The evolution of left ventricular assist devices (LVADs) has drastically changed the management of end-stage heart failure and the practice of heart transplantation. Although outcomes with continuous flow LVADs have continued to improve the survival and quality of life in advanced heart failure patients, adverse events remain problematic. Thrombosis is one such life-threatening complication of LVADs that significantly impacts short- and long-term outcomes. Multiple contributing factors of pump thrombosis have been identified, including a combination of patient characteristics,1 compliance with anticoagulation,2,3 hypercoagulability disorders,2,3 as well as technical issues related to implantation.4–6 Advancements in pump technologies, as well as refinements in medical management and surgical techniques, have contributed to reducing the incidence of pump thrombosis.7–9
Pump thrombosis can occur in three different sites of LVADs: the inflow cannula, the pump itself, or, less commonly, the outflow graft.10 Causes of post pump occlusion may include a distal thrombus within the outflow graft11 as well as kinking or external compression of the outflow graft.12 Polytetrafluoroethylene (PTFE) covering of the outflow graft has customarily been used by surgeons to reduce adhesions and facilitate reentry sternotomy at the time of heart transplantation or LVAD exchange (Figure 1). Herein, we report the presentations, diagnosis, and treatment of four cases of outflow graft obstruction because of formation of thrombus between the outflow graft and its protective PTFE cover.
Our first case (patient 1) is a 54-year-old male with end-stage ischemic cardiomyopathy who underwent placement of a Heartware HVAD (Heartware International, Inc. Framingham, MA). A 20 mm PTFE shield graft (Gore Medical, Flagstaff, AZ) was placed around the outflow graft during LVAD implantation. Antithrombotic management was based on oral anticoagulation (coumadin) targeted to international normalized ratio (INR) 2.5–3.5 and aspirin (81 mg/day). He presented to clinic 41 months after implantation with dizziness and blurred vision. Interrogation of his LVAD revealed low-flow alarms and a low pulsatility index (< 1). His laboratory data did not reveal any evidence of hemolysis, and his INR was within therapeutic range of 2.5–3.5 (Table 1).13 Computed tomography angiography (CTA) of the chest showed an eccentric thrombus at the distal portion of the outflow cannula (Figure 2). A Heartware pump exchange procedure was undertaken via reentry sternotomy and left anterior thoracotomy. The outflow graft was resected in its entirety at the time of pump exchange. Direct examination intraoperatively revealed extensive clot formation in the space between the PTFE covering and outflow graft causing impingement of the outflow graft. During pump exchange, we elected to not use PTFE for fear of recurrent complication. Following successful pump exchange, the patient was bridged with heparin infusion to therapeutic coumadin (INR 2.5–3.5) and aspirin (81 mg). One year after pump exchange, the patient is without further complications.
Our next case (patient 2) is a 36-year-old male with end-stage ischemic cardiomyopathy who underwent placement of a HeartMate II (Abbott, Abbott Park, IL). Similar to our first case, an 18 mm PTFE graft was placed around the outflow graft during LVAD implantation. Antithrombotic management was based on oral anticoagulation (coumadin) targeted to INR 2–3 and aspirin (81 mg/day). Twenty-three months after implantation, he presented with dyspnea and dark-colored urine. He was noted to have low flows on interrogation of his LVAD. Lactate dehydrogenase (LDH) and free hemoglobin were elevated consistent with hemolysis, and his INR was within the therapeutic range of 2–3 (Table 1).13 Pump thrombosis was suspected. and CTA of the chest was performed. CTA imaging indicated partial obstruction of the outflow graft beginning at the anastomosis of the outflow graft with the ascending aorta and extending over a length of 5 cm proximally. Close inspection of the images suggested that the thrombus was external to the lumen and appeared to be located between the PTFE covering and outflow graft (Figure 3). To avoid redosternotomy and pump exchange, a percutaneous approach was planned. In the cardiac catheterization laboratory, a single LD Mega 12 × 36 mm stent (Medtronic, Plymouth, MN) was deployed in the distal portion of the outflow graft. Following the stenting procedure, there was minimal improvement in LVAD flows, and LDH levels continued to rise to a maximum of 2,710 units/L. A repeat CTA revealed residual stenosis of the outflow graft proximal to the stent. Five additional overlapping LD Mega 12 mm × 36 mm stents were deployed along the length of the outflow graft, resulting in normalization of LVAD flows. Similarly, LDH levels normalized over the following few days. After a 3-day hospitalization, the patient was discharged home on warfarin therapy (INR 2–3) and aspirin (81 mg). He subsequently was hospitalized 4 months later with recurrent hemolysis despite therapeutic INRs on routine monitoring follow-up visits. Repeat CTA demonstrated patency of the stented outflow graft and stable alignment of the inflow cannula. Thrombosis within the pump impeller at this time was suspected, and he underwent subcostal exchange of the Heartmate II pump while retaining the outflow graft and inflow cannula components. Hemolysis resolved with pump exchange, and thereafter he was bridged with heparin infusion to therapeutic warfarin (INR 2–3) and aspirin (81 mg). His postoperative course was uneventful, and 4 months following pump exchange, he is undergoing close surveillance of anticoagulation and LDH levels.
