Clinical guidelines for a dry land: challenges in developing recommendations for DOAC management in emergent cardiac surgery

Clinical practice guidelines form the cornerstone of optimal patient management and substantially rely on evidence-based medicine, with updated research and rigorous methodology serving as their foundation [1, 2]. In 2024, two major EACTS guidelines on perioperative care in adult cardiac surgery underwent comprehensive updates [3, 4]. These updates also addressed one of the most pivotal advancements in the prevention and management of cardioembolic diseases in modern cardiac surgery: the implementation of direct oral anticoagulants (DOACs) in daily cardiac surgery practice. Since their widespread adoption between 2010 and 2015, DOACs have significantly impacted the field of cardiac surgery, presenting challenges akin to a tsunami [5]. Initially, the cardiac surgery community was unprepared to manage patients receiving this novel therapy. However, as the current guidelines illustrate, the field has gradually adapted to managing surgical patients treated with these anticoagulants, considering their major pharmacological aspects and the availability of calibrated assays to measure drug levels. These monitoring assays have confirmed that in patients with severe renal impairment, residual therapeutic drug concentrations can persist even 48 h after discontinuation, depending on the drug’s classification and half-life. Additionally, critical thresholds for residual drug levels have been defined to ensure safety before elective cardiac surgery [3], although these thresholds require further clinical validation in the context of cardiac surgery. Emergent cardiac surgery, however, remains a significant challenge, especially in patients with (supra)therapeutic plasma levels of the predominantly prescribed anti-Xa DOACs. Reversal options are limited, and evidence in cardiac surgery is scarce in terms of efficacy and especially safety [3, 4]. The inactive recombinant factor Xa, andexanet alfa, binds directly and indirectly to anti-Xa drugs, including DOACs, but its use is associated with increased thrombotic risk and acquired, transient heparin resistance [6]. Consequently, as highlighted in the guidelines, the European Medicines Agency advises against andexanet alfa use in cardiac surgery [7]. An alternative strategy involves restoring haemostatic activity through four-factor prothrombin complex concentrates at doses of up to 50 IU/kg [3, 4, 6]. Data from non-cardiac surgery settings suggest this approach might be effective, and its reversal capacity appears to depend on residual DOAC concentration [6]. Moreover, the safety profile and the rather high dosing strategy in high-thrombotic-risk patients [8], such as those undergoing cardiac surgery, remain undefined [6]. Perioperative haemoadsorption, developed initially to eliminate cytokines in patients with sepsis or systemic inflammatory response syndrome, has been explored as a potential device for removing DOACs during cardiac surgery. However, the elimination of cytokines via haemoadsorption in a landmark study of patients with infective endocarditis who underwent cardiac surgery, along with recent meta-analyses of various clinical settings involving critically ill patients with hyperinflammatory conditions, failed to show clear benefits in terms of the primary end points of postoperative organ dysfunction, inflammatory biomarkers and/or mortality [9–11]. Comprehensive and rapid intraoperative DOAC removal, rather than DOAC reversal, presents an intriguing approach. However, the haemoadsorption use during emergent cardiac surgery for DOAC removal has yielded conflicting results. Data from the START registry revealed that 16% of patients experienced BARC type 4 bleeding, and the variability in DOAC discontinuation interval (44.6 ± 33.1 h) further complicated interpretation, in addition to the intrinsic limitations of observational studies. Additionally, preoperative drug levels were often not measured, leading to uncertainty regarding the plasma concentrations before surgery [3, 12]. An in vitro study using a closed-loop bovine blood model with a flow rate simulating intraoperative practice demonstrated that rivaroxaban and apixaban could be efficiently removed, with a 60-min haemoadsorption treatment achieving a 95% reduction in drug levels [13]. These compelling results await clinical translation, as interventions hypothesized on the basis of in vitro experimental settings have frequently failed to demonstrate efficacy in randomized controlled trials (RCTs) [14, 15]. Importantly, the STAR-T U.S. Pivotal Trial did not demonstrate a reduction in moderate-to-severe postoperative bleeding complications with the intraoperative use of DrugSorb-ATR in patients undergoing cardiothoracic surgery with cardiopulmonary bypass performed within 48 h of the last ticagrelor dose (Puskas et al. presented at 104th AATS Annual Meeting) [16]. Thus, there is a clear need for well-designed RCTs to evaluate the clinical safety and efficacy of haemoadsorption for DOAC elimination during cardiac surgery. Consistent with the available evidence, recently released EACTS guidelines on perioperative medication and patient blood management have refrained from providing specific recommendations but have included this aspect in the ‘gaps in knowledge’ section [1, 2]. While haemoadsorption presents a potentially viable option in this ‘dry and hazardous landscape’ of limited data, its role deserves further sound research. Adhering to a rigorous methodology, including critical evidence assessment, is essential to ensure that guideline recommendations are based on robust clinical evidence, ultimately providing patients with the most effective and safest treatment. Overreliance on observational or preclinical studies can be inappropriate and even misleading in the context of clinical guidelines. While promising preclinical data can generate hypotheses, true clinical efficacy must be demonstrated through well-designed, prospective, controlled studies. Translating these hypotheses into high-quality research with clinically relevant outcomes is essential to ensure that recommendations are credible, with benefits clearly outweighing risks and futility. None declared. This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Conflict of interest: none declared. No data were produced for the content of this paper.