Secondary Logo

Journal Logo

Adult Circulatory Support

Long-Term Continuous-Flow Left Ventricular Assist Device Support After Left Ventricular Outflow Tract Closure

Kurihara, Chitaru*,†,‡; Cohn, William E.*,†; Kawabori, Masashi*,†; Sugiura, Tadahisa*,†; Civitello, Andrew B.*,†; Morgan, Jeffrey A.*,†

Author Information
doi: 10.1097/MAT.0000000000000856


Implantation of long-term continuous-flow left ventricular assist devices (CF-LVADs) for the treatment of advanced, refractory end-stage heart failure has increased substantially over the past decade.1–3 Patients with severe aortic insufficiency (AI) require a concomitant aortic valve (AV) procedure at the time of CF-LVAD implantation to address the AI. Potential surgical options for addressing AI include an AV replacement, central AV closure (Park stitch), partial AV closure, and patch closure of the left ventricular outflow tract (LVOT).4,5 Although there are several reports describing the experience with AV repair or replacement, there is a paucity of data regarding the long-term safety of LVOT closure. We previously reported short-term outcomes in five patients who underwent LVOT closure at the time of CF-LVAD implantation.6 In this current analysis, we examined the long-term outcomes of patients at our center who underwent surgical closure of the LVOT concomitantly with CF-LVAD implantation, and we compared these outcomes with those of patients who did not undergo concomitant LVOT closure during CF-LVAD placement.


This was a single-center retrospective review, with a study cohort consisting of all patients who underwent primary implantation of a HeartMate II CF-LVAD (Thoratec Corp., Pleasanton, CA) or HeartWare ventricular assist device (HVAD; HeartWare Inc., Framingham, MA) between November 2003 and March 2016. Patient data, including demographics, preoperative characteristics, postoperative complications, and outcomes, were collected retrospectively from the Texas Heart Institute/Baylor College of Medicine clinical LVAD database. Approval to conduct this review was obtained from our institutional review board, which waived the consent requirement because the study was retrospective.

Regarding postoperative complications, readmission was defined as return to the hospital within 30 days of discharge from the index admission. Neurologic dysfunction (ND) was defined as a new neurologic deficit associated with abnormal neuroimaging findings and was classified as either ischemic ND or hemorrhagic ND according to these findings. Right heart failure (RHF) was defined according to the International Mechanically Assisted Circulatory Support (INTERMACS) Registry as follows: the need for a right ventricular assist device, an inotrope, or an intravenous or inhaled pulmonary vasodilator (e.g., prostaglandin E1) for more than 1 week at any time after LVAD implantation. Additionally, patients had to meet two of the four following clinical criteria: 1) central venous pressure (CVP) >18 mm Hg or mean right atrial pressure >18 mm Hg; 2) cardiac index <2.3 L/min/m2; 3) ascites or evidence of moderate to worse peripheral edema; and 4) evidence of elevated CVP on echocardiography (dilated superior or inferior vena cava, with collapse) or physical examination (signs of increased jugular venous pressure). Patients were considered to have gastrointestinal bleeding if they had one or more of the following symptoms: guaiac-positive stool, hematemesis, melena, active bleeding at the time of endoscopy or colonoscopy, or blood within the stomach at endoscopy or colonoscopy. Patients were considered to have developed acute kidney injury (AKI) if they met at least the “Injury” criterion of the Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease (RIFLE) end-stage kidney disease classification within 7 days of implantation. Pump thrombosis was defined as visible pump thrombus that necessitated pump exchange. Patients were considered to have an infection if they had either sepsis or a driveline or pump infection that required surgical treatment. Driveline and pump infections were defined as a patient having a positive culture from the skin or tissue surrounding the driveline or pump and clinical evidence of infection, such as pain, fever, drainage, or leukocytosis, coupled with the need for surgical treatment by incision and debridement or by removal or exchange of the device. Sepsis was defined as positive repeat blood cultures. Sepsis was considered non-LVAD related when a noncardiac source for the infection was documented, such as a urinary tract infection or pneumonia.

