An evidence summary of the management of patients taking direct oral anticoagulants (DOACs) undergoing dental surgery

Abstract

Direct oral anticoagulant (DOAC) drugs (dabigatran, rivaroxaban, and apixaban) have emerged in the last decade to overcome some of the drawbacks of existing medications. To date, little is known about the dental management of patients taking these drugs. This study was undertaken to establish the evidence for the management of patients undergoing dental procedures while taking these medications. A rapid review approach was used to identify clinical and scientific research related to dental surgery performed in patients taking DOACs in order to produce an evidence summary. The rapid review did not identify any systematic reviews or original clinical trials and the overall quality of evidence found was poor. Most of the literature consisted of non-structured review articles and guidance documents based on assumptions from non-dental data and expert opinion, and recommendations on best practice varied throughout. The findings from the review of the literature varied considerably. Currently, recommendations are based on poor quality scientific data and clinical trials are required to establish best evidence-based practice guidance.

There are many dental procedures during and after which one would expect a degree of bleeding. The surgeon should take measures to ensure haemostasis is established and the risk of the patient developing postoperative bleeding is minimized. One factor that can place the patient at increased risk is the use of anticoagulant medication. A number of anticoagulant drugs have been developed in recent years, and the implications for dentistry are not yet clear. This study aimed to establish current approaches to the management of patients taking one of these agents.

Secondary haemostasis relies on fibrinogen being converted to fibrin, which is then cross-linked to form a blood clot ; this process occurs simultaneously with primary haemostasis. The synthesis of factor II, VII, IX, and X in the coagulation cascade is vitamin K-dependent, and the most common oral anticoagulant, warfarin, interferes with vitamin K recycling, affecting the intrinsic, extrinsic, and common pathways. Heparin, a commonly prescribed parenteral anticoagulant, affects the common pathway by indirectly inhibiting both factor Xa and factor IIa to varying degrees.

Although warfarin has been a popular oral anticoagulant drug for over 60 years, it has limitations such as dietary and drug interactions, narrow therapeutic range, and the need for monitoring. One of the major drawbacks of heparin is the need for parenteral administration. Therefore, in recent years new anticoagulant agents have been introduced that claim improved safety and superior therapeutic value compared to their predecessors.

Many of the initial studies described the new anticoagulant drugs as either ‘new’ or ‘novel’ oral anticoagulants (NOACs); however using these terms has limitations. For example, the use of the term ‘No AC’ is used in medicine to denote that the patient is not taking an anticoagulant, which could lead to confusion. Also, naming the drugs new or novel is time-limited. Therefore the terms ‘direct oral anticoagulant’ (DOAC) and ‘target-specific oral anticoagulant’ (TSOAC) have begun to appear in the literature. The former has been adopted by the International Society on Thrombosis and Haemostasis Scientific and Standardisation Committee and therefore will be used throughout this article.

The DOACs have two sub-groups: direct thrombin inhibitors (DTI) and factor Xa inhibitors (FXaI), both of which target one specific factor in the coagulation cascade. These drugs are becoming more frequently prescribed and their use has significant implications for the management of patients requiring anticoagulant therapy. The most striking difference compared to traditional anticoagulants is the elimination of the need for monitoring, even if the patient is to undergo surgery. Secondly, these drugs are administered orally at a fixed daily dose, making patient compliance more straightforward.

Dabigatran etexilate (Pradaxa) was the first oral target-specific anticoagulant to be approved for use, and although other DTIs (argatroban, bivalirudin, desirudin, and lepirudin) have been developed, they are administered parenterally and are reserved for patients with, or at risk of, heparin-induced thrombocytopenia. Dabigatran causes reversible inhibition of both free and fibrin-bound factor IIa (thrombin), a potent agonist of platelet activation and aggregation. The drug spares a small amount of free enzymatically active thrombin for haemostasis. Dabigatran is taken twice daily and reaches a predictable peak concentration 1–2 h after dosing, allowing fixed-dose administration without monitoring. Consequently, although the drug itself is considerably more expensive, dabigatran therapy is more cost-effective than that of warfarin. In hepatic failure, the absorption and excretion of dabigatran is largely unaffected, whereas moderate and severe renal failure leads to drug accumulation. Therefore kidney function needs to be considered when prescribing dabigatran, but also if considering a period of discontinuation prior to invasive surgery.

