Discussion
In this retrospective analysis, we demonstrate that the use of apixaban for prophylactic and therapeutic indications can be safe and effective in patients awaiting heart transplantation. We report no significant preoperative bleeding or thrombotic complications while receiving apixaban therapy. Postoperatively the rate of major bleeding was comparable to that previous reported, given patient complexity and likely unrelated to our anticoagulation strategy. The benefit of DOAC therapy includes reliability, predictability, ease of oral administration and the need for less frequent monitoring. In total, apixaban therapy accounted for 2 178 patient days in our patient population, all of whom would have otherwise received twice daily LMWH injections, or associated rigorous monitoring and dosing parameters of vitamin K antagonists.
Apixaban levels are not routinely recommended for monitoring therapy in adult patients. However, given the lack of apixaban data in pediatrics, and our use of apixaban in a high-risk population, intermittent monitoring may help guide management and reduce the risk of complications from bleeding or thrombosis while receiving apixaban. Most importantly, baseline peak and random apixaban levels allow for assessment of drug clearance prior to invasive procedures, such as heart transplantation. Drug monitoring is also beneficial for identifying causality in situations with significant post-operative bleeding. While our center can receive same day results for apixaban levels, we do realize that is not an option at all institutions. The described prophylactic and therapeutic apixaban levels have been demonstrated to have appropriate efficacy, safety, and low events rates in an analysis 219 pediatric patients from our institution [16], along with a large meta-analysis in adults [17]. It has been demonstrated that chromogenic anti-factor Xa assay calibrated with heparin standards shows a correlation between apixaban levels and LMWH levels [18]. These previously described findings correlate with our experience as well, with goal LMWH of greater than 1.5units/mL suggesting appropriate apixaban anticoagulation.
We acknowledge the seemingly high percentage of significant post-operative bleeding. Unfortunately, there is no standard definition of what should be considered significant postoperative bleeding in patients requiring cardiopulmonary bypass. The limited pediatric literature available is applicable to infants under the age of 1 year, where the author also highlights the discrepancy in definitions used for other studies [19]. Looking at the rates of significant post-operative bleeding, our rate of 32% appears to correlate with what has been described from previously published studies in pediatric patients requiring cardiopulmonary bypass, with incidence ranging from 21.5% to 42% [20-24]. These studies all utilize different values based on chest tube output in the intensive care unit for their definition of significant bleeding. We utilized a more stringent definition of significant bleeding in our study. If we were to utilize the definitions described in the above studies, our rate of significant bleeding would range from 10.5% to 32%. Additionally, these patients have a high prevalence of previous sternotomies, presence of cyanosis, and hemodilution from cardiopulmonary bypass [25].
We do acknowledge that this lack of a formal definition is a limitation to our retrospective analysis given the subjectivity of formulating our own definition. This definition was determined in discussion with a multidisciplinary team including cardiac anesthesiologists, cardiac surgeons, transplant cardiologists and cardiac intensivists at our institution. Further analysis shows that in 3 of the 5 patients defined as having significant post-operative bleeding by chest tube output, anti-Xa heparin levels obtained within the first 18 hours post-operatively were under 0.2units/mL. This further supports the idea that, in those patients, there was little to no appreciable apixaban effect remaining post-operatively, based on available serum testing.
After extensive institutional review, we did not attribute any of the mortalities described to the use of apixaban pre-transplant. As described, one patient, later noted to have pulmonary emboli while on ECMO, received recombinant Factor Xa prophylactically early in our use of apixaban. Given this potential complication we advise against empiric use of recombinant Factor Xa in the absence of life threatening bleeding, as reversal can be a potential trigger of thrombosis. Data from a recent meta-analysis on reflex thrombosis secondary to reversal showed a rate of thromboembolism after utilization of andexanet alfa of 10.7% [26], making this a well-described and not insignificant complication. No other patients required recombinant Factor Xa.
We recognize there are several limitations to our study, including the previously mentioned lack of definition for post-operative bleeding, as well as potential inability to monitor apixaban levels promptly. Lastly, our single center cohort was small, retrospective and descriptive in nature. We hope that our experience can lead to randomized controlled studies comparing apixaban use to the more traditional anticoagulants.
Conflict of Interest: The authors have no disclosures to report.
Funding: Supported by the Grousbeck Fazzalari Fund for Cardiology Research and the Furber Advanced Cardiac Therapies Fellowship Fund
Acknowledgements: None