Share this article on:

Novel Thrombosis Risk Index as Predictor of Left Ventricular Assist Device Thrombosis

Trivedi, Jaimin R.; Sobieski, Michael A.; Schwartz, Samuel; Williams, Matthew L.; Slaughter, Mark S.

doi: 10.1097/MAT.0b013e318291d0c1
Adult Circulatory Support

Left ventricular assist devices (LVADs) are an effective therapy for patients with advanced heart failure, increasing patient survival and quality of life. Left ventricular assist devices are associated with the risks of bleeding and thrombosis. We used lactate dehydrogenase (LDH) and brain natriuretic peptide (BNP) as biomarkers for developing a thrombosis risk index. Data from a single center was retrospectively queried between January 1, 2008, and 10/15, 2011, to identify LVAD patients and related complications. Of 75 consecutive patients implanted with an LVAD, nine had device thrombosis. Analysis of the relationship among international normalized ratio (INR), BNP, LDH, and device thrombosis was performed. Of nine patients with thrombosis, seven had at least 30% of INR readings below 1.5 and two had at least 22% below the therapeutic range from the time of LVAD implant to date of thrombosis. Timeline variability in INR, BNP, and LDH were used to calculate thrombosis risk index (TRI). The TRI retrospectively was able to accurately predict patients with impending device thrombosis. Indexed rise in BNP and LDH in LVAD patients may be useful to identify early device dysfunction and possible future thrombosis. Fibrinogen and D-dimer assays are currently being evaluated in these patients for their possible inclusion and added value to the TRI.

From the Division of Thoracic and Cardiovascular Surgery, University of Louisville, Louisville, Kentucky.

Submitted for consideration January 2013; accepted for publication in revised form February 2013.

Presented as a poster at the 32nd annual meeting of the International Society of Heart and Lung Transplantation, April 2012, Prague, Czech Republic.

Disclosures: The authors have no conflicts of interest to report.

Reprint Requests: Mark S. Slaughter, 201 Abraham Flexner Way, Suite 1200, Louisville, KY 40202. Email:

Despite advances in medical therapy, mortality and morbidity remain high in patients with advanced heart failure.1 An increasing number of end-stage heart failure patients and limited availability of donor organs have increased demand for left ventricular assist device (LVAD) therapy for both bridge to transplant and permanent use.2 Over the years, LVADs have improved patient survival and quality of life in end-stage heart failure patients.3 Newer devices using continuous flow (CF) are smaller and more durable resulting in fewer complications of infection, stroke, device malfunction, and thrombosis compared with previous technology.2,4

Currently, patients on CF LVADs are instructed to take aspirin and warfarin with target international normalized ratio (INR) range of 2.0–3.0,2 and many programs allowing the INR to be as low as 1.5. Although CF LVADs have fewer adverse events than their predecessors, the newer devices are still at risk of thromboembolism, resulting in stroke and/or pump thrombosis.5 Pump thrombus results in device malfunction and subsequent replacement in most cases. Diagnosing pump thrombus early can potentially avert pump thrombosis, thus preventing the need for device replacement. D-dimer and fibrinogen (specific to fibrinolysis) levels increase early after the LVAD implant and return to baseline around 6–12 months and are not routinely screened in LVAD patients.6,7 Brain natriuretic peptide (BNP) and lactate dehydrogenase (LDH) (not specific to fibrinolysis) are routinely tested in patients post-LVAD implant as a measurement of heart failure and an evidence of hemolysis, respectively. We evaluated the interaction of INR, BNP, and LDH levels in patients diagnosed with CF pump thrombus to identify their association with pump thrombosis.

Back to Top | Article Outline


The LVAD database from our center was queried between January 1, 2008, and 10/15, 2011, to identify patients diagnosed with LVAD pump thrombosis. All LVAD pump thrombosis patients were then retrospectively evaluated to identify association of thrombus activity with INR, BNP, and LDH levels during their follow-up care postimplant. The levels of INR, BNP, and LDH in these patients were obtained from the date of LVAD implantation until diagnosis of thrombosis. Thrombosis was diagnosed with imaging studies and clinical performance of the pump. All patients had the thrombosed device explanted and sent to the manufacturer where pump thrombosis was confirmed.

