Periprocedural Management of Antithrombotic Agents

Procedural bleeding risk

Thromboembolism risk




Continue antithrombotic medications

Continue antithrombotic medications


Temporarily discontinue antithrombotic medications without bridging therapy

Continue antithrombotic medications or temporarily discontinue antithrombotic medication with bridging therapy

Bleeding Risk of Endoscopic Procedures


Procedures considered high risk for bleeding (Table 2.2) are those associated with ≥1 % risk of causing clinically significant hemorrhage (i.e., requiring hospitalization, transfusion, endoscopic treatment, or surgery) [4, 5]. Low-risk procedures include diagnostic endoscopy with or without biopsy, endoscopic retrograde cholangiopancreatography (ERCP) without sphincterotomy, endoscopic ultrasound (EUS) without fine-needle aspiration (FNA), and capsule endoscopy. High-risk procedures include polypectomy at any location (≥1 cm), ERCP with biliary/pancreatic sphincterotomy, and endoscopic hemostasis, among others. The bleeding risk associated with enteral stent placement but without dilation and ERCP with papillary balloon dilation but without sphincterotomy remains controversial. As a general rule, elective high-risk procedures should be delayed until the patient’s risk for thromboembolism is reduced and/or antithrombotic medications are optimized to minimize bleeding complications. In the setting where emergent endoscopic intervention is required, every effort should be made to conservatively manage these patients (e.g., transfusions) until their periprocedural bleeding and thromboembolic risks are reduced.

Table 2.2
Bleeding risks of endoscopic procedures

Low-risk procedures (<1 %)


High-risk procedures (≥1 %)

Diagnostic (EGD, colonoscopy, flexible sigmoidoscopy, BAE) ± biopsy

Enteral stent placement without dilation

Polypectomy (any location, ≥1 cm)

ERCP ± stenting without sphincterotomy

ERCP papillary balloon dilation without sphincterotomy

ERCP with biliary/pancreatic sphincterotomy

EUS without FNA

EUS with FNA

Capsule endoscopy

PEG placement

Therapeutic BAE

Pneumatic or bougie dilation

Endoscopic hemostasis

Treatment of varices


EMR, ESD, ampullectomy

Ablation of tumor or vascular lesion by any technique

BAE balloon-assisted enteroscopy, EGD esophagogastroduodenoscopy, ERCP endoscopic retrograde cholangiopancreatography, EUS endoscopic ultrasound, FNA fine-needle aspiration, PEG percutaneous endoscopic gastrostomy, EMR endoscopic mucosal resection, ESD endoscopic submucosal dissection

Endoscopic Sphincterotomy

The majority of ERCP-related bleeding is intraluminal and is primarily related to sphincterotomy, including precut papillotomy. In a pooled analysis of 21 prospective cohort studies involving 16,855 patients who underwent ERCP, clinically significant bleeding occurred in 226 patients (1.34 %, 95 % confidence interval [CI] 1.16–1.52 %) [6]. Independent predictors of post-ERCP hemorrhage include sphincterotomy, coagulopathy before the procedure (partial thromboplastin or prothrombin time >2 s above normal, platelet count <80,000 mm3, or ongoing hemodialysis), anticoagulant therapy within 3 days post procedure (oral warfarin or intravenous heparin), cholangitis before the procedure, intraprocedural bleeding (ranging from oozing to requiring endoscopic hemostasis), precut papillotomy, obstruction/stenosis of the orifice of the papilla of Vater, low endoscopist case volume (≤1 sphincterotomy/week), and low center volume (<200 ERCPs/year) [79]. Freeman et al. showed that while cirrhosis was not an independent predictor of post-sphincterotomy bleeding (p = 0.06), the two patients with fatal bleeding complications had Child–Pugh class C cirrhosis [7]. Neither extension of previous sphincterotomy nor the size of sphincterotomy was associated with increased post-sphincterotomy bleeding [7].

