Antithrombotic and Thrombolytic Therapy for Valvular DiseaseAntithrombotic Therapy for Valvular Disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines FREE TO VIEW

Note on Shaded Text: Throughout this guideline, shading is used within the summary of recommendations sections to indicate recommendations that are newly added or have been changed since the publication of Antithrombotic and Thrombolytic Therapy: American College of Chest Physicians Evidence- Based Clinical Practice Guidelines (8th Edition). Recommendations that remain unchanged are not shaded.

2.0.1. In patients with rheumatic mitral valve disease and normal sinus rhythm with a left atrial diameter < 55 mm we suggest not using antiplatelet or vitamin K antagonist (VKA) therapy (Grade 2C).

2.0.2. In patients with rheumatic mitral valve disease and normal sinus rhythm with a left atrial diameter > 55 mm, we suggest VKA therapy (target international normalized ratio [INR], 2.5; range, 2.0-3.0) over no VKA therapy or antiplatelet (Grade 2C).

2.0.3. For patients with rheumatic mitral valve disease complicated by the presence of left atrial thrombus, we recommend VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy (Grade 1A).

2.0.4. For patients with rheumatic mitral valve disease complicated singly or in combination by the presence of atrial fibrillation or previous systemic embolism, we recommend VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy (Grade 1A).

2.1.1. For patients being considered for percutaneous mitral balloon valvotomy (PMBV) with preprocedural transesophageal echocardiography (TEE) showing left atrial thrombus, we recommend postponement of PMBV and that VKA therapy (target INR, 3.0; range, 2.5-3.5) be administered until thrombus resolution is documented by repeat TEE over no VKA therapy (Grade 1A).

2.1.2. For patients being considered for PMBV with preprocedural TEE showing left atrial thrombus, if the left atrial thrombus does not resolve with VKA therapy, we recommend that PMBV not be performed (Grade 1A).

6.2.1. In patients with asymptomatic patent foramen ovale (PFO) or atrial septal aneurysm, we suggest against antithrombotic therapy (Grade 2C).

6.2.2. In patients with cryptogenic stroke and PFO or atrial septal aneurysm, we recommend aspirin (50-100 mg/d) over no aspirin (Grade 1A).

6.2.3. In patients with cryptogenic stroke and PFO or atrial septal aneurysm, who experience recurrent events despite aspirin therapy, we suggest treatment with VKA therapy (target INR, 2.5; range, 2.0-3.0) and consideration of device closure over aspirin therapy (Grade 2C).

6.2.4. In patients with cryptogenic stroke and PFO, with evidence of DVT, we recommend VKA therapy for 3 months (target INR, 2.5; range, 2.0-3.0) (Grade 1B) and consideration of device closure over no VKA therapy or aspirin therapy (Grade 2C).

7.1.1. In patients with infective endocarditis (IE), we recommend against routine anticoagulant therapy, unless a separate indication exists (Grade 1C).

7.1.2. In patients with IE, we recommend against routine antiplatelet therapy, unless a separate indication exists (Grade 1B).

7.2. In patients on VKA for a prosthetic valve who develop IE, we suggest VKA be discontinued at the time of initial presentation until it is clear that invasive procedures will not be required and the patient has stabilized without signs of CNS involvement. When the patient is deemed stable without contraindications or neurologic complications, we suggest reinstitution of VKA therapy (Grade 2C).

7.3. In patients with nonbacterial thrombotic endocarditis and systemic or pulmonary emboli, we suggest treatment with full-dose IV unfractionated heparin (UFH) or subcutaneous low-molecular-weight heparin (LMWH) over no anticoagulation (Grade 2C).

8.2.1. In patients with aortic bioprosthetic valves, who are in sinus rhythm and have no other indication for VKA therapy, we suggest aspirin (50-100 mg/d) over VKA therapy in the first 3 months (Grade 2C).

8.2.2. In patients with transcatheter aortic bioprosthetic valves, we suggest aspirin (50-100 mg/d) plus clopidogrel (75 mg/d) over VKA therapy and over no antiplatelet therapy in the first 3 months (Grade 2C).

8.2.3. In patients with a bioprosthetic valve in the mitral position, we suggest VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy for the first 3 months after valve insertion (Grade 2C).

8.3. In patients with bioprosthetic valves in normal sinus rhythm, we suggest aspirin therapy over no aspirin therapy after 3 months postoperative (Grade 2C).