The RE-LY multicentre trial of 18,113 patients, with blinded evaluation of outcomes, demonstrated fewer thromboembolic events and a reduced bleeding risk with dabigatran compared to warfarin in the management of atrial fibrillation (AF). Additionally, dabigatran has been demonstrated to be as effective as both low molecular weight heparin (LMWH) and warfarin in the prevention of venous thromboembolism (VTE) following hip and knee replacement surgery and in the prevention of recurrent VTE. The National Institute for Health and Care Excellence (NICE) in the UK has recommended dabigatran as an option in the prevention of deep vein thrombosis (DVT) and pulmonary embolism (PE) in patients undergoing hip or knee surgery and in the management of nonvalvular AF. Despite promise, negative findings have also been published in the literature, such as an increased risk of myocardial infarction (MI) in patients taking dabigatran for AF, and dabigatran is contraindicated in patients with mechanical heart valves due to an increased risk of MI or stroke.

Although monitoring is not required for dabigatran, if presented with a bleeding patient it may be desirable to establish the level of anticoagulation. The prothrombin time (PT) is not sensitive to dabigatran-induced anticoagulation, but the activated partial thromboplastin time (aPTT) may be used with caution in establishing over-coagulation in emergency situations or in centres where other tests are not available. Thrombin (clotting) time (TT/TCT) and ecarin clotting time (ECT) are more sensitive methods of testing anticoagulation with DTIs, but are not widely available. A reversal agent is not currently available, but attempts to develop one are showing promise.

Finally, thrombin also plays a part in angiogenesis, the immune response, infection, tumour growth, and endothelial function, and it is not yet clear whether DTIs will also interfere with these other processes.

The FXaIs reversibly bind to and inhibit free and clot-bound factor Xa directly. Rivaroxaban (Xarelto), normally taken once daily, and apixaban (Eliquis), taken twice daily, reach peak plasma concentration after around 2–4 h and 3–4 h, respectively, and are better tolerated than dabigatran. Both drugs can be used with caution (possibly at a reduced dose) in mild or moderate hepatic or renal failure.

Monitoring is not required, but if an assessment of anticoagulant effect is desired, the following investigations may be useful. Apixaban can prolong PT (depending on the laboratory reagent used) and increase aPTT, but does so to varying degrees. Rivaroxaban has a dose-dependent effect with PT with most laboratory reagents. However, to achieve an accurate assessment of anticoagulant effect, chromogenic anti Xa assays, which are not universally available, are required. As with dabigatran, work is being undertaken to develop an antidote to the Xa inhibitors.

The ROCKET and ARISTOTLE clinical trials demonstrated that rivaroxaban and apixaban were as effective as warfarin in the management of nonvalvular AF, with fewer bleeding complications. Both have also been demonstrated to be as effective as LMWH in the prophylaxis and treatment of VTE in orthopaedic surgery patients. NICE recommends rivaroxaban and apixaban as options in the management for each of the latter indications. The FXaIs have also been studied in patients with acute coronary syndrome, but increased major bleeding events were found.

Following a number of clinical trials, another FXaI–edoxaban (Savaysa (USA) or Lixiana (EU))–produced by Daiichi Sankyo, has recently been approved by the US Food and Drug Administration (FDA), has been granted approval marketing authorization in Europe, and is currently being considered by NICE for use in the management of patients with AF and in the treatment of VTE. Betrixaban (proprietary name TBC) from Portola Pharmaceuticals, has completed phase 2 trials for the prevention of VTE in AF patients and for the prevention of VTE following orthopaedic surgery, and is entering phase 3 trials for its use in the long-term prevention of VTE in medically ill patients.