Evidence is conflicting as to the risk of post-sphincterotomy bleeding in the setting of recent aspirin or nonsteroidal anti-inflammatory drug (NSAID) use. Freeman et al. showed no increased risk of bleeding if aspirin or NSAID was used within 3 days of endoscopic sphincterotomy [7]. In another case–control study, there was no increased risk of clinically significant bleeding related to the use of antiplatelet agents [10]. Conversely, one study demonstrated an increased incidence of post-sphincterotomy bleeding in aspirin users relative to nonusers (9.7 % vs. 3.9 %, p = 0.01), and the withholding of aspirin for 7 days prior to endoscopic sphincterotomy did not decrease the risk for bleeding (9.5 % vs. 3.9 %, p = 0.01) [11]. Unfortunately, data are lacking regarding the safety of endoscopic sphincterotomy in patients on dual antiplatelet agents and/or anticoagulants or in those who are coagulopathic due to cirrhosis or hemodialysis.

In one study, endoscopic balloon dilation of the biliary sphincter was as effective as biliary sphincterotomy for the removal of common bile duct stones, with significantly reduced bleeding complications (0 % vs. 2.0 %, p = 0.001) [12]. However, the rate of post-ERCP pancreatitis was higher in the balloon dilation group (7.4 % vs. 4.3 %, p = 0.05). Therefore, it cannot be advocated for routine use [12]. There are no well-designed, head-to-head comparisons of the two methods at this time in patients who are on antithrombotic therapy.

An endoscopist performing ERCP on an emergent basis is likely already dealing with a patient at high risk for post-procedural bleeding. Based on current evidence, ERCP can be performed with low risk of post-procedural bleeding if sphincterotomy is not necessary or can be deferred until the patient’s bleeding risk is reduced. If the patient is medically stable, transfer to a high-volume center for ERCP should be considered.

Endoscopic Hemostasis

Contribution of Antithrombotic Medications to Gastrointestinal Bleeding

In the setting of antiplatelet use, recurring patient-related risk factors for gastrointestinal bleeding (GIB) include prior history of GIB, history of H. pylori infection, and advanced age. Concurrent use of anticoagulants, steroids, or NSAIDs is also a consistent predictor of GIB. GIB risk increases with the number of risk factors present in the patient [13]. Among patients using low-dose aspirin (75–325 mg daily), a meta-analysis of placebo-controlled trials for vascular protection demonstrated a relative risk of 2.07 (95 % CI, 1.61–2.66), conferring an increased annual incidence of 0.12 % (95 % CI 0.07–0.19 %) for major GIB attributed to low-dose aspirin use [14].

The risk of GIB with combination antithrombotic agents is increased when compared with low-dose aspirin alone. A meta-analysis showed an increased risk of major GIB when aspirin was combined with clopidogrel (odds ratio [OR] 1.86, 95 % CI, 1.49–2.31) or with an anticoagulant (OR 1.93, 95 % CI, 1.42–2.61) compared with aspirin alone [15]. In the same study, proton-pump inhibitor (PPI) therapy significantly reduced the risk of GIB events in patients given low-dose aspirin [15]. The routine use of PPI with clopidogrel is controversial due to impairment of antiplatelet effects of clopidogrel by PPI in in vitro studies [13]. Although findings from clinical studies are inconsistent, product labeling of omeprazole and esomeprazole includes warnings about possible interactions with clopidogrel.

Patients who undergo careful monitoring of anticoagulant intensity have a 0.3–0.5 % increased annual risk of major bleeding compared with controls [16]. Independent predictors of anticoagulant-related bleeding include intensity of anticoagulant effect, age >75, concomitant use of antiplatelets, and length of therapy [17].

