Direct oral anticoagulants (DOACs) such as apixaban, dabigatran, rivaroxaban, and edoxaban have demonstrated a safe and effective profile in treating non-valvular atrial fibrillation (AF). Currently, they are considered the first-choice treatment for this patient profile. Non-valvular AF is an evolving concept, presently referring to AF in the absence of a mechanical heart prosthesis, rheumatic mitral stenosis, and/or moderate to severe mitral stenosis. DOACs have been widely discussed in previous blog entries, reviews, and recent meta-analyses. American guidelines on valvular diseases recommend warfarin over DOACs during the first three months following a bioprosthetic implant (II-A recommendation). After this period, DOACs may be used as an alternative. However, off-label use of DOACs within these first three months is becoming increasingly frequent, despite limited evidence on the use of these anticoagulants in the initial three months post-cardiac intervention. 

Today’s article aims to generate evidence by evaluating clinical practice in Alberta, Canada. To this end, data from all patients operated on in two hospitals from July 2014 to June 2021 were retrospectively collected. Patients under 18, with mechanical prostheses, transcatheter valves, those who died in the hospital, without discharge data, or missing anticoagulant dispensation information for the first 90 days post-intervention were excluded from the analysis. Data were accessed using Alberta’s surgical registry, linking identifiable data to locate patients in Alberta’s pharmaceutical dispensation database. The primary efficacy outcome was a composite measure of mortality, stroke, transient ischemic attack, and systemic embolism within the first three months post-intervention. The primary safety outcome included intracranial hemorrhage, cardiac tamponade, gastrointestinal bleeding, clinically relevant bleeding at other sites, or a 20 g/L decrease in hemoglobin. Secondary analysis included a detailed comparison of primary outcomes, temporal anticoagulation patterns, and 30-day readmission rates. 

Information was collected on a total of 1,743 patients. Among the 570 patients receiving DOACs, 17 cases (2%) presented an efficacy event and 55 (10%) a safety event. Of the 1,173 patients treated with warfarin, 41 (3%) had an efficacy event and 114 (10%) a safety event. The secondary analysis did not reveal significant differences between the two regimens in terms of safety, efficacy, or 30-day readmissions. 

The authors suggest that the use of DOACs within the first three months after valvular cardiac surgery involving repair or replacement with a bioprosthesis could be as safe and effective as warfarin. However, confirmation of these findings requires adequately powered randomized prospective studies. 

COMMENTARY: 

In the immediate postoperative period following cardiac surgery, we must choose an anticoagulant that does not increase bleeding rates and can be quickly reversed if necessary. Furthermore, it should achieve therapeutic ranges rapidly, as the first three months carry the highest embolism risk. Dicoumarin anticoagulants offer the advantage of extensive experience due to their presence in pharmacopoeias since the 1950s. However, the Achilles’ heel of this medication is its interpatient variability, which meant that only a quarter of the study cohort was within therapeutic anticoagulation ranges. DOACs, on the other hand, exhibit more predictable pharmacokinetics by selectively inhibiting specific coagulation factors. Their effect is immediate and requires no monitoring. The downside of these medications is the lack of validation of anticoagulant effect in this type of patient. 

It is interesting to note that from 2019 to 2020, DOAC prescriptions doubled compared to warfarin, becoming the most commonly used anticoagulation regimen. This shift can be explained by the impact of the coronavirus pandemic, which restricted hospital access and overwhelmed laboratories. The most logical solution was to opt for anticoagulation regimens requiring less frequent monitoring. 

DOACs have limited representation in the immediate postoperative period in major clinical trials. In the RIVER trial (Rivaroxaban in Patients with Atrial Fibrillation and Bioprosthetic Mitral Valve), events were evaluated one year post-intervention. Only 19% of patients underwent surgery within three months prior to inclusion, and the results for this subgroup were not reported. The ENAVLE trial (Efficacy and Safety of Edoxaban in Patients Early After Surgical Bioprosthetic Valve Implantation and Valve Repair), previously discussed in a blog entry, remains the largest study with 220 patients evaluating a DOAC in the immediate postoperative period, but only five patients in the study we analyzed were treated with edoxaban. We await results from the DANCE trial (Direct Oral Anticoagulation vs. Warfarin After Cardiac Surgery) with over 6,000 patients for robust data. 

Finally, we must mention the study’s limitations. Despite being the largest study investigating DOAC use in the context of atrial fibrillation in the immediate postoperative period, it still has the inherent limitations of a retrospective, registry-based study: coding gaps, incomplete data, prescription bias, or the absence of post-discharge clinical data. The study population included both aortic and mitral valve patients, which are distinct profiles with differing morbidity and mortality, requiring different anticoagulation regimens. The study was affected by the pandemic, where adequate patient follow-ups were not conducted. We must interpret these results cautiously, as they do not fully reflect routine practice. 

REFERENCE: 

Moser N, Omar MA, Koshman SL, Lin M, Youngson E, Kent W, et al. Direct oral anticoagulants for atrial fibrillation in early postoperative valve repair or bioprosthetic replacement. J Thorac Cardiovasc Surg. 2024 Aug;168(2):523-532.e3. doi: 10.1016/j.jtcvs.2023.03.004.