9.1. In patients with mechanical heart valves, we suggest bridging with unfractionated heparin (UFH, prophylactic dose) or LMWH (prophylactic or therapeutic dose) over IV therapeutic UFH until stable on VKA therapy (Grade 2C).

9.2. In patients with mechanical heart valves, we recommend VKA therapy over no VKA therapy for long-term management (Grade 1B).

9.3.1 In patients with a mechanical aortic valve, we suggest VKA therapy with a target of 2.5 (range, 2.0-3.0) over lower targets (Grade 2C).

9.3.2. In patients with a mechanical aortic valve, we recommend VKA therapy with a target of 2.5 (range 2.0-3.0) over higher targets (Grade 1B).

9.4. In patients with a mechanical mitral valve, we suggest VKA therapy with a target of 3.0 (range, 2.5-3.5) over lower INR targets (Grade 2C).

9.5. In patients with mechanical heart valves in both the aortic and mitral position, we suggest target INR 3.0 (range 2.5-3.5) over target INR 2.5 (range 2.0-3.0) (Grade 2C).

9.6. In patients with a mechanical mitral or aortic valve at low risk of bleeding, we suggest adding over not adding an antiplatelet agent such as low-dose aspirin (50-100 mg/d) to the VKA therapy (Grade 1B).

Remarks: Caution should be used in patients at increased bleeding risk, such as history of GI bleeding.

9.7. For patients with mechanical aortic or mitral valves we recommend VKA over antiplatelet agents (Grade 1B).

10.1. In patients undergoing mitral valve repair with a prosthetic band in normal sinus rhythm, we suggest the use of antiplatelet therapy for the first 3 months over VKA therapy (Grade 2C).

10.2. In patients undergoing aortic valve repair, we suggest aspirin at 50 to 100 mg/d over VKA therapy (Grade 2C).

11.1. For patients with right-sided prosthetic valve thrombosis (PVT), in the absence of contraindications we suggest administration of fibrinolytic therapy over surgical intervention (Grade 2C).

11.2.1. For patients with left-sided PVT and large thrombus area (≥ 0.8 cm²), we suggest early surgery over fibrinolytic therapy (Grade 2C). If contraindications to surgery exist, we suggest the use of fibrinolytic therapy (Grade 2C).

11.2.2. For patients with left-sided PVT and small thrombus area (< 0.8 cm²), we suggest administration of fibrinolytic therapy over surgery. For very small, nonobstructive thrombus we suggest IV UFH accompanied by serial Doppler echocardiography to document thrombus resolution or improvement over other alternatives (Grade 2C).

Thromboembolic complications of valvular heart disease are often devastating. Antithrombotic therapy can reduce the risk of thromboembolism, but at the cost of increased bleeding. This article seeks to provide recommendations based on the optimal balance of these competing factors.

Table 1 describes the population, intervention, comparator, and outcome (PICO) elements for the questions addressed in this article and the design of the studies used to address them. We define only patient characteristics relevant to our questions. This article does not make recommendations specific to atrial fibrillation (AF); for this issue, we direct you to the article by You et al¹ on AF in this supplement. In areas of overlap with the AF article, where newer anticoagulants such as dabigatran may be considered for nonvalvular AF, caution must be used when extrapolating their use to the populations described in this article. This article continues to consider vitamin K antagonists (VKAs) as the first-line oral anticoagulant until evidence of efficacy and safety within the valve population is generated. For recommendations on the management of parenteral anticoagulation (dosing and monitoring), oral anticoagulation (dosing and monitoring), and bleeding complications, please refer to the article by Holbrook et al² about management of anticoagulation in this guideline. Finally, there are very few data directly addressing the antithrombotic management of right-sided prosthetic valves. Indirect evidence from mitral and aortic valves provides the best evidence and the basis for recommendations regarding tricuspid and pulmonic prostheses.

The development of the current recommendations followed the general approach of Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines. Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines.³ In brief, literature searches to update the existing database from the AT8 guidelines were performed (January 1, 2005 to October 2009). The literature was rated according to the Grades of Recommendations, Assessment, Development, and Evaluation (GRADE) framework. The panel considered quality of information, balance of risk and harm, and patients’ values and preferences to determine the strength of recommendation.

In making recommendations, we have taken a primum non nocere approach, placing the burden of proof with those who would claim a benefit of treatment. In other words, when there is uncertain benefit and an appreciable probability of important harm associated with treatment, we recommend against such treatments.