Holster et al. performed a meta-analysis of 43 randomized trials comparing bleeding risk of the new oral anticoagulants versus standard therapy [18]. While all the studies included bleeding events as a safety outcome, only 19 of these trials assessed GIB as a separate subgroup (Table 2.3). The overall OR for GIB among patients taking the new oral anticoagulants was 1.45 (95 % CI, 1.07–1.97), and the OR for clinically relevant bleeding (as defined by the International Society of Thrombosis and Haemostasis [ISTH] and Thrombolysis in Myocardial Infarction [TIMI] study group) was 1.16 (95 % CI, 1.00–1.34). Subgroup analyses demonstrated significantly increased bleeding risk of the new oral anticoagulants versus standard therapy if the indications included acute coronary syndrome (ACS) and treatment of venous thrombosis, but not atrial fibrillation (AF) or thromboprophylaxis after orthopedic surgery (OS). Dabigatran and rivaroxaban were also associated with significantly increased risk for GIB. The meta-analysis was limited by substantial heterogeneity between studies with an I2 of 60.8 % (p < 0.05) for studies assessing GIB and I2 of 83.5 % (p < 0.05) for studies assessing clinically relevant bleeding. Further studies assessing specific GIB-related outcomes in patients taking the new oral anticoagulants are warranted.

Table 2.3
Bleeding risk of new oral anticoagulants [18]


OR (95 % CI)

Clinically relevant bleeding

1.2 (1.0–1.3)

Gastrointestinal bleeding

1.5 (1.1–2.0)



5.2 (2.6–10.5)

 Venous thrombosis

1.6 (1.0–2.4)


1.2 (0.9–1.6)

 OS thromboprophylaxis

0.8 (0.3–2.0)

Drug-specific GIBa


1.6 (1.3–1.9)


1.5 (1.2–1.8)


1.2 (0.6–2.7)


0.3 (0.0–7.7)

OR odds ratio, CI confidence interval, ACS acute coronary syndrome, AF atrial fibrillation, OS orthopedic surgery, GIB gastrointestinal bleeding

aResults based on three studies for dabigatran, five studies for rivaroxaban, eight studies for apixaban, and one study for edoxaban

Considerations Regarding Hemostatic Techniques

Most studies evaluating endoscopic hemostasis in anticoagulated patients are retrospective in nature. In these studies, identifying the site of GIB was successful in >80 % of patients [19, 20]. Gastroduodenal ulcers and erosions accounted for >50 % of lesions causing upper GIB. Studies evaluating specific lower GI sources of bleeding are lacking, although common causes include polyps, diverticula, and angiodysplasia. Among patients with GIB on antiplatelets or anticoagulants, 17–29 % will have no mucosal abnormality on endoscopic evaluation [21].

Endoscopic clips are safe and effective in the treatment of bleeding peptic ulcers, Dieulafoy lesions, and Mallory–Weiss tears, as well as for prophylaxis or treatment of post-polypectomy bleeding and diverticular hemorrhage [22]. Clip placement has been demonstrated to be superior to injection alone and comparable to thermal coagulation for the treatment of non-variceal upper gastrointestinal bleeding [23]. Endoscopic clip placement, when technically feasible, may be preferable to thermal therapies in patients on antithrombotic therapy for several reasons. Thermal therapies induce or extend ulcer formation and may exacerbate bleeding from tissue injury. Clips have the theoretical advantage of applying mechanical compression to bleeding lesions and can be applied with minimal tissue injury. Additionally, clips can serve as angiographic or surgical markers if bleeding cannot be controlled endoscopically. Clips achieve high rates of primary hemostasis (85–100 %) with low rebleeding rates (2–20 %), although their effectiveness in the setting of antithrombotic therapy is unclear [22]. Studies comparing the different modalities for endoscopic hemostasis in patients on antithrombotic agents are lacking.


Polypectomy is usually performed in the elective setting with outpatient antithrombotic medications optimized prior to the procedure. Moreover, immediate post-polypectomy bleeding (PPB) can usually be treated effectively with traditional hemostatic techniques. However, severe delayed PPB (1–14 days post procedure) may require emergent endoscopic intervention and often occurs in patients on antithrombotic therapy [24, 25]. Independent predictors of delayed PPB include resumption of anticoagulation following polypectomy, polyp diameter (≥10 mm), number of polyps removed, proximal colonic location, history of cardiovascular disease, and hypertension [2428].

Aspirin/NSAID use alone has not been shown to increase the risk of delayed PPB [24, 29]. Current data suggest that there is an increased risk of PPB in patients who continue clopidogrel alone or in combination with aspirin, with an event rate ranging from 2.4 to 3.5 % [27, 28, 30]; however, bleeding was controlled without the need for angiographic or surgical intervention. Thus, in patients who are at high risk for cardiovascular complications, such as those with recent ACS or stent placement, continuation of dual antiplatelet therapy may be reasonable.