Materials and methods

The ‘evidence summary’ approach used was adapted from the rapid review method of the Knowledge-to-Action (KTA) research programme at the Ottawa Hospital Research Institute in order to avoid some of the disadvantages of full systematic review. Where systematic reviews can take 6–24 months to carry out, a rapid review generally takes less than 5 weeks to complete, and although the latter assembles a smaller volume of information, the search strategy is structured and involves rigorous appraisal of papers. Therefore, it is thought that there is little difference in the overall findings of rapid review when compared to systematic review.

The Web of Science database was searched for each of the generic and proprietary drug names ( Table 1 ) against information specific to dental surgery using a comprehensive list of terms. A second similar search of the PubMed database was performed for completeness. Relevant papers were identified, with meta-analyses and systematic reviews being deemed the most reliable sources, followed by individual randomized control trials, cohort studies, case–control studies, cross-sectional studies, case series/case reports, and expert opinion, respectively.

Table 1
Generic and proprietary names of DOACs.
Generic drug name Proprietary drug name
Dabigatran Pradaxa
Rivaroxaban Xarelto
Apixaban Eliquis

DOAC, direct oral anticoagulant.

Specifically, the literature sought to find information on the management and outcomes of dental treatment in patients taking these drugs. On first pass, abstracts were studied and any that gave reference to both (1) dabigatran, rivaroxaban, or apixaban, and (2) any type of surgical procedures, were included in the next stage. Then, the full text of each remaining paper was examined for information specific to the management and outcomes of any form of dental surgery in patients taking these drugs–the main inclusion criteria. More generic search terms were also applied to widen the search and locate papers missed using the specific drug names. Relevant citations identified within these papers that were not identified in the database searches were also pursued. Only English language papers and human studies were included. Any paper not available electronically was sourced in hardcopy, providing it was held within University of Glasgow libraries. The quality of each eligible full text was evaluated using Critical Appraisal Skills Program (CASP) tools. Three broad questions are sought when using CASP: (1) Is the study valid? (2) What are the results? (3) Are the results useful? Each section is broken down into further questions in order to assess whether the study has value for the reader’s purpose. Finally, sources of grey matter were searched and the information was gathered to generate a narrative of the evidence.

Results

A search of Web of Science for the use of dabigatran in patients undergoing oral surgery or invasive dental procedures yielded 33 papers. On review of the abstracts, four studies were excluded as they were not relevant, one duplicate entry was removed, and one non-English paper was excluded. Of the remaining 27 findings, two were letters to the editor of journals and most were review or guidance documents. A similar search of the same database for the use of rivaroxaban in dental patients found 28 results. Following abstract review, 13 studies were excluded as they were not relevant and one duplicate result was removed. Of the remaining search results, two were letters to the editor of a journal, as described above, and two papers were excluded as they had no reference to dental surgery within the full text. Each of the 10 remaining studies included was a review article. Finally, a search of the literature for dental management of patients taking apixaban found 14 results. On review of the abstracts, two studies were excluded, one of which was a duplicate entry within the search results and one a non-English language paper.

Full texts of the remaining studies were screened and a further five were excluded as they did not give reference to dental or oral surgery procedures, leaving mainly review papers. A further five studies were located using a second search performed on PubMed and by searching both databases using more general search terms for the drug type. The total number of papers and study types found for each of the drugs is outlined in Table 2 .

Table 2
Summary of literature review findings.
Study type Drug
Dabigatran Rivaroxaban Apixaban
Meta-analysis/systematic review 0 0 0
RCT 1 a 0 1 a
Cohort 1 1 1
Case–control 0 0 0
Cross-sectional 0 0 0
Case reports/series 2 0 0
Review (non-systematic) 16 13 11
Other 17 14 12
Total 37 28 25
RCT, randomized controlled trial.

a The randomized controlled trials found were not original studies, rather studies that used the data from RE-LY (dabigatran) and ARISTOTLE (apixaban) clinical trials to obtain dental data.