The value given to the harmful effect of an extracranial bleeding event (as compared with that of valve thrombosis, peripheral thromboembolism, or stroke) greatly impacts the balance of benefits and harms of a given therapy. There are limited data to guide us with respect to the relative value of these outcomes. For this article, we used the result of the preference-weighting exercise carried out by MacLean et al⁴ as part of these guidelines, which attributes approximately three times the disutility (aversiveness, negative weight) to a stroke vs an extracranial bleeding event; a valve thrombosis carries slightly greater disutility than an extracranial bleeding event.

Rheumatic mitral valve disease carries the greatest risk of systemic thromboembolism of any common form of acquired valvular disease. Wood⁵ cited a prevalence of systemic emboli of 9% to 14% in several large early series of patients with mitral stenosis. In 1961, Ellis and Harken⁶ reported that 27% of 1,500 patients undergoing surgical mitral valvotomy had a history of clinically detectable systemic emboli. Among 754 patients followed up for 5,833 patient-years, Szekely⁷ observed an incidence of emboli of 1.5% per year, whereas the number was found to vary from 1.5% to 4.7% per year preoperatively in six reports of rheumatic mitral valve disease.⁸ Although the risk may increase in the elderly and those with lower cardiac indices,^9‐12 these findings have been inconsistent across studies.^5,13‐21 Other characteristics that may increase the risk of systemic embolism include the presence of a left atrial thrombus and significant aortic regurgitation.²²

The relationship between thromboembolism and left atrial size remains unclear. Early studies^5,13,14 of rheumatic mitral valve disease reported a weak correlation. However, several studies have now demonstrated an association between larger left atrial size and left atrial thrombus or spontaneous echocardiographic contrast.^15‐17

In those patients with rheumatic mitral valve disease who suffer a first embolus, recurrent emboli occur frequently (one-third to two-thirds of cases) and early (two-thirds within the first year).^5,23‐25 A hypercoagulable state in mitral stenosis might contribute to the risk of thromboembolism.^26,27 No randomized trial has been completed in this population, but observational data suggest that the risk of recurrent emboli may be reduced by VKA therapy. Szekely⁷ found a recurrence rate of 9.6%/y with no anticoagulation and 3.4%/y with warfarin (relative risk [RR], 0.36; 95% CI, 0.08-1.6). Similar estimates have been reported by others.^14,28 Among patients with mitral stenosis and left atrial thrombus on transesophageal echocardiography (TEE), VKA therapy results in a 62% thrombus disappearance over an average of 34 months.²⁹

The onset of AF increases the risk of systemic embolization in patients with rheumatic mitral valve disease.^7,13 As in those with recurrent embolism, observational studies suggest a large decrease in risk with warfarin administration.^13,30 Indirect evidence from randomized trials in nonvalvular AF provide further support for the impact of warfarin in the prevention of thromboembolism in patients with rheumatic mitral valve with AF.

Recommendations

2.0.1. In patients with rheumatic mitral valve disease and normal sinus rhythm with a left atrial diameter < 55 mm, we suggest not using antiplatelet or VKA therapy (Grade 2C).

2.0.2. In patients with rheumatic mitral valve disease and normal sinus rhythm with a left atrial diameter > 55 mm, we suggest VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy or antiplatelet (Grade 2C).

2.0.3. For patients with rheumatic mitral valve disease complicated singly or in combination by the presence of left atrial thrombus, we recommend VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy (Grade 1A).

2.0.4. For patients with rheumatic mitral valve disease complicated singly or in combination by the presence of AF or previous systemic embolism, we recommend VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy (Grade 1A).

During the percutaneous mitral balloon valvotomy (PMBV) procedure, when the catheter is pushed through the septum it often goes into the left atrial appendage, the usual site of thrombus. Thus, the presence of left atrial thrombus precludes PMBV. Accurate detection of thrombus requires a transesophageal echocardiogram (TEE). Silaruks et al³¹ have demonstrated that 24.2% of left atrial thrombi will resolve within 6 months of anticoagulation. Further, Kang et al³² have demonstrated that after thrombus resolution, PMBV can be safely performed. Predictors of thrombus resolution are New York Heart Association (NYHA) functional class II or better, left atrial appendage thrombus size ≤ 1.6 cm², less dense spontaneous echocardiographic contrast, and an INR ≥ 2.5. Patients with all of these predictors had a 94.4% chance of complete thrombus resolution at 6 months.³¹

In those patients without left atrial thrombus and no other indication for anticoagulation, Abraham et al³³ demonstrated PMBV can be performed in the absence of anticoagulation, with no patients in 629 procedures performed having an embolism within 3 months post procedure.