Endoscopic clip placement over the polypectomy defect may decrease the risk of delayed PPB. In the only randomized controlled trial to evaluate this intervention, no difference was seen in the rates of delayed PPB in the prophylactic clip placement group compared with the group that received no clip; however, the polyps removed were generally low-risk, small (mean size 7.8 ± 4.0 mm) lesions [31]. On the other hand, a large retrospective study of patients with resected polyps of ≥2 cm showed that prophylactic clip closure significantly reduced the risk of PPB compared with no clip closure (1.8 % vs. 9.7 %) [32].

Data are limited on the effectiveness of prophylactic clip placement after polypectomy in the setting of uninterrupted anticoagulation. A small retrospective study of 21 patients (41 polypectomies) on uninterrupted warfarin (mean international normalized ratio [INR] 2.3, range 1.4–4.9) who underwent hot snare resection of small polyps (≤10 mm) had no PPB events when one or two clips were placed immediately after polyp resection. Warfarin was withheld for 36 h before the procedure, while patients remained on a modified diet to avoid supra-therapeutic INR and without concomitant antiplatelet agents. Warfarin was resumed according to the patient’s standard schedule [33]. Prophylactic clip placement after polypectomy may be effective in preventing PPB in select patients on uninterrupted anticoagulation, although confirmatory data are needed.

Left Ventricular Assist Devices

Left ventricular assist devices (LVADs ) are increasingly being used in patients with advanced cardiac failure as a bridge to cardiac transplantation or destination therapy (i.e., ineligible for transplantation). Bleeding complications after LVAD implantation are common, with 30 % requiring surgery and 50–80 % requiring at least 2 units of packed red blood cells [34, 35]. Risk factors for GIB after LVAD implantation include use of nonpulsatile device and history of GIB prior to device placement [36, 37]. Retrospective studies show rates of GIB varying from 8 to 40 %, likely due to differences in the definition of GIB, and rebleeding is common [3741]. Endoscopy is safe in LVAD patients and identifies the etiology of GIB in 60–70 % of cases, with peptic ulcer bleeding and vascular malformations of the upper GI tract being the more common sources [39, 42]. Endoscopic hemostasis is generally successful, but data are limited to small studies [42]. The cardiologist and/or cardiac surgeon should be involved in any plan to modify antithrombotic medications.

Endoscopic Bleeding Risks for Other Situations

Foreign Body Ingestion/Food Impaction

Data from two large retrospective studies found bleeding related to endoscopic foreign body removal ranging from 1 to 3 % [43, 44]. Bleeding associated with endoscopic esophageal food disimpaction ranged from 0 to 1 % in two retrospective studies [45, 46].

Colonic Decompression

The risk of causing bleeding from endoscopic decompression of colonic pseudo-obstruction is uncommon [47].

Luminal Stents

A systematic review of gastroduodenal self-expanding metal stents (SEMS) found a 0.5 % risk of bleeding in a pooled analysis of 606 patients [48]. Data regarding bleeding complications from placement of esophageal and colonic SEMS are scant.

Assessing Risk for Thromboembolism

Bleeding complications from endoscopy can be problematic but are rarely catastrophic. Conversely, thromboembolic events are associated with high rates of morbidity and mortality. The following is an approach to risk stratify patients according to their risk of thromboembolic events. Patients with prosthetic heart valves, AF, and venous thromboembolism (VTE) frequently require chronic anticoagulation therapy. A strategy has been proposed for risk stratifying patients susceptible to perioperative thromboembolism according to indication for anticoagulant therapy (Table 2.4) [49]. Patients with a >10 % annual risk for thromboembolism are classified as “high risk,” 5–10 % annual risk as “moderate risk,” and <5 % annual risk as “low risk.” While this classification system can provide some guidance for the risk of developing a thromboembolic event, a patient’s risk assessment should be individualized according to patient- and procedure-related factors.

