Aortic valve closure associated with HeartMate left ventricular device support: Technical considerations and long-term results

http://www.jhltonline.org

Aortic valve closure associated with HeartMate left ventricular device support: Technical considerations and long-term results Robert M. Adamson, MD,a Walter P. Dembitsky, MD,a Sam Baradarian, MD,a Joseph Chammas, MD,a Karen May-Newman, PhD,b Suzanne Chillcott, RN,c Marcia Stahovich, RN,c Vicki McCalmont, NP,c Kristi Ortiz, NP,c Peter Hoagland, MD,d and Brian Jaski, MDd From the aDepartment of Cardiovascular Surgery, Sharp Memorial Hospital, San Diego; bDepartment of Bioengineering, San Diego State University; and Departments of cNursing and dCardiology, Sharp Memorial Hospital, San Diego, California.

KEYWORDS: left ventricular assist device; HeartMate; aortic valve closure; LVAD

BACKGROUND: Aortic valve integrity is crucial for optimal left ventricular assist device (LVAD) support. Pre-existing native aortic insufficiency, aortic valve incompetence acquired during support, as well as previously placed prosthetic aortic valves present unique problems for these patients. METHODS: We reviewed and analyzed data for 28 patients who underwent left ventricular outflow tract closure associated with HeartMate I (n ⫽12) and HeartMate II (n ⫽ 16) LVAD insertion or exchange. Indications for valve closure, surgical technique, LVAD function, survival rates and complications were retrospectively analyzed. Survival rates were compared with those of HeartMate LVAD patients (n ⫽ 104) who did not undergo aortic valve closure. RESULTS: Indications for closure included native aortic valve insufficiency (10 patients), aortic valve deterioration after prolonged LVAD support (8 patients) and previously placed mechanical (9 patients) or bioprosthetic aortic prostheses (1 patient). There were 2 operative and 5 late deaths (mean 227 days post-operatively). Of the deaths, none were due to aortic valve closure. Actuarial survival was 78% at 1 year and 53% at 3 years, which was statistically better than for our patients with an intact aortic outflow (61% at 1 year, 45% at 3 years; p ⬍ 0.05). Five patients had transplants, 1 patient was successfully bridged to recovery, and 15 patients remain on LVAD support. No patient with outflow closure developed regurgitation, embolization or compromised LVAD support. CONCLUSION: Outflow tract closure in LVAD-supported patients is safe, often necessary and well tolerated. J Heart Lung Transplant 2011;30:576–82 © 2011 International Society for Heart and Lung Transplantation. All rights reserved.

Clinical experience with left ventricular device (LVAD) support as a bridge to heart transplantation and for destination therapy has demonstrated survival and quality-of-life benefits for patients in advanced stages of heart failure.1–3 However, persistent adverse events, including valve dysfunction in the HeartMate I device, have limited the effectiveness of this therapy.4,5 Native valve dysfunction can also adversely affect the ability of an LVAD to adequately supReprint requests: Robert M. Adamson, MD, Department of Cardiovascular Surgery, Sharp Memorial Hospital, 7910 Frost Street, Suite 330, San Diego, CA 92123. Telephone: 858-300-4747. Fax: 858-300-4740. E-mail address: [email protected]

port the circulation.6 Aortic insufficiency creates central recirculation, and inefficient LVAD support. Therefore, the assessment and correction of native valve dysfunction at the time of LVAD implant is important for patients’ long-term survival. Both native and prosthetic aortic valves can be problematic for patients undergoing LVAD implantation. Native aortic valve insufficiency compromises LVAD support because regurgitation diminishes systemic perfusion, elevates left heart filling pressures and increases LVAD flow, which may result in congestive heart failure or accelerated mechanical deterioration of the HeartMate I LVAD. Patients

1053-2498/$ -see front matter © 2011 International Society for Heart and Lung Transplantation. All rights reserved. doi:10.1016/j.healun.2010.11.007

Adamson et al.

