Minding the Gap: Bridge Therapy in A-fib

We all have an innate desire for clarity in our decisions. Often, however, we are faced with controversy and uncertainty.  Grey areas where there is a gap between clinical practice, logical reasoning, and evidence.  One of these grey areas is the “right” approach to the peri-procedural management of chronic anticoagulation.  There are so many clinical decisions involved (See Table 1) and little clarity about any of them!

Table 1:  Clinical questions regarding the peri-procedural management of patients on chronic anticoagulation

There are guidelines (See Table 2) that provide direction and support the use of bridge therapy, such as low-molecular weight heparin or unfractionated heparin, in patients at high risk for thromboembolism during the immediate pre- and post-procedural time period.^1,2  However, much of the “evidence” used to support these recommendations are based on observational studies … or simply expert opinion.¹

Table 2: Summary of Thrombotic Risk Stratification and Bridge Therapy recommendations from the Perioperative management of antithrombotic therapy: Antithrombotic Therapy and Prevention of Thrombosis Guidelines¹

In the 2014 AHA/ACC/HRS Guideline for the Management of Patients with Atrial Fibrillation, bridge therapy with UFH or LMWH is a 1C recommendation for patients with a-fib AND a mechanical heart valve.  For all other patients with atrial fibrillation, the recommendation is to balance the risk of thrombosis with risk of bleeding.²  In other words, use your best judgment!

For many of us the decision to use bridging therapy in a patient with atrial fibrillation revolves around the CHADS₂  score.  CHADS₂ and CHA₂DS₂-VASc might be helpful when quantify a patient’s estimated annual adjusted percent risk of stroke but they are not validated tools to stratify stroke risk in patients on chronic anticoagulation requiring treatment interruption for procedure.^3,4

To better understand the “real-world” clinical care of patients with atrial fibrillation, including the use of bridge therapy and patient outcomes during anticoagulation interruption, the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) registry was formed. In this prospective, observational study of 7372 patients age 18 years and older with ECG-documented atrial fibrillation, the occurrence of myocardial infarction, stroke or systemic embolism, major bleeding, cause-specific hospitalization, and death in the 30 days after the temporary interruption of oral anticoagulation for a procedure was analyzed.^5,6  The study did not collect information regarding the clinician’s reasoning for anticoagulation therapy interruption but did collect the date and type of procedure, whether bridging anticoagulation was used, and the adverse events that occurred.  No data was collected regarding the peri-procedural anticoagulation plan – such as timing of discontinuation of chronic anticoagulant, timing of initiation and discontinuation of bridging anticoagulant, use of reversal therapies such as vitamin K, or the drug, dose, frequency, or duration of bridging anticoagulants.  All of these factors would obviously influence the risk of bleeding and thrombotic complications.  Also baseline patient characteristics that might influence bleeding such as uncontrolled blood pressure (systolic >160mmHg), abnormal renal or liver function, or alcohol use were not reported.

Bridge therapies, most often low-molecular-weight heparin (73%) and unfractionated heparin (15%), were utilized during 24% of the interruptions.  Patients who received an anticoagulant bridge were more likely to have had a prior stroke or TIA, congestive heart failure, significant valve disease, or a mechanical valve.  Not surprisingly, the mean CHADS₂ and CHA₂DS₂-VASc scores were statistically higher in patients who received bridge therapy compared to those who were not bridged (CHADS₂: 2.53 versus 2.34; p=0.004 and CHA₂DS₂-VASc 4.25 versus 4.03; p=0.01 respectively), but whether this difference is clinical meaningful is another matter.⁵  The mean risk score for the ORBIT-AF population translates into an annual predicted stroke risk of 4 to 5% per year.  Thus, many of the patients include in this analysis were low risk.

The primary outcome was assessed during two different time frames: during the period of chronic anticoagulation therapy interruption and the 30 days after interruption.  Patients who received bridge therapy compared to those who did not, as expected, had a significantly higher rate of major bleeding (3.6% versus 1.2%; p=0.0007; NNH 41).  Patients who received bridging were also significantly more likely to be hospitalized for bleeding (2.2% versus 0.7%; p=0.006; NNH 66) during the 30 days after treatment interuption.⁵  Because the bridging therapy and no bridging therapy groups were different at baseline, the authors adjusted the 30-day outcomes based on bleeding and thrombosis risk factors (i.e. age, prior cerebrovascular events, prior gastrointestinal bleeding, congestive heart failure, CHADS₂ score, procedure type, and baseline oral anticoagulant).  After adjustment, major bleeding and bleeding hospitalizations remained significantly higher in those who received bridging therapy (adjusted odds ratio, 3.84; p<0.0001).  After adjustment, patients in the bridge therapy group also had more cardiovascular events (including stroke, systemic embolism, myocardial infarction, or cardiovascular hospitalization) but this did not reach statistical significance (adjusted odds ratio, 1.62; p=0.07).  Overall the author’s concluded that interruptions in anticoagulant therapy are common in patients with atrial fibrillation and recommended against the routine use of bridge therapy.

The data from this study are likely contrary to the perceptions and intent of bridging therapy.  Although the data in the study were prospectively collected, the two patient populations — patients who received bridging therapy and those who did not — likely had intrinsically different risks.  No amount of statistical adjustments can truly account for these differences.  Only a randomized, controlled trial can truly inform us of the potential risks and benefits of anticoagulation bridge therapy.  Fortunately, there are two large randomized trials – the BRIDGE and PERIOP2 studies — that will hopefully provide us more definitive answers.  Published results are expected as early as June 2015.^8,9 

This study does provide us with some “real-world” data regarding peri-procedural anticoagulant management in patients with atrial fibrillation. Interruptions in anticoagulation in patients with atrial fibrillation are common and during those interruptions about 25% of patients receive some type of bridge therapy.  When looking at the procedures which led some clinicians to interrupt chronic anticoagulation therapy in ORBIT-AF, such as catheter ablation, endoscopic procedures, and dental work, many may not have needed treatment interruption.^1,2  Unfortunately, this study does not report when and how the chronic and bridging anticoagulants were discontinued, initiated, or dosed.⁷  If the chronic anticoagulant was discontinued either too early or too late in relation to the procedure then adverse events could have been increased.  Also, the bridge anticoagulant, if not timed and dosed appropriately would also have increased the risk of adverse events.  Moreover, It would have been helpful to see the outcome events stratified by bleeding and thromboembolic risk scores.

In my opinion, patients with atrial fibrillation at high risk for thromboembolism should receive anticoagulation bridge therapy during interruptions.  The ORBIT-AF study affirms what current guideline recommend — patients who are at low risk of thromboembolism should not routinely receive bridge therapy.  This still leaves us at a quandary what to do for those patients at moderate risk of thromboembolism.  This remains a grey area and requires shared decision-making.  No risk stratification system captures all elements of risk and they don’t consider patient-specific preferences.  Many practitioners may wish to error on the side of using bridge anticoagulation in moderate risk patients.  Perhaps most importantly, when the decision to use anticoagulation bridge therapy is made, a clear written plan that all parties agree on is needed in order to minimize adverse outcomes.

What have we learned from these data? How does this study influence your practice? How do you determine whether peri-procedural bridge therapy is needed in a moderate risk patient?  Is there a role for target-specific oral anticoagulants as bridge therapy?  Tell us what you think!