Patient Data

Patient demographics and preoperative characteristics included sex, body surface area, body mass index, cause of heart failure, previous sternotomy, preoperative inotrope use and short-term mechanical circulatory support, preoperative hemoglobin level, several preoperative laboratory values (i.e., preoperative hemoglobin level, white blood cell count, plasma level of platelets, sodium concentration, creatinine level, blood urea nitrogen level, liver enzyme levels, albumin level, and international normalized ratio), and associated comorbidities, including hypertension, diabetes mellitus, smoking history, and chronic obstructive pulmonary disease. Hemodynamic and echocardiographic data included pre-LVAD CVP, pulmonary artery pressure, pulmonary capillary wedge pressure, left ventricular ejection fraction, cardiac index, left ventricular end-diastolic diameter, and aortic, mitral, and tricuspid regurgitation. Operative characteristics included type of CF-LVAD implanted (HeartMate II or HVAD), cardiopulmonary bypass (CPB) use and time, cross-clamp use and time, and right ventricular assist device use. Outcome variables were postoperative survival at 1 month, 6 months, 1 year, and 2 years.

Statistical Analysis

Statistical analysis was performed with SAS 9.2 software (SAS Institute Inc., Cary, NC). The Kaplan–Meier method was used to estimate survival. To compare the patients who underwent concomitant LVOT closure with those who underwent CF-LVAD placement only, we performed a univariate Cox analysis on all variables listed in Table 1. Those variables with a univariate p <0.20 were included in our multivariate analysis. Cox proportional hazard regression was used to derive hazard ratios and 95% CIs.

Table 1.
Table 1.:
Characteristics of CF-LVAD Patients


Clinical Characteristics

During the study period, 526 patients underwent CF-LVAD implantation at our center. Of these, 16 underwent surgical closure of the LVOT at the time of CF-LVAD implantation: nine with native AV insufficiency, two with previous bioprosthetic AVs, and five with prior mechanical AVs (Table 2).

Table 2.
Table 2.:
Types of Valves and Operative Techniques Used in Continuous-Flow Left Ventricular Assist Device Implantation with Left Ventricular Outflow Tract Closure

Clinical characteristics of the 16 patients who underwent concomitant LVOT closure and the 510 patients who did not are outlined in Table 1. Seven of the 16 LVOT-closure patients received a CF-LVAD as a bridge to transplant, and nine patients received it as destination therapy. Six of the 16 patients used short-term mechanical circulatory support before CF-LVAD implantation; of these, three patients received a TandemHeart, and the other three patients received an intra-aortic balloon pump.

There were no significant differences between the CF-LVAD–only group and the LVOT-closure group in sex, body mass index, body surface area, cause of heart failure, hypertension, diabetes mellitus, smoking history, chronic obstructive pulmonary disease, device type, preoperative left ventricular ejection fraction, left ventricular end-diastolic diameter, cardiac index, pulmonary artery pressure, pulmonary capillary wedge pressure, or right ventricular assist device use. Likewise, the laboratory values did not differ significantly between the groups. The LVOT-closure group had a greater mean age (p = 0.009) and a higher rate of previous cardiac surgery (p < 0.001) than the CF-LVAD–only group (Table 2).

Mean operative time for CF-LVAD insertion and concomitant LVOT closure was 351 ± 127 minutes, mean CPB time was 143.5 ± 76.4 minutes, and mean cross-clamp time was 33.5 ± 35.1 minutes (Table 3). Two LVOT-closure patients required concomitant short-term right ventricular assist device implantation at the time of CF-LVAD implantation. Operative time, CPB use and time, and cross-clamp use and time were greater in the LVOT-closure group than in the CF-LVAD–only group (p < 0.001).

Table 3.
Table 3.:
Operative Characteristics of CF-LVAD Patients

Two of the 16 bridge-to-transplant patients underwent heart transplantation at 326 and 1,039 days of CF-LVAD support. Total support time was 26.1 years.

