Abstract
Background:
There is a lack of robust data evaluating outcomes of enoxaparin “bridge” therapy in left ventricular assist device (LVAD) patients.
Methods:
We performed a retrospective study of HeartMate II (HM II) and HeartWare HVAD recipients that received therapeutic enoxaparin as “bridge” therapy to describe bleeding and thrombotic events and compare outcomes between devices. The primary endpoint was the incidence of bleeding within 30 days of “bridge” episode. Major bleeding was defined by INTERMACS criteria.
Results:
We evaluated 257 “bridge” episodes in 54 patients, 35 with a HM II device and 19 with an HVAD device that underwent 176 and 81 bridging episodes, respectively. The median INR prior to “bridge” was lower in the HM II group compared to the HVAD group (1.5 vs 1.7, P < .01), however, there was no difference in the median duration of “bridge” therapy (7 vs 7 days, P = .42). There were a total of 30 (12%) bleeding episodes, with the majority in the HM II group vs HVAD (26 [15%] vs 4 [5%], P = .02). We observed 3 (1%) thromboembolic events in 2 (4%) patients with an HVAD device. On multivariate analysis, the presence of a HM II device was associated with a 4-fold increased risk of bleeding.
Conclusion:
We found the use of enoxaparin “bridge” therapy to be associated with a higher incidence of bleeding in patients with a HM II device compared with an HVAD device. Assessment of device- and patient-specific factors should be evaluated to minimize bleeding events.
Introduction
Patients with left ventricular assist devices (LVAD) often receive “bridge” therapy with a parenteral anticoagulant such as intravenous heparin or low molecular weight heparin (LMWH) such as enoxaparin, in the setting of subtherapeutic INR or when temporary therapy interruption is necessary for a procedure. However, the optimal management of subtherapeutic INRs in LVAD patients remains unknown, especially due to varying hemocompatibility and bleeding outcomes among device types. 1 -3 Initial anticoagulation strategies often differ between device types at individual centers, with higher intensity anticoagulation and higher antiplatelet dosing utilized for the HVAD device due to a greater risk of thromboembolic events, specifically ischemic stroke. 4 -7 Therefore, anticoagulation “bridge” strategies with enoxaparin may similarly need to be adjusted based on the device implanted.
Clinical outcomes of enoxaparin “bridge” therapy in LVAD recipients have yielded mixed results. In the largest published experience of LVAD recipients with subtherapeutic INRs, enoxaparin “bridge” therapy was associated with an increased risk of bleeding, without a significant increase in thromboembolic events. 8 However, other centers, with a mix of HMII and HVAD devices, using a similar strategy have not consistently demonstrated an increased bleeding risk with “bridge” therapy. 9 -11 In one small report, clinical outcomes compared between HMII and HVAD found no difference in bleeding. 9 Given the limited data evaluating the safety and efficacy of “bridge” therapy in LVAD patients, we aimed to evaluate our device-specific enoxaparin “bridge” protocol in HM II and HVAD recipients.
Materials and Methods
Study Cohort
We performed a single center, retrospective cohort study to assess the incidence of bleeding and thromboembolic events associated with enoxaparin “bridge” therapy in patients with a durable LVAD. Patients who underwent LVAD implantation with either HM II (Thoratec, Pleasanton, CA, USA) or HVAD (HeartWare, Framingham, MA, USA) at NYU Langone Health (NYULH) between October 2011 and November 2018 were identified from an internal database. We analyzed all patients who received enoxaparin as “bridge” therapy in the setting of a subtherapeutic INR, as defined below, excluding the initial perioperative period. We excluded 1 patient who underwent device exchange from a HM II to an HVAD. This study was approved by the NYULH Institutional Review Board (IRB) and informed consent was waived due to the retrospective design.
Anticoagulation Management
Following implantation of a HM II device, patients are initially prescribed aspirin 81 mg daily and warfarin with a goal INR range of 2-2.5. Patients with an HVAD implant are prescribed aspirin 325 mg daily and warfarin with a goal INR range of 2-3. An enoxaparin “bridge” protocol is used in the setting of a subtherapeutic INR during routine outpatient follow-up or for temporary warfarin therapy interruption in the setting of an elective procedure. All outpatient anticoagulation is managed by advanced practice providers as part of the LVAD team. Enoxaparin “bridge” is initiated at an INR ≤ 1.9. However, if the patient is ≥65 years old and is implanted with a HM II device, “bridge” therapy is initiated at an INR ≤ 1.6. Our center’s protocol recommends therapeutic enoxaparin (1 mg/kg every 12 hours if creatinine clearance is ≥30 mL/min, or 1 mg/kg every 24 hours if clearance is <30 mL/min) and warfarin is dose adjusted to achieve a device-specific therapeutic INR. A repeat INR is checked within 3 days, and enoxaparin is continued if the INR remains subtherapeutic, with a follow-up INR as determined by the outpatient LVAD team. Patients within 3 weeks of LVAD implant or those with a history of pump thrombosis or major thromboembolic event are excluded from the enoxaparin “bridge” protocol and are admitted to the hospital for IV heparin “bridge” therapy, to ensure closer monitoring.