Table 2.4
Proposed perioperative risk stratification for patients at risk for thromboembolism on anticoagulation [49]

Annual risk for thromboembolism


Low (<5 %)

Moderate (5–10 %)

High (>10 %)

Mechanical heart valve

– Bileaflet aortic valve without atrial fibrillation or risk factorsa

– Bileaflet aortic valve with at least 1 risk factora

– Any mechanical mitral valve

– Older aortic mechanical valve (caged ball, tilting disk)

– Recent (<6 months) stroke/TIA

Atrial fibrillation

– CHADS2 score 0–2 without previous stroke/TIA

– CHADS2 score 3 or 4

– CHADS2 score 5 or 6

– Rheumatic or severe valvular disease

– Recent (<3 months) stroke/TIA

Venous thromboembolism

– VTE >12 months previously without other risk factors

– VTE within the past 3–12 months

– Non-severe thrombophiliab

– Recurrent VTE

– Active cancer (diagnosis <6 months or undergoing treatment)

– Recent (<3 months) VTE

– Severe thrombophiliac

CHADS2 score (range 0–6): congestive heart failure, hypertension, age >75 years, and diabetes mellitus are assigned 1 point apiece, while previous stroke or TIA is assigned 2 points

CHADS 2 cardiac failure–hypertension–age–diabetes–stroke, TIA transient ischemic attack, VTE venous thromboembolism

aRisk factors for stroke without atrial fibrillation: congestive heart failure, hypertension, age >75 years, diabetes, prior stroke/TIA

bNon-severe thrombophilia: heterozygous factor V Leiden or prothrombin gene G20210A mutation

cSevere thrombophilia: deficiency of protein C, protein S, or antithrombin, antiphospholipid syndrome (presence of antiphospholipid antibodies or lupus anticoagulant), homozygous for factor V Leiden, homozygous for prothrombin gene G20210A, compound heterozygous mutations of latter two genes

Atrial Fibrillation

In patients with AF, the CHADS2 score is useful to risk stratify a patient’s annual risk for stroke, although it has not been validated in the perioperative setting [50]. The CHADS2 score scheme is based on a scale of 0–6. Congestive heart failure, hypertension, age >75 years, and diabetes mellitus are assigned 1 point apiece, while previous stroke or transient ischemic attack (TIA) is assigned 2 points. Patients with AF at highest risk for stroke (>10 % annual stroke risk) include a CHADS2 score of 5 or 6, recent (<3 months) ischemic stroke or TIA, or the presence of rheumatic or severe valvular heart disease. Patients with a CHADS2 score of 3 or 4 are considered moderate risk (5–10 % annual risk) and 0–2 are low risk (<5 % annual risk) for stroke [49].

Mechanical Heart Valves

Patients with mechanical heart valves who are at high risk for thromboembolic events include a prosthesis in the mitral position, any caged-ball or tilting disk aortic valve prosthesis, and recent (<6 months) ischemic stroke or TIA. Patients with bileaflet aortic valve prostheses with one or more risk factors, including AF, prior stroke or TIA, hypertension, diabetes, congestive heart failure (CHF), or age >75 years, are at moderate risk. Patients with bileaflet aortic valve prostheses without AF or other risk factors for stroke are at low risk [49].

Deep Vein Thrombosis/Pulmonary Emboli

Patients with recent (<3 months) VTE and severe thrombophilia are considered high risk for additional thromboembolic events. Those at moderate risk are patients with VTE within the past 3–12 months, recurrent VTE, active cancer (diagnosis < 6 months or undergoing treatment), and non-severe thrombophilia. Remote VTE (>12 months) with no other risk factors is considered low risk (Table 2.4) [49].

