to 3.0. The INR value increased by more than twice as much in 1 of 110 patients (0.9%).

According to the Guidelines for Patients on Antithrombotic Therapy Requiring Dental Extraction 15, tooth extraction can be safely performed without interrupting warfarin when the INR value is below 3.0 [7]. In the present study, when the INR values were below 3.0 in all patients, there were no cases of hemorrhages that required systemic treatment (e.g., vitamin K or clotting factor), and no thromboses (e.g., cerebral embolisms). Post-extraction hemorrhage that required additional treatment such as re-suturing, was observed in seven out of 110 patients (6.4%) (Table 1). In the past, many reports have investigated post-extraction hemorrhages in patients taking anticoagulants, and reported that the incidence of post-extraction hemorrhages was 0–26% [11‐15]. Our results were similar to those of other reports [11‐15].

Some reports have shown that an increased age (> 75 years old), male, high doses of warfarin, renal failure or liver failure, diarrhea, and drug interactions can all cause increases in INR values [16, 17]. Regarding the interaction between antibiotics and warfarin, Rice et al. [1] conducted a review of many reports and reported that many antibiotics increased INR values, although the duration of administration varied. The mechanism by which antibiotics increase the action of warfarin is known to alter the intestinal flora and decrease the production of vitamin K, thereby enhancing the action of warfarin. This mechanism was applicable to cephalosporins (e.g., CFPN-PI) and penicillin antibiotics (e.g., AMPC). Antibiotics also inhibit cytochrome P-450 (CYP) in the liver, increasing the concentration of warfarin in the blood [1]. This mechanism was applicable to macrolide antibiotics (e.g., AZM). No reports have investigated the effect of surgical invasion by tooth extraction on the INR. Several studies have shown that infection and inflammation decrease the expression and activity of CYP, resulting in decreased drug clearance [1, 3]. Other studies have reported that infection itself affects the metabolism of warfarin [3, 4]. These reports may suggest that tooth extraction affects the INR values. In the present study, Case B, whose INR value more than doubled 1 week after tooth extraction, was the only one patient who underwent additional treatments for two post-extraction hemorrhages. Case B might have been the most invasive case and had evidence of an infection accompanied by necrotic tissue and delayed wound healing 1 week after tooth extraction.

Several reports have investigated the relationship between various antibiotics and INR values [6, 9, 18]. Ghaswalla et al. [18] reported that there was a significant interaction between time and antibiotics on the INR values of elderly (> 65 years old) patients who were on stable warfarin therapy (CFPN-PI, AMPC, AZM, levofloxacin [LVFX]). AZM in particular, which has a significantly long half-life, has been widely discussed in this context [6, 9]. Glasheen et al. reported that the INR value increased beyond the therapeutic range in 31% of AZM cases and 33% of levofloxacin LVFX cases 1 week after oral administration. This was seen in patients on stable warfarin therapy [9]. On the other hand, Kusafuka et al. [6] reported that changes in INR values 1 week before and after tooth extraction were not statistically significant (2-factor analysis of variance, not significant [NS]) when 18 patients taking warfarin were administered AZM. In the present study, the changes in INR values throughout the study were not statistically significant (2-factor analysis of variance, NS) for all antibiotics, including AZM. However, in 3 patients whose INR values were close to 3.0 before tooth extraction, the INR values increased beyond the therapeutic range. This result indicates that surgeons have to pay attention to medications when the INR value is close to 3.0 before the tooth extraction.

Although most NSAIDs are known to enhance the action of warfarin [5], APAP also requires discussion [19]. Cardeira et al. [19] conducted a review of many reports and reported that taking APAP was associated with a mean INR increase of 0.62 compared to placebo, for patients taking warfarin. However, in all reports used in this review, the duration of APAP treatment was longer than 4 weeks. In the present study the APAP treatment was just for 3–7 days after tooth extraction, so there may have been no significant association between APAP and the increase in the INR values 1 week after extraction. Surgeons need to be wary about prescribing APAP long term.

This study has some limitations. First, there is a possibility of unknown confounding factors and factors not studied (e.g., the presence of diarrhea) due to the retrospective nature of this study. Second, the number of patients was small depending on the antibiotics prescribed. Third, frequent PT-INR measurements are highly invasive for patients, so we have to set criteria for measurements based on patients. In the present study, the INR value increased beyond the therapeutic range in 3 out of 110 patients (2.7%). The INR values were close to 3.0 before tooth extraction in these three patients. Our results suggest that surgeons have to take precautions before performing tooth extraction when INR values are close to 3.0. In addition, in those cases, measuring INR values 1 week later may be useful. Finally, in our hospital, the patients have been routinely prophylactically administered antibiotics for a few days after tooth extraction. However, in recent years, we have been trying to reduce the administration of unnecessary antibiotic prophylaxis to prevent antimicrobial resistance.

Our results suggest that prophylactic antibiotic administration has little effect on INR values when patients on stable warfarin therapy undergo tooth extraction. Surgeons have to take attention if the patients whose INR values are close to 3.0 before their extraction.