Recommendations

2.1.1. For patients being considered for PMBV with preprocedural TEE showing left atrial thrombus, we recommend postponement of PMBV and that VKA therapy (target INR, 3.0; range, 2.5-3.5) be administered until thrombus resolution is documented by repeat TEE over no VKA therapy (Grade 1A).

2.1.2. For patients being considered for PMBV with preprocedural TEE showing left atrial thrombus, if the left atrial thrombus does not resolve with VKA therapy, we recommend that PMBV not be performed (Grade 1A).

Mitral valve prolapse (MVP) is a common congenital form of valve disease. Although early evidence from case series and control studies suggested an association with stroke,^34‐40 Gilon et al⁴¹ and the Framingham Heart Study⁴² failed to replicate the results. More recently, Avierinos et al⁴³ found that people with MVP had an excess lifetime risk of stroke or transient ischemic attack (TIA) (RR, 2.2; P < .001). Thus, it is as yet unclear whether MVP truly increased the risk of thromboembolic process. Mitral valve strands, also known as Lambl’s excrescences, have also been implicated as a potential embolic source, but they do not seem to increase the risk of stroke recurrence.⁴⁴ We therefore suggest that patients with MVP or strands who have not experienced systemic embolism, unexplained TIAs, or ischemic stroke, and do not have evident vascular disease should be managed as other patients considering primary prevention of vascular events. Given the risk of bleeding complications with anticoagulation and the lack of data to demonstrate a benefit in terms of reducing (recurrent) thromboembolic events, patients with MVP or strands and a history of ischemic stroke or TIA should be treated with antiplatelet agents following the recommendations by Landsberg et al⁴⁵ for patients with noncardioembolic stroke. In those patients with MVP or strands who have recurrent thromboembolic events despite antiplatelet agent (APA) therapy, the likelihood of a cardiac source increases.

Mitral annular calcification (MAC), like MVP, may be a source of cardioembolic stroke. The best estimate of the embolic potential of MAC comes from the Framingham Heart Study.⁴⁶ Among 1,159 individuals with no history of stroke at the index echocardiographic examination, the risk of stroke in those with MAC was 2.1 times greater than those without MAC (5.1% without MAC vs 13.8% with MAC, P = .006), independent of traditional risk factors for stroke. There was a continuous relationship between the risk of stroke and the severity of the MAC. A major issue in this condition is that emboli may represent thrombus or calcific spicules, the latter of which antithrombotic therapy will not prevent.^46‐48 From the available literature, we suggest that patients with MAC who have not experienced systemic embolism, unexplained TIAs, or ischemic stroke, and do not have evident vascular disease should be managed as other patients considering primary prevention of vascular events. It would be reasonable to manage patients with MAC and evidence of thromboembolic process with no other identifiable source as patients with TIAs without MAC.⁴⁵ Failure of this antithrombotic therapy or evidence of multiple calcific emboli should prompt consideration of valve replacement.

Clinically significant systemic emboli in isolated aortic valve disease are uncommon. A lack of association between aortic valve calcification and clinical emboli has been supported by several studies.^49‐51 Thus, in the absence of other indications, antithrombotic therapy does not have a role in calcified aortic valve disease.

The presence of aortic plaque is associated with stroke risk.^52,53 In a TEE substudy of the Stroke Prevention in Atrial Fibrillation (SPAF) trial, the risk of stroke at 1 year in patients with AF with complex aortic plaque was 12% to 20% vs 1.2% if no plaque was observed.⁵⁴ Cohen et al⁵⁵ demonstrated that aortic plaques > 4 mm in thickness increased the risk of vascular events, and this risk was further increased by lack of plaque calcification (RR = 10.3; 95% CI, 4.2-25.2). There are no randomized trials assessing the effectiveness of anticoagulation therapy for the prevention of ischemic embolic events in patients with aortic plaque.

Ferrari et al⁵⁶ examined the effects of antithrombotic therapy in an observational study of 129 patients with aortic atheroma on TEE. They found that patients treated with APAs rather than VKAs had more combined vascular events and a higher mortality rate (RR = 5.9; 95% CI, 1.4-15). However, Tunick et al⁵⁷ reported that antithrombotic therapy did not significantly reduce recurrent events results in 519 patients with severe aortic plaque ( ≥ 4 mm) identified during TEE evaluation for embolic events.

