Optimal anticoagulation strategy for left ventricular assist device patients undergoing thoracentesis: a retrospective study

Abstract Background Left ventricular assist devices (LVADs) are associated with thromboembolic risks necessitating anticoagulation, yet balancing these risks with bleeding concerns remains challenging. Current practice involves warfarin with a target INR of 2-3 alongside anti-platelet therapy. However, consensus on anticoagulation management is lacking, particularly during sub-therapeutic INR or procedural settings. Common strategies include bridging anticoagulation with heparin or enoxaparin, temporary cessation of anticoagulation during the procedure or uninterrupted anticoagulation. Purpose To evaluate and compare the incidence of bleeding and thromboembolic events in LVAD patients undergoing thoracentesis, with a focus on determining the optimal anticoagulation strategy during procedures. Methods Retrospective data was collected over one year from LVAD patients undergoing thoracentesis at a large academic center. Patients were categorized into ‘bridge’ (received intravenous heparin for bridging during thoracentesis) and ‘no-bridge’ (did not receive bridging anticoagulation) groups. Warfarin was paused before procedure and resumed after procedure in both groups. Baseline characteristics, including age, sex, race, BMI, LVAD type, and INTERMACS profile, were recorded. Primary outcomes were major bleeding and thromboembolic events within 30 days of warfarin cessation. Statistical analysis used t-test/Wilcoxon for continuous variables and Chi-square/Fisher exact test for categorical variables. A p-value <0.05 was considered significant. Results During the study, twenty LVAD patients underwent thirty-two thoracenteses (nineteen bridge, thirteen no-bridge). The demographics were similar among the two groups. Warfarin was held for an average of 5.5 ± 4.6 days before and resumed 4.1 ± 4.6 days after the procedure in the bridge group and 3.8± 3.2 days before and 1.7 ± 2.5 days after in the no-bridge group. Major bleeding occurred in 73.7% of bridge group cases versus 30.8% of no-bridge group cases (p=0.059). 63.2% of bridge group patients had a bloody pleural effusion on thoracentesis compared to 46.2% of no-bridge group patients (p value= 0.34). Average transfusion requirements were 2.8 units and 2 units in the bridge and the no-bridge groups respectively (p=0.658). Thromboembolic events included a 10.5% ischemic CVA rate in the bridge group, with none in the no-bridge group. No other arterial thromboembolic events, pump thrombosis or DVT/PE were observed in either group. Conclusion In LVAD patients undergoing thoracentesis, a non-bridging strategy does not appear to elevate the risk of thrombosis compared to a bridging approach. Moreover, the incidence of bleeding, while not statistically significant, trended higher in the bridging group.