Outcomes
The primary outcome was the incidence of bleeding within 30 days of a “bridge” episode. The 30-day timeframe was selected to report both bleeding and thromboembolic complications, in accordance with recommendations for standardized reporting in patients with arterial indications for chronic oral anticoagulant therapy. 12 Major bleeding, as defined by INTERMACS criteria, includes an episode of suspected internal or external bleeding that results in one or more of the following: death, re-operation, hospitalization, or transfusion of red blood cells. 13 Consistent with prior literature, hemorrhagic stroke was classified as a major bleed event, despite being considered a central nervous system event by INTERMACS criteria. 8 Minor bleeding includes all episodes that do not meet the criteria for major bleeding. Secondary endpoints included the incidence of thromboembolic events including stroke, venous thromboembolism, myocardial infarction, or pump thrombosis within 30 days of “bridge” episode completion. Pump thrombosis was defined by INTERMACS criteria and classified as either suspected or confirmed pump thrombosis. 13
Statistics
All statistical analyses were performed using IBM SPSS Statistics Software for Windows, version 25. Categorical variables were compared using Chi-square test or Fisher’s exact test and reported as frequencies and proportions. Continuous variables were analyzed using Mann-Whitney U test and reported as median with interquartile range (IQR). Statistical analysis was performed by comparing patients with a HM II versus an HVAD. Rate of bleeding was calculated as events/1,000-patient days and was compared between LVAD types using a Poisson generalized linear model to generate a rate ratio. Binary logistic regression was used to estimate the association between “bridge” and patient characteristics and the risk of a bleeding event. Covariates for the multivariate model were selected based on significance on univariate analysis at P-value of <.2. Kaplan-Meier analysis was performed using log-rank test to compare time to bleeding in patients with a HM II versus an HVAD device. A 2-sided P-value of <.05 was considered statistically significant.
Results
Study Cohort
A total of 54 patients received enoxaparin as “bridge” therapy and were included in the analysis. One patient was excluded due to a device exchange. There were 35 patients with a HM II device and 19 patients with an HVAD device. The median age for HM II patients was older as compared to HVAD patients (62 vs 53 years, P = .01), along with higher prevalence of atrial fibrillation (49% vs 5%, P < .01). All other baseline characteristics were similar between the 2 groups (Table 1).
Baseline Demographics.
Abbreviations: BTT: bridge to transplant; DT: destination therapy.
* Age collected on first date of enoxaparin bridge.
Bridge Characteristics
There were a total of 257 “bridge” episodes amongst the 54 patients. The median number of “bridge” episodes per patient was 4 (IQR 2, 7). The time to first enoxaparin “bridge” episode after the implant date was 105 days in the HM II group compared to 42 days in the HVAD group (P < .01). The median INR at the time of “bridge” initiation was 1.5 (IQR 1.4, 1.7) in the HM II group compared to 1.7 (IQR 1.4, 1.8) in the HVAD group (P < .01). The median duration of “bridge” episode in the total cohort was 7 days (IQR 4, 9), and the median time to first INR check was 4 days (IQR 3, 6). The majority of “bridge” episodes (76%) utilized enoxaparin 1 mg/kg every 12 hours. Once daily dosing was utilized infrequently, but was similar between groups. The median dose in once daily regimen was 1.33 mg/kg (IQR 1, 1.49) in the HM II group compared to 0.95 mg/kg (IQR 0.93, 0.99) in the HVAD group (P < .01). Further details of “bridge” characteristics are described in Table 2. Concomitant medications and laboratory parameters at the time of the “bridge” episode are described in Supplemental Table S1.
Bridge Characteristics.
* All values presented as median (IQR), unless otherwise specified.