Postimplantation Survival

For the LVOT-closure patients, survival at 30 days, 6 months, 1 year, and 2 years was 81.3%, 81.3%, 75.0%, and 68.8%, respectively (Figure 1). There were no deaths related to LVOT closure. There were two deaths because of sepsis at 39 and 561 days, one death from hemorrhagic shock at 1,218 days, and one death from noncardiac causes at 298 days. Survival was similar for the LVOT-closure and CF-LVAD–only groups (Figures 2 and 3).

Figure 1.
Figure 1.:
Kaplan–Meier curve showing survival of patients after continuous-flow left ventricular assist device (CF-LVAD) implantation with left ventricular outflow tract (LVOT) closure.
Figure 2.
Figure 2.:
Kaplan–Meier curves showing survival of patients after continuous-flow left ventricular assist device (CF-LVAD) implantation with (red line; n = 16) and without (blue line; n = 510) concomitant closure of the left ventricular outflow tract (LVOT).
Figure 3.
Figure 3.:
Echocardiograms obtained before and after continuous-flow left ventricular assist device (CF-LVAD) implantation. Aortic regurgitation is evident in the preoperative image (A) but was no longer detected after CF-LVAD implantation (B).

Postimplantation Complications and Neurologic Deficit

There were no significant differences between the two groups in the rate of any postimplantation complication except for gastrointestinal bleeding (13 in LVOT-closure patients [0.50 event rate per patient-year {EPPY}], 203 in CF-LVAD–only patients [0.22 EPPY], p = 0.002). In addition, there was no significant difference between the two groups in pump thrombosis: In the LVOT-closure group, two patients had two events (0.08), and in the CF-LVAD–only group, 79 patients had 91 events (0.10) (p = 0.80; Table 4).

Table 4.
Table 4.:
Incidence of Postoperative Complications

A total of 195 ND events (6 in LVOT-closure patients, 189 in CF-LVAD–only patients) occurred in 142 patients (6 LVOT-closure patients, 136 CF-LVAD–only patients). Total support time was 26.1 years in the LVOT-closure group and 938.6 years in the CF-LVAD–only group. The rate of EPPY was 0.23 in the LVOT-closure group and 0.20 in the CF-LVAD–only group (p = 0.97). Three deaths (0.11 EPPY) occurred in the LVOT-closure group, and 47 deaths (0.05 EPPY) occurred in the CF-LVAD–only group (p = 0.20; Supplemental Table 1, Supplemental Digital Content,

Cox Logistic Regression Analysis of Association Between Left Ventricular Outflow Tract Closure and Outcome

The multivariable Cox model showed that surgical closure of the LVOT was not an independent predictor of postoperative survival (hazard ratio = 1.50, 95% CI = 0.87–2.57, p = 0.14). In contrast, preoperative age, previous cardiac surgery, and severe tricuspid valve regurgitation were independent predictors of postoperative survival (Table 5).

Table 5.
Table 5.:
Cox Multivariate Logistic Regression Analysis: Predictors of Postoperative Mortality


We analyzed the records of 16 patients who underwent concomitant LVOT closure along with CF-LVAD implantation. The primary finding was that surgical closure of the LVOT for AV insufficiency in a native, prosthetic, or mechanical valve was not associated with an increase in mortality or postoperative complications. None of the deaths in patients who underwent LVOT closure were related to the LVOT closure.

The method of LVOT closure was tailored to each patient according to the patient’s particular AV anatomy and pathophysiology. For nine patients with severe AI of their native AV, we sutured a circular glutaraldehyde-treated bovine pericardial patch circumferentially to the AV annulus with a running suture. Glutaraldehyde-treated bovine pericardium was used for the patch because it is readily available, easy to handle, and acceptably thromboresistant. The thin, mobile valve leaflets did not interfere with exposure of the annulus, which provided a robust, easy-to-sew circumferential attachment site for the patch. Primary closure of the native valve was not attempted in these patients because we believed that the diaphanous leaflet material would not hold a suture line and, therefore, would not allow robust closure, posing a high risk of failure.