Aortic Valve Closure with LVAD Support

577

who have an existing prosthetic valve and require LVAD support also present a management dilemma because of their increased risk of thromboembolism during periods when series flow conditions are episodically replaced with parallel conditions through the native outflow tract.7 Historically, uncorrected moderate or greater aortic insufficiency and/or a pre-existing prosthetic aortic valve were considered contraindications to LVAD support. However, successful aortic valve closure at the time of LVAD implant has been reported without untoward effects.8,9 Also, patients who develop aortic insufficiency and concurrent LVAD failure can have the aortic outlet closed safely during LVAD exchange, with a reasonable long-term outcome.10 In this report, we present our single-center experience with aortic valve closure and LVAD insertion or replacement in patients with incompetent native or prosthetic aortic valves.

Methods Patients We retrospectively reviewed 132 consecutive patients supported between February 1991 and February 2009 with HeartMate I (n ⫽ 92) and HeartMate II (n ⫽ 40) LVADs (both Thoratec Corp., Pleasanton, CA) at our hospital. Of the 132 patients, 28 (21%) had their aortic valves closed at the time of LVAD implant or exchange. For these patients we analyzed indications for valve closure, surgical closure technique and clinical function of the LV– LVAD complex.11 To assess the possible adverse effects of valve closure we analyzed survival rates and adverse events, especially thromboembolism. Survival rates of this group were compared with the rest of our LVAD patients (n ⫽ 104), who did not undergo aortic valve closure. Patients were censored for transplant or death. Survival was calculated from the date of first LVAD implant or date of aortic closure. The study was approved by the institutional review board of our institution, and all patients gave informed consent to participate in the study.

Indications for aortic valve closure All patients with a mechanical prosthesis in the aortic position underwent valve closure using a variety of techniques detailed in what follows. Closure of the regurgitant native aortic valve has evolved over time. Currently, patients with transesophageal echocardiography (TEE) demonstrating greater than trace insufficiency once cardiopulmonary bypass is initiated (coincident with arterial waveform pulsatility on full bypass), patients in whom significant aortic insufficiency is seen through the apical coring site during LVAD implant, and patients with TEE-documented aortic insufficiency immediately after LVAD transition (after exiting cardiopulmonary bypass) undergo native valve closure using a technique using 3 felt strips, described in what follows.

Techniques used for aortic valve closure Multiple closure techniques have evolved since 1994 when we first closed the native aortic outflow.8 In 2 early HeartMate I patients with native aortic insufficiency, we replaced the valves with Han-

Figure 1 Suture technique for native aortic valve closure with felt strips: horizontal mattress suture with a second layer of overand-over stitch anchored to the aortic wall. cock bioprostheses (Medtronic, Inc., Minneapolis, MN). At transplantation in both of these patients, the valves were fibrosed closed and had developed subvalvular thrombus, which, viewed from the ventricular aspect, was endothelialized. The aortic aspect was free of thrombus. We subsequently modified the procedure by sewing a pericardial disk beneath the implanted Hancock valve to assure immediate closure (3 patients). In our early experience, we exchanged a mechanical prosthesis for a bioprosthesis in 2 patients, sewing the pericardial disk beneath the valve. We subsequently modified this procedure because we believed a valve exchange was unnecessary. After removing the metal prosthetic valve, we simply closed the outflow with a pericardial disk sewn directly to the annulus (1 patient). More recently, we sewed the pericardial disk to the valve sewing ring (6 patients). At explantation, these valves retained reservoirs of thrombus on the ventricular side, but the aortic side was well healed. In our last 14 patients with native aortic insufficiency, we used a simpler method: 3 thin felt strips were sewn along each aortic valve leaflet in 2 layers, anchored to the aortic wall at the commissures (Figure 1). In patients currently undergoing LVAD insertion we evaluate the aortic valve competence using TEE before, during and after initiation of cardiopulmonary bypass. If greater than mild aortic insufficiency is detected, our tendency is to close the outflow tract.

