Thromboembolic risk associated with the indication for anticoagulant therapy (a mechanical heart valve, AF, or venous thromboembolism) is classified as high when the annual risk of arterial or venous thromboembolism is greater than 10%, moderate when it is between 5% and 10%, and low when it is less than 5%.18 Recent recommendations that use CHA2DS2-VASc rather than CHADS2 scores for stroke risk assessment were taken into account for patients with AF.16 The risk stratification for thromboembolism according to the indication for anticoagulant therapy is presented in Table 3.

The following aspects should be considered when defining thrombotic risk associated with atherosclerotic conditions that require antiplatelet therapy: time to intervention, disease presentation (acute vs stable), the patient's clinical characteristics, and the treatment received (PCI and stent [type], surgery, or medical treatment only).19 It is important to emphasize that these risks have not been evaluated in clinical trials and that the recommendations presented are based on observational and retrospective studies and consensus documents by other groups.13 Based on a combination of the above factors, risk of thrombosis in patients taking antiplatelet agents is classified as high, moderate, or low, as shown in Table 3.13,20

Time from the ischemic event to the intervention is the most important determinant of events related to the interruption of dual antiplatelet therapy (DAPT) prior to surgery or an invasive procedure.21 Recent findings indicate that patients who undergo emergency surgery or surgery within 6 months of stent placement have an increased risk of thrombosis.22 The recommendations of the Cardiovascular Thrombosis Working Group of the Spanish Cardiology Society on the optimal duration of DAPT in patients with atherothrombotic disease are summarized in Figure 1.23,24

Figure 1.

Recommendations on the optimal duration of dual antiplatelet therapy for patients with atherothrombotic disease. A, aspirin; BMS, bare metal stent; BVS, bioabsorbable vascular scaffold; C, clopidogrel; CABG, coronary artery bypass grafting; DEB, drug-eluting balloon; DES, drug-eluting stent; P, prasugrel; PCI, percutaneous coronary intervention; T, ticagrelor. Adapted with permission from Gómez-Polo et al.23.

(0.81MB).

Stent type is another important factor when evaluating thrombotic risk. PCI with a first-generation drug-eluting stent (DES) vs a bare metal stent is associated with a higher incidence of late and very late stent thrombosis.25 Second-generation DESs, however, offer greater safety as they incorporate improved design features and polymer coatings. They have even been associated with a lower incidence of thrombosis than metal stents.26 In addition, the emergence of new platforms and PCIs with drug-eluting balloons is driving the search for devices that offer increased safety without compromising effectiveness.27 Although experience with bioabsorbable vascular scaffolds is still limited, several studies have reported a higher incidence of long-term stent thrombosis with their use.5,23 The recommendation in such cases thus is to maintain DAPT for at least 12 months. Finally, patients who have undergone coronary artery bypass grafting or noninvasive medical treatment have a much lower risk of complications, as the risk of stent thrombosis, which is frequently catastrophic, is eliminated.

Another important consideration is acute vs stable disease. Patients with acute coronary syndrome have a greater risk of thrombosis than those with stable coronary artery disease. Comorbidity should also be evaluated. Conditions such as diabetes mellitus, chronic kidney disease, severe left ventricular dysfunction, and a history of stroke and/or transient ischemic attack have traditionally been associated with an increased risk of thrombosis.26 Patients who have undergone a complex PCI (placement of long, multiple, or overlapping stents, bifurcation or left main stents, or stents in vessels < 2.5mm or saphenous vein grafts) should also be considered to have an increased risk of thrombosis.28

Surgical intervention in patients with a recent history of stroke is a risk factor for new events, particularly in the 30 days following the stroke and after discontinuation of antiplatelet therapy.29 Stent thrombosis associated with peripheral artery disease has mainly been described in the first month. The risk is greatest in patients with DESs or stents used to treat chronic occlusions. The recent guidelines on the diagnosis and treatment of peripheral arterial diseases recommend DAPT for at least 1 month in patients who have undergone stenting for lower-extremity peripheral artery disease, regardless of stent type.24