Coronary Stents and Recent Acute Coronary Syndrome

Dual antiplatelet therapy with combination aspirin and thienopyridine has been shown to reduce adverse events in patients receiving coronary artery stents. Premature discontinuation of antiplatelet therapy is associated with increased risk of stent thrombosis, myocardial infarction, and death. Stent thrombosis can have catastrophic consequences, with incidence of death ranging from 20 to 45 % and myocardial infarction in up to 64 % of cases [51]. Patients at highest risk for stent thrombosis are those with bare-metal stents (BMS) placed within 6 weeks and drug-eluting stents (DES) placed within 12 months [3]. Guidelines vary in regards to when dual antiplatelet therapy can be interrupted (while aspirin is continued) for elective procedures: 4–6 weeks after placement of BMS and 6–12 months after placement of DES [25]. Individuals at higher risk for thrombotic events (diabetes, renal failure, cancer, heart failure, complex coronary disease, or history of coronary stent thrombosis) or with stent placement in the setting of ACS may need longer periods of uninterrupted dual antiplatelet therapy prior to elective/urgent procedures [52]. Dual antiplatelet therapy should be resumed after bleeding risk is minimized from the endoscopic intervention and continued for the recommended duration (up to 12 months for patients with BMS and at least 12 months for patients with DES) [53].

Non-cardioembolic Stroke and Transient Ischemic Attack Prevention

Risk factors for non-cardioembolic stroke include hypertension, diabetes, and hyperlipidemia. Aggressive control of risk factors and lifestyle changes (smoking and alcohol cessation) are recommended to prevent a stroke [54]. Aspirin reduces the risk for secondary stroke by 15 % (95 % CI, 6–23 %) compared with placebo. Aspirin monotherapy, combination aspirin/dipyridamole, and clopidogrel monotherapy are all acceptable options for stroke prevention . Use of an antiplatelet agent is preferred over oral anticoagulants for non-cardioembolic stroke prevention [54].

Left Ventricular Assist Devices

LVADs induce hypercoagulability and persistent platelet activation through various mechanisms, frequently requiring combination anticoagulation and antiplatelet therapy depending on the device implanted [55]. Two randomized controlled trials investigating one of the most common LVADs (HeartMate II, Thoratec, Pleasanton, CA) found low rates of thrombotic complications (ischemic stroke ranging from 3 to 8 %; device thrombosis ranging from 2 to 4 %) in patients on combination warfarin and aspirin [34, 35]. Ischemic strokes are more common with lower INR (<1.5), and hemorrhagic strokes are more common with higher INR (>3.0) [56].

Management of Antithrombotic Medications


Overview of Anticoagulants

Indications for anticoagulation therapy encompass a heterogeneous group of conditions that have varying risks of developing into thromboembolism, including patients with prosthetic heart valves, AF, VTE, and hypercoagulable states (e.g., thrombophilia, active cancer). Anticoagulants exert their effects at various points in the coagulation cascade, which include coagulation initiation and propagation, as well as fibrin formation (Fig. 2.1). An overview of currently available anticoagulants is provided in Table 2.5.


Fig. 2.1
Simplified diagram of coagulation cascade with sites targeted by anticoagulant drugs

Table 2.5
Current anticoagulant agents


Main indications


Mechanism of action

Time to maximal effect

Elimination half-lifea

Return of normal coagulation after cessation

Reversal agent or antidote

Warfarin (Coumadin, Bristol-Myers Squibb) [5759]

VTE treatment; AF, post-MI, mechanical valve, bioprosthetic valve, others


Vitamin K-dependent inhibition of clotting factors II, VII, IX, and X

5–7 days for therapeutic INR

36–42 h

~5 days to normalize INR

Vitamin K, FFP, PCC, rFVIIa

Unfractionated heparin (Fresenius Kabi USA) [60, 61]

ACS, VTE treatment or prophylaxis, bridge therapy for AF/cardioversion

IV or SC

AT-mediated indirect inhibition of factors XIIa, IXa, XIa, and Xa and thrombin

Immediate (IV)

Within 6 h (SC)

30–120 min

4 h

Hold or protamine sulfate

Low-molecular-weight heparin (enoxaparin [Lovenox, Sanofi Aventis], dalteparin [Fragmin, Eisai]) [60, 62, 63]

ACS, VTE treatment or prophylaxis, bridge therapy for AF/cardioversion


AT-mediated indirect inhibition of factors XIIa, IXa, XIa, and Xa and thrombin

3–5 h

3–6 h

24 h

Hold or protamine sulfate

Fondaparinux (Arixtra, GlaxoSmithKline) [60, 64]

