Evaluation of Anticoagulation Bridging Strategies in LVAD Patients

Purpose Warfarin is the established anticoagulant for LVAD patients. For patients with a subtherapeutic INR, bridging with low molecular weight heparin (LMWH) or unfractionated heparin (UFH) is common, but safety and efficacy of these strategies have not been established. Methods We performed a retrospective longitudinal cohort study to evaluate safety and efficacy of bridging strategies including all adult LVAD patients implanted at the Medical University of South Carolina from 8/2014 to 7/2017. Patients were followed from LVAD implantation to transplant, death, or study termination. Bridging episodes occurred when a patient received LMWH or UFH; non-bridging episodes were defined as any time period a patient did not receive LMWH or UFH for a subtherapeutic INR. Primary outcomes were major bleeding and thrombotic events evaluated for each bridging or non-bridging episode in an intention to treat analysis. Multivariable Cox regression survival models were used for analysis. Results There were 563 episodes evaluated in 50 patients (182 bridging episodes and 381 non-bridging episodes). Of the 182 bridging episodes, 129 were with LMWH (70.9%) and 53 (29.1%) with UFH. Overall, the median INR (IQR) was 1.8 (1.5-2.1) at bridging and was similar between all bridging episode types. Compared to non-bridging episodes, UFH (HR 3.75, [95%CI 1.45-9.73], P = 0.007) and LMWH (HR 2.25, [1.03-4.94], P = 0.04) were both associated with an increased risk of bleeding. There was a trend toward reduced bleeding risk when comparing LMWH to UFH (HR 0.4, [0.13-1.19], P = 0.1). Among bridging episodes, all 17 thrombotic events occurred in the LMWH group and were primarily pump thrombosis events. Compared to non-bridging events, LMWH was not associated with a decreased risk of clotting events (HR 1.56, [0.28-8.73], P = 0.62). Patients with an INR ≤ 1.5 at the time of bridging had a trend towards increased risk of thrombotic events (HR 2.43, [0.88-6.69], P = 0.09). For patients with an INR ≥ 2.0 at the time of bridging, a significant increase in bleeding events was observed (HR 2.34, [1.06-5.14], P = 0.03). Conclusion Our data suggests that bridging is not necessary for certain patients with a subtherapeutic INR and that the efficacy particularly of LMWH bridging in LVAD patients warrants further investigation. Furthermore, INR at time of bridging may significantly affect subsequent risk of bleeding and clotting.