Pump thrombosis (PT) is a well-established complication after left ventricular assist device (LVAD) implantation associated with increased mortality with a 24% probability of death within 3 months after a PT event occurs.1 This highly morbid event can be seen regardless of the type of continuous flow device support chosen with studies showing a higher rate of PT with axial compared with centrifugal flow devices and completely absent with the fully magnetic levitated centrifugal LVAD.2,3 Current diagnostic and management strategies for LVAD-PT have been developed for both axial and centrifugal flow devices with the latter being the prevalent LVAD technology most implanted today. Although much has been written about the management of LVAD-PT, no current consensus exists on the best single strategy to achieve PT resolution including timing of thrombolytic use, drug delivery method, or duration of therapy. Moreover, little is known about the evaluation and management of PT for right ventricular assist devices (RVAD). The use of an RVAD is commonly reserved as salvage therapy for those patients who present with severe biventricular failure for which limited options exists. More recently, our group and others have described the use of the HVAD specifically in the right atria (RA-HVAD).4–6 In this report, we now describe a case series of six patients with RA-HVAD PT and various treatment strategies employed.
Materials and Methods
Six of 13 patients with history of biventricular support using an HVAD and RA-HVAD developed PT of the RA-HVAD between June 2014 and December 2017 according to the INTERMACS definition of suspect PT.7 All patients were classified as INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support), levels 1–2 and no LVAD-PT was seen or suspected on any of the six patients. All biventricular-supported patients received postoperatively aspirin 325 mg. RA-HVAD–associated PT was suspected when 2 of the 3 criteria were met: 1) presence of hemolysis determined by an elevated lactate dehydrogenase (LDH) >2.0 times the upper limit of normal (ULN), or plasma-free hemoglobin (pfHgb) >40 mg/dl or hemoglobinuria; 2) unexpected power elevation with higher than expected flow estimation or a precipitous drop in both flow and power; or 3) presence of heart failure not explained by changes in structural heart disease. Medical success of thrombus resolution was defined by improvement in biochemical markers of hemolysis, and return to normal pump parameters, without recurrence or need for exchange, urgent United Network of Organ Sharing (UNOS) status 1A transplant, or death within 1 month of thrombus resolution. All PT cases occurred in men with ages ranging from 21 to 58 years, and all patients were designated as bridge-to-transplantation (BTT) therapy. Table 1 summarizes RA-HVAD PT patient characteristics and outcomes. All RA-HVAD pumps were placed with inflow cannula in the right atrium, the tricuspid valve was left intact and none of the outflow grafts were narrowed but had an average length of 15 cm. When an RVAD was decommissioned, it was simply turned off and the driveline was cut at the skin. No attempt at occluding the outflow graft was performed as no evidence of retrograde flow was observed by echocardiogram.
A 55 year old male with nonischemic cardiomyopathy (NICM) underwent HVAD implant with postoperative course complicated by severe RV failure (RVF) requiring RA-HVAD implant. Because of active gastrointestinal (GI) bleeding, his anticoagulation regimen was withheld and resumed 7 days after. Pump power and flows increased precipitously (17.4 W and >10 L/min, respectively) while the patient developed concomitant clinical signs of RVF. His LDH elevated to 1,632 U/L and creatinine to 3.37 mg/dl in the setting of a subtherapeutic international normalized ratio (INR). The patient was placed on heparin without success, therefore, a decision was made to use catheter-directed thrombolysis with tissue plasminogen activator (tPA) into the inflow cannula for 15 mins followed by an additional dose over a period of 40 mins utilizing an ultrasound-assisted catheter delivery system. The latter approach was used because of the ease of delivery of the tPA as experienced by our interventional cardiologists. Pump power decreased immediately after the intervention from 17.4 to 2.6 W with subsequent creatinine and LDH decreasing. However, his course was further complicated by hemopericardium and cardiac tamponade post-tPA infusion requiring surgical evacuation. His anticoagulation was held again leading to a recurrent RA-HVAD PT event which prompted a device exchange.
