Left ventricular assist devices (LVADs) are being used widely to support patients with advanced systolic heart failure (HF) as both the durability and compatibility with the human circulation improve. However, these devices are not risk free, and adverse events are encountered such as pump thrombosis that may require exchange of the LVAD.1 , 2 Further, some of these patients will go through more than one thrombosis event in their life span and may even require multiple pump exchange procedures.3
A recent study demonstrated that a novel device is associated with reduced thrombosis.4 In this article, we report three patients who experienced repeated pump thrombosis, of other rotary flow LVADs, in which we performed pump replacement with HeartMate III (Abbott Inc, Chicago, IL). The first case was that of a 42-year-old woman who was implanted with an axial flow device (Heartmate II, Abbott Inc, Chicago, IL) in 2012 for a dilated cardiomyopathy. Because of obesity, she was not considered as a transplant candidate. After 3 years of support, she presented with hemolysis and HF and underwent a subcostal HMII exchange procedure.5 Less than 1 year later, she was admitted to the hospital again due to shortness of breath. Hemolysis was present, and echocardiography demonstrated inability of the HMII pump to decompress the left ventricle at higher speeds, suggesting pump thrombosis (Figure 1A). Hemolysis failed to improve with intensified anticoagulation.
In the second case, a 51-year-old woman with ischemic cardiomyopathy was implanted with the centrifugal flow LVAD, HVAD (Medtronic, Minneapolis, MN). Three months later due to hemolysis and increased power consumption, she was admitted and ultimately required device replacement to a second HVAD. This first replacement was conducted via a left anterior thoracotomy. Unfortunately, less than 1 year later, she was readmitted with hemolysis and increased pump power; she underwent treatment with heparin anticoagulation and thrombolytic therapy. Thrombolytic therapy was associated with an intracranial hemorrhage, but the patient had only a minimal neuro deficit. Unfortunately, she had persistent HF secondary to pump thrombosis.
The third case was a 74-year-old male patient with a history of ischemic cardiomyopathy (ICM) and a previous coronary artery bypass grafting (CABG), who went redo sternotomy for implantation of a HMII LVAD. Regrettably, 3 months later re-required HMII exchange due to pump thrombosis, conducted via a left subcostal approach. Unfortunately, only a month later, he was hospitalized again due to pump thrombosis and hemolysis.
In all three cases, the decision was made to exchange to the HeartMate III device as a compassionate use. Although the surgical experience for pump replacement to the same model/brand is extensive,5 , 6 the surgical experience in replacing an food and drug administration (FDA)-approved device with the HeartMate III is limited.
In all three cases, increased anticoagulation failed to resolve hemolysis and HF. We chose to replace all elements of the previous devices for the HMIII; therefore, redo sternotomy was chosen (patients had previous sternotomies and left anterior thoracotomy or left subcostal incisions). The preoperative evaluation included chest CT scanning with intravenous contrast to help plan the redo-sternotomy and establish that HF symptoms are not due to outflow graft thrombus or outflow graft external compression.
Before redo sternotomy, peripheral cannulation for cardiopulmonary bypass (CPB) is established via right axillary artery and right femoral vein. Dissection is focused on the graft attachment to the ascending aorta and to the LV apical cannulation. The right atrial area does not require extensive dissection. The entire old pump is mobilized. CPB is initiated, and the LVAD is turned down to minimal setting. While on CPB, the entire apical attachment and aortic attachment are mobilized. The LVAD is then turned off and removed from the apical sewing ring. The sewing ring is taken down completely with removal of all pledget or felt material. Importantly, the left ventricle is thoroughly inspected. This is an important step as others have reported clot formation or pannus at the apical cannula, and this could be a source of recurring thrombosis of the new device. The new HMIII sewing cuff can be secured with either interrupted pledgetted sutures, or a running suture, or both (Figure 1B–D). The HMIII pump is then placed into the sewing cuff and secured with the locking mechanism. Care is taken to remove adjacent apical scar tissue, which would interfere with the proper seating of the pump into the sewing ring; this will result in an inability to properly lock the pump in.
Using a side biting aortic clamp, the entire old outflow graft is removed. Removal of the entire outflow graft removes the possibility of obstruction, which could effect the new pump and lead to inadequate pump flow. We have seen cases of both Heartmate II and Heartware outflow graft obstruction. After completion of the outflow graft, the left ventricle and the LVAD pump are deaired, using intraoperative TEE guidance (throughout the procedure, the field was flooded with carbon dioxide). The LVAD pump is initially activated at low pump speeds with care to have the heart filled by allowing volume to enter the circulation from the CPB circuit. The left ventricle and LVAD pump are deaired by placing vents in the ascending aorta and in the outflow graft of the pump. After all air is dispersed as evidenced by the intraoperative TEE, the patient is carefully weaned off a CPB. Once hemodynamically stable, the CPB cannulas are removed, and the wound closure is performed per-usual.
None of the three patients have had any signs for repeat thrombosis; LDH levels are less than 200 u/l (follow-up of up to 1 year).
An alternative strategy, when exchanging the HM II for the HM IIII, is to leave the HM II sewing cuff and place the HM III cannula into this old retained sewing cuff (Figure 2 A, B). This is feasible since the diameter for the HM II and HM III cannulas are the same. The HM III cannula, however, is shorten and does not stand fully into the ventricle when placed through the HM II sewing cuff (Figure 2 A, B). Therefore, we believe that there is an increased risk for pannus overgrowth of the cannula in this position. Also, there is a concern that the cannula cannot be properly secured in the sewing cuff, given the shorten length. An additional strategy would be to cut the HMII cuff and thus shorten its length, but again the safety of this approach is unknown. These patient’s future anti coagulation regimen is strict; patients are bridged with intra-vascular (IV) heparin until the international normalized ratio (INR) values are therapeutic. Aspirin antiplatelet therapy is started immediately after the exchange.
Given the low incidence of pump thrombosis with the HMIII, this device should be considered for replacement in cases of multiple prior pump thromboses of FDA-approved devices, this practice should be validated in future studies of long-term follow-up for the HMIII.4 Complete removal of all old device parts, while more invasive, may represent the best strategy to avoid future device thrombosis.
1. Najjar SS, Slaughter MS, Pagani FD, et alHVAD 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,
2. Starling RC, Moazami N, Silvestry SC, et alUnexpected abrupt increase in left ventricular assist device thrombosis. N Engl J Med 2014.370: 33–40,
3. Kirklin JK, Naftel DC, Pagani FD, et alSeventh INTERMACS annual report: 15,000 patients and counting. J Heart Lung Transplant 2015.34: 1495–1504,
4. Mehra MR, Naka Y, Uriel N, et alMOMENTUM 3 Investigators: A fully magnetically levitated circulatory pump for advanced heart failure. N Engl J Med 2017.376: 440–450,
5. Schechter MA, Patel CB, Blue LJ, et alImproved early survival with a nonsternotomy approach for continuous-flow left ventricular assist device replacement. Ann Thorac Surg 2015.99: 561–566,
6. Milano CASurgical technique for replacement of centrifugal left ventricular assist device through a left thoracotomy. Ann Thorac Surg 2016.101: e55–6,