With the evolution of therapies available for the treatment of aortic stenosis there has been an increase in the utilization rates of surgical bioprosthetic aortic valves (bio-SAVR). Patients are favoring bioprosthetic valves, in part due to their low thromboembolic risks and the ability to avoid long-term anti-coagulation with a vitamin K antagonist (VKA). The historical durability concerns of bio-SAVR have been mitigated with the growing opportunities for catheter-based therapies, such as "valve-in-a-valve" for failing bio-SAVRs. However, due to the concerns of early thromboembolic (TE) complications, current American College of Cardiology/American Heart Association (ACC/AHA) guidelines advocate a period of 3 months to protect against this early risk. The basis for this recommendation was from early experiences in biologic valves in which there was a theoretical concern that until the cloth surfaces of the valves endothelialize the patients are at increased risk for TE events and that such events can be minimized with a short period, typically 3 months, of anticoagulation with a VKA. The level of evidence to support this recommendation is classified as Class IIa (therapy is reasonable, can be useful, with Benefit >> Risk), Level of Evidence C (limited data and/or expert opinion, including meta-analyses of existing, but limited, randomized or nonrandomized observational or registry studies).¹ In addition, the most recent European Society Guidelines (2017) indicate that "low-dose aspirin (75 - 100 mg/day) should be considered for the first 3 months after surgical implantation of an aortic bioprosthesis or valve-sparing aortic surgery" (also Class IIa, Level of Evidence C) and "oral anticoagulation may be considered for the first 3 months after surgical implantation of an aortic bioprosthesis" (IIb, C).² However, given the implications of these recommendations (i.e., balancing embolic vs. bleeding risks), maybe now is the time to critically review the data and experiences that have accumulated over the years since the initial papers on this topic to see if there are opportunities to ensure that current guidelines reflect current science and clinical experiences.

The landmark paper by Heras et al. published in 1995 specifically explored the risk of thromboembolic events in patients undergoing bioprosthetic valve replacement.3 They retrospectively reviewed 812 patients who underwent either an aortic (n = 424), mitral (n = 326), or mitral and aortic (n = 66) biologic valve implanted between 01/1975 and 12/1982. The patients with neurologic deficits within 24 hours of surgery were excluded. One hundred and twelve patients experienced a thromboembolic event, a rate of 2.3% per year during their almost 10 years of follow-up. Most (58%) recovered, but 4% died and 38% had permanent neurologic damage. They noted that only 10% of aortic, 44% of mitral and 17% of double valve recipients were anticoagulated with a VKA at the time of their event. When exploring the time course of such events, Heras et al. observed that the greatest risk for a TE was during the first 10 days after surgery (not including intra-operative events). Specifically, for aortic valves, the annualized rate of events was 44% per year. However, this rate decreased to 3.5%/year for days 11-90, and 1.9%/year after 90 days. These differences were statistically significant. They concluded that the risk for thromboembolic events was greatest in the first 11 days after surgery and decreased significantly after 90 days in patients who were either not anticoagulated or were sub-therapeutic. While they did notice an increased rate of bleeding related complications, such events did not appear to be related to the use of anti-platelet agents or anti-coagulants. They advocated the following protocol and called for further studies on this topic:³

1. Initiation of heparin 6 hours after operation to prolonged PTT to "slightly" above normal
2. Increase heparin to target a PTT of 1.5 to 2.0 x control 6 hours after chest tube removal
3. Initiate low dose aspirin (80-100 mg) 48 hours post-op
4. Initiative anticoagulation the night BELOW surgery and continue for 3 months post-op
   1. Target INR should be 3.0 to 3.5

The conclusions of this article, as discussed below, have typically served as the basis for the guidelines recommending a 3-month interval of anticoagulation with a VKA. Of note, rarely discussed is that despite the dramatic statement of a 44% event rate per year, the actual number of events in the entire cohort during the first 10 days does not appear to support their conclusions, let alone appear (at least in the perspective of these authors) to be robust enough to serve as the basis for decades of International management guidelines (Table 1).