A 48 year old male with NICM underwent biventricular support at initial VAD implantation. After more than 320 days of durable biventricular support, he was hospitalized with hematochezia and supratherapeutic INR. A Dieulafoy lesion was identified and treated with hemostatic clips. Thirteen days after discharge, he presented with elevated RA-HVAD power and flows with clinical signs of RVF (>10 W and 12.3 L/min, respectively). His LDH was greater than 2500 U/L and his creatinine raised to 3.47 mg/dl in the setting of a therapeutic INR. The patient was placed on heparin initially with thrombolytic therapy being considered but deemed to high risk given his recent major GI bleeding episode. Because RV function had partially recovered as noted on a follow-up echocardiogram, a decision was made to have the RA-HVAD decommissioned. The patient’s LDH, serum creatinine, and total bilirubin decreased soon after turning the RA-HVAD off. No further heart failure episodes occurred and remain asymptomatic on follow-up clinic visits. Six and a half months later, the patient unfortunately was deceased from a large intracranial hemorrhage.
A 44 year old with NICM underwent biventricular support implantation at initial VAD implantation. After more than 620 days of biventricular support with RA-HVAD, the patient was hospitalized with RVF and upper GI bleeding from a Mallory Weiss tear which was treated with hemostatic clips. His anticoagulation was withheld and resumed 5 days after. While resuming his anticoagulation, his hemolysis biomarkers, pump power, and flow started to gradually increase (>10 W and 6 L/min, respectively). His LDH was greater than 2500 U/L and developed significant elevated bilirubin (6.75 mg/dl). The patient was started on bivalirudin with no improvement in pump parameters or biomarkers. A discussion was made to decommission his RA-HVAD as he was deemed to high risk for bleeding with thrombolytic therapy. His LDH and bilirubin decreased soon after device turn off. He was placed on inotropic support for RVF management during the remainder of his hospital stay and was discharged home on oral therapy only without further complications. On follow-up clinic visits, the patient remains with adequate exercise tolerance to perform his routine physical activities.
A 57 year old with NICM underwent biventricular support implantation at initial VAD implantation. Twenty days after implant while preparing for hospital discharge, a peripherally inserted central catheter (PICC) was removed. Immediately after a sudden drop in his RA-HVAD parameters was noted, with power changing from 2.7 to 2.1 W and flow decreasing from 4.6 to 2.5 L/min. No changes in LDH or other markers of hemolysis were noted. No clinical signs of RVF were noted at the time. His pump waveforms at the time were suggestive of occlusion likely from ingested thrombus associated with the recently removed PICC line. The patient underwent catheter-directed thrombolysis with tPA directly into the inflow cannula for 15 mins followed by a second dose infused for 30 mins utilizing the ultrasound-assisted catheter delivery system with prompt normalization of pump parameters. In this particular case, the unique HeartWare waveforms were utilized to diagnose his PT event. The patient had no further PT events and was ultimately transplanted on day 264 of biventricular assist device (BIVAD) support.
A 21 year old male with NICM underwent biventricular support at the time of initial VAD implantation. The patient presented to the hospital with elevated liver enzymes, ascites, and supratherapeutic INR. After initial management for RVF, large spikes in power and flow were noted on the RA-HVAD monitor. His LDH elevated to 1,106 U/L and total bilirubin to 1.53 mg/dl. The patient was treated with bivalirudin with partial improvement in flow parameters. Subsequently, he underwent catheter-directed thrombolysis with tPA for 6 mins. Immediately after power and flow decreased. The following day, power and flow were noted to increase again requiring a second dose to be administered over 20 mins. After the second intervention, all pump parameters normalized without recurrence of PT with power lowering from 4 to 2.7 W and flow decreasing from >10 to 4.7 L/min. His LDH and total bilirubin decreased. The patient remained on bivalirudin for 13 days while transitioning to oral anticoagulation. On day 192 of BIVAD support with no further episodes of PT, he underwent heart transplantation.