All of the papers are included in the narrative below, but none of the above studies met all of the CASP criteria, mainly as the studies were not valid and the authors, in the absence of good scientific research, ultimately gave recommendations that were based largely on expert opinion and therefore subject to bias. The majority of the findings were review papers, which were of varying quality, with only three providing details of the search methodology. The conclusions and recommendations found varied and conflicted between papers. An additional 24 sources of information were gathered through searches of a substantial, but not exhaustive, list of alternative or grey literature sources.

Results

A search of Web of Science for the use of dabigatran in patients undergoing oral surgery or invasive dental procedures yielded 33 papers. On review of the abstracts, four studies were excluded as they were not relevant, one duplicate entry was removed, and one non-English paper was excluded. Of the remaining 27 findings, two were letters to the editor of journals and most were review or guidance documents. A similar search of the same database for the use of rivaroxaban in dental patients found 28 results. Following abstract review, 13 studies were excluded as they were not relevant and one duplicate result was removed. Of the remaining search results, two were letters to the editor of a journal, as described above, and two papers were excluded as they had no reference to dental surgery within the full text. Each of the 10 remaining studies included was a review article. Finally, a search of the literature for dental management of patients taking apixaban found 14 results. On review of the abstracts, two studies were excluded, one of which was a duplicate entry within the search results and one a non-English language paper.

Full texts of the remaining studies were screened and a further five were excluded as they did not give reference to dental or oral surgery procedures, leaving mainly review papers. A further five studies were located using a second search performed on PubMed and by searching both databases using more general search terms for the drug type. The total number of papers and study types found for each of the drugs is outlined in Table 2 .

Table 2
Summary of literature review findings.
Study type Drug
Dabigatran Rivaroxaban Apixaban
Meta-analysis/systematic review 0 0 0
RCT 1 a 0 1 a
Cohort 1 1 1
Case–control 0 0 0
Cross-sectional 0 0 0
Case reports/series 2 0 0
Review (non-systematic) 16 13 11
Other 17 14 12
Total 37 28 25
RCT, randomized controlled trial.

a The randomized controlled trials found were not original studies, rather studies that used the data from RE-LY (dabigatran) and ARISTOTLE (apixaban) clinical trials to obtain dental data.

All of the papers are included in the narrative below, but none of the above studies met all of the CASP criteria, mainly as the studies were not valid and the authors, in the absence of good scientific research, ultimately gave recommendations that were based largely on expert opinion and therefore subject to bias. The majority of the findings were review papers, which were of varying quality, with only three providing details of the search methodology. The conclusions and recommendations found varied and conflicted between papers. An additional 24 sources of information were gathered through searches of a substantial, but not exhaustive, list of alternative or grey literature sources.

Discussion

Ideally the review of the literature would have been performed by more than one person, as this would have reduced the risk of missing investigator errors, such as incorrectly omitting relevant papers. Additionally, there were potentially relevant papers found in other languages that were not included due to limited resources.

The low quality of evidence found meant that the CASP tool was perhaps not the most appropriate means of appraising each paper, given that no study fulfilled the full criteria. However, the narrative in the discussion below serves to portray not just the findings but also the quality of the evidence for each of the studies.

