Perioperative bridging in patients treated with vitamin K antagonists

In what patients treated with vitamin K antagonists should we use perioperative bridging therapy?

Wendy Lim, MD: Typically, anticoagulation is used for 3 broad indications. Worldwide, probably the most common is chronic atrial fibrillation for stroke prevention. The second major category is for either treatment or secondary prevention of venous thrombosis, so that would be either deep vein thrombosis typically of the lower extremities or pulmonary embolism. The third main category is mechanical heart valves.

Typically, if we are thinking about bridging in the perioperative period, again, one has to balance the risk of thromboembolic events versus the risk of bleeding with the procedure [see Continuing antithrombotic treatment in patients undergoing invasive procedures]. An important study that was published a number of years ago was called the BRIDGE (Bridging Anticoagulation in Patients who Require Temporary Interruption of Warfarin Therapy for an Elective Invasive Procedure or Surgery) study. In those patients who had chronic atrial fibrillation, they compared the utility of providing bridging anticoagulation, that is, they stopped the vitamin K antagonist about 5 days in advance of the procedure or surgery and then provided low-molecular-weight heparin therapy to bridge that period of time leading up to the procedure, and they provided bridging anticoagulation post procedure. They compared that with not providing bridging, that is, just stopping the vitamin K antagonist. Interestingly, the patients who received the bridging therapy had a higher incidence of bleeding, and there was no difference in the rate of thromboembolic events.

That study is specific for patients who have atrial fibrillation, and the vast majority of the patients who were enrolled in that study had what we would call a fairly low to moderate thromboembolic risk when we look at various scoring systems, the CHADS₂ score for example. If one were to extrapolate that, one would say, “We probably do not need to bridge a low CHADS₂ score atrial fibrillation.” One may extend that to a high CHADS₂ score of 5 to 6, but I think that was really the first study that gave us some randomized data as to what to do in that setting. I would say that for atrial fibrillation, bridging therapy might be considered for a high CHADS₂ or high CHA₂DS₂-VASc score.

In the setting of venous thromboembolism, probably the most key feature is how recent the thromboembolic event was. If it has just happened, the first question would be, “Can we defer that procedure or that surgery at all?” And if one cannot, then one would like to try to at least get in maybe a month, or perhaps 3 months, of anticoagulation, because for most venous thromboembolic events a minimum duration of time that would be needed is 3 months. I would say that bridging therapy would be considered if the individual was within a month or perhaps up to 3 months of their event or they had some thrombotic risk factors that made them particularly high risk. Perhaps they have high-risk thrombophilia or have accompanying risk factors for thrombosis that might make that reasonable.

In the third category, mechanical heart valves, it largely depends on the location of the valve and the age of the valve. Typically, the thrombotic risk associated with mitral valves is higher than with valves in the aortic position, and modern generation valves probably carry a lower thrombotic risk. Putting that together, I would probably consider bridging anticoagulant therapy in individuals who have a mitral valve, have more than one valve, or particularly if the valve is an older valve.

Aside from the BRIDGE study, we do not have a lot of trials which can guide our practice. A lot of times patients bring their own experiences into this and physicians bring their own experiences as well. Oftentimes, one considers the patient’s preference and the feasibility of using bridging low-molecular-weight heparin therapy because some patients who received bridging therapy in the past have done well and do not fall into a category where one would clearly not bridge. I think it is up for discussion between the physician and the patient to decide whether that is appropriate. The other thing that I would add is that the introduction of direct oral anticoagulants has changed our approach to the need for perioperative bridging because of the short half-life of direct oral anticoagulants. It is evolving. There are studies that are ongoing, looking at the appropriate perioperative management of direct oral anticoagulants, so I think there is more information to come.