Bleeding Events
Bleeding within 30 days occurred in 30 (12%) “bridge” episodes among 22 (41%) patients in the total cohort. There were more bleeding episodes in the HM II group compared to the HVAD group (26 [15%] vs 4 [5%], P = .02). The incidence of bleeding per 1,000-patient days was higher in patients in the HM II group compared to the HVAD group (4.48 vs. 1.74 bleeds/1,000-patient days), which resulted in a rate ratio of 2.57 (95% CI 0.90-7.73, P = .08) adjusting for days at risk. There were 17 (57%) major bleeds, the majority of which occurred in the HM II group (14/17 [82%]). There was one major central nervous system bleed in the HM II group. The median INR at the time of bleeding event was 2.5 (IQR 2.1, 3.2) in the HM II group versus 2.1 (IQR 2.0, 2.5) in the HVAD group (P = .62). The median time to a bleed event after “bridge” initiation was 5 days in the HM II group compared to 14 days in the HVAD group (P = .20). In the HM II group, 13 (50%) of bleeding events took place during a “bridge” episode and the other 13 (50%) took place after a “bridge” episode, whereas all 4 bleeding events in the HVAD group took place after a “bridge” episode. The median time to a bleeding event occurring after “bridge” completion was 7 days in both groups. There was a trend toward increased bleeding-free survival, adjusted for the number of days at risk, in the HVAD group (Figure 1). Further information regarding bleeding events is in Table 3.
Time to first bridge-related bleeding event.
Bleeding Events.
Abbreviations: GI: gastrointestinal; CNS: central nervous system; PRBC: packed red blood cells; FFP: fresh frozen plasma; 4F-PCC: four-factor prothrombin complex concentrate; EGD: esophagogastroduodenoscopy.
* All values presented as n (%), unless otherwise specified.
Univariate and multivariate analyses were performed to identify “bridge”- and patient-specific risk factors associated with an increased risk of bleeding (Supplemental Table S2 and Supplemental Table S3, respectively). The presence of a HM II device was associated with an increased risk of bleeding in both univariate (OR 3.97, 95% CI: 1.10-14.38, P = .03) and multivariate analyses (OR 3.98, 95% CI: 1.01-15.63, P = .05). Age, atrial fibrillation, antiplatelet therapy, drug interactions, magnitude of change in INR at the time of first INR check, and duration of “bridge” therapy did not significantly affect the bleeding risk on univariate analysis.
Thromboembolic Events
There were 3 (1%) thromboembolic events among 2 (4%) patients, both with an HVAD device that occurred within 30 days of “bridge” episode. No thromboembolic events occurred in the HM II group. One patient had an ischemic stroke 10 days after discontinuing enoxaparin, at an INR of 2.1. Intervention was deferred due to the time frame of symptom onset and the patient’s clinical improvement. The second patient had suspected pump thrombosis and a large vessel occlusion stroke. The pump thrombosis occurred 4 days after discontinuing enoxaparin, at an INR of 2.9. The pump thrombosis was successfully treated with thrombolysis. Two weeks after the pump thrombosis, the patient was found to have a right MCA stroke, at an INR of 2.4, with recanalization achieved by thrombectomy. Further information regarding thrombotic outcomes is listed in Table 4.
Thromboembolic Events.
Abbreviation: TPA: tissue plasminogen activator.
* All values presented as n (%), unless otherwise specified.
Discussion
Maintaining a therapeutic INR is often challenging in LVAD patients and “bridge” therapy with parenteral anticoagulation is often necessary. Sub- and supratherapeutic INRs in LVAD patients have been associated with thromboembolic and bleeding events, respectively. 14,15 However, the risk for these events differs based on type of LVAD implanted, but it is unknown if this difference extends to periods necessitating “bridge” therapy. Currently, our protocol stratifies the threshold for when to use enoxaparin to “bridge” based on INR, age, and device type, but does not further modify enoxaparin dose intensity based on risk factors for bleeding or thrombosis. We found a nearly 4-fold increase in bleeding risk related to enoxaparin “bridge” therapy in patients implanted with a HM II device, but no thromboembolic events. In contrast, patients implanted with an HVAD device had a significantly lower incidence of bleeding, but 2 patients had severe thromboembolic complications. Despite a decline in HM II implantation with the introduction of the HeartMate 3 (HM 3), these results remain relevant as over 6,000 patients were still supported with HM II devices as of December 2018, highlighting divergent device-specific outcomes with enoxaparin “bridge” therapy. 16
The HM II and HVAD devices differ in design (axial vs. centrifugal flow) and adverse effect profile despite similar overall survival. 1,17 In practice, gastrointestinal bleeding complications and pump thrombosis predominate in HM II recipients, whereas stroke is the most feared complication in HVAD recipients. 1,17 -19 This has led to differing anticoagulation strategies with many centers initially targeting a lower INR goal and prescribing a lower aspirin dose for patients implanted with a HM II device. However, in HVAD recipients, thromboembolic risk has demonstrated to be higher in those with a lower INR or aspirin dose. 4,5,7 Despite these differences, including intensity of anticoagulation, there remains limited data comparing outcomes of anticoagulation “bridge” therapy between device types.