No systematic reviews or original clinical trials assessing the dental management of patients taking DOACs were found. Healey et al. took data from the RE-LY trial and evaluated the risk of peri-procedural bleeding for two doses of dabigatran (110 mg and 150 mg) compared to warfarin. Approximately 25% ( n = 4591) of patients in each group underwent surgery over a 2-year period and 10% of these underwent dental procedures. The drugs were discontinued prior to surgery, with the last dose of dabigatran given 35–85 h (mean 49 h) prior to the procedure, compared to 87–144 h (mean 114 h) for warfarin. Although the drug-free duration was shorter for dabigatran, there was no difference in peri-procedural bleeding between the two groups. Once haemostasis was established and deemed stable, the anticoagulant was restarted. The main results were subjected to appropriate statistical analysis and presented using the relative risk and a χ 2 test; no significant difference in bleeding was found between the two drugs. Heparin bridging was used in 15% of patients taking 110 mg dabigatran, 17% of those taking 150 mg dabigatran, and 29% of warfarinized patients, but similar rates of both bleeding and thromboembolic events were seen for the two drugs. One potential flaw with the data was that procedures were deemed ‘major’ if they lasted over 1 h, which is not a true indicator of the risk of bleeding. However, although the RE-LY trial was not designed to analyze peri-procedural bleeding, it had the advantage of having a large number of subjects randomized to each condition and this improved the reliability of the study. The main disadvantage identified for dabigatran was the lack of a reversal agent. There were two main advantages demonstrated in the trial: the period of discontinuation was shorter (which minimizes the risk of a thromboembolic event and allows surgery to commence sooner) and heparin bridging was less frequently required. Also, the short half-life of dabigatran allows a relatively rapid return to normal haemostasis; this in turn eliminated the desire for an antidote in the majority of patients. Healey et al. concluded that for standard bleeding risk procedures, dabigatran should be stopped 2–3 half-lives before the procedure, and for high bleeding risk procedures, dabigatran should be stopped 4–5 half-lives prior to the procedure in addition to testing aPTT or TT.

Garcia et al. also used data from a previous large clinical trial to evaluate the outcomes, in this case for patients taking either apixaban or warfarin who underwent a surgical procedure(s). ARISTOTLE was a double-blind, double-dummy, randomized control trial comparing apixaban with warfarin in 18,201 patients with AF that demonstrated superiority of apixaban over warfarin in the management of nonvalvular AF, with fewer bleeding complications. Garcia et al. used data from this trial to identify all patients who “underwent a procedure for which anticoagulant therapy would, in some clinical situations, be interrupted”. Dental extraction/oral surgery (which would not normally lead to warfarin interruption) accounted for 14.6% of all procedures in ARISTOTLE. In 37.5% of the 9260 procedures included, there was no drug interruption, and in the remaining 62.5%, the duration of interruption ranged from 1 to 7 days, with the majority stopping for between 3 and 5 days. Due to the ARISTOTLE study design, bridging therapy in apixaban patients was used more than may be expected (11.7%) and the results suggested that perioperative bridging is not necessary. One drawback of the analysis of Garcia et al. arises due to the fact that the peri-procedural anticoagulation had not been randomized or controlled in ARISTOTLE. The duration of interruption of apixaban was variable and therefore conclusions cannot be drawn as to the ideal duration of interruption. They concluded that the duration of interruption of apixaban remains dependent on an assessment of the risk of bleeding and the risk of a thromboembolic event. Lopes et al. also assessed peri-procedural complications on analysis of data taken from the ARISTOTLE clinical trial, and although similar results were found to those of Garcia et al., they presented the findings only in a short meeting abstract.

Beyer-Westendorf et al. used information from a prospective data registry of over 2100 patients to study the peri-procedural management of DOACs, allegedly the first of its kind, supporting DOAC interruption without heparin bridging. Although they relied partially on individual patient accounts of issues, on the whole the study was well conducted and well written. Procedures were categorized into minimal procedures (including non-extraction dental procedures) and minor procedures (including dental extractions). The majority of patients in the study were taking rivaroxaban (76%), followed by dabigatran (23.5%) then apixaban (0.5%). The prevalence of bleeding was found to be higher in the groups of patients who had received heparin bridging, as might be expected, without affecting the risk of a thromboembolic event. Heparin bridging was therefore not recommended in the peri-procedural management of patients taking DOACs, and short-term interruption appeared to be safe in the minimal and minor surgery groups. The authors admitted that the study was open to selection bias, and no protocol was in place for the decision on when to use a vitamin K agonist over a DOAC. Therefore, the patients may have had different bleeding and thromboembolic event risks than those found in a random selection. Also, a larger sample size is required, as many of the risk factors were found to be non-statistically significant with wide confidence intervals.