Table 1

Events	Anti-coagulated	Not anti-coagulated
Aortic (days 0-10)	0	5
Aortic (days 11-90)	1	2
Aortic (days >90)	3	40
Mitral (days 0-10)	2	3
Mitral (days 11-90)	5	1
Mitral (days >90)	17	27

A retrospective review of the Society of Thoracic Surgeons (STS) database of 25,656 over the age of 65, who received a bio-SAVR demonstrated that despite these guideline recommendations, only 12% of patients were discharged on warfarin and 23% were discharged on warfarin and aspirin with 49% receiving aspirin only.⁴ In the aspirin only group, at 3-month (the study period chosen to align with the current guidelines) the mortality was 3%, embolic risk was 1.0%, and bleeding risk was 1.0%. While those patients discharged with warfarin and aspirin tended to have a lower risk of death (relative risk: 0.80, 95% CI: 0.66-0.96), lower embolic risk (RR: 0.52, 95% 0.35-0.76), but a high risk of bleeding (RR: 2.8, 95% CI: 2.18-3.60). For those discharged with warfarin but without aspirin, the rates of death, bleeding and embolism were all similar when compared to aspirin early. While advocates of the use of anti-coagulation point to the survival advantage and decreased embolic risks as justification for this practice, opponents argue that there were substantial differences in the different populations that suggest a strong selection bias towards different treatment strategies that account for the outcomes beyond the decision to anti-coagulate or not. In other words, therapy with a VKA after bio-SAVR might be a marker for an inherently 'healthier' patient population who are less likely to experience life-threatening post-operative events independent of their anti-coagulation status. A similar argument is used to explain the small but significant survival advantage of patients who receive a mechanical valve versus a bioprosthetic valve.⁵ Nevertheless, the overall findings do not demonstrate a significant (statistical or clinical) benefit to an early interval of VKA therapy, but there is the observation that such a practice can increase the risk of bleeding complications.

Acknowledging the variability, often at the individual surgeon level, of the practice to anticoagulated after aortic tissues valves, Sundt et al. from the Mayo Clinic (this is the same group who published the original paper raising the concern of the embolic risk after surgery which served as the basis for the ACC/AHA guidelines in reference 3) reviewed their contemporary practice and outcomes.⁶ Retrospectively reviewing 1151 (1993-2000) patients who underwent isolated bio-SAVR (641) or bio-SAVR+CABG (51), they compared the 624 who were treated with anticoagulation (AC+) to the 527 who were not (AC-). Interestingly, in the AC+ group the rate of post-operative stroke was 2.4% versus 1.9% in the AC- group. In a multi-variate analysis, they determined that the risk of post-operative stroke was independent of the use of anticoagulation (p = 0.32). The rates of bleeding (i.e., need for surgical re-exploration) and bleeding related complications at 90 days were also similar between the two groups. Recognizing their role in authoring the initial paper that was used to establish the current guidelines, they emphasized the inherent limitations of their previous work. Although while the operative mortality rates were much higher in the AC- group (7.8%) versus the AC+ group (1.0%, p <0.001) this was explained that the patients who died early did not live long enough to be considered for anti-coagulation and may have had existing co-morbidities that heavily biased against the use of anti-coagulation, such as endocarditis with pre-operative stroke. Furthermore, they advocated that the observed stroke risk tended to be the consequence of intra-operative events and, therefore, not impacted by the decision to anti-coagulate. While the discussion for the paper focuses on the challenges of definitively answered the question regarding the role of an early interim of anticoagulation after bio-SAVR, the authors strongly make the point that the current guidelines are predominately based upon their previously reported manuscript with far fewer patients, fewer events and an overall questionable methodology when compared to their current study and the existing literature. Nevertheless, a key point of their paper was questioning the appropriateness of current guidelines, especially in light of the growing body of literature (of which only a few manuscripts have been highlighted in this review) that also questions the efficacy and safety of this practice.