Given the paucity of data evaluating management of a subtherapeutic INR in patients with LVADs, several centers have evaluated their site-specific practices, with varying results in the incidence of bleeding with enoxaparin “bridge” therapy. 8 -11,20 All of these centers reported a mixed population of LVADs implanted. However, only one center compared outcomes based on device type. 9 Sloan, et al compared bleeding episodes between HM II, HVAD and HM 3 recipients and found no difference in the incidence of bleeding events; however the number of bleeding episodes was overall lower than other centers. 9 In our analysis, we observed greater bleeding episodes in patients with a HM II device compared to those with an HVAD device. The predominant site of bleeding was the gastrointestinal (GI) tract and occurred a median of 5 days from the initiation of enoxaparin “bridge” anticoagulation despite a lack of supratherapeutic INR at the time of the bleeding event. We hypothesize that this may be related to the older age, history of bleeding events in HM II recipients, and also due to the time to first bridging episode after LVAD implant date, which was much later in the HM II group compared with the HVAD group (105 vs 42 days, P < .01). Arteriovenous malformations (AVMs) are a common source of GI bleeding in LVAD patients, but have been reported to first present at a median of 67 days after LVAD implant. 19 In addition, we observed a shorter time to a bleeding event after “bridge” initiation in the HM II group compared with the HVAD group (5 vs 14 days, P = .20), however, the clinical significance of this finding is unclear. Both devices have distinct effects on hemolysis and activation of the coagulation system, 21 but significantly higher levels of lactate dehydrogenase (LDH) were observed in HM II compared to HVAD patients, likely due to the greater degree of sheer stress produced by the axial-flow design of a HM II device. The association between higher LDH levels and the observed bleeding rates in the HM II group remains speculative at this point, but may reflect underlying coagulopathy. 21
There are several limitations applicable to our study related to single center and retrospective design. The majority of “bridge” episodes took place in the outpatient setting, limiting monitoring of true adherence to enoxaparin and/or warfarin. However, there were no documented cases of non-adherence. Although we did not assess the impact of warfarin dosing on the observed bleeding events, our outpatient LVAD team uses a standardized anticoagulation protocol to decrease the incidence of variability in the magnitude of INR changes. Furthermore, the first INR measured after bridge initiation and at the time of bleeding event did not differ between groups. Reasons for deviation from standard antithrombotic recommendations were unknown, but were at the discretion of the attending LVAD cardiologist. Additionally, differences in bleeding during versus after “bridge” therapy could suggest that bleeding was related to patient factors or anticoagulation management rather than the device itself. However, patient characteristics were similar at baseline except for age and antiplatelet therapy, but these factors did not correlate with bleeding events on univariate or multivariate analysis. Anticoagulation and bridge management was also similar in terms of dosing, duration, and INR at first check or time of bleeding event. This increases the likelihood that device-related factors may have contributed to the observed differences in bleeding independent of patient factors or antithrombotic management. Lastly, we observed a low incidence of renal insufficiency in our cohort (5%), limiting applicability to this patient population at a greater baseline risk for bleeding.
In conclusion, we found a higher incidence of bleeding events during an enoxaparin “bridge” episode in patients with a HM II device compared with an HVAD device, despite a lower threshold for “bridge” initiation in the HM II cohort. Assessment of device- and patient-specific risk factors for bleeding and thrombosis may be necessary to individualize the enoxaparin “bridge” strategy for the management of subtherapeutic INR in LVAD patients. Larger studies evaluating a reduced intensity of enoxaparin “bridge” therapy, specifically in HM II recipients should be evaluated, to determine if bleeding risk can be mitigated, without an increase in thromboembolic events. This strategy alone, or paired with home INR monitoring to potentially minimize the duration of “bridge” with enoxaparin should be evaluated further.
Supplemental Material
Supplemental Material, sj-docx-1-cpt-10.1177_10742484211006998 - Comparison of Outcomes of Enoxaparin Bridge Therapy in HeartMate II versus HeartWare HVAD Recipients
Supplemental Material, sj-docx-1-cpt-10.1177_10742484211006998 for Comparison of Outcomes of Enoxaparin Bridge Therapy in HeartMate II versus HeartWare HVAD Recipients by Mitulkumar Patel, Tania Ahuja, Serena Arnouk, Claudia Gidea, Alex Reyentovich, Deane E. Smith, Nader Moazami, John Papadopoulos and Tyler C. Lewis in Journal of Cardiovascular Pharmacology and Therapeutics
Footnotes
Author Contributions
MP, TA, SA, JP, and TCL contributed to the study concept and design. MP and TCL were responsible for data acquisition and statistical analysis. MP wrote the initial manuscript draft. TA, SA, CG, AR, DES, NM, JP, and TCL all provided critical review of the manuscript. All authors approved of the final manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