VTE treatment and prophylaxis


AT-mediated indirect inhibition of factor Xa

3–5 h

17–21 h

2–4 days

No antidote; consider rFVIIa

Bivalirudin (Angiomax, The Medicines Company) [60, 65]

PCI; ACS; HITT treatment and prophylaxis


Reversible direct thrombin inhibition


20–30 min

1 h

No antidote; consider hemodialysis

Desirudin (Iprivask, Canyon Pharmaceuticals) [66, 67]

VTE prophylaxis


Reversible direct thrombin inhibition

60–90 min

2–3 h

16–36 h

No antidote; consider hemodialysis

Argatroban (Eagle Pharmaceuticals) [60, 68]

PCI (patients with heparin allergy); HITT treatment and prophylaxis


Reversible direct thrombin inhibition


40–50 min

2–4 h

No antidote; consider hemodialysis

Dabigatran (Pradaxa, Boehringer Ingelheim) [57, 58, 69]

Non-valvular AF


Reversible direct thrombin inhibition

0.5–2 h

12–17 h

24–36 h

No antidote; charcoal for overdose (ingestion <2 h); consider hemodialysis (~60 % removal), rFVIIa, PCC, or FEIBA

Rivaroxaban (Xarelto, Janssen Pharmaceuticals) [57, 58, 70]

VTE treatment and prophylaxis; non-valvular AF


Reversible direct factor Xa inhibition

1–4 h

5–13 h

24 h

No antidote; charcoal for overdose (ingestion <2 h); consider PCC

Apixaban (Eliquis, Bristol-Myers Squibb) [71]

Non-valvular AF; phase III studies: VTE treatment and prophylaxis


Reversible direct factor Xa inhibition

1–4 h

8–15 h

24 h

No antidote; charcoal for overdose (ingestion <2 h); consider PCC

Edoxaban (Daiichi Sankyo) (investigational) [72, 73]

Phase III studies: VTE treatment and prophylaxis (Japan); non-valvular AF (USA)


Reversible direct factor Xa inhibition

1–2 h

6–11 h

24–36 h

No antidote; charcoal for overdose (ingestion <2 h); consider PCC

VTE venous thromboembolism, AF atrial fibrillation, MI myocardial infarction, PO per oral, INR international normalized ratio, FFP fresh frozen plasma, PCC prothrombin complex concentrates, rFVIIa recombinant activated factor VIIa, ACS acute coronary syndrome, IV intravenous, SC subcutaneous, AT antithrombin, PCI percutaneous coronary intervention, HITT heparin-induced thrombocytopenia and thrombosis

a Note: elimination half-life is dose dependent

Vitamin K antagonists (VKAs), such as warfarin, are the mainstay of chronic anticoagulation therapy. VKAs inhibit γ-carboxylation of vitamin K epoxide reductase in the liver, which inhibits the production of factors II, VII, IX, and X in the coagulation cascade. While VKAs are effective at reducing thromboembolic events, they have several limitations, including slow onset of action (~5 to 7 days to therapeutic INR), need for regular monitoring, variability in drug metabolism, narrow therapeutic window (usually an INR between 2.0 and 3.0), and several drug and dietary interactions. Approximately 5 days are needed for the INR to normalize after VKA cessation. The effects of VKAs can be reversed more rapidly with administration of vitamin K and fresh frozen plasma (FFP) primarily.

Unfractionated heparin (UFH) can be administered in intravenous (IV) and subcutaneous (SC) forms. Its mode of action is through antithrombin (AT) III-mediated inhibition of factor Xa and thrombin (factor IIa) of the coagulation cascade. Intravenous formulations are used for treatment of VTE, ACS, and bridging anticoagulation for AF and cardioversion. Subcutaneous formulations are used for VTE prophylaxis. The IV UFH anticoagulant response is monitored by measuring the activated partial thromboplastin time (aPTT) at 6 h intervals. UFH is favored over low-molecular-weight heparin (LMWH) in certain clinical situations given its short half-life, reversal capabilities, and safe use in patients with renal dysfunction. Urgent reversal can be achieved with protamine sulfate [57].