There is currently insufficient evidence to support the use of antithrombotic therapy for the prevention of ischemic events in patients with severe thoracic aortic atheroma.⁵⁸ We, therefore, suggest that patients with aortic atheroma who have not experienced systemic embolism, unexplained TIAs, or ischemic stroke, and do not have evident vascular disease should be managed as other patients considering primary prevention of vascular events.⁴⁵ Patients with atherosclerotic aorta and evidence of thromboembolic process with no other identifiable source should be managed as those with TIAs and no atheromatous disease.⁴⁵

In patients with patent foramen ovale (PFO) and atrial septal aneurysm who suffer an ischemic stroke, the source is unclear in approximately 40%.⁵⁹ The interatrial septum has received considerable attention as a possible source of cryptogenic stroke. Paradoxical embolism through a PFO is well described, and atrial septal aneurysm with thrombus has been demonstrated.⁶⁰ However, PFO and septal aneurysm are weak risk factors for stroke.

Patient characteristics that have been associated with ischemic stroke in PFO include larger-sized PFO, hemodynamic states that result in right atrial pressure overload with right-to-left shunting, hypercoagulability, the presence of eustachian valve, Chiari network, and atrial septal aneuryms.^59,61 More recent studies^62,63 from Olmsted County, Minnesota and the Stroke Prevention: Assessment of Risk in a Community (SPARC) study have suggested that after adjusting for age and other comorbidities associated with stroke, PFO is not an independent risk factor for stroke. This may be a function of how PFOs are detected. TEE with saline contrast injection is the diagnostic technique of choice for demonstrating a PFO.⁶⁴ However, since the sensitivity of saline contrast TEE is greater than that of transthoracic echocardiography, whether smaller PFOs identified only by TEE are clinically relevant remains uncertain.

Mas et al⁶⁵ suggest that patients with both a PFO and atrial septal aneurysm who have had cryptogenic ischemic stroke are at particularly high risk for recurrence. At 4 years of follow-up, the risk of recurrent stroke in the presence of an isolated PFO was 2.3% (95% CI, 0.3%-4.3%), 15.2% (95% CI, 1.8%-28.6%) among the patients with both PFO and atrial septal aneurysm, and 4.2% (95% CI, 1.8%-6.6%) among the patients with neither of these abnormalities. All patients within this study received aspirin, with no comparator group. Therefore, conclusions on appropriate antithrombotic therapy vs no antithrombotic therapy are currently not possible. Patients with evidence of thromboembolic process and no other identifiable source should be managed as those with cryptogenic TIA or stroke.⁴⁵

Homma et al⁶⁶ reported on the subgroup of 203 patients with PFO in the Warfarin-Aspirin Recurrent Stroke Study (Table 2, Table S1). (Tables that contain an “S” before the number denote supplementary tables not contained in the body of the article and available instead in an online data supplement. See the “Acknowledgments” for more information.) For the outcome of stroke or death at 2 years, the results neither establish nor exclude a difference between aspirin therapy and VKA therapy (RR, 0.80; 95% CI, 0.41-1.55).

In patients with cryptogenic systemic thromboembolism, the demonstration of right-to-left shunting through a PFO warrants a search for DVT. Evidence for venous thrombosis (or pulmonary embolism) together with systemic embolism and a PFO provides a strong indication for anticoagulation, and when technically feasible, closure of the PFO. There are several ongoing trials of device closure which will better inform this area in the near future.

Recommendations

6.2.1. In patients with asymptomatic PFO or atrial septal aneurysm, we suggest against antithrombotic therapy (Grade 2C).

6.2.2. In patients with cryptogenic stroke and PFO or atrial septal aneurysm, we recommend aspirin (50-100 mg/d) over no aspirin (Grade 1A).

6.2.3. In patients with cryptogenic stroke and PFO or atrial septal aneurysm, who experience recurrent events despite aspirin therapy, we suggest treatment with VKA therapy (target INR, 2.5; range, 2.0-3.0) and consideration of device closure over aspirin therapy (Grade 2C).

6.2.4. In patients with cryptogenic stroke and PFO, with evidence of DVT, we recommend VKA therapy for 3 months (target INR, 2.5; range, 2.0-3.0) (Grade 1B) and consideration of device closure over no VKA therapy or aspirin therapy (Grade 2C).