A systematic review of the role of anti-thrombotic therapy after bio-SAVR in 2006 raised concerns about the clinical appropriateness of the current (at the time) recommendations of the American and European society's guidelines.⁷ In their review, Nowell acknowledged the initial work by Heras et al., who advocated the importance of 3 months of anticoagulation with a VKA based upon their linearized rate of TE complications of 44% per year (based on five events in the first 10 days in 424 patients). The review also highlighted a paper by Orszulak, who performed a retrospective study of 561 patients in which 5% of patients experienced a neurologic event during a 12-year follow-up period. Two percent of the patients (or 40% of all events) who experienced a neurologic event did so in the first 2 weeks after surgery – and none were on a VKA at the time of their event. This observation led them to conclude a potential role of VKA for an interim after surgery to help reduce the neurologic event rate.⁸ The review emphasizes that the current guidelines by the ACC/AHA, American College of Chest Physicians, and the European Society of Cardiology (several other national society guidelines were also referenced) all advocate some degree of anti-coagulation (with at least aspirin) in either low-risk patients or those who already have existing indications for anti-coagulation, such as atrial fibrillation. Nevertheless, despite widespread variability in this practice, the consensus was that the overall benefits outweigh the risks—real or theoretical. Of note, each guideline appears to justify their recommendation based upon the others' guidelines and each base the overall clinical strength of the recommendation primarily on the retrospective review by Nowell or the original paper by Heras. Obviously, there are concerns about the circular logic used and marginal concrete evidence in the development of these guidelines.

A more recent meta-analysis by Masri et al. reviewed 14 studies which included 31,740 patients and divided patients into two groups: those who received post-operative anticoagulation and those who did not. Overall, their findings, as summarized in Table 2, were consistent with previous reviews on the topic—essentially there was no observed benefit or protection from thromboembolic complications with the use of anti-coagulation with a VKA, but there was a substantial increased risk of major bleeding related complications and a slight, non-significant increase in 30-day mortality. In the seven of 14 studies that specifically excluded patients in atrial fibrillation, there was no observed difference in thromboembolic complications between those who received anti-coagulation vs those who did not (2.4% vs. 2.6%, OR: 1.28; 95% CI: 0.79-2.06).

Table 2

Outcome (n, %)	Anti-coagulation	No Anti-coagulation	RR, 95% CI
Thromboembolism	145 (1%)	262 (1.5%)	0.96; 0.60-1.52
Major bleeding	292 (2.6%)	189 (1.1%)	2.26; 1.67-3.05
All-cause mortality	351 (3.5%)	415 (3%)	1.32; 0.76-2.30
Re-op for bleeding*	47 (3.3%)	55 (3.2%)	0.81; 0.42-1.58

One topic that has been rarely addressed has been the basic mechanism of TE after bio-SAVR. The proposed mechanism has been, as mentioned above, based upon the raw prosthetic cloth surfaces of the valves themselves that epithelialize over time. The Sundt paper does suggest that many of peri-operative TE events are related to intra-operative events (i.e., lose calcium debris from incomplete debridement of the valvular structures) and questions the overall importance of the cloth surface serving as a nidus for TE.⁶ Nevertheless, if the mechanism of epithelialization is believed, then one would hypothesize that the surface area of cloth would have some relationship to the rate of TE events. Many of the early generation of tissue valves had a much larger relative surface area of cloth material in the sewing cuff than some of the current generation of valves (see pictures of older vs. newer valves). As such, it might be reasonable that older valves (i.e., those used in the Heras study: Hancock, 48%; Carpentier-Edwards, 18%; and Ionescu-Shiley, 34%) might have a greater cloth surface area in some of the more contemporary valves and, hence, potentially explain the difference in TE rates observed in older studies versus newer studies. Rarely are specific valve types (or manufactures) referenced in the studies, meta-analyses or reviews on this topic and it is unclear if a study specifically addressing the early TE risk of specific types of bio-SAVR has ever been performed. Nevertheless, if the amount of cloth surface area of the sewing cuff is implicated as the source of TE material then, as the amount of material used has decreased over time, so should the risk, and so should the strength of the recommendation to anticoagulate. The extension of this concept is seen in the extremely low TE risk observed studies that specifically explore such events in the newer transcatheter valves, which tend to also have relatively minimal cloth surface areas. To further cloud this issue are the recent concerns of subclinical leaflet thrombosis in patients treated with either a TAVR or bio-SAVR. Advanced 4D volume rendering CT scanning has shown that reduced aortic valve leaflet mobility that resolved with anticoagulation. While the clinical significance of this—as such events only in a pooled registry dataset confer a small, but significant, increased risk for neurologic events (1/117 with normal leaflets vs. 3/17 with reduced leaflets, p = 0.007)—remains unclear, such findings illustrate that there is still much to learn regarding the mechanisms of early valvular dysfunction, neurologic events, and potential opportunities for the diagnosis and treatment.⁹