LMWHs have increased bioavailability over UFH when administered subcutaneously. LMWHs inhibit factor Xa and, to a lesser degree, thrombin (IIa) to achieve their anticoagulant effects. Laboratory monitoring is usually not needed, but anti-Xa assays are used in select patients. Clinical indications are similar to UFH, and urgent reversal of anticoagulant effects can partially be achieved with protamine sulfate [57].

Fondaparinux is administered subcutaneously and inhibits factor Xa [57, 60]. This agent is approved for use in the prophylaxis and treatment of VTE and may be employed in situations where UFH and LMWH cannot be used, such as in the setting of heparin-induced thrombocytopenia and thrombosis (HITT). Monitoring is not usually necessary, but an anti-Xa assay may be used to identify if activity is present [74]. Recombinant activated factor VII (rFVIIa) can be considered for emergent reversal [60].

Bivalirudin and desirudin are synthetic analogs of r-hirudin. They reversibly bind to the enzymatic catalytic site and anion binding site of thrombin [57]. The short half-life of bivalirudin enables its use in the periprocedural setting. It is an accepted alternative anticoagulant to UFH for percutaneous coronary interventions (PCI) and ST elevation myocardial infarction (STEMI), as well as in select patients with unstable angina (UA) and non-ST elevation myocardial infarction (NSTEMI) [53, 75]. There is some evidence that bleeding complications are lower with bivalirudin than with combination UFH and glycoprotein IIb/IIIa inhibitors (GPI) in the setting of ACS [53, 76]. Bivalirudin may be monitored by activated clotting time (ACT) [74]. Desirudin has been used mainly for VTE prophylaxis [66]. Monitoring can be done by following the aPTT [74]. Serious bleeding complications with desirudin are comparable to SC UFH and LMWH [66]. There are no known reversal agents for bivalirudin and desirudin.

Argatroban is an IV anticoagulant derived from the amino acid arginine and reversibly binds to the thrombin active site. It has a short half-life, and coagulation parameters normalize within hours of infusion cessation but may take longer in patients with hepatic impairment. The aPTT or ACT should be followed for appropriate dosing. It is used primarily in the management of HITT and as a potential alternative to UFH during PCI in patients with heparin allergy [52, 57, 60]. There is no known reversal agent for argatroban.

Several novel oral anticoagulants have recently been marketed for use or are in late phases of clinical trials. These new agents provide the convenience of oral administration and avoid many of the limitations of warfarin. However, there are reports of increased clinically relevant bleeding complications, including GIB, with the new oral anticoagulant agents compared with standard therapies [18]. Dabigatran is a direct thrombin inhibitor approved for use in non-valvular AF [77]. Time to maximal effect is 0.5–2 h with a terminal half-life of 12–17 h at steady-state levels [58]. There is no specific reversal agent. Because dabigatran is a direct thrombin inhibitor, administration of FFP or prothrombin complex concentrate (PCC) may not be completely effective in reversing its effects. Hemodialysis may be effective at removing dabigatran (~60 %) from the bloodstream, and activated charcoal may be helpful in the setting of overdose [77]. Rivaroxaban is a direct factor Xa inhibitor and is approved for treatment and prophylaxis of VTE and stoke prevention in the setting of non-valvular AF [57, 58]. Time to maximal inhibition is 1–4 h. Its half-life is 5–13 h [58]. There is no specific reversal agent. Activated charcoal may be useful in the setting of overdose. However, given that rivaroxaban is highly protein bound, hemodialysis will not be effective in removing it from plasma. As it is an upstream inhibitor of coagulation, administration of FFP, PCC, or rFVIIa may reverse its effects [57]. Apixaban (recently FDA approved) and edoxaban (in phase III clinical trials) are both direct factor Xa inhibitors with similar indications and pharmacologic properties as rivaroxaban [57].

Only gold members can continue reading. Log In or Register to continue

May 30, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on Periprocedural Management of Antithrombotic Agents
Premium Wordpress Themes by UFO Themes