Native valve infective endocarditis (IE) is a serious infectious entity, the morbidity of which is primarily related to the consequences of systemic embolism from valve vegetations. The risk of embolization is proportional to the size of the vegetation and the type of organism (eg, Staphylococcus aureus increases risk).^67,68 The majority of clinically apparent emboli from left-sided lesions involve the CNS resulting in catastrophic stroke. The incidence of pulmonary emboli in right-sided endocarditis is also high, and this complication may contribute to significant respiratory complications, including lung abscess and pulmonary hypertension.⁶⁹

Antibiotics are the most important medical therapy to decrease the incidence of emboli from native bacterial endocarditis. Whereas in the preantibiotic era, clinically detectable emboli occurred in 70% to 97% of patients, since that time the prevalence has ranged between 12% and 40%.^69‐73 Further, the incidence of embolic complications, highest at the onset of disease, falls precipitously after 2 weeks of appropriate antibiotic therapy, from approximately 15 embolic events per 1,000 patient-days to fewer than two events per 1,000 patient-days.⁶⁹

The use of anticoagulant therapy in IE was initially introduced as a mechanism to improve the penetration of antibiotics into infected vegetations.⁷⁵ When closely examined, the effect of anticoagulants on the incidence of embolism was not evident.⁷⁶ In a retrospective study of 61 patients with native valve endocarditis, Paschalis et al⁷⁷ reported that 18 patients suffered embolic neurologic complications. The incidence of embolism was the same with and without anticoagulation. Subsequent reports have demonstrated that patients treated with anticoagulant therapy were at significant risk of intracerebral hemorrhage (ICH). Thill et al⁷⁸ described 22 patients taking combined penicillin and dicumarol with a high incidence of fatal cerebral hemorrhage. Other groups have reported an alarming incidence of cerebral hemorrhage.^79‐83

A trial of 115 patients with IE who were randomized to aspirin treatment reported the effect of aspirin therapy on the risk of embolic events in IE (n = 60, 325 mg/d) or placebo (n = 55) for 4 weeks (Table 3, Table S2).⁸⁴ The addition of aspirin did not reduce the risk of embolic events, with 17 (28.3%) such events in the aspirin group vs 11 (20.0%) in the placebo group (OR, 1.62; 95% CI, 0.68-3.86). These data provide moderate-quality evidence (wide CI includes benefit from aspirin) that there is no role for APA therapy in IE unless another indication exists.

In summary, there is no convincing evidence that prophylactic anticoagulant therapy reduces the incidence of emboli in this disorder. The evidence to date further suggests that anticoagulant therapy in this setting increases the rate of neurologic complications related to cerebral hemorrhage.

Recommendations

7.1.1. In patients with IE, we recommend against routine anticoagulant therapy, unless a separate indication exists (Grade 1C).

7.1.2. In patients with IE, we recommend against routine antiplatelet therapy, unless a separate indication exists (Grade 1B).

The risk of thromboembolic events in prosthetic valve endocarditis (PVE) is higher than that in native valve endocarditis, with reports of rates between 50% and 88% of patients.^71,73,85,86 Antimicrobial therapy remains the mainstay of embolization prevention; delay in therapy is related to the frequency of embolic stroke within 3 days of diagnosis.⁸⁷

Only one article has reported benefit of antithrombotic therapy in PVE. In an observational study, Wilson et al⁸⁶ described CNS complications in only three of 38 patients with PVE who received adequate anticoagulant therapy compared with 10 of 14 patients who received either inadequate or no anticoagulation. The majority of studies suggest the risk of continuing anticoagulation in this disorder outweighs the potential benefits. In another observational study, Yeh et al⁸⁸ reported that therapeutic anticoagulation not only failed to control emboli in PVE, but the risk of bleeding appeared to be greater among patients with infected prostheses. Others have published similar observational results.^73,89 Some authors continue to suggest that anticoagulant therapy should be continued in patients with PVE,^72,85,86 whereas others do not.^69,73

In conclusion, the use of anticoagulants in PVE must steer a path between the potential for thromboembolism and the risk of serious bleeding, including ICH. Although one might expect that the incidence of thromboembolism will be reduced by anticoagulant therapy, there is no evidence that embolic vegetations are controlled by this therapy. Further, the consequences of ICH may be irreversible and not infrequently fatal. Embolic events in PVE may represent dislodged vegetations or, alternatively, true thromboembolism unrelated to the valve infection.