A 43 year old male with NICM underwent biventricular support at the time of initial VAD implantation. He presented to emergency room with high power alarms on his RA-HVAD associated with worsened abdominal distention, increased dyspnea on exertion, weight gain fatigue, and dark-tea colored urine. Initial biomarker analysis showed therapeutic INR levels but an elevated LDH greater than 2,500 U/dl, serum creatinine of 7.1 mg/dl, and potassium of 7 mEq/L. His RA-HVAD device interrogation showed several power spikes and high flows 3 days prior to his hospital presentation (peaked at 6 W and 10 L/min, respectively). Bivalirudin was started followed by catheter-directed thrombolysis into the RA-HVAD inflow cannula with normalization of all pump parameters soon after lytic administration. His power decreased from 6 to 2.9 W and flow decreased from 10 to 6 L/min. His hemolysis biomarkers downtrend through his hospital course while remaining on bivalirudin and transitioning him to oral anticoagulation. The patient had no further PT events and after 320 days on BIVAD support, he underwent a successful heart transplantation with no further complications.
The overall clinical response to thrombolytic occurred in three of the six patients with PT with an average time for LDH level to reach close to normal values of 14.5 days (range 5–18 days) and resolution of other hemolysis biomarkers taking place in all patients. The median LDH after PT intervention decreased from 2,066 to 269.5 U/L, median creatinine decreased from 2.57 to 0.965 mg/dl, and total bilirubin from 2.865 to 1.13 mg/dl. In responders to thrombolysis, the duration of thrombolytic therapy was 1–2 days.
We present the first case series of RA-HVAD patients treated for PT describing the consequences of PT and outcomes in six patients with durable biventricular support. We and others have demonstrated the feasibility of providing biventricular support as a BTT strategy in critically ill patients.4,5 As the use of durable BIVADs has increase in recent years as an alternative for biventricular failure, there is a growing need for guidance on the management of device complications of patients on dual mechanical support.
Pump Thrombosis Risk Factors
The etiology for PT in VAD patients is multifaceted with recent literature showing an increase in thrombotic events directly related to pump-specific properties (flow dynamics, pump speed, shear stress, and device hemocompatibility), patient characteristics (comorbidities, hypercoagulability, low flow states from low cardiac output (CO) and stasis), and anticoagulation practices can impact the incidence of PT.8 In our case series, pump-specific factors that could have potentially contribute may have included the pump speed and region of low flow stasis of the right atrium. Patient-related factors could have also contributed because half of our patients’ anticoagulation was withheld because of active GI or other major bleeding. In the ADVANCE (Evaluation of the HeartWare Left Ventricular Assist Device for the Treatment of Advanced Heart Failure) Bridge to transplant and Continued Access Protocol Trial, the risk factors associated with LVAD-PT included mean arterial pressure ≥90 mm Hg, aspirin dose of ≤81 mg, INR ≤2, and INTERMACS profile ≥3.9 In our case series, none of the patients where INTERMACS ≥3 or on low dose aspirin at the time of HVAD implant patient suggesting that such factors did not play a role in RA-HVAD PT development.
Pump Thrombosis Management Strategy
Our institution anticoagulation protocol follows guideline-recommended INR goals for mechanically supported patients. We attempt to maintain INR goals in the 2.5–3 range and check daily LDH levels while patients are hospitalized and biweekly after discharge in our hospital laboratory. All patients receive aspirin 325 mg postoperatively with a second antiplatelet agent added (i.e., dipyridamole), in case of documented aspirin resistance by aspirin sensitivity assay. Initial suspicion of PT is based on a combination of clinical signs of RVF, elevated hemolysis biomarkers with LDH levels (>2.0 times upper limits of normal), and changes in pump parameters. Pump thrombosis detection of the RA-HVAD however can be challenging as the distinctive cyclical waveform variation seen in the LV-HVAD system is absent in the RA-HVAD, thus relying only on pump power and flow variations to determine PT. These power changes have been proven to be useful in detecting PT and determining success to medical therapy especially when the rate of power increase and percent of expected value is low.10 In our study, those patients with a percent peak expected power greater than 200 were less likely to respond to thrombolytics and thus required device exchange or decommissioned their pump.
