Improved Mechanical Reliability of the HeartMate XVE Left Ventricular Assist System

Transcription

1 Improved Mechanical Reliability of the HeartMate XVE Left Ventricular Assist System Francis D. Pagani, MD, PhD, James W. Long, MD, PhD, Walter P. Dembitsky, MD, Lyle D. Joyce, MD, PhD, and Leslie W. Miller, MD Section of Cardiac Surgery, University of Michigan Health System, Ann Arbor, Michigan; Utah Artificial Heart Program, University of Utah School of Medicine and LDS Hospital, Salt Lake City, Utah; Division of Cardiovascular Surgery, Sharp Memorial Hospital, San Diego, California; and Division of Cardiothoracic Surgery and Cardiovascular Division, University of Minnesota, Minneapolis, Minnesota Background. The HeartMate XVE left ventricular assist device is a valuable treatment option for patients with end-stage heart failure. During the past several years, the XVE has undergone a series of design enhancements to improve reliability. We compared the reliability of the two most recent design iterations of the XVE pump (stitch modification to the inflow valve assembly and new inflow valve housing redesign) to the earlier VE version. Methods. A retrospective evaluation of device reliability was performed for 268 devices implanted in 245 patients (VE: n 167 devices, 147 patients, implant dates October 16, 1998, to December 19, 2003; XVE: n 101 devices, 98 patients, implant dates August 1, 2002, to April 14, 2004). Results. Median duration of device support for the VE and XVE was 159 days (range, 0 to 1,206 days) and 229 days (range, 0 to 693 days), respectively (p 0.495). Significantly fewer major device malfunctions occurred within the XVE group as compared with the VE group (6 versus 36, respectively; p ). The number of major device malfunctions per patient-year of support for inflow valve dysfunction, bearing wear, and other failures for the VE and XVE were 0.2 versus 0.04 (p 0.006), 0.16 versus 0.01 (p 0.005), and 0.06 versus 0.04 (p 1.000), respectively. The freedom from major device malfunction at 1 year was 76% 6% for the VE and 97% 2% for the XVE device (p < 0.001). The freedom from death as a result of major device malfunction at 1 year was 97% 2% for the VE and 98% 2% for the XVE (p 0.698). Conclusions. Design enhancements to the HeartMate XVE have significantly reduced the incidence of major device malfunctions compared with the earlier VE model because of a reduction in failure modes from bearing wear and inlet valve dysfunction. Further follow-up is necessary to establish the long-term durability of the most recent XVE pump version. (Ann Thorac Surg 2006;82:1413 9) 2006 by The Society of Thoracic Surgeons During the past decade, implantable left ventricular assist device (LVAD) systems have gained clinical acceptance as an appropriate therapy for patients with end-stage heart failure refractory to medical therapy. The majority of experience with LVAD therapy has occurred in patients who were supported as a bridge to cardiac transplantation [1 3]. Although cardiac transplantation is the ideal treatment modality for some patients with end-stage heart failure, fewer than 3,000 donor organs are available worldwide [4]. This scarcity of donor organs coupled with the growing epidemic of heart failure [5 7] has led researchers to explore the use of LVAD therapy as an alternative to heart transplantation in patients with end-stage heart failure [8]. In situations in which LVAD therapy is considered for long-term treatment (ie, years), device durability becomes an important consideration. The HeartMate LVAD system (Thoratec Corp, Pleasanton, CA) was originally conceived as a permanent Accepted for publication April 19, Address correspondence to Dr Pagani, Section of Cardiac Surgery, 2124 Taubman Center, Box 0348, 1500 E. Medical Center Dr, Ann Arbor, MI 48109; fpagani@umich.edu. LVAD; however, to date it has been used primarily as a temporary bridge to heart transplantation [9]. Recently (2003), the device received US Food and Drug and Administration approval for destination therapy in addition to its prior approved indication