Linear graphical analyses and descriptive statistical tests were used for analyses. Baseline levels of BNP and LDH after the device placement for each individual patient were compared with their levels at the time of diagnosis of thrombosis. Timeline variability of INR, BNP, and LDH levels since device placement till diagnosis of thrombosis was plotted on individual graphs for each patient. The X-axis represented time from implant, Y1-axis represented LDH and BNP levels, and Y2-axis represented INR. The linear correlation between INR values and LDH and BNP changes were observed in all patients individually.

Back to Top | Article Outline


Seventy-five consecutive patients were implanted with CF LVAD (HeartMate II, Thoratec Inc., Pleasanton, CA) and nine (seven males, age range 34–70 years) were diagnosed with pump thrombosis. Of the nine patients, seven had 30% of INR readings below the low therapeutic range of 1.5. The remaining two patients had 22% of the INR readings below the therapeutic range. Table 1 shows average baseline levels of BNP and LDH in individual patients, and the time of the first significantly detected increase in their levels.

Figure 1 shows graphical representation of variability in INR, LDH, and BNP of a typical patient with pump thrombosis. It shows variability in INR readings of a pump thrombosis patient, marking one of the low therapeutic readings as a possible triggering point of thrombosis activity. The figure also shows notable simultaneous increase in baseline LDH and BNP levels corresponding to the subtherapeutic INR level (triggering point). Figure 2 shows graphical representation of the pump hemodynamics as well as power monitoring in a thrombosis patient showing reduced VAD flow, increasing VAD pulsatility index, and power decrease. Figure 3 shows graphical representation of the INR, BNP, and LDH in a nonthrombosis VAD patient.

Using the observations and values of BNP, LDH, and INR triggers, we developed the following formula to predict device thrombus formation:

Based on the proportional theory, we used BNP, LDH, and INR (inversely) to derive the formula above. The constant “k” represents the collective constant of the contributing variables of BNP, LDH, and INR to the thrombosis risk index. The constant “f” is the frequency of INR triggers (below therapeutic values). Using this formula, we identified six of nine patients. Of three missed patients, two had insufficient BNP readings available and one patient had formula optimization issue.

Back to Top | Article Outline


Left ventricular assist device pump thrombosis is not very common but is a significant adverse event resulting in device replacement, the need for thrombolytic therapy, or a potentially fatal outcome.8 Boyle et al.5 established an evidence-based INR target by evaluating the risk of thromboembolism and hemorrhage in CF LVAD patients. They showed low occurrence of stroke and pump thrombosis with INR range above 1.5 in cohort of 331 patients. Although suggested INR target range is 1.5–2.5 in CF LVAD patients, not all laboratory values below the therapeutic INR range will act as a triggering point of thrombosis activity. In our study, all patients (thrombosis and nonthrombosis) had at least one INR reading below therapeutic range of 1.5, but only nine patients had thrombosis. Low therapeutic INR alone is not a good indicator of pump thrombus formation. Pump thrombosis post-LVAD implantation can be better correlated to subtherapeutic INR readings (acting as a triggering point) coupled with variability in the BNP and LDH levels from baseline.

Increase in plasma LDH levels is not specific to thrombosis but characterizes tissue breakdown. Increase in BNP is a measure of worsening heart failure.9 Pump thrombosis leads to reduction in VAD flow, thus causing increase in strain on heart and increase in BNP.10 Sudden simultaneous increase in both LDH and BNP from baseline can be suggestive of ongoing thrombosis activity (Figure 1). Usually a patient has reduced LVAD flow because of pump or outflow cannula obstruction (thrombosis) but not all nine patients in our study had reduced flow or change in LVAD power. Sudden increases in baseline values of LDH and BNP should be followed up further by tests more specific to pump thrombosis activity, such as 2D echo, D-dimer, and fibrinogen measurements.