We initially used heparin as the first intervention to mitigate thrombus burden but given its suboptimal response, we decided to use bivalirudin as an alternative approach. This decision was supported by recent literature demonstrating clinical responsiveness to this agent by reduction in hemolysis markers in patients presenting with LVAD-PT.11 The dose utilized in our cohort ranged from 0.05 to 0.15 mg/kg/hr to achieve a goal activated partial thromboplastin time (aPTT) 60–90. Despite this early intervention, neither of our cases benefited from heparin or bivalirudin alone prompting us to use thrombolytic therapy. We do not consider IIb/IIIa inhibitors before thrombolytics because of their low effectiveness in PT resolution.2 Prior studies have shown the effectiveness in using intravenous thrombolytic infusion to treat HVAD-PT. Success rates of 48–56% are reported, although it has been associated with a significant risk of intracranial hemorrhage in up to 21% of cases.9,12,13 This risk has been mitigated by direct intracavitary lytic administration at the inflow cannula.14 In our cohort, all patients who were treated with catheter-directed thrombolysis into the inlet demonstrated prompt resolution of PT by improvement in pump parameters. The initial dose utilized was 1 mg/min to lessen the risk of systemic bleeding. While the delivery of tPA into the RA-HVAD inflow cannula should, in theory, have less risk for intracranial hemorrhage, we monitored for this possible complication which did not occur. Intravenous anticoagulation is resumed safely when no overt bleeding is seen after tPA administration as a bridge strategy while adding and up titrating oral anticoagulation.
Two important observations can be made with our series. First, those treated successfully with thrombolytic therapy were the ones with PT early on during support (approximately within 6 months of support). Second, the two cases in which the RVAD was decommissioned occurred in patients who have had right-sided support for more than 10 months (312 days in case 2 and 621 days in case 3). Furthermore, both patients’ echocardiographic RV indices of systolic function had improved while on support and this was sustained after stopping the RA-HVAD post-PT. It is plausible that the effects of chronic unloading seen in both of these patients may allowed for partial RV systolic function recovery and subsequent decreased flow through the RA-HVAD. Others have shown that among patients with RVF requiring acute temporary RVAD support, only half can be weaned and are associated with higher mortality, while those with planned BIVAD have lower mortality and improvement in hepatic and renal function biomarkers, possibly indicating an improvement in RV function.15 In both of these cases, device exchange was considered but ultimately not pursued because of recent GI bleeding and the increased mortality risk associated with the exchange. Decommissioning RV support after PT has been previously reported by others utilizing dual HVAD similarly to our report, patients with longer duration on support showed evidence of RV remodeling by normalizing systolic function and RV end-diastolic diameter.16 After the RVAD pump was stopped, no hemodynamic changes were noted and all patients remained on single LV support only.
Our PT experience is comparable with a recent study showing the burden of PT in biventricular-supported patients with INTERMACS 1–2 score. No patient developed LVAD thrombosis at the time of the PT.5 Similarly, these findings have been confirmed in a recent study from the INTERMACS registry where 32 patients received intracorporeal biventricular support and only six had suspected thrombosis with four requiring device exchange (three RV position and one RA position). A trend toward better survival was noted in those with RA cannulation.6
Patients receiving mechanical circulatory support with BIVAD are more critically ill than those receiving an LVAD alone. As such, these patients have a significant risk for adverse events compared with LVAD patients. RA-HVAD PT is a serious but potentially treatable complication of HVAD BIVAD therapy. Our experience suggests that prompt administration of thrombolytics is indicated, especially within the 6 months following BIVAD implantation with log file analysis serving as an indicator of treatment response. Furthermore, we have also demonstrated that RA-HVAD PT is frequently associated with RV recovery. In fact, in three of our patients, the RA-HVAD was decommissioned without an impact on RV hemodynamics and none of the patients exhibited evidence of LVAD-PT events. Hence, at the time of RA-HVAD PT, an assessment of RV function is mandated. The use of log file characteristics, clinical signs, echocardiographic features of RVF, and biomarkers can result in prompt detection of PT and allow use of effective medical therapies with safe and favorable outcomes (Figure 1). Although our single-center retrospective case series is limited, all six patients survived the PT events suggesting that further study is indicated of this therapeutic approach.