for bridge to heart transplantation. This approval for the HeartMate VE device for destination therapy indication was based on data from a prospective, multicenter randomized trial, REMATCH, which compared either LVAD therapy or optimal medical management in patients not considered candidates for heart transplantation [8]. Data from this trial demonstrated a significant survival benefit and improvement in quality of life for patients treated with LVAD therapy compared with optimal medical management [8]. However, in the REMATCH trial, device failure of the HeartMate VE contributed to significant patient morbidity and mortality [10]. The probability of Heart- Mate VE device failure was 35% at 24 months, and device replacement was required in 10 patients [9]. System Dr Miller discloses a financial relationship with Thoratec; Dr Long with Medquest and Thoratec by The Society of Thoracic Surgeons /06/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 1414 PAGANI ET AL Ann Thorac Surg MECHANICAL RELIABILITY OF THE HEARTMATE XVE 2006;82: Fig 1. The previous inflow graft assembly unit of the HeartMate XVE was modified by the addition of a stitch at the commissure of the inlet valve (arrow) to strengthen the attachment of the valve to the graft material to prevent valve dysfunction and reduce the risk of disconnection. failure defined as the failure of the device to support the circulation or death after surgery for replacement was 0.13 per patient-year, whereas the confirmed LVAD malfunction rate was 0.90 per patient-year [10]. Freedom from device replacement was 87% at 1 year and 37% at 2 years [10]. The most common failure modes of the device were inlet valve dysfunction and bearing wear [10]. Significant improvements and system enhancements to the original engineering design of the HeartMate VE device have been implemented after the experience from the REMATCH trial. The HeartMate extended-lead, vented electric (XVE) device became available in September 2001 with the goals of improving pump reliability and durability, and of decreasing device-related complications. The purpose of this study is to determine whether the most recent design enhancements of the XVE device have improved device reliability and durability compared with the HeartMate VE device used in the REMATCH trial. Material and Methods Device Incremental improvements were made to the current HeartMate XVE device in comparison with the previous HeartMate VE model. The first significant change was to the inflow valve assembly with the addition of sutures at the valve commissures to strengthen the attachment to the graft material, prevent valve dysfunction, and reduce the risk of disconnection (Fig 1). High pressure in the pump chamber would tend to displace the original valve and cause graft distortion. In addition, sutures that attach the vascular graft to the metal valve cage would break because of the pressure and allow the valve to oscillate in the cage. This oscillation could produce abrasion as well as distortion, leading to valve incompetence. Most recently, a completely new inflow valve housing, incorporating bilateral flex joints and an inflow cannula quickconnect screw ring, has been introduced (Fig 2). Another significant improvement to the HeartMate XVE system was the incorporation of new software for the controller, called Opti-Fill. This software was designed to reduce stress on the inflow valves, diaphragm, and motor bearings by optimizing pump filling volumes, thereby reducing internal pump chamber pressure. Enhancements to the HeartMate XVE also included a new percutaneous lead that was longer, narrower, and more flexible, which was designed to reduce the risk of driveline infections, improve strength, and resist kinking. Fig 2. (Left) Original inflow valve housing design. (Right) New inflow valve housing design with polytetrafluoroethylene (PTFE) sleeve to reduce suture abrasion and preclotting surface area, titanium cylinder for protection of the tissue valve and conduit, and two-point articulation to protect against tissue valve deformation.