Breik et al. presented a case series of five patients taking dabigatran, summarized in Table 3 . They speculated, following a short review of the literature, that continuation of dabigatran is safe for scaling, restorative treatment, endodontics, and single tooth extractions, providing additional local haemostatic measures (ALHM) are used, but that for multiple dental extractions (number not specified), there should be dialogue with the patient’s physician. The authors suggest that dabigatran cessation should only be performed after liaising with the prescribing physician, as it may be safe for a patient with AF with no history of stroke to stop for a short period of time, but in a patient with recurrent DVT this carries a greater risk.

Table 3
Case series described by Breik et al.
Case no. Details of dental treatment carried out in patients taking dabigatran
1 78-year-old male had extraction of tooth 24 under LA without drug interruption. The socket was sutured and no adverse bleeding was encountered.
2 80-year-old male had extraction of tooth 36 under LA without drug interruption. The socket was sutured and there was no excessive bleeding at the time of the procedure, but the patient reported minor bleeding at home, which was controlled with pressure, and the patient did not feel the need to return to hospital.
3 76-year-old male had extraction of tooth 17 under LA without drug interruption. The socket was sutured and Surgicel was placed. No adverse bleeding was encountered.
4 72-year-old female had extraction of tooth 35, 44, and 46 under LA. Dabigatran was stopped 48 h (no note of restart and/or ALHM used). No adverse bleeding was noted.
5 84-year-old male had extraction of 18 teeth performed under GA along with drainage of swelling without drug interruption. Preoperative INR was 1.5 and aPTT was 47 s. Significant bleeding was noted intraoperatively and wounds were “sutured tightly”. The patient returned to theatre for control of “significant haemorrhage” and dabigatran was then stopped with bleeding under control 24 h later.
LA, local anaesthetic; ALHM, additional local haemostatic measures; GA, general anaesthetic; INR, international normalized ratio; aPTT, activated partial thromboplastin time.

Restarting the drug postoperatively is recommended by the drug manufacturer as soon as haemostasis is achieved, whereas Breik et al. were more cautious with respect to bleeding risk, suggesting 24–48 h postoperatively. Case number five in the series mentioned above, suggests that when significant bleeding is encountered, a positive effect may be gained relatively rapidly following cessation of the drug, in combination with local surgical measures. The bleeding risk may be further increased in patients over 75 years of age taking dabigatran. In this emergency situation, although withholding or delaying a dose of dabigatran may not have been possible, the removal of 18 teeth in one sitting seemed to be a risky strategy that may not have been entirely necessary, particularly in the absence of a reversal agent or well-established protocols for the management of bleeding patients taking dabigatran. This case also highlights the limitations of the international normalized ratio (INR) in patients taking dabigatran, as an INR of 1.5 in a warfarinized patient might be considered a safe level to proceed with dental extractions, whereas the prolonged aPTT may raise more concerns about bleeding risk.

Romond et al. presented a case report of a patient taking dabigatran for AF who underwent eight extractions, alveoplasty, and tuberosity reduction. The patient’s renal function was normal and dabigatran was withheld the night before surgery and resumed the day after. Resorbable gelatine sponges and sutures were placed, following curettage of the tooth sockets. No postoperative bleeding was observed, although the immediate denture may have assisted in maintaining postoperative haemostasis. The patient was reviewed 1 week following the procedure and reported only minimal bleeding. The authors discussed whether the decision to stop the drug was correct and concluded that they did so for three reasons: (1) due to the lack of a reversal agent, (2) because the patient’s physician was consulted and agreed that interruption was appropriate, and (3) due to the type of surgery.