1. Smedira NG, Blackstone EH, Ehrlinger J, et al. Current risks of HeartMate II pump thrombosis
: Non-parametric analysis of Interagency Registry for Mechanically Assisted Circulatory Support data. J Heart Lung Transplant 2015.34: 1527–1534.
2. Dang G, Epperla N, Muppidi V, et al. Medical management of pump-related thrombosis in patients with continuous-flow left ventricular assist devices: A systematic review and meta-analysis. ASAIO J 2017.63: 373–385.
3. Uriel N, Colombo PC, Cleveland JC, et al. Hemocompatibility-related outcomes in the MOMENTUM 3 Trial at 6 months: A randomized controlled study of a fully magnetically levitated pump in advanced heart failure. Circulation 2017.135: 2003–2012.
4. Tran HA, Pollema TL, Silva Enciso J, et al. Durable biventricular support using right atrial placement of the HeartWare HVAD. ASAIO J 2018.64: 323–327.
5. Shehab S, Macdonald PS, Keogh AM, et al. Long-term biventricular HeartWare ventricular assist device
support–Case series of right atrial and right ventricular implantation outcomes. J Heart Lung Transplant 2016.35: 466–473.
6. Arabía FA, Milano CA, Mahr C, et al. Biventricular support with intracorporeal, continuous flow, centrifugal ventricular assist devices. Ann Thorac Surg 2018.105: 548–555.
7. Kirklin JK, Pagani FD, Kormos RL, et al. Eighth annual INTERMACS report: Special focus on framing the impact of adverse events. J Heart Lung Transplant 2017.36: 1080–1086.
8. Shah P, Tantry US, Bliden KP, Gurbel PA. Bleeding and thrombosis associated with ventricular assist device
therapy. J Heart Lung Transplant 2017.36: 1164–1173.
9. 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: 23–34.
10. Jorde UP, Aaronson KD, Najjar SS, et al. Identification and management of pump thrombus in the HeartWare left ventricular assist device
system: A novel approach using log file analysis. JACC Heart Fail 2015.3: 849–856.
11. Sylvia LM, Ordway L, Pham DT, DeNofrio D, Kiernan M. Bivalirudin for treatment of LVAD thrombosis: A case series. ASAIO J 2014.60: 744–747.
12. Stulak JM, Dunlay SM, Sharma S, et al. Treatment of device thrombus in the HeartWare HVAD: Success and outcomes depend significantly on the initial treatment strategy. J Heart Lung Transplant 2015.34: 1535–1541.
13. Scandroglio AM, Kaufmann F, Pieri M, et al. Diagnosis and treatment algorithm for blood flow obstructions in patients with left ventricular assist device
. J Am Coll Cardiol 2016.67: 2758–2768.
14. Kiernan MS, Pham DT, DeNofrio D, Kapur NK. Management of HeartWare left ventricular assist device
thrombosis using intracavitary thrombolytics. J Thorac Cardiovasc Surg 2011.142: 712–714.
15. Takeda K, Naka Y, Yang JA, et al. Outcome of unplanned right ventricular assist device
support for severe right heart failure after implantable left ventricular assist device
insertion. J Heart Lung Transplant 2014.33: 141–148.
16. Potapov E, Schweiger M, Vierecke J, et al. Discontinuation of HeartWare RVAD support without device removal in chronic BIVAD patients. ASAIO J 2012.58: 15–18.