Statistical analyses Kaplan–Meier (Gehan–Breslow) survival analyses were performed using SIGMAPLOT, version 10.0 (Systat Software, Inc., Chicago IL). Values were considered significant at p ⬍ 0.05. The pre-operative patient characteristics (see Table 2) compare patients with and without aortic closure. One-way analysis of variance (ANOVA) was used to compare the 2 arms of treatment for the continuous-level demographic (i.e., age), whereas a z-test for independent proportions, using a macro for SPSS (version 18.0.2), was used to compare proportions between groups. Keeping in

578 Table 1

The Journal of Heart and Lung Transplantation, Vol 30, No 5, May 2011 Aortic Valve Closure in Patients Undergoing LVAD Support (n ⫽ 28)

Characteristic

Patients

Duration of LVAD support in mean days (range)

Pre-existing aortic insufficiency Prosthetic aortic valve

10

214 (9–464 days)

10

382 (9–883 days)

Aortic insufficiency with LVAD exchange

8

507 (0 1,274 days)

mind the inflated Type I error rate, a level of significance of 0.05 was used.

Results Of the 28 patients, 1 was supported by the HeartMate IP LVAD, 11 by the HeartMate VE/XVE LVAD and 16 by the HeartMate II. There were 22 men and 6 women (mean age 63 years, range 16 to 87 years). Indications for aortic valve closure included native aortic valve insufficiency at the time of LVAD implant (10 patients), aortic valve deterioration with mixed stenosis and insufficiency during prolonged HeartMate LVAD support (8 patients), a pre-existing mechanical aortic prosthesis (9 patients), and a previously placed aortic bioprosthesis (1 patient). The number of patients, duration of LVAD support and outcomes are grouped by indication for closure (see Table 1). Of the 28 patients, 15 (54%) are still being supported by an LVAD at a mean of 461.2 days (range 9 to 1,274 days). Five patients (18%) underwent heart transplantation at a mean of 163 days (range 63 to 334 days), and 1 patient was successfully bridged to recovery at 363 days. Overall, there were 7 deaths (25%) during support (mean 165 days, range 7 to 270 days), 2 operative deaths (sepsis in 1, anoxic encephalopathy in 1) and 5 late deaths (mean 227 days) (multiple-organ failure in 2, sepsis in 2, allergic reaction to medication in 1). The actuarial survival for the group (78% at 1 year, 53% at 3 years) was significantly better than that of a group of all other HeartMate LVAD patients who did not have aortic valve closure (61% at 1 year, 45% at 3 years; p ⫽ 0.03) (Figure 2). Of the 28 patients, 8 had their native aortic valves closed due to progressive, symptomatic aortic insufficiency during LVAD support, and closure was concomitant with device exchange. Aortic insufficiency was not present in any of these patients at their initial LVAD implant. The average duration of support after the LVAD exchange was 524 days (range 100 to 1,274 days). Of the 28 patients, 1 had undergone an aortic root enlargement and bioprosthetic aortic valve replacement. This valve was left in place, and warfarin anti-coagulation was given after LVAD implantation. The valve subsequently closed spontaneously. Another patient had suture closure of an existing mechanical valved conduit with preservation of previously placed coronary artery bypass grafts. This pa-

Outcome Transplant (2 patients), ongoing LVAD (4 patients), died (4 patients) Transplant (3 patients), ongoing LVAD (4 patients), bridge to recovery (1 patient), died (2 patients) Transplant (0 patients), ongoing LVAD (7 patients), died (1 patient)

tient was successfully bridged to recovery after replacing the valved conduit. There were significant differences in the baseline patient characteristics and types of LVADs implanted between the aortic closure and intact groups. Comparisons of types of pumps implanted, frequency of LVAD exchange, reasons for censoring and total time with a closed aortic valve are presented in Table 2. Patients requiring aortic closure were more likely to have had a previous aortic prosthesis (p ⬍ 0.05) with more re-do sternotomies (p ⬍ 0.05). Although not reaching statistical significance these patients underwent more frequent LVAD exchanges (29% vs 18%) and had a higher percentage of HeartMate II devices (53% vs 35%) implanted. Conversely, patients with intact aortic valves were more likely to have only a HeartMate I device implanted (p ⬍ 0.001) and, although not statistically significant, had a tendency for more frequent censoring for transplantation (30% vs 18%). The native and prosthetic aortic valves have remained closed in all patients, with no recurrence of aortic regurgitation on echocardiography or excessive LVAD flows. At transplant or autopsy, all valve closures were intact and

Figure 2 Kaplan–Meier survival curves comparing patients who had aortic valve closure (n ⫽ 28) with our group of LVADsupported patients who did not have aortic valve closure (n ⫽ 103).