Van Diermen et al. reviewed the literature via a systematic approach to find research relating to dental treatment in patients taking anti-thrombotic medication. The three main DOACs were not specifically included, but results relating to these drugs were found as the search term ‘anticoagulant’ was used. The authors recommended that DOACs should not be interrupted prior to the extraction of up to three teeth, placement of up to three dental implants, scaling and root planing, probing, flap surgery, and alveoplasty, but that the patient should not take their medication between 1 and 3 h prior to dental treatment. A review article, by Mingarro-de-Leon et al., identified only three articles in the dental literature via a structured search method pertaining to DOACs. One of these was a review by Little, published in 2012 when the DOACs were not yet in widespread use; consequently, no dental recommendations were made. The second (non-systematic) review paper by Firriolo and Hupp assumed the bleeding risk with dabigatran and rivaroxaban to be similar to LMWH and is discussed further below. The third, by Fakhri et al., was a review paper with a structured methodology searching for studies with dental considerations for patients on anti-thrombotic therapy. The latter proposed four main considerations in the risk assessment of a patient taking a DOAC undergoing surgery: (1) the half-life of the drug, (2) risk of bleeding from the procedure, (3) co-morbidities, such as renal impairment, and (4) risk of thromboembolic complications. The authors went on to recommend that for high-risk procedures, two doses of dabigatran should be missed preoperatively, reducing to 1 day for moderate bleeding risk procedures, but the definition of moderate and high risk was not clear.

Several non-systematic review articles have suggested the continuation of drug therapy for non-surgical dental extractions. O’Connell and Stassen stated that “Turpie et al. suggest that interruption of rivaroxaban is not required for simple dental extractions”. However, this interpretation appears a little stronger than Turpie et al. may have intended. They provided a recommendation, stressing that it was the opinion of the authors, and stated that “consideration should be given to maintaining the anticoagulant treatment”. Weitz et al. gave no reference source for recommending drug continuation for dental extractions and therefore it must be assumed that the recommendation was based on opinion. The latter suggested that treatment should be carried out more than 10 h after the last dose of dabigatran to ensure levels are at their lowest; however, in practical terms this could be difficult if the patient normally takes their dose early in the morning. A short review by Syyed et al. on the dental implications of patients taking dabigatran placed emphasis on safely carrying out up to three extractions without interrupting dabigatran. This was based on Belgian guidance produced in early 2013, discussed further below, which is based on expert opinion.

A review article by Davis et al. took a reasonable methodological approach and made good use of references. They noted that a Cochrane review carried out in 2010 comparing dabigatran to LMWH or warfarin in the management of VTE in over 27,000 patients undergoing total hip or knee replacement, found no statistically significant difference in bleeding events. Davis et al. concluded, on analysis of the data that they studied, that treating a patient taking dabigatran is comparable to treating a patient taking warfarin with an INR of between 2.0 and 3.0. However, this was based on the target INR for these patients, not the true INR, which may not be reliable bearing in mind that a considerable proportion of patients will often be out with their therapeutic range. Also, the plasma concentration of the DOACs rapidly reaches a peak after administration and then reduces more gradually over the subsequent several hours, but it is not clear at what time following dose administration the patient is thought to have this alleged level of anticoagulation equivalent to an INR of 2.0–3.0. Finally, Davis et al. suggested that consideration should be given to interruption of the drug where multiple surgical extractions or maxillofacial procedures are required.

A review article by Scott et al. studying the management of dental patients taking dabigatran, rivaroxaban, and apixaban, highlighted that restarting a DOAC does not lead to rebound hypercoagulability as with warfarin, and the short half-lives are favourable in that they allow more rapid removal of the anticoagulant effect if this is desired. They also estimated that post-procedural bleeding, if appropriate local measures are taken, “will, as with, warfarin, be relatively rare”. Finally, an investigation by the authors into the health board and health planning unit found two main recommendations being offered: (1) do not interrupt the drug and treat as a warfarinized patient with an INR of less than 4, or (2) avoid invasive treatment until as long as possible following previous dose (24 h for rivaroxaban and 12 h for dabigatran and apixaban, assuming normal renal function) and wait 4 h before recommencing.