Over the years there has been extensive discussion and research into this area, and as the above-mentioned articles and meta-analysis reviews indicate, the answer to the role of an early period of therapy with a VKA after bio-SAVR remains unclear. Such a practice does not appear to conclusively eliminate, or even reduce, the risk of TE and in fact may increase the risk of bleeding complications. The role of guidelines is to summarize current evidence-based practices into recommendations that can be used to help direct clinical decision-making or provide a framework for shared decision making with the goal of providing direction towards what should hopefully be in the best interests of the patient.

Without a doubt, with the development of catheter-based therapies, the paradigms and options for the management of aortic valve disease has changed dramatically. The changes in how we select patients for a specific therapy have also prompted a greater appreciation of the role of post-intervention decision making and the potential impact on short- and long-term outcomes. Unfortunately, with the rapid development of technologies, sometimes guidelines for management do not always evolve in parallel and at the same rate. As such, the use of anti-coagulation, typically with a VKA, after bio-SAVR for a brief interval of time is one of those guidelines that does not appear to reflect either the years of data and experiences in this area – nor current safe clinical practice. As such, hopefully with growing interest and concerns about the appropriateness of this practice that there will be increased enthusiasm in sponsoring a clinical trial (as currently under development by the CTSNET investigators). If nothing else, perhaps there needs to be at least a critical review of the current data, with emphasis on modern experiences and outcomes, and how it should be integrated into current American (ACC/AHA), European (ESC), and various professional or national society guidelines that address the management of patients with bio-SAVR.

References

1. Nishimura RA, Otto CM, Bonow RO, et al. 2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2017;70:252-89.
2. Baumgartner H, Falk V, Bax JJ, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J 2017.
3. Heras M, Chesebro JH, Fuster V, et al. High risk of thromoemboli early after bioprosthetic cardiac valve replacement. J Am Coll Cardiol 1995;25:1111-9.
4. Brennan JM, Edwards FH, Zhao Y, et al. Early anticoagulation of bioprosthetic aortic valves in older patients: results from the Society of Thoracic Surgeons Adult Cardiac Surgery National Database. J Am Coll Cardiol 2012;60:971-7.
5. Chiang YP, Chikwe J, Moskowitz AJ, Itagaki S, Adams DH, Egorova NN. Survival and long-term outcomes following bioprosthetic vs mechanical aortic valve replacement in patients aged 50 to 69 years. JAMA 2014;312:1323-9.
6. Sundt TM, Zehr KJ, Dearani JA, et al. Is early anticoagulation with warfarin necessary after bioprosthetic aortic valve replacement. J Thorac Cardiovasc Surg 2005;129:1024-31.
7. Nowell J, Wilton E, Markus H, Jahangiri M. Antithrombotic therapy following bioprosthetic aortic valve replacement. Eur J Cardiothorac Surg 2007;31:578-85.
8. Masri A, Gillinov AM, Johnston DM, et al. Anticoagulation versus antiplatelet or no therapy in patients undergoing bioprosthetic valve implantation: a systematic review and meta-analysis. Heart 2017;103:40-8.
9. Makkar RR, Fontana G, Jilaihawi H, et al. Possible subclinical leaflet thrombosis in bioprosthetic aortic valves. N Engl J Med 2015;373:2015-24.

Keywords: Thoracic Surgery, American Heart Association, Aortic Valve, Aortic Valve Stenosis, Anticoagulants, Aspirin, Atrial Fibrillation, Biological Products, Bioprosthesis, Chest Tubes, Blood Platelets, Debridement, Decision Making, Embolism, Endocarditis, Follow-Up Studies, Expert Testimony, Heart Defects, Congenital, Heart Valve Diseases, Heparin, International Normalized Ratio, Prospective Studies, Registries, Research Personnel, Retrospective Studies, Risk, Selection Bias, Research, Stroke, Surgeons, Thromboembolism, Tomography, X-Ray Computed, Transcatheter Aortic Valve Replacement, Thrombosis, Vitamin K, Warfarin

< Back to Listings