Patient 3 is a 56-year-old male with ischemic cardiomyopathy who underwent placement of a HeartMate II (Abbott, Abbott Park, IL) as a bridge to transplantation. At the time of implantation, an 18 mm PTFE tube graft was placed around the outflow graft. Antithrombotic management was based on oral anticoagulation (coumadin) targeted to INR 2–3 and aspirin (81 mg/day). He presented to clinic 23 months after implantation with worsening fatigue. Interrogation of his LVAD demonstrated low flows and low pulsatility indices. Laboratory analysis revealed no evidence of hemolysis and an INR within therapeutic range of 2–3 (Table 1).13 Chest CTA indicated partial obstruction of the outflow graft proximal to the aortic anastomosis. Similar to patient 2, the thrombus appeared to be located between the PTFE shield and outflow graft (Figure 4). Intravascular ultrasound (IVUS) was employed during catheterization which confirmed obstruction external to the outflow cannula (Figure 5). The patient then underwent placement of four overlapping LD Mega 12 mm × 36 mm stents resulting in immediate improvement of flows to 5.5 L/min. He was discharged in stable condition after an overnight hospital stay and received a heart transplant 1 month following stenting of the outflow graft.
Patient 4 is a 57-year-old female with end-stage nonischemic cardiomyopathy who underwent placement of an HVAD with a 20 mm PTFE graft covering. Antithrombotic management was based on oral anticoagulation (coumadin) targeted to INR 2.5–3.5 and aspirin (81 mg/day). On a scheduled follow-up visit 29 months after implantation, she was noted to have multiple episodes of suboptimal flows (2.5–3.0 L/min) despite optimization of her blood pressure and volume status. This problem worsened with low-flow levels of 1.8 L/min. She was otherwise asymptomatic, and her INR levels remained in the therapeutic range (2.5–3.5). A chest CTA revealed evidence of thrombus between the PTFE covering and outflow graft. This finding was subsequently confirmed by IVUS during invasive angiography. The patient underwent successful percutaneous stenting of the outflow graft with 7 overlapping LD Mega 12 mm × 36 mm stents with improvement in LVAD flows (2.5–3.0 L/min). She remains asymptomatic and with no recurrent complications after 7 months of follow-up.
In this series, we are reporting a novel mechanism of LVAD outflow graft obstruction that is distinctly different from commonly recognized pump thrombosis. Of 49 total LVADs implanted with PTFE shields at our institution from the years 2014 to 2017, the four cases described in this series were caused by obstructive thrombus between the outflow graft and the PTFE shield. This complication was identified in cases of both Heartware (2/17 patients) and HeartMate II LVADs (2/24 patients) and is not a pump-specific complication but rather a consequence of the surgical practice of covering the outflow graft with a PTFE shield. Specifically in these cases, clot accumulation was identified near the aortic anastomosis with impingement on the outflow graft at the distal portion of the PTFE shield. Patients in our series presented after a median duration of 26 months (range 23–41 months) from LVAD implantation with symptoms of fatigue, poor exercise tolerance, and low LVAD flows. Hemolysis was typically absent; however, there may be concomitant pump thrombosis as was observed in one of our cases (patient 2). In this instance, a high-grade obstruction of the outflow graft presumably resulted in upstream thrombosis within the pump manifested by clinically apparent hemolysis and elevated LDH levels. Unfortunately, this was not recognized initially because LDH levels resolved following outflow graft stenting; however, the patient developed recurrent hemolysis 4 months later requiring pump exchange because of overt pump thrombosis.
LVAD thrombosis is a well-recognized complication with a reported incidence of 0.04–0.09 events per patient year.9,10 Interruption of blood flow can occur at the inflow cannula, the outflow graft, or within the pump itself giving rise to different clinical signs and disturbances to LVAD parameters.10 The most commonly reported site of LVAD thrombosis is clot formation between the impeller and the pump housing.10 Affected patients typically present with overt hemolysis and high LVAD power spikes. In cases of inflow cannula thrombosis and obstruction, hemodynamic instability is the typical presentation, particularly when there is minimal native left ventricular function. In contrast, outflow graft occlusion is the most clinically subtle and can often be asymptomatic.10 Although several cases of outflow graft obstruction have been reported, the mechanisms of obstruction were poorly described (see Table, Supplemental Digital Content 1, http://links.lww.com/ASAIO/A367).11,12,14–21 An observation similar to ours was recently noted at the time of transplant in cases of HeartMate 3 malfunction where thrombotic impingement was identified between the outflow graft and bend relief.14
There is no published data on the frequency of PTFE graft usage during LVAD implantation. Surgeons at our center frequently applied a PTFE graft over the entire length of the LVAD outflow graft to reduce adhesion formation and facilitate sternal reentry when LVAD explant was anticipated. Although this application of PTFE appears to be effective in reducing adhesions and facilitating reoperations, complications secondary to the use of PTFE have not been previously reported. Our first recognition of this unusual cause of outflow graft obstruction was during pump exchange. Despite the initial suspicion for a thrombus within the outflow graft, intraoperative examination of the outflow graft revealed the presence of white thrombus formation between the outflow and PTFE grafts. We deduce that this mechanism of white thrombus formation between the outflow and PTFE grafts is analogous to type IV endoleakage with stent endografting of aortic aneurysms. Seepage of blood components through the endograft material into the aneurysm sac leads to gradual aneurysm enlargement. This is a well-recognized phenomenon in endovascular surgery and is attributed to the inherent porosity of the endograft. Similarly with LVADs, the outflow graft is constructed of a knitted polyethylene terephthalate material that has its own inherent porosity permitting seepage of blood components that are then trapped in between the impermeable PTFE graft, thereby forming an extravascular thrombus. Gradual enlargement of this thrombus eventually leads to outflow graft impingement. Interestingly, these cases all developed in patients with therapeutic INR levels and without underlying hypercoagulable disorders as determined by preimplant hypercoagulable evaluation which included testing for protein C, protein S, lupus anticoagulant, antithrombin 3, G6PD, factor VIII, factor V, factor V by polymerase chain reaction (PCR), and rheumatoid factor. On the other hand, HVAD patients were managed with low dose aspirin. Indeed, our program is using 81 mg in an attempt to reduce the incidence of Gastrointestinal (GI) bleeding at our institution. Nevertheless, we believe that this thrombotic complication can occur despite therapeutic anticoagulation because of the intrinsic thrombogenicity of the extravascular space where stagnation of blood flow occurs and a milieu of thrombotic tissue factors prevail.
Our cases outline important features related to LVAD outflow graft obstruction. First, clinical presentation can be varied and nonspecific. Second, hemolysis appears to be minimal, and low flow on LVAD interrogation appears to be the most consistent clinical presentation. In terms of diagnostic testing, CTA provides excellent information about thrombus location and burden. In our cases, careful examination of the images revealed thrombus between the PTFE grafts, externally compressing the LVAD outflow graft. Since first recognizing this mechanism of external outflow graft obstruction during LVAD pump exchange, subsequent cases were treated successfully with endoluminal stenting. As such, we have adopted endoluminal stenting as first-line therapy, reserving surgical LVAD exchange as a secondary treatment measure. The use of IVUS in these cases was a valuable confirmatory testing modality and was felt to provide more enhanced imaging of the obstruction than was observed with standard angiography of the outflow conduit.
Since many of our LVAD patients have been previously implanted with a PTFE covering, we anticipate that we will continue to encounter cases of outflow graft obstruction from PTFE usage. As such, we have preemptively created a registry of patients implanted with PTFE and have begun performing surveillance CTAs to detect outflow graft obstructions before the onset of symptoms. Furthermore, given the potential for this complication, we have abandoned the use of PTFE grafts during LVAD implantation. Although leaving the LVAD outflow grafts bare is associated with increased adhesion formation, the shift towards intrapericardial pump usage at our institution (Heartware HVAD and Heartmate 3) has permitted the outflow grafts to be routed posteriorly along the diaphragmatic pericardium, thereby reducing the anterior adhesions that are problematic with sternal reentry. Additionally, with increased implementation of the mini-thoracotomy approaches to Heartware HVAD implantation, the anterior pericardium remains intact, also minimizing troublesome anterior adhesions. As such, we no longer appreciate a need to cover the outflow grafts on initial LVAD implantation.
Outflow graft obstruction is a relatively uncommon complication of ventricular assist devices. Our case series highlights the varied and subtle clinical presentation of outflow graft obstruction, the utility of CTA and IVUS in establishing the diagnosis, and the growing role of percutaneous stenting in its management. Most importantly, however, our findings indicate that PTFE covering of the outflow graft can result in extraluminal thrombus formation, and its use during LVAD implantation should be reconsidered. Health care providers should consider this complication when evaluating patients with LVAD pump thrombosis, particularly patients who have been on support for an extended period of time.
1. Consolo F, Sferrazza G, Motolone G, et al. Platelet activation is a preoperative risk factor for the development of thromboembolic complications in patients with continuous-flow left ventricular assist device
. Eur J Heart Fail 2018.20: 792–800.
2. de Biasi AR, Manning KB, Salemi A. Science for surgeons: understanding pump thrombogenesis in continuous-flow left ventricular assist devices. J Thorac Cardiovasc Surg 2015.149: 667–673.
3. 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: 924–930.
4. Taghavi S, Ward C, Jayarajan SN, Gaughan J, Wilson LM, Mangi AA. Surgical technique influences HeartMate II left ventricular assist device thrombosis
. Ann Thorac Surg 2013.96: 1259–1265.
5. Chivukula VK, Beckman JA, Prisco AR, et al. Left ventricular assist device
inflow cannula angle and thrombosis
risk. Circ Heart Fail 2018.11: e004325.
6. Aliseda A, Chivukula VK, Mcgah P, et al. LVAD outflow graft angle and thrombosis
risk. ASAIO J 2017.63: 14–23.
7. Mahr C, Chivukula VK, McGah P, et al. Intermittent aortic valve opening and risk of thrombosis
in ventricular assist device
patients. ASAIO J 2017.63: 425–432.
8. Maltais S, Kilic A, Nathan S, et al.; PREVENT Study Investigators: PREVENtion of HeartMate II Pump thrombosis
through clinical management: the PREVENT multi-center study. J Heart Lung Transplant 2017.36: 1–12.
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. 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.
11. Hubbert L, Forssell C, Baranowski J, et al. Endovascular stenting of a LVAD outflow graft thrombosis
. ASAIO J 2017.63: e3–e5.
12. Grüger T, Kaufmann F, Dreysse S, Falk V, Krabatsch T, Potapov E. Late post-pump blood flow obstruction in a novel left ventricular assist device
: The unusual case of a twisted outflow graft. J Thorac Cardiovasc Surg 2018.155: e33–e35.
13. Shah P, Mehta VM, Cowger JA, Aaronson KD, Pagani FD. Diagnosis of hemolysis and device thrombosis
with lactate dehydrogenase during left ventricular assist device
support. J Heart Lung Transplant 2014.33: 102–104.
14. Duero Posada JG, Moayedi Y, Alhussein M, et al. Outflow graft occlusion of the HeartMate 3 left ventricular assist device
. Circ Heart Fail 2017.10: e1–e3.
15. Pieri M, Scandroglio AM, Kukucka M, et al. Heart failure
after 5 years on LVAD: diagnosis and treatment of outflow graft obstruction
. ASAIO J 2017.63: e1–e2.
16. Hanke JS, Rojas SV, Poyanmehr R, et al. Left ventricular assist device
implantation with outflow graft tunneling through the transverse sinus. Artif Organs 2016.40: 610–612.
17. Abraham J, Remick JD, Caulfield T, et al. Left ventricular assist device
outflow cannula obstruction treated with percutaneous endovascular stenting. Circ Heart Fail 2015.8: 229–230.
18. Pham DT, Kapur NK, Dermody M, Halin N. Stenting of an outflow graft obstruction
after implantation of a continuous-flow, axial-flow left ventricular assist device
. J Thorac Cardiovasc Surg 2015.150: e11–e12.
19. Retzer EM, Tannenbaum SA, Fedson SE, et al. Successful percutaneous trans-catheter treatment of left ventricular assist device
outflow graft stenosis with a covered stent. ESC Heart Fail 2015.2: 100–102.
20. Cheng RK, Aboulhosn J, Nsair A. Percutaneous angioplasty of stenotic outflow graft anastomosis of HeartMate II. JACC Cardiovasc Interv 2014.7: 700–703.
21. Ganapathi AM, Andersen ND, Prastein DJ, et al. Endovascular stent grafting of a left ventricular assist device
outflow graft pseudoaneurysm. Circ Heart Fail 2013.6: e16–e18.