Continuous-flow left ventricular assist devices (LVADs) are increasingly used as a bridge to heart transplantation and as destination therapy.1 However, bleeding and thrombotic complications are common adverse events associated with LVADs. The continuous-flow physiology creates high shear stress that leads to loss of von Willebrand factor multimers, activation and damage of platelets, altered perfusion of the gastrointestinal (GI) and intestinal mucosa, and formation of arteriovenous malformations that all contribute to an increased risk of bleeding.2 The artificial texture of blood-contact surfaces in the LVAD can activate the coagulation system and increase the thrombotic risk.2
Anticoagulation with a vitamin K antagonist, such as warfarin, is used to mitigate the thrombotic risk; however, anticoagulation management is often a challenge because of the complex patient and device physiology.2 The intensity of warfarin anticoagulation at a given time is traditionally measured by a single international normalized ratio (INR) value. Anticoagulation can also be assessed over a given time period by the time in therapeutic range (TTR), a measure of the percentage of INR values within a goal range.3,4 Deviating above or below a therapeutic range in the LVAD population has been associated with an increased risk of bleeding or thrombotic complications, respectively.5–8 We hypothesize that TTR, rather than a single INR value, may be a more accurate assessment of the quality of warfarin anticoagulation for LVAD patients. The goal of this analysis was to correlate a 30-day TTR with clinical bleeding and thrombotic events.
A chart review was completed of LVAD patients at a large tertiary care center from implant date to September 30, 2014 or the date of mortality, heart transplantation, LVAD explant, or transfer to another institution (whichever occurred first). All patients who received the HeartMate II or HeartWare Ventricular Assist System (HVAD) LVAD after January 1, 2012 and had the LVAD placed for greater than 3 months were included in the study. Exclusion criteria included mortality or transplant during implant admission, anticoagulation monitored entirely by chromogenic factor X levels, or anticoagulation managed by an outside facility. This retrospective cohort analysis received institutional review board approval. The anticoagulation strategy for this institution is described below.
Institution-Specific Anticoagulation Strategy
At this institution, the target INR range was 2 to 2.5 for the HeartMate II and 2 to 3 for the HVAD. Aspirin doses varied from 81 to 325 mg for the HeartMate II, and aspirin 325 mg was used for the HVAD. Before discharge after LVAD implantation, aspirin was started, and a therapeutic INR was achieved. If a bleeding or thrombotic event occurred, an adjustment of the aspirin dose or target INR was considered. Intravenous (IV) heparin or bivalirudin after LVAD implant was initiated at the provider and surgeon’s discretion once a patient was clinically stable and in the absence of coagulopathy. The decision for IV heparin or bivalirudin bridging for a subtherapeutic INR was patient specific and based on factors, such as previous thrombotic or bleeding history. Outpatient anticoagulation bridging was not standard practice.
Anticoagulation was managed by a team of LVAD nurse coordinators, physicians, and clinical pharmacists. Before 2013, anticoagulation management was completed by the LVAD nurse coordinator under the guidance of a provider with no standard dosing guideline. In 2013, an anticoagulation dosing guideline was developed by the clinical pharmacists at the institution. LVAD nurse coordinators utilized the guideline to manage anticoagulation and relied on physician input for any INR greater than 0.5 above or below INR goal, or two consecutive subtherapeutic INRs.
Time in Therapeutic Range Calculation
The TTR was calculated for all outpatient INR values using linear interpolation calculated by the Rosendaal method.3 The Rosendaal method assumes a linear relationship between consecutive INR measurements and calculates an INR for days in between INR checks based on linear interpolation.