Adamson et al. Table 2

Aortic Valve Closure with LVAD Support

579

Pre-operative Patient Comparisons for Aortic Closure vs Intact Aortic Valve

Patients (n) Total pumps implanted Exchanges Age (mean) Male (%) Previous aortic prosthesis Redo sternotomy HeartMate I only HeartMate II (% total pumps) Censored for transplant Censored for recovery Total time with aortic closure

Aortic closure

Intact aortic valves

p-value

28 36 8/28 (29%) 61.2 years 22/28 (78%) 10/28 (36%) 28/36 (78%) 8/28 (29%) 19/36 (53%) 5/28 (18%) 1 11,994 patient-days, 32.86 patient-years

104 123 19/104 (18%) 59.5 years 89/104 (86%) 0/104 71/123 (58%) 65/104 (63%) 43/123 (35%) 31/104 (30%) 0 0

0.230 (NS) .555 (NS) .368 (NS) ⬍0.05 .029 0.001 0.054 (NS) 0.208 (NS)

NS, not statistically significant.

without aortic-side thrombus or deterioration; however, red thrombus was frequently seen on the ventricular side of the closed aortic valve or prosthesis. Complications specifically related to aortic closure were not seen. There were no right ventricular assist devices (RVADs) required, no right heart failure with the need for prolonged inotropic support (⬎7 days) and no cases of recurring heart failure. No cerebrovascular accidents, coronary embolizations or other thromboembolic events were associated with closure of native or prosthetic valves. LVAD function was similar to that of patients without aortic closure. All patients were adequately supported using standard post-LVAD protocols and their follow-up care was indistinguishable from that of patients with intact aortic outflow.

Discussion Patients with end-stage heart failure and aortic valve disease are considered to be at high risk of serious complications after LVAD implantation. During the bridge-to-transplant clinical trials in the 1980s, the presence of a prosthetic aortic valve or significant aortic insufficiency was considered a contraindication to LVAD implantation. However, clinical experience in recent years has led to treatment strategies for closing the aortic valve that have resulted in acceptable survival rates for patients who have had valve surgery and subsequent LVAD implantation.10,12,13 The overall results of our series of LVAD-supported patients who underwent aortic valve closure were good: 2 operative deaths (7%) and a 1-year survival rate of 78%. Our results concur with those of others in that closing the aortic valve at LVAD implant optimizes the functional interaction between the native heart and the LVAD, which is important for successful support.10 Furthermore, surgical treatment of aortic insufficiency when an LVAD is exchanged optimizes flow and extends survival, because several patients undergoing LVAD exchange with coincident native aortic insufficiency had pre-operative refractory, life-threatening heart failure that was not correctable with increasing LVAD flow. Overall, in our series,

22% of the patients would have been denied access to life-saving LVAD support if the presence of a prosthetic valve or native aortic insufficiency were considered contraindications to support. Aortic insufficiency before LVAD implant may be difficult to detect because of the low pressure gradient between the aorta and the left ventricle.8 Severe heart failure is often characterized by a low aortic diastolic pressure and a high left ventricular end-diastolic pressure (LVEDP), which results in minimal regurgitant flow across the abnormal aortic valve. Aortic insufficiency may be observed to be trivial; however, with the initiation of LVAD support, LVEDP decreases and aortic pressure increases, resulting in an increased transvalvular gradient and significant aortic insufficiency. Aortic insufficiency can be detected during LVAD implantation by persistent left ventricular ejection after initiation of cardiopulmonary bypass or observing significant blood flow from the left ventricle through the apical core site. For patients with an LVAD in place, TEE can be used to assess aortic insufficiency and should be performed prebypass, shortly after bypass is initiated, and after exiting bypass. The etiology of aortic insufficiency during LVAD support is unknown. We suspect that increased stress in the leaflets due to prolonged diastole under high transvalvular pressures and infrequent valve opening may be responsible. The near constant coaptation of the leaflets predisposes them to fusion. Furthermore, we suspect that the lack of systolic intervals and the associated elimination of transmural leaflet gradients generate valve leaflet dysfunction by reducing the time that the valve requires to maintain its integrity. These and other factors may cause commissural fusion and elongation (Figure 3).6,14 –17 These progressive aortic valve abnormalities that develop during support have been observed in patients being supported for longer LVAD durations, with both the HeartMate XVE and the HeartMate II.18 As aortic insufficiency increases in severity, it is characterized by an increase in LVAD flow, reduced systemic flow, and, eventually, increasing symptoms of heart failure.