Recommendation

7.2. In patients on VKA for a prosthetic valve who develop IE, we suggest VKA be discontinued at the time of initial presentation until it is clear that invasive procedures will not be required and the patient has stabilized without signs of CNS involvement. When the patient is deemed stable without contraindications or neurologic complications, we suggest reinstitution of VKA therapy (Grade 2C).

The clinical picture of nonbacterial thrombotic endocarditis (NBTE) has been well characterized by Lopez et al.⁹⁰ These lesions generally are rounded, sessile, > 3 mm, and heterogeneous in shape compared with excrescences, which are smaller ( < 2 mm) and sometimes elongated. Excrescences are seen exclusively near leaflet closure lines.

Although most authors agree that key principle for dealing with NBTE is to control underlying disease, anticoagulants were recommended due to the general belief that Trousseau syndrome and NBTE represent a continuum and that disseminated intravascular coagulation is the basis for treating most patients with this disorder.⁹¹ In support of anticoagulant therapy, recurrent thromboembolic complications have been reported after heparin therapy was discontinued, although little apparent benefit has been observed with VKA therapy.^90‐92

Recommendation

7.3. In patients with NBTE and systemic or pulmonary emboli, we suggest treatment with full-dose IV UFH or subcutaneous LMWH over no anticoagulation (Grade 2C).

There are no studies examining early bridging therapy such as UFH or LMWH prior to antiplatelet therapy or VKA initiation in the bioprosthetic valve population. Therefore, we are currently unable to make recommendations on this topic.

The first 3 months after valve implantation are a high-risk period for thromboembolic events, particularly in the mitral valve population.^93,94 Because the risk of a thromboembolic event varies by valve location, we have generated separate evidence profiles by this division.

Evidence comparing VKA to no VKA is available from observational studies with a focus on stroke and major hemorrhage. The quality of the evidence is low due to study limitations and imprecision (Table 4, Table S3).^95,96 Moinuddeen et al⁹⁵ failed to demonstrate or exclude an effect of oral anticoagulation therapy on stroke (RR, 1.1; 95% CI, 0.38-3.28). Blair et al⁹⁶ demonstrated a trend toward increased risk of major hemorrhage but failed to establish or refute effect on thrombosis. Indirect supporting evidence that VKA therapy leads to an increased risk of hemorrhage compared with aspirin or no therapy also comes from studies in patients with AF.¹

Two randomized trials have compared antiplatelet therapy with VKA for the initial antithrombotic management of patients with bioprosthetic heart valves (Table 5, Table S4).^97,98 Aramendi et al⁹⁷ randomized 191 patients to either antiplatelet therapy with triflusal 600mg/d (an antiplatelet agent similar to aspirin that irreversibly inhibits cyclooxygenase) or acenocoumarol (target INR range, 2.0-3.0) using an open-label design. The study included patients with aortic (93.8%) and mitral (5.2%) bioprosthetic valves and is thus most applicable to patients with aortic valve prostheses. Study treatments were started within 48 h of surgery and were continued for 3 months with follow-up to 180 days. The primary outcome was the composite of thromboembolism, hemorrhage, and valve-related death. Results failed to demonstrate or exclude a beneficial effect or detrimental effect on the primary efficacy outcome (RR, 0.89; 95% CI, 0.38-2.09) or treatment-related bleeding (RR, 0.50; 95% CI, 0.13-1.92).

Colli et al⁹⁸ performed a pilot randomized trial of 75 aortic valve patients who received either warfarin (n = 34) or aspirin (n = 35) therapy. The warfarin group, starting on postoperative day 1, received VTE prophylaxis-dose LMWH, and warfarin was started on day 2. Warfarin was dosed to reach an INR of 2.0 to 3.0. This was continued for 3 months. The aspirin group also received VTE prophylaxis-dose LMWH starting on postoperative day 1 but were then given 100 mg daily of aspirin for 3 months. Given the size of the study, there are predictably no differences in the main outcomes of cerebral ischemic events, bleeding, and death. These studies were not blinded and reported few events. Therefore, the quality of this evidence is low.

Recommendation

8.2.1. In patients with aortic bioprosthetic valves, who are in sinus rhythm and have no other indication for VKA therapy, we suggest aspirin (50-100 mg/d) over VKA therapy in the first 3 months (Grade 2C).