The primary outcomes for this study were TTR during the study time period and TTR 30 days before a bleeding or thrombotic event. A bleeding event was defined as a GI bleed confirmed by endoscopy, hemorrhagic stroke confirmed by computed topography (CT), receipt of a transfusion, or any bleeding requiring admission. A thrombotic event was defined as a lactate dehydrogenase (LDH) level greater than three times the upper limit of normal (ULN), ischemic stroke confirmed by CT, transient ischemic attack confirmed by CT, venous thromboembolism confirmed by ultrasound dopplers or ventilation–perfusion scan, LVAD exchange for suspected thrombus, or a change in LVAD pump power precipitating admission. If a second bleeding or thrombotic event occurred within 30 days of the first primary outcome, it was excluded from data analysis. For example, if a patient presented with an LDH greater than three times the ULN and 10 days later underwent an LVAD exchange only the elevated LDH was used for data analysis. In addition, the INR target range, aspirin dose, and number of inpatient hospital days 30 days before a bleeding or thrombotic event were recorded.
Descriptive statistics are displayed as means and SDs for continuous variables, and number and percentage for categorical variables. Where continuous variables had a skewed distribution (time data), data are described as median and interquartile range (IQR). Fisher’s exact or χ2 tests were used to assess the statistical significance of categorical variables between groups and one-way ANOVA or Kruskal–Wallis tests were used for continuous variables. A p value of 0.05 or less was considered statistically significant, and p values are two-sided when possible. All statistical calculations were done in Stata 14.1 (StataCorp LP, College Station, TX).
Baseline patient characteristics before LVAD implantation are shown in Table 1. The study population consisted of 51 patients who underwent LVAD implantation. The mean age at the time of LVAD implant was 57.0 ± 14.6 years old, and 78.4% of patients were male. Thirty patients (58.8%) had ischemic cardiomyopathy. Thirty-five patients (68.7%) had a left ventricular ejection fraction of less than 20%. Sixteen patients (31.4%) had atrial fibrillation, 6 (11.7%) had a history of a deep venous thrombosis or pulmonary embolism, and two patients (3.9%) had a history of a GI bleed.
Device and anticoagulation characteristics are shown in Tables 2 and 3. Forty-seven (92.2%) patients had a HeartMate II device, and 34 (66.7%) patients received their LVAD as destination therapy. The median total LVAD duration was 462 days. There were 72 total bleeding or thrombotic events in 30 patients during the study period: 38 bleeding events in 19 patients and 34 thrombotic events in 17 patients. The overall bleeding or thrombotic event rate was 1.02 events/patient-year, including 0.54 events/patient-year for bleeding events and 0.48 events/patient-year for thrombotic events. The median time from implant to a bleeding event was 169 days, and the median time to a thrombotic event was 259 days. The average time between INR checks during the study period was 8.12 days. The median TTR for all patients during the study period was 52.0%, time above range was 27.4%, and time below range was 18.3%.
Table S1 (Supplemental Digital Content, http://links.lww.com/ASAIO/A110) in the supplemental index outlines the bleeding and thrombotic events by subtype. The majority of thrombotic events were LDH values greater than three times the ULN (23 events), 10 of which ultimately ended in the patient having their LVAD exchanged. Six events were LVAD power spikes that prompted admission. Gastrointestinal bleeding was the most common bleeding event (26 events), and there were four hemorrhagic strokes.
Table 4 describes patient anticoagulation characteristics 30 days before the bleeding or thrombotic events. Compared with thrombotic events, there were more patients on low-dose aspirin (≤81 mg) 30 days before a bleeding event (36.8% vs. 11.8%; p = 0.018). Forty-two patients had a hospital admission in the 30 days before an event with a median length of stay of 12 days. There was no statistically significant difference in the number of admissions or length of stay before bleeding or thrombotic event. The majority of bleeding (28 of 38) or thrombotic (18 of 34) events occurred with a target INR of 2–2.5 before that event.
Table 5 outlines a comparison of TTR, time above therapeutic range, and time below the therapeutic range 30 days before bleeding or thrombotic event, as well as the INR immediately preceding an event. Patients with bleeding events spent a higher percentage of time above range in the 30 days before the event compared with thrombotic patients (41.2% vs. 16.7%; p = 0.007). There was no statistically significant difference in the TTR or time below range (p = 0.22 and p = 0.29, respectively) before bleeding or thrombotic event. Median INR preceding a bleeding event was higher compared with a thrombotic event (2.7 vs. 2.2; p = 0.049).