580

The Journal of Heart and Lung Transplantation, Vol 30, No 5, May 2011

Figure 3 Aortic valve, showing fusion of all leaflets that developed during LVAD support.

Symptomatic aortic insufficiency during support with the HeartMate XVE LVAD may contribute to motor failure. As aortic insufficiency increases, LV filling increases. The XVE’s automatic control responds by increasing the pump rate. Thus, a “racing” pump rate with high flows that are unexplained by other physiologic changes are likely related either to deterioration of the native aortic valve or of the LVAD inflow valve. The high pump rate increases wear on the motor bearings and, eventually, the pump will fail. Therefore, patients who present with high pump rates and flows should always be assessed for aortic insufficiency, and an analysis of the pump life should be performed.4 When aortic insufficiency is present and the HeartMate XVE motor shows signs of bearing failure, the aortic valve should be closed and the device exchanged. The surgical techniques for treating aortic insufficiency have varied, but most experienced LVAD users agree that closing both native and prosthetic valves is necessary for successful long-term support. The use of a central leaflet coapting stitch as reported by Park et al,19 a bioprosthetic valve, or other techniques to improve aortic valve function are often sub-optimal and not recommended for patients undergoing LVAD support. We observed a case of progressive aortic insufficiency in a HeartMate I LVAD recipient from an outside institution whose central leaflet closure remained intact. The remaining leaflets appeared to be elongated producing insufficiency between the central closed

area and the commissures. Our group and others observed that bioprosthetic valves will normally close completely from fibrosis, but there is a risk of thromboembolism.9,13 Therefore, we now recommend that pre-existing bioprosthetic valves be removed and the outflow tract closed with a pericardial patch at the time of LVAD implant. If the bioprosthetic valve cannot be easily removed, early anticoagulation may be successful for avoiding embolization. If greater than mild aortic insufficiency exists at the LVAD implant, then we generally close the valve. Our preferred technique is to sew 3 thin felt strips along each aortic valve leaflet in 2 layers, with anchoring to the aortic wall at the commissures. Pre-existing mechanical aortic valves are associated with a risk of thrombus formation above the valve and thromboembolism when the valve opens. We have effectively closed mechanical valves by sewing a pericardial disk to the sewing ring. In a retrospective review, it is impossible to determine why survival was better in patients who had aortic outflow closure. Although this was a consecutive series with a single discriminator, aortic closure, these patient groups were very different and these inherent differences may be more predictive of survival than aortic closure. A significant number of patients in the closure group (36%) had a previously placed aortic prosthesis with the possible survival advantage of a chronically conditioned right ventricle. There was a higher incidence, although not statistically significant, of patients who required aortic closure associated with LVAD exchange (29% vs 18%). In our experience, LVAD exchange is associated with excellent survival.20 The fact that they survived long enough to require an exchange could also contribute to the survival advantage in the aortic closure group. There was also a higher percentage of aortic closure occurrences in patients with a HeartMate II implant (53% vs 35%) and, accordingly, a higher percentage of patients with an intact aortic valve who only had a HeartMate I (63% vs 29%; p ⬍ 0.05). In our experience, and in randomized trials,21,22 HeartMate (HM) II patients have shown better survival than HM I patients. This characteristic alone could be responsible for the survival difference. More patients were censored for transplantation in the intact group, artificially limiting their time on the device. The combined variables of more HM I-only patients, increased censoring for transplantation and fewer exchanges may have disadvantaged the survival of patients with an intact aortic valve. Conversely, the combination of a previously placed aortic prosthesis, a greater percentage of HeartMate II and exchange procedures, and decreased censoring for transplantation may be responsible for the aortic closure’s survival advantage. With such a small sample size and the presence of confounding variables it is not possible to ascribe the improved survival to aortic closure. Although it is important to remember that after 32.9 patient-years with a closed aortic outflow there has not been a single closure that developed insufficiency, a systemic embolization or LVAD support failure. The theoretical advantages of native aortic outflow closure include: (1) mandatory ejection of all blood through the

Adamson et al.

Aortic Valve Closure with LVAD Support

LVAD, which guarantees flow through the pump during pump off situations and may prevent thrombus formation; (2) a single outlet left ventricle, which may produce beneficial flow characteristics in the native ventricle; (3) elimination of intermittent aortic valve opening and ejection of sub- or supravalvular thrombus; and (4) prevention of aortic valvular deterioration and resultant insufficiency, obviating the risk of future closure. Although we cannot delineate the etiology or mechanism of improved survival, we can say that aortic closure as described is durable, does not negatively affect survival, and is not associated with long-term complications of emboli, aortic insufficiency or inadequate LVAD function and support. There remain concerns related to aortic valve closure, most notably the potential for thromboembolism or the potential difficulty of mandatory LV ejection through the LVAD during periods of device failure or disconnect. Fortunately, we did not see these events in this series. Systemic embolism has been seen in patients with LVADs with intact native aortic valves. During long-term LVAD support, thrombus formation on the aortic valve has been observed by us and others.23,24 Because transmitting a thrombus through a rotary pump is unlikely to occur without first causing stoppage due to inflow occlusion, the source of these emboli in continuous-flow LVADs is probably the aortic root. Pre-emptive outflow tract closure may circumvent this potentially devastating complication. It is clear that the native aortic valve deteriorates in both HeartMate I and HeartMate II patients. We have seen uncorrected, mild native aortic insufficiency progress and cause congestive heart failure necessitating re-do sternotomy aortic valve closure in 3 HM II–supported patients. We do not know if the incidence of aortic insufficiency is greater in HM II than in HMI patients, although longer support times for HM II vs HM I patients are likely to result in an increased incidence of aortic insufficiency. The intrinsic pumping characteristics of the devices may play a role. Each device creates different loading conditions for the valve.

Limitations Due to the small number of patients in this retrospective review, coupled with the cross-over nature of HM I to HM II exchanges, subset analysis is not powered to reach any meaningful conclusions when comparing differences in HM I vs II patients or the precise etiology of survival differences.

Clinical implications An intact mechanical aortic prosthesis remains a contraindication to LVAD support due to the risk of thromboembolism associated with intermittent opening. Closure of these valves was not associated with thromboembolism or LVAD dysfunction and, therefore, patients with a previous aortic prosthesis that can be closed at the time of LVAD insertion should not be excluded from LVAD support. Tissue aortic prosthesis will spontaneously close during LVAD support. There is a theoretical possibility of thromboembo-

581 lism of sub-valvular debris before complete valve closure; therefore, they should be left in place with early anti-coagulation or a pericardial patch sewn into the annulus at the time of LVADs implant. The most controversial situation is with regard to the regurgitant native aortic valve. It has been shown that the native aortic valve deteriorates with the unique physiology of LVAD support. If the valve is already regurgitant before LVAD implant it can be assumed that this floppy valve will continue to deteriorate, increasing the leak and compromising LVAD support. The 3-felt-strip technique of closure is reliable, does not deteriorate over time, and does not compromise LVAD function. It is unknown what level of regurgitation warrants closure, but we recommend closure for any leak that demonstrates more than mild insufficiency once cardiopulmonary is initiated. Interestingly, a mild leak during full cardiopulmonary bypass can become significantly greater once LVAD flow is initiated. In conclusion, aortic outflow closure during LVAD support appears safe. Therefore, native aortic insufficiency or the presence of a prosthetic aortic valve should not be considered contraindications for LVAD support.

Disclosure statement This study was funded by an unrestricted research grant from the J. Douglas and Marian R. Pardee Foundation. The authors have no conflicts of interest to disclose.

References 1. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med 2007; 357:885-96. 2. Frazier OH, Rose EA, Oz MC, et al. Multicenter clinical evaluation of the HeartMate vented electric left ventricular assist system in patients awaiting heart transplantation. J Thorac Cardiovasc Surg 2001;122: 1186-95. 3. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009;361:2241-51. 4. Haft JW, Suzuki Y, Aaronson KD, et al. Identification of device malfunction in patients supported with the HeartMate XVE left ventricular assist system. ASAIO J 2007;53:298-303. 5. Pagani FD, Long JW, Dembitsky WP, et al. Improved mechanical reliability of the HeartMate XVE left ventricular assist system. Ann Thorac Surg 2006;82:1413-8. 6. Letsou GV, Connelly JH, Delgado RM III, et al. Is native aortic valve commissural fusion in patients with long-term left ventricular assist devices associated with clinically important aortic insufficiency? J Heart Lung Transplant 2006;25:395-9. 7. Jaski BE, Miller DA, Hoagland PM, et al. Assessment of recurrent heart failure associated with left ventricular assist device dysfunction. J Heart Lung Transplant 2005;24:2060-7. 8. Adamson RM, Dembitsky WP, Jaski BE, et al. Left ventricular assist device support of medically unresponsive pulmonary hypertension and aortic insufficiency. ASAIO J 1997;43:365-9. 9. Baradarian S, Dembitsky WP, Jaski B, et al. Left ventricular outflow tract obstruction associated with chronic ventricular assist device support. ASAIO J 2002;48:665-7.

582

The Journal of Heart and Lung Transplantation, Vol 30, No 5, May 2011

10. Rao V, Slater JP, Edwards NM, et al. Surgical management of valvular disease in patients requiring left ventricular assist device support. Ann Thorac Surg 2001;71:1448-53. 11. Branch KR, Dembitsky WP, Peterson KL, et al. Physiology of the native heart and Thermo Cardiosystems left ventricular assist device complex at rest and during exercise: implications for chronic support. J Heart Lung Transplant 1994;13:641-50. 12. Barbone A, Rao V, Oz MC, et al. LVAD support in patients with bioprosthetic valves. Ann Thorac Surg 2002;74:232-4. 13. Feldman CM, Silver MA, Sobieski MA, et al. Management of aortic insufficiency with continuous flow left ventricular assist devices: bioprosthetic valve replacement. J Heart Lung Transplant 2006;25:1410-2. 14. Connelly JH, Abrams J, Klima T, et al. Acquired commissural fusion of aortic valves in patients with left ventricular assist devices. J Heart Lung Transplant 2003;22:1291-5. 15. Mudd JO, Cuda JD, Halushka M, et al. Fusion of aortic valve commissures in patients supported by a continuous axial flow left ventricular assist device. J Heart Lung Transplant 2008;27:1269-74. 16. Rose AG, Park SJ, Bank AJ, et al. Partial aortic valve fusion induced by left ventricular assist device. Ann Thorac Surg 2000;70:1270-4. 17. May-Newman K, Abulon D, Joshi M, et al. Morphology and tissue characterization of aortic valve geometry and fusion in LVAD patients. J Heart Valve Disease. In press.

18. Matthews JC, Aaronson KD, Jain R, et al. Aortic insuffciency—trends over time in LVAD supported patients. J Heart Lung Transplant 2009;28(suppl):S306. 19. Park SJ, Liao KK, Segurola R, et al. Management of aortic insufficiency in patients with left ventricular assist devices: a simple coaptation stitch method (Park’s stitch). J Thorac Cardiovasc Surg 2004; 127:264-6. 20. Adamson RM, Dembitsky WP, Baradarian S, et al. HeartMate left ventricular assist system exchange: results and technical considerations. ASAIO J 2009;55:598-601. 21. Miller LW, Pagani FD, Russel SD, et al. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med 2007; 357:885-95. 22. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow ventricular assist device. N Engl J Med 2009;361:2241-51. 23. Pu M, Stephenson ER Jr, Davidson WR Jr, et al. An unexpected surgical complication of ventricular assist device implantation identified by transesophageal echocardiography: a case report. J Am Soc Echocardiogr 2003;16:1194-7. 24. Crestanello JA, Orsinelli DA, Firstenberg MS, et al. Aortic valve thrombosis after implantation of temporary left ventricular assist device. Interact Cardiovasc Thorac Surg 2009;8:661-2.