Studies show active fibrinolysis in CF LVAD patients early after implantation, indicated by increase in D-dimer and fibrinogen levels, which start declining at 3 months and return to baseline at 6–12 months after device placement.6,7 These tests are not as frequently measured in LVAD patients as BNP and LDH. Confirming thrombosis activity using D-dimer and fibrinogen levels early after device placement can be misleading (because of initial protein absorption by device blood contact surface and active fibrinolysis), but their continued increase or new evidence of their increase after returning to baseline should indicate ongoing fibrinolysis activity and warrant further investigations using imaging studies. As previously mentioned, a sudden increase in BNP and LDH from baseline 3 months post-device placement should also trigger D-dimer and fibrinogen measurements to rule out thrombosis activity. Thrombosis activity if diagnosed early can prevent fatal outcomes and reoperation for device explant in these patients. Sudden increase in baseline BNP and LDH levels cannot confirm thrombosis activity but may indicate further investigations based on clinical evidence in the form of imaging studies and thrombosis-specific laboratory tests to rule out thrombosis activity.

This is a retrospective analysis of patients already with diagnosed thrombosis. This study is limited by its retrospective nature, small patient population, and data points. It would be useful to conduct a multicenter study to evaluate BNP and LDH measurements in a larger population of patients who were diagnosed with pump thrombosis using the formula generated above.

Back to Top | Article Outline


Left ventricular assist device thrombus formation results in pump dysfunction and eventual pump thrombosis and pump failure. Earlier diagnosis of pump dysfunction from thrombus formation could lead to treatment that might restore pump function and avoid the morbidity and mortality associated with LVAD failure from thrombosis. Our new thrombosis risk index could be a useful tool to follow patients after implant to guide future therapy and early treatment of pump thrombus formation.

Back to Top | Article Outline


1. Ezekowitz JA, Lee DS, Tu JV, Newman AM, McAlister FA. Comparison of one-year outcome (death and rehospitalization) in hospitalized heart failure patients with left ventricular ejection fraction >50% versus those with ejection fraction <50%. Am J Cardiol. 2008;102:79–83
2. Miller LW, Pagani FD, Russell SD, et al.HeartMate II Clinical Investigators. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
3. Rogers JG, Aaronson KD, Boyle AJ, et al.HeartMate II Investigators. Continuous flow left ventricular assist device improves functional capacity and quality of life of advanced heart failure patients. J Am Coll Cardiol. 2010;55:1826–1834
4. Griffith BP, Kormos RL, Borovetz HS, et al. HeartMate II left ventricular assist system: From concept to first clinical use. Ann Thorac Surg. 2001;71(3 suppl):S116–S120; discussion S114
5. Boyle AJ, Russell SD, Teuteberg JJ, et al. Low thromboembolism and pump thrombosis with the HeartMate II left ventricular assist device: Analysis of outpatient anti-coagulation. J Heart Lung Transplant. 2009;28:881–887
6. Slaughter MS, Sobieski MA, Gallagher C, Graham J, Brandise J, Stein R. Fibrinolytic activation during long-term support with the HeartMate II left ventricular assist device. ASAIO J. 2008;54:115–119
7. John R, Panch S, Hrabe J, et al. Activation of endothelial and coagulation systems in left ventricular assist device recipients. Ann Thorac Surg. 2009;88:1171–1179
8. John R, Kamdar F, Liao K, et al. Low thromboembolic risk for patients with the HeartMate II left ventricular assist device. J Thorac Cardiovasc Surg. 2008;136:1318–1323
9. Jourdain P, Jondeau G, Funck F, et al. Plasma brain natriuretic peptide-guided therapy to improve outcome in heart failure: The STARS-BNP Multicenter Study. J Am Coll Cardiol. 2007;49:1733–1739
10. Sareyyupoglu B, Boilson BA, Durham LA, et al. B-type natriuretic peptide levels and continuous-flow left ventricular assist devices. ASAIO J. 2010;56:527–531

LVAD; left ventricular assist device; thrombosis

Copyright © 2013 by the American Society for Artificial Internal Organs