Curtin et al. recited the manufacturer recommendation to stop a DOAC 24 h before a surgical procedure that carries a low bleeding risk, restarting as soon as haemostasis is achieved. Rider and Rider gave a brief overview of the DOACs but had few references, presumably largely down to a lack of available literature at the time it was written in mid-2013. The paper focused mostly on dabigatran, giving very little information for rivaroxaban and apixaban. The authors speculated that, due to the rapid off- and onset of dabigatran, cessation of the drug prior to dental surgery is likely to be adopted in the future and given that there is no reversal agent available, careful planning is required for invasive dental procedures. Rider and Rider also stressed that the suggestion to stop an anticoagulant may be an uncomfortable step for the dental profession after many years of proceeding with dental surgery without interruption of warfarin.

A brief review by Cohen et al. failed to use a structured method and was found to be poorly reference and generally poorly written. Two recommendations were offered: (1) for basic oral surgery (with “a few dental extractions or localized periodontal surgery” given as examples in this case), continue the drug but “limit the extent of procedures so as to be able to evaluate the potential for excessive bleeding” in combination with the use of ALHM, and (2) in procedures where “excessive bleeding is anticipated” (with multiple extraction, alveoplasty, and implant placement provided as examples), discontinuation of the drug for 48 h preoperatively may be advisable, in collaboration with the patient’s physician, restarting soon after, providing haemostasis is stable. However, no reference source or justification for these recommendations was provided, bringing into question their validity.

Wynn discussed dabigatran and rivaroxaban in an early review of the literature. The author stressed that stopping dabigatran carries a risk of thromboembolic events and that the drug should be restarted as soon as possible. With respect to rivaroxaban, Wynn recommended consultation with the prescribing physician prior to embarking on invasive dental procedures. The author did highlight, however, that patients taking a DOAC for prophylaxis of DVT/PE following orthopaedic surgery will be taking the drug for a short period of time, allowing elective invasive dental treatment to be deferred. McCormick et al. failed to add to the literature, as the source used as the main reference for recommendations on dental treatment in patients taking DOACs was based on a paper by Gomez-Moreno et al. that was published in 2010. The latter, written prior to widespread use of the drugs, discussed rivaroxaban and dabigatran only and was based largely on speculative expert opinion rather than scientific fact.

A very recent review article, published in September 2015, summarizes many of the papers already discussed above. Many of the important papers discussed above have been omitted, possibly due to the lack of a structured search methodology, and two papers that give no reference to dental procedures have been included. Of the 11 papers found in the study that offer guidance, seven do not recommend interrupting the DOAC prior to minor invasive procedures, including extractions. Finally, the authors state that the risk of a thromboembolic event from stopping a DOAC is comparable to the risk when stopping a vitamin K agonist such as warfarin. However, given that the onset and clearance of each of the DOACs is much faster than warfarin, the drug-free perioperative period is much shorter, and there is no rebound effect, this risk is suggested elsewhere to be a lower risk for the DOACs.

A number of guidance documents were found from various organizations across the world. When faced with a patient taking a DOAC, National Health Service (NHS) Tayside guidance recommends continuing dental treatment without interrupting the drug, limiting the number of extractions to three to four teeth in one session, and carrying out an assessment of the level of bleeding following the first extraction. However, no reference source or justification for this decision is provided within the guidance document.

Guidelines produced by the University of California Anticoagulant Service advise continuing anticoagulation for low bleeding risk procedures, but for ‘extensive surgery’ (apicectomy or alveolar surgery involving bone removal), the bleeding risk is considered high and therefore anticoagulant interruption is recommended, as outlined in Table 4 . The authors have taken a more conservative approach than the manufacturers advise, with higher upper limits given for the period of discontinuation of each drug. This approach is surprising, particularly for the FXaIs, as they do not exert such an effect as dabigatran in patients with renal failure and the effects of drug accumulation are not as profound. The authors then recommend that the DOAC is started as soon as possible after the procedure and ALHM should be adopted.

Jan 16, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on An evidence summary of the management of patients taking direct oral anticoagulants (DOACs) undergoing dental surgery

VIDEdental - Online dental courses

Get VIDEdental app for watching clinical videos