In general, anticoagulant therapy should be stopped before the vast majority of surgical and other invasive procedures. There are, however, cases in which maintaining treatment carries a very low risk of bleeding that may be acceptable to the operator.35 The BRUISE CONTROL trial36 showed that maintaining VKA therapy (INR < 3) in patients with a high risk of thromboembolism who required pacemaker or defibrillator surgery was associated with a lower risk of bleeding than bridging therapy with low-molecular-weight heparin (LMWH).34 The recent results of the BRUISE CONTROL-2 trial showed no differences in the incidence of ischemic and/or bleeding events between patients who stopped and who continued taking direct oral anticoagulants (DOACs).37 In the COMPARE trial of patients with pulmonary vein catheter ablation, those who discontinued anticoagulant therapy with an oral VKA (IRN 2-3) had a lower risk of bleeding and thromboembolic events than those who switched to LMWH.38 Although recent evidence suggests that certain procedures are safe when DOAC therapy is not withdrawn, more studies are needed to allow a definitive recommendation.39 In conclusion, this working group recommends considering not stopping anticoagulant therapy for low-risk procedures with a risk of very minor or clinically insignificant bleeding deemed acceptable by the operator (Table 1 of the supplementary material).

The moment at which anticoagulant therapy should be stopped is largely determined by the drug's elimination route. VKAs are metabolized mainly through the liver and only minimally through the kidneys. With the exception of patients with a liver disease that alters drug behavior (and in whom anticoagulants are generally contraindicated), we recommend stopping acenocoumarol 3 days before an intervention and warfarin 5 days before (Figure 2).9,16 This recommendation, however, is only applicable to patients with INR values within the therapeutic range (INR 2-3) 7 days before the intervention. In patients with supratherapeutic INR values or scheduled for a procedure with a high risk of bleeding, the moment for withdrawing anticoagulant therapy will depend on the exact figure and on the target figure for the operation or procedure. In patients with subtherapeutic INR values, VKAs can be discontinued 1 day later (2 days before the procedure for acenocoumarol and 4 days for warfarin). INR values should always be checked before the procedure. In general, the target should be an INR < 1.5. As DOACs have predictable pharmacokinetics and depend on renal function, the decision on when to withdraw them will depend on the creatinine clearance rate and the risk of bleeding associated with the scheduled procedure.34,40,41

In general, there is a low risk of thromboembolism when oral anticoagulant therapy is discontinued without bridging therapy with heparin.42 In addition, recent data suggest that heparin bridging is associated with an increased risk of bleeding but does not confer any benefits in terms of the incidence of thromboembolic events.43 In the BRIDGE trial, which analyzed bridging anticoagulation in 1884 patients with AF,44 no significant differences were observed for rate of thromboembolic events, but patients in the bridging therapy group experienced more bleeding episodes than those in the placebo group. The main limitation of the trial was that it included patients with a low risk of thromboembolism (exclusion of patients with mechanical heart valves, a history of stroke in the previous 3 months, and a mean CHADS2 score of 2.3) and a low risk of bleeding (only 11% of the procedures carried a moderate to high risk). A similar trial is currently underway to assess patients with a high risk of thromboembolism (PERIOP-2, NCT00432796).

Based on the available evidence, we recommend only using heparin bridging for patients with a high thromboembolic risk (Table 2). In patients being treated with a VKA, it is sufficient to start LMWH (or unfractionated heparin in patients with kidney failure and creatinine clearance < 30mL/min) when the INR is less than 2 or, where this value is not available, when 2 to 3 doses of the drug have been omitted. We do not recommend DOACs as bridging therapy (although some protocols still contemplate their use after the omission of 2 to 3 doses of the VKA). The last dose of LMWH should be administered 12hours before the operation or procedure in the case of prophylactic doses and 24hours in the case of therapeutic doses. Unfractionated heparin should be administered 4 to 6hours in advance.

Achieving adequate hemostasis during the procedure and absence of postprocedural bleeding are the 2 most important factors to consider when restarting a patient on anticoagulant therapy. The moment at which the drug is to be reintroduced should therefore be decided in conjunction with the surgical/intervention team. As a general rule, and under the previous proviso, we recommend resuming anticoagulant therapy 24hours after the procedure. As the anticoagulant effect of VKAs starts to set in at 24 to 72hours, we recommend using heparin bridging therapy in patients with a high thromboembolic risk. This is not necessary with DOACs.41 Reintroduction of oral anticoagulant therapy should be postponed for 48 to 72hours only in patients with a high risk of postoperative bleeding (Figure 2). If it is not possible to restart oral anticoagulant therapy, we recommend evaluating the option of parenteral anticoagulation, preferably with balanced doses of LMWH to cover the risk of thrombosis and minimize that of bleeding.

