Search

Background: There is limited data to guide prescribing of all available anticoagulants in obese patients, in particular those weighing > 150kg or with a BMI > 40kg/m2. We aimed to examine anticoagulant prescribing patterns in obese patients in Australia and New Zealand including choice of agent and dosing, determine whether patients with obesity achieve appropriate direct oral anticoagulant (DOAC) levels and evaluate the efficacy and safety of anticoagulation in obese patients. Methods: Patients with a BMI > 35kg/m2 or weight > 120kg receiving anticoagulant therapy are prospectively identified at 7 sites in Australia and New Zealand. Demographic data, indication for anticoagulation and choice of anticoagulant are collected at baseline. At follow-up visits, drug specific levels, dose adjustments and clinical progress (symptoms, imaging, recurrent thrombosis and adverse events) are recorded. Results: 155 patients have been recruited across seven sites (recruitment ongoing). The median age of patients was 53 (range 22-85) and 49% of patients were male. 77 (50%) of patients were anticoagulated for pulmonary embolus (PE) while 40% were anticoagulated for deep vein thrombosis (DVT) without PE. Initial anticoagulant prescribed was apixaban in 37 patients, dabigatran in 4 patients, rivaroxaban in 57 patients, low molecular weight heparin (LMWH) in 29 patients and warfarin in 28 patients (Table 1). A subset of patients initially received LMWH for 5-14 days prior to transition to a DOAC (rivaroxaban in 7, dabigatran in 6). There was cross-over between agents during the study period. Table 2 demonstrates median peak and trough DOAC levels. All peak apixaban levels and 25 of 37 (68%) peak rivaroxaban levels were within the therapeutic ranges published by the International Council for Standardization in Haematology (Gosselin, Thromb Haemost. 2018). Assessment of appropriate trough levels was limited by the limit of detection at different laboratories. 27 patients receiving enoxaparin had at least one peak Anti-Xa level (target peak 0.5-1.2U/ml). The median enoxaparin dose to achieve therapeutic peak Anti-Xa was 0.9mg/kg (range 0.52-1.3). 15 patients weighing > 150kg had a recorded peak Anti-Xa, 10 of these patients were prescribed a capped dose of 150mg BD (weight range 155-290kg) and nine had a therapeutic Anti-Xa and one had a supratherapeutic Anti-Xa measurement. One patient had a PE on apixaban 5mg twice daily (in the setting of suspected non-compliance), one patient had a new confirmed PE after 4 days of non-compliance with rivaroxaban 20mg daily and another patient had a confirmed recurrence on therapeutic warfarin. 1 patient had a PE when warfarin was withheld for emergency surgery. An equivocal recurrence occurred on enoxaparin with a supratherapeutic Anti-Xa (1.76 U.ml) and 1 DVT occurred on dabigatran for a cardiac indication (no drug level available). One patient had recurrent superficial thrombophlebitis on rivaroxaban 20mg daily (peak 301ng/L and trough Conclusion: There is limited data to guide prescribing of all anticoagulants in obese patients, in particular in patients weighing > 150kg or BMI > 40kg/m2. There is increasing use of DOACs in Australia and New Zealand in patients with obesity given the ease of administration and lack of monitoring required. Despite the recommendation to perform DOAC levels in this group of patients, interpretation of levels is difficult due to a lack of well validated therapeutic ranges with clinical correlation. Further data regarding the clinical utility of DOAC levels, appropriate dosing strategy for enoxaparin, and clinical outcomes in this population is required. Our registry provides data regarding drug levels in this population, prescribing characteristics and clinical outcome data. Disclosures No relevant conflicts of interest to declare.

Protein Z dependent protease inhibitor (ZPI) is a human plasma serpin that inhibits factors Xa and XIa in the coagulation pathway. Kinetic studies have shown that ZPI is an efficient inhibitor of factor Xa. This interaction is accelerated by protein Z in the presence of calcium and phospholipid. The inhibition of factor XIa is by an alternative mechanism that does not require phospholipid but is influenced by both heparin and protein Z; heparin increases inhibition whereas protein Z /calcium delays inhibition. The latter maybe physiologically relevant to XIa inhibitory activity as ZPI and protein Z circulate in complex in plasma, with ZPI in slight excess. The kinetics of ZPI/XIa/heparin interactions have not been fully characterised. In this study we report the kinetic analysis of the interaction between factor XIa and recombinant ZPI expressed in Escherichia coli, in the presence of unfractionated heparin, pentasaccharide and protein Z. Results: Recombinant ZPI (rZPI) had characteristics similar to native protein as previously described; serpin-protease complexes were non-covalent, and reactive centre loop proteolytic cleavage of ZPI was noted during this reaction. On SDS gel rZPI had a molecular weight of 50kDa. Circular dichroism of rZPI showed an unfolding transition on thermal denaturation typical of a serpin in native conformation with a Tm of 58.8°C. In contrast loop cleaved rZPI had increased thermal stability as expected for a serpin. A discontinuous method was used to measure the rate of inhibition of target proteases by rZPI; the first order rate constant was measured over time under pseudo first order conditions, and was used to derive the second order rate constant (ka) as shown in Table 1. Table 1: Second order rate constants (ka) Xa (Ms−1) XIa (Ms−1) * And calcium/phospholipid vesicles ZPI 1.9 x 103 9.0 x 104 ZPI + protein Z* 6.6 x 105 1.8 x 104 ZPI + unfractionated heparin (1IU/mL) - 1.0 x 106 ZPI + pentasaccharide (500nM) - 1.0 x 105 ZPI + protein Z* + unfractionated heparin (1IU/mL) - 6.0 x 105 In the presence of calcium and phospholipid vesicles the ka for Xa in the absence and presence of protein Z was similar to previous reports. The ka for ZPI inhibition of XIa increased 11 fold in the presence of unfractionated heparin. Heparin titration curves suggest that a template mechanism is likely; the heparin pentasaccharide had no impact on inhibition rate consistent with this observation. It has previously been reported that protein Z impedes the ZPI inhibition of XIa; our results support this with a 5-fold reduction in ka. However the addition of heparin to ZPI in the presence of Protein Z leads to a significant 33-fold increase in inhibitory activity with a rate similar to the inhibitory activity of ZPI against Xa in the presence of protein Z. Conclusions: There are several physiological inhibitors of factor XIa. In previous reports ZPI inhibition only increased 2-fold in the presence of heparin. Our results confirm that heparin increases the ZPI inhibition of XIa but with a significantly higher increase than previously reported. The increase in activity is even more marked in the presence of protein Z. Although protein Z alone inhibits ZPI activity against XIa the combination with heparin achieves a 33 fold increase in activity. Of all the serpin inhibitors of XIa, antithrombin is considered a major inhibitor in the presence of heparin. Our results show that the ka for FXIa inhibition in the presence of ZPI, protein Z and heparin is more than 20 fold higher than the ka reported for the antithrombin/heparin inhibition of XIa (2-3x104Ms−1). We have previously reported a higher incidence of venous thrombosis in patients with nonsense mutations in the ZPI gene suggesting the clinical significance of this protein. The suggestion that the inhibition of XIa maybe more relevant than that of Xa has been raised in recent murine work where ZPI knockouts have a more severe phenotype than protein Z knockouts. ZPI inhibition of XIa in the presence of endogenous heparins maybe an important contributor to its physiological significance.