Transcatheter aortic bioprosthesis is a new technology that compresses a tissue valve onto an expandable balloon, permitting placement without the traditional open-chest approach. The first implant in a human was in 2002.⁹⁹ There are no studies that compare antithrombotic strategies for these valves. The approach that has been adopted is an extension of the therapy used in coronary stenting: clopidogrel and aspirin for 3 to 6 months, followed by long-term aspirin therapy.^100,101 Further studies addressing antithrombotic management are required.

Recommendation

8.2.2. In patients with transcatheter aortic bioprosthetic valves, we suggest aspirin (50-100 mg/d) and clopidogrel (75 mg/d) over VKA therapy and over no antiplatelet therapy in the first 3 months (Grade 2C).

The risk of a stroke from a mitral bioprosthetic valve in the first postoperative month has been reported to be as high as 40 events per 100 patient-years.^93,94,102 The direct evidence on the effects of early anticoagulation on this risk is of low quality. Unlike the aortic bioprosthesis, randomized data comparing VKA to antiplatelet in the setting of mitral valve bioprosthesis is lacking. The best evidence comes from Heras et al⁹³ (Table 6, Table S5). Their observational study reports a trend toward reduced thromboembolic events in those receiving warfarin over no warfarin. However, the regimen of anticoagulation, target INR, and comparator are not clearly reported and the estimates are based on 11 events. Further, bleeding will increase relative to aspirin or no antithrombotic therapy.

Turpie et al¹⁰³ randomized patients with bioprosthetic valves to receive warfarin at a target INR range of 2.5 to 4.0 (n = 108) or 2.0 to 2.25 (n = 102) using an open-label design (Table 7, Table S6). This trial included patients with aortic, mitral, and tricuspid valves but was not large enough to present the results by subgroup of valve position. Patients started warfarin after surgery as soon as they were able to tolerate oral medications. Results failed to demonstrate or exclude a beneficial effect or detrimental effect of the different INR targets on thrombosis (1.9% vs 2.0% for major embolism and 10.2% vs 10.8% for minor embolism, P value not reported). There were significantly more bleeding events in patients treated with high- compared with low-intensity VKA therapy (13.9% vs 5.9%, P = .04). Different laboratory methods were used to monitor the intensity of VKA therapy in the two groups, and it is unclear whether this impacted results.

There is currently little evidence regarding the addition of APA to oral anticoagulation (OAC) in the early treatment of bioprosthetic valves. A Cochrane review provides indirect evidence in patients with mechanical valves that suggests a significant reduction in mortality and thromboembolic outcomes at the cost of increased risk of bleeding (section 9.6). However, the trials included were dominated by mechanical valves and we are not confident that the results are generalizable to the bioprosthetic valve population.

Recommendation

8.2.3. In patients with a bioprosthetic valve in the mitral position, we suggest VKA therapy (target INR, 2.5; range, 2.0-3.0) over no VKA therapy for the first 3 months after valve insertion (Grade 2C).

The long-term risk of thromboemboli with a bioprosthetic valve is in the range of 0.2% to 2.6%/y.¹⁰⁴ The risk is lower for patients with aortic valve position (0.2%/y) and sinus rhythm.¹⁰⁵ There is currently no evidence supporting the long-term use of oral anticoagulation in patients with bioprosthetic valves. Case series have reported very low event rates in patients receiving APA.^106‐109 For aortic bioprosthesis, Goldsmith et al¹⁰⁸ report that 145 patients in normal sinus rhythm treated with aspirin 75 mg daily suffered no major thromboembolic events and no major bleeding events in 254 patient-years of follow-up. Nunez et al¹⁰⁹ report that aspirin therapy in 185 patients receiving a mitral porcine bioprosthesis and in normal sinus rhythm resulted in no thromboembolic events.

Thromboembolism in patients with bioprosthetic valves and AF presumably relates to both the bioprosthetic valve and to the AF.¹ The incidence of thromboembolism in these patients was reported to be as high as 16% at 31 to 36 months.^110,111 Other factors such as lower ejection fraction and large left atrium have also been suggested to increase thromboembolic risk in the setting of bioprostheses¹⁰⁵; however, this evidence in not compelling and there is no evidence to support the use of oral anticoagulation in the presence of normal sinus rhythm.

Recommendation

8.3. In patients with bioprosthetic valves in normal sinus rhythm, we suggest aspirin therapy over no aspirin therapy after 3 months postoperative (Grade 2C).