Aspirin should be maintained in patients on single antiplatelet therapy, as it has been found to reduce ischemic risk without significantly increasing bleeding risk.45 Withdrawal for 3 days should only be considered for procedures with a very high risk of bleeding (eg, neurosurgical procedures).9 Unless contraindicated, a low dose (100mg/d) should be administered and maintained, although detection of higher doses (eg, 300mg) before the procedure does not justify deferral. In patients on P2Y12 inhibitor monotherapy, we recommend withdrawing the drug 3 to 7 days before the procedure (3-5 days for ticagrelor, 5 days for clopidogrel, and 7 days for prasugrel) and replacing it with aspirin100mg where possible.45

In patients receiving DAPT and scheduled for an elective procedure, the first step should be to consider postponing the procedure if the patient has a moderate-to-high risk of thrombosis. Where possible, the procedure should be deferred until the risk is low. In other cases, we recommend the following:

• •

  Maintain aspirin for all procedures unless contraindicated (eg, neurosurgery).

• •

  In patients with a moderate risk of thrombosis (and with the exception of procedures with a low bleeding risk), discontinue the P2Y12 inhibitor 3 to 7 days before the procedure as follows: 3 to 5 days for ticagrelor, 5 days for clopidogrel, and 7 days for prasugrel.5,9 For emergency procedures that can be deferred for 72hours (eg, hip fracture surgery), evaluate the indication for surgery 3 days after withdrawing ticagrelor or clopidogrel and 5 days after withdrawing prasugrel.5,9,30

• •

  The strategy in patients with a high thrombotic risk will depend on the bleeding risk associated with the procedure. If this is low, we recommend not stopping DAPT. Cases of moderate-to-severe risk are more controversial and we therefore recommend evaluating patients on a case-by-case basis within a multidisciplinary team. In general, it is important not to stop DAPT in the 30 days after the index event for which it was prescribed.5

Unlike the case of anticoagulation, there are few publications and little clinical experience with the use of bridging therapy in patients on antiplatelet therapy. It is a common mistake to consider heparin in this setting, as heparin can increase platelet reactivity and have the opposite effect to that desired.13 If bridging therapy is necessary, we recommend using an antiplatelet rather than an anticoagulant, as heparin can be harmful.12 Of note among the intravenous antiplatelet agents studied to date are the glycoprotein IIb/IIIa inhibitors, tirofiban and eptifibatide, and the P2Y12 adenosine diphosphate receptor inhibitor, cangrelor.

The perioperative use of intravenous tirofiban and eptifibatide, 2 fast-acting glycoprotein IIb/IIIa inhibitors, has been evaluated in small series of patients with a high thrombotic risk (mainly patients treated with PCI and DES placement), and the results have been controversial.46 Cangrelor, an intravenous reversible P2Y12 inhibitor, has been evaluated in a phase II trial of patients on DAPT who underwent coronary artery bypass grafting surgery.47 Compared with placebo, the drug was associated with a similar incidence of thrombotic events and a slightly higher number of minor bleeding episodes. The trial, however, was designed to study platelet reactivity and not thrombotic or bleeding events, and therefore the findings should be interpreted with caution. Furthermore, the perioperative use of cangrelor has not yet been approved. For the moment, the drug is only indicated for PCIs.

Accordingly, and considering the little evidence available for bridging therapy in patients on antiplatelet therapy, we recommend restricting its use to patients with a high thrombotic risk (essentially patients receiving DAPT for an event that occurred in the previous 30 days) who are scheduled for an unpostponable operation or procedure with a moderate to severe risk of bleeding. The risk-benefit relationship in such cases should be analyzed on a case-by-case basis by a multidisciplinary team. If it is decided that bridging therapy is indicated, the patient should be started on a glycoprotein IIb/IIIa inhibitor 72hours after withdrawal of the P2Y12 inhibitor. The recommended doses are 0.1μg/kg/min for tirofiban and 2μg/kg/min for eptifibatide and these should be maintained for up to 4 to 6hours before the procedure. A loading dose is not required.

The 2 most important considerations when restarting a patient on antiplatelet therapy are having achieved adequate hemostasis during the procedure and absence of postprocedural bleeding. The moment at which the drug is to be reintroduced should thus be decided in conjunction with the surgical/intervention team. As a general rule, and under this proviso, we recommend resuming antiplatelet therapy 24hours after the procedure. If the patient has a high risk of thrombosis and is receiving DAPT, consider restarting the P2Y12 inhibitor with a loading dose of 300 to 600mg for clopidogrel, 60 mg for prasugrel, and 180 mg for ticagrelor. Restarting of oral anticoagulant therapy should only be delayed (for 48 to 72hours) in patients with a high risk of postoperative bleeding (Figure 3).