3 Ann Thorac Surg PAGANI ET AL 2006;82: MECHANICAL RELIABILITY OF THE HEARTMATE XVE 1415 Table 1. Patient Characteristics Variable VE XVE p Value Number of patients Median age (range) 55 (19 84) 54 (16 77) Male (%) Etiology of heart failure (%) Idiopathic:ischemic:other 37:59:4 47:52: Device indication (%) Bridge to transplant Location of implanting center (%) LDS:Michigan:Minnesota: Sharp a 23:37:29:11 21:35:34: a Latter Day Saints Hospital (LDS), University of Michigan Medical Center (Michigan), University of Minnesota (Minnesota), and Sharp Memorial Hospital (Sharp). The HeartMate XVE additionally includes modifications made to a later version of the HeartMate VE, most notably, the addition of an outflow graft bend relief and locking screw rings. The bend relief is positioned over the outflow graft to prevent graft kinking and abrasion of the outflow graft. Outflow graft kinking and abrasion is believed to cause graft erosion and blood loss as well as high pump chamber pressure, which contributes to increased stress on the inlet valves, bearings, and diaphragm. Another significant improvement was the locking screw rings, a ratchet-type locking mechanism that eliminates the need for conventional sutures to secure the grafts and valves as well as conduits to the inflow and outflow body of the pump. The locking screw rings were designed to reduce the possibility of blood loss caused by accidental loosening or dislodgement. Patients A retrospective review of all medical records was performed for all patients undergoing implantation of the HeartMate VE or XVE device with stitch modification to the inflow valve assembly or new inflow valve housing design at the Latter Day Saints Hospital (Salt Lake City, UT), University of Michigan Health System (Ann Arbor, MI), University of Minnesota or Fairview Medical Center (Minneapolis, MN), and Sharp Memorial Hospital (San Diego, CA) from September 28, 1998, through May 31, Institutional review board approval was obtained at each institution through waiver of the informed consent process (Latter Day Saints Hospital, Fairview Medical Center, and Sharp Memorial Hospital) or through the informed consent process (University of Michigan). All patients were supported as a bridge to cardiac transplantation or for destination therapy. Major device malfunctions were defined as an internal electrical or mechanical component failure resulting in one or more of the following: (1) death; (2) need for readmission to the hospital for urgent transplant listing; (3) need for pneumatic actuation of the device; (4) need for reoperation; and (5) inflow valve incompetence confirmed by transesophageal echocardiography and hemodynamic monitoring resulting in device replacement, death, or hospitalization. Failure of external device components that were replaceable and correctable, including system controller, power base unit, power base cable, batteries, and battery clips, were not included as major device malfunctions. Statistical Methods Comparisons of baseline demographics between patients supported with the VE or XVE HeartMate device were performed using Fisher s exact test or unpaired Student s t tests as appropriate. Probability was two-tailed with a probability value less than 0.05 regarded as statistically significant. A Wilcoxon rank-sum test was used to determine the differences in duration of support between the XVE and VE devices. Freedom from device malfunction was analyzed using the product-limit method of Kaplan and Meier with a log rank analysis used to detect differences between groups. Events per patient-year were compared using Poisson regression. A Cox regression analysis was performed to identify variables associated with device malfunction. Results From September 28, 1998, through May 31, 2004, 147 patients underwent implantation of 167 HeartMate VE devices and 98 patients underwent implantation of 101 HeartMate XVE devices. There were no significant differences between groups with regard to age, sex, cause of the heart failure, percent of patients undergoing implantation for bridge to transplant indication, and volume of implants at each participating center (Table 1). The median duration of support for the VE was 159 days (range, 0 to 1,206 days) and 229 days (range, 0 to 693 days) for the XVE (p ). Total use was 103 patient-years for the VE and 67 patient-years for the XVE. A total of 36 major device malfunctions occurred with the VE device and 6 major device malfunctions occurred Table 2. Comparison of the Type and Number of Major Device Malfunctions Occurring in the VE and XVE Groups Malfunction VE (n 167) XVE (n 101) Any major device 36 6 malfunction Bearing failure (main bearing 12 1 or cam follower bearing) Inflow valve dysfunction 19 2 Outflow valve dysfunction 2 0 Diaphragm rupture 2 0 Other 1 a 3 b a Failure of rotor band improperly contacting stator. b One event was attributed to failure of the commutator, and two malfunctions were attributed to internal air flow restriction.

4 1416 PAGANI ET AL Ann Thorac Surg MECHANICAL RELIABILITY OF THE HEARTMATE XVE 2006;82: Table 3. Comparison of the Type of Major Device Malfunctions Between VE and XVE Groups Expressed as the Number of Events per Patient-Year of Device Support Variable VE XVE Risk Ratio (95% Confidence Interval) Inflow dysfunction ( ) Bearing failure ( ) Other failure ( ) in the XVE cohort (p ). A summary of the number and types of failure modes for each group is listed in Table 2. Importantly, no malfunctions attributable to diaphragm rupture or outflow graft dysfunction occurred in the XVE cohort. The number of major device malfunctions per patient-year of support attributable to inflow valve dysfunction or bearing failure was significantly reduced in the XVE group as compared with the VE group (p and p 