Bivalirudin Superiority in Pediatric Berlin Hearts: Too Early to Draw Conclusions?

Abstract

To the Editor: We commend Freniera et al. for their single-center retrospective review involving 31 pediatric Berlin Heart Ventricular Assist Devices anticoagulated with either bivaluridin or unfractionated heparin (UFH).1 We would like to highlight the need for robust performance assessment of the effects of these systematic anticoagulants, particularly during mechanical support. As a measure of inhibition of a single enzyme (activated coagulation factor X), the anti-factor-Xa (anti-Xa) assay represents a more direct measure of UFH activity than the activated partial thromboplastin time (aPTT), with reduced variability and minimal interference from biologic factors, such as lupus anticoagulants or elevated factor VIII.2 As such, the College of American Pathologists defines therapeutic reference ranges as direct measurements of heparin activity, such as anti-Xa. This is important, as divergent assay occurrence rates of up to 40%–60% may exist between groups whose systemic anticoagulant is titrated upon functionally different assays.3 The greater chest tube output and trend toward increased transfusion requirement in those receiving UFH may be explained by more intense anticoagulation while the rates of elevated D-dimer and plasma-free hemoglobin may be reflective of less vigorous anticoagulant effect. Corroborating this postulated causation, Freniere et al. found no differences in time to reach therapeutic aPTT between groups (5.7 vs. 12.6 hours, p = 0.149) despite profound variance when comparing bivalirudin titrated with aPTT and UFH managed with anti-Xa, which is derived from the markedly prolonged time to therapeutic range for UFH (69.5 hours).1 Accordingly, the concordance between aPTT and anti-Xa was found to be moderate (r = 0.63). Additionally, the impact of kidney impairment on bivalirudin pharmacokinetics needs to be accounted for. Although it remains impossible to determine the presence, nor the functional impact, of disparate anticoagulant intensities between groups with the data provided in this study, the lack of elucidation necessitates consideration before drawing definitive conclusions. Although we share the authors passion and enthusiasm for the administration of bivalirudin in the context of mechanical circulatory support due to the ease of dosing, the potential promise of enhanced outcomes,4 and favorable cost profile,5 caution must be urged before these advantages can be translated into the causation of yet to be determined superiority of clinical outcomes. We would also like to caution on the authors introduction of a term so-called VAD-itis seemingly on the basis of retrospective record entry or receipt of immunosuppressants. Although introducing foreign surfaces with mechanical devices triggers the inflammatory system, this “condition” should be identified via actions and judgment, not by any isolated immunological variable. For example, c-reactive protein was not different between groups.1 Considering that biofilm is formed, and the immune system evolves over time, an analysis of simple averages is of little value, even for the highly nonspecific c-reactive protein. While the authors discussion focused on methylprednisone, aspirin is also a potent immunomodulator whose dosing differed significantly between groups. However, to imply that higher aspirin dosing in bivalirudin group is related to more “VAD-it is” is an example of action taken as evidence, while not evidence or causation of any sort.