We identified an overall TTR during the study period and TTR in the 30-day period preceding a bleeding or thrombotic event in our LVAD study population. Our results found an overall TTR of 52% within this patient population, which is similar to other LVAD cohorts with a reported average TTR between 40% and 50%.9,10 There is a demonstrated relationship between TTR and a lower risk of morbidity and mortality for other indications of anticoagulation.11–13 The TTR has been widely used as a quality measure for warfarin use in atrial fibrillation populations, with a median TTR of about 65% in oral anticoagulation trials and with well-controlled treatment proposed as a TTR greater than 70%.14,15 It is difficult to compare anticoagulation effectiveness in atrial fibrillation and LVAD patient populations, but there is a clear difference between this study population and the TTR in atrial fibrillation population. This difference may be explained by the more narrow INR target range of 2–2.5 with the LVAD patient population compared with 2–3 with the atrial fibrillation population. This LVAD population also had an INR checked on average every 8 days, which increases the likelihood of identifying an out of range compared with the longer amount of time between INR checks for the atrial fibrillation population. LVAD patients have a different level of complexity considering comorbidities, such as heart failure and pump physiology. A definition for well-controlled anticoagulation management for LVAD patients has not been established and anticoagulation strategies vary greatly between institutions. However, our results suggest monitoring TTR as a component of LVAD anticoagulation could provide added value above the measurement of single INR values alone. Techniques to improve the TTR in the atrial fibrillation population, such as the use of anticoagulation clinics, pharmacist managed anticoagulation, and patient self-testing may be valuable in the LVAD patient population.16,17
We did not find a significant difference in the 30-day TTR between bleeding and thrombotic events; however, bleeding events were preceded by a higher percentage of time above the therapeutic range. This highlights the well-documented association between periods of supratherapeutic anticoagulation and the risk of bleeding events. In addition, there was a significant difference in a single INR immediately preceding a bleeding or thrombotic event. An INR of 2.7 before bleeding event and an INR of 2.2 before a thrombotic event were close to the most common target INR range of 2 to 2.5 in our study population. We observed a lower aspirin dose 30 days before a bleeding event in our study population. This is likely a reflection of multiple bleeding events occurring in the same patient and a decreased aspirin dose as the first adjustment in anticoagulation to limit future bleeding events.
There are limitations to acknowledge in our study. We performed a retrospective study at a single-LVAD center with a small sample size and short time period. In addition, the study population included two device types, a variation in aspirin dose between patients and variability in the use of IV heparin bridging. Although many of these variabilities are common in clinical practice, it does create heterogeneity in the study population. Institutional anticoagulation strategies and practices changed during the course of the study time period. Specifically, a protocol for warfarin dosing and monitoring developed by the clinical pharmacists may have resulted in minor practice changes by standardizing warfarin management. Our analysis relied on accurate documentation of patient information, and rarely some INR values were not included because of the lack of documentation in the electronic medical record. There are also innate limitations with the Rosendaal Method as a measurement of the TTR, including variability from the target INR range and frequency of INR checks. These variabilities also limit the ability to compare TTR between patient groups.
We conclude an overall TTR of 52% serves as a baseline for our LVAD program and as a comparison for other LVAD programs. The time above range was shown to have a correlation with bleeding events, and this underscores the need to avoid periods of supratherapeutic anticoagulation in the LVAD patient population to prevent bleeding events. Additional studies are needed to establish a definition of well-controlled anticoagulation treatment for the LVAD patient population.
As the LVAD patient population continues to grow, we believe that the TTR will be a valuable indicator of the overall quality of warfarin anticoagulation in LVAD patients, and the relative time above and below the TTR may correlate with adverse events related to anticoagulation. Furthermore, the TTR, time below and time above range can be used to identify specific patients at risk of bleeding or thrombotic events.
1. Trivedi JR, Cheng A, Singh R, Williams ML, Slaughter MS: Survival on the heart transplant waiting list: impact of continuous flow left ventricular assist device as bridge to transplant. Ann Thorac Surg 2014.98: 830834.
2. Eckman PM, John R: Bleeding and thrombosis in patients with continuous-flow ventricular assist devices. Circulation 2012.125: 30383047.
3. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briët E: A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993.69: 236239.
4. Veeger NJ, Piersma-Wichers M, Tijssen JG, Hillege HL, van der Meer J: Individual time within target range in patients treated with vitamin K antagonists: main determinant of quality of anticoagulation and predictor of clinical outcome. A retrospective study of 2300 consecutive patients with venous thromboembolism. Br J Haematol 2005.128: 513519.
5. Uriel N, Nahumi N, Colombo PC, et al.: Advanced heart failure in patients infected with human immunodeficiency virus: is there equal access to care? J Heart Lung Transplant 2014.33: 924930.
6. Najjar SS, Slaughter MS, Pagani FD, et al.; HVAD Bridge to Transplant ADVANCE Trial Investigators: An analysis of pump thrombus events in patients in the HeartWare ADVANCE bridge to transplant and continued access protocol trial. J Heart Lung Transplant 2014.33: 2334.
7. Ravichandran AK, Parker J, Novak E, et al.: Hemolysis in left ventricular assist device: a retrospective analysis of outcomes. J Heart Lung Transplant 2014.33: 4450.
8. Goldstein DJ, John R, Salerno C, et al.: Algorithm for the diagnosis and management of suspected pump thrombus. J Heart Lung Transplant 2013.32: 667670.
9. Bishop MA, Streiff MB, Ensor CR, Tedford RJ, Russell SD, Ross PA: Pharmacist-managed international normalized ratio patient self-testing is associated with increased time in therapeutic range in patients with left ventricular assist devices at an academic medical center. ASAIO J 2014.60: 193198.
10. Jennings D, McDonnell J, Schillig J: Assessment of long-term anticoagulation in patients with a continuous-flow left-ventricular assist device: a pilot study. J Thorac Cardiovasc Surg 2011.142: e1e2.
11. Gallagher AM, Setakis E, Plumb JM, Clemens A, van Staa TP: Risks of stroke and mortality associated with suboptimal anticoagulation in atrial fibrillation patients. Thromb Haemost 2011.106: 968977.
12. Morgan CL, McEwan P, Tukiendorf A, Robinson PA, Clemens A, Plumb JM: Warfarin
treatment in patients with atrial fibrillation: observing outcomes associated with varying levels of INR control. Thromb Res 2009.124: 3741.
13. Wan Y, Heneghan C, Perera R, et al.: Anticoagulation control and prediction of adverse events in patients with atrial fibrillation: a systematic review. Circ Cardiovasc Qual Outcomes 2008.1: 8491.
14. Connolly SJ, Pogue J, Eikelboom J, et al.; ACTIVE W Investigators: Benefit of oral anticoagulant over antiplatelet therapy in atrial fibrillation depends on the quality of international normalized ratio control achieved by centers and countries as measured by time in therapeutic range. Circulation 2008.118: 20292037.
15. Camm AJ, Lip GY, De Caterina R, et al.; ESC Committee for Practice Guidelines (CPG): 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J 2012.33: 27192747.
16. Entezari-Maleki T, Dousti S, Hamishehkar H, Gholami K: A systematic review on comparing 2 common models for management of warfarin
therapy; pharmacist-led service versus usual medical care. J Clin Pharmacol 2016.56: 2438.
17. DeSantis G, Hogan-Schlientz J, Liska G, et al.: STABLE results: warfarin
home monitoring achieves excellent INR control. Am J Manag Care 2014.20: 202209.