In contrast, in patients whose native AV leaflets were markedly thickened and fibrotic, such as those patients with a congenital bicuspid AV, the strength of the leaflet tissue seemed more than adequate for closure, so a primary closure was performed with a running 4.0 polypropylene suture from commissure to commissure to seal the LVOT.

In patients with bioprosthetic valves, primary closure was not attempted for the same reason as for native valves with fragile leaflet tissue: concern about inadequate tissue strength and high risk for failure. Instead, we closed the LVOT by suturing a circular bovine pericardial patch circumferentially to the AV annulus. For the five patients with mechanical AVs, the sandwich plug technique7 (in which the mechanical valve is “sandwiched” between layers of felt that are then sutured together; Figure 4) was used to close the valve in four of them, and in one patient, the mechanical AV was removed, and the AV annulus was closed with a bovine pericardial patch (Figure 5).

Figure 4.
Figure 4.:
Sandwich plug technique for aortic valve closure. After mechanical valve leaflets are pivoted open, first circular patch and two pledgets are passed through central orifice. Sutures are manipulated to allow patch to drop into left ventricle, then lightly tensed to pull patch up against underside of valve and pivot leaflets closed. From Cohn and Frazier7 used with permission.
Figure 5.
Figure 5.:
A circular patch of bovine pericardium is sutured circumferentially to the aortic annulus above the native aortic valve leaflets. From Cohn et al.6

When AV insufficiency is severe, the CF-LVAD must recirculate the regurgitant volume in addition to maintaining the systemic circulation. This extra workload can result in poor systemic perfusion, inadequate left ventricular unloading, greater power consumption (resulting in a shorter battery life), and possibly accelerated pump wear. Several techniques have been reported for treating or preventing AV insufficiency by definitively closing the LVOT because closing the LVOT can prevent late and recurrent AV insufficiency after LVAD implantation.6,8,9

The optimal method of managing AV insufficiency at the time of device implantation remains controversial. Several procedures intended to treat AV insufficiency have been recommended.9,10 We have been using different methods according to individual patients’ AV anatomy and pathophysiology; these methods include simple closure of the AV, multiple horizontal mattress pledget sutures, bovine pericardial patch, and Teflon sandwich. For example, in patients whose valve leaflets are thin and fragile, using the bovine pericardial patch is preferable; on the other hand, in patients with a thickened and fibrotic AV, primary closure is better. Another possibility is replacement with a bioprosthetic valve. Some suggest that valve replacement is preferable to closure in patients with continuous-flow pumps, for whom retaining some degree of native left ventricular ejection is desirable.11 However, predicting AV opening in patients with adequate left ventricular ejection is difficult, and compared with AV repair, AV replacement requires a longer aortic cross-clamp time, which could cause RHF.

Thromboembolic risks associated with mechanical and bioprosthetic AV insufficiency have been reported; the incidence of prosthetic valve thrombosis has been reported as 0.1% to 0.5%.8,9,12 Thromboembolism can result from stasis or from the intermittent opening of the AV, even in anticoagulated patients. Patients in whom a periprosthetic thrombus develops are at risk for embolic events. If the persistently closed prosthetic valve begins to open intermittently because of ventricular recovery or a change in pump speed or volume status, portions of the thrombus may escape and cause thromboembolic ischemia, resulting in ND. Our data do not show a difference in the rate of ND between CF-LVAD–only patients and patients who underwent LVOT closure along with CF-LVAD placement.

A previous study found that LVOT was strongly related to pump thrombosis and stroke events.13,14 In contrast, our postimplantation rates of thrombotic events and stroke were low in both LVOT-closure and CF-LVAD–only patients. Our sample of LVOT-closure patients was small, so it is possible that the statistical tests for differences between these groups were insufficiently powered. However, we carefully chose the method of LVOT closure, which was tailored to each patient according to the patient’s particular AV anatomy and pathophysiology. This might have reduced this group’s vulnerability to adverse outcomes.2,6,7

As previously mentioned, a disadvantage to LVOT closure is that the patient will not be able to maintain hemodynamic stability if the device fails. Thus far in our experience, however, we have not observed any device failure in these patients. Another disadvantage of LVOT closure is that afterward, bridge to recovery is no longer an option, because a closed LVOT does not allow weaning the patient off CF-LVAD support to assess for native heart recovery.


Long-term outcomes were acceptable in patients who underwent LVOT closure for treatment of AI during their CF-LVAD implantation. These results suggest that LVOT closure is a valid option for select patients with AI who are undergoing CF-LVAD implantation. Longer-term studies are necessary to determine whether aortic root thrombus and subsequent thromboembolic complications eventually become an issue in these patients.


The authors would like to thank Stephen N. Palmer, PhD, Editor in the Life Sciences, of the Section of Scientific Publications at the Texas Heart Institute for editorial support.


1. John R, Kamdar F, Liao K, Colvin-Adams M, Boyle A, Joyce L. Improved survival and decreasing incidence of adverse events with the HeartMate II left ventricular assist device as bridge-to-transplant therapy. Ann Thorac Surg 2008.86: 1227–1234; discussion 1234.
2. Slaughter MS, Pagani FD, Rogers JG, et al. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant 2010.29: S1–S39.
3. Slaughter MS, Rogers JG, Milano CA, et al.; HeartMate II Investigators: Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009.361: 2241–2251.
4. Morgan JA, Brewer RJ. Modified central closure technique for treatment of aortic insufficiency in patients on left ventricular assist device support. ASAIO J 2012.58: 626–628.
5. Park SJ, Liao KK, Segurola R, Madhu KP, Miller LW. Management of aortic insufficiency in patients with left ventricular assist devices: A simple coaptation stitch method (Park’s stitch). J Thorac Cardiovasc Surg 2004.127: 264–266.
6. Cohn WE, Demirozu ZT, Frazier OH. Surgical closure of left ventricular outflow tract after left ventricular assist device implantation in patients with aortic valve pathology. J Heart Lung Transplant 2011.30: 59–63.
7. Cohn WE, Frazier OH. The sandwich plug technique: Simple, effective, and rapid closure of a mechanical aortic valve prosthesis at left ventricular assist device implantation. J Thorac Cardiovasc Surg 2011.142: 455–457.
8. Pelletier MP, Chang CP, Vagelos R, Robbins RC. Alternative approach for use of a left ventricular assist device with a thrombosed prosthetic valve. J Heart Lung Transplant 2002.21: 402–404.
9. Rao V, Slater JP, Edwards NM, Naka Y, Oz MC. Surgical management of valvular disease in patients requiring left ventricular assist device support. Ann Thorac Surg 2001.71: 1448–1453.
10. Samuels LE, Thomas MP, Holmes EC, et al. Insufficiency of the native aortic valve and left ventricular assist system inflow valve after support with an implantable left ventricular assist system: Signs, symptoms, and concerns. J Thorac Cardiovasc Surg 2001.122: 380–381.
11. Bryant AS, Holman WL, Nanda NC, et al. Native aortic valve insufficiency in patients with left ventricular assist devices. Ann Thorac Surg 2006.81: e6–e8.
12. Rose AG, Connelly JH, Park SJ, Frazier OH, Miller LW, Ormaza S. Total left ventricular outflow tract obstruction due to left ventricular assist device-induced sub-aortic thrombosis in 2 patients with aortic valve bioprosthesis. J Heart Lung Transplant 2003.22: 594–599.
13. Dobarro D, Urban M, Booth K, et al. Impact of aortic valve closure on adverse events and outcomes with the HeartWare ventricular assist device. J Heart Lung Transplant 2017.36: 42–49.
14. Robertson JO, Naftel DC, Myers SL, et al. Concomitant aortic valve procedures in patients undergoing implantation of continuous-flow left ventricular assist devices: An INTERMACS database analysis. J Heart Lung Transplant 2015.34: 797–805.

left ventricular assist device; heart failure; aortic valve; aortic insufficiency; left ventricular outflow tract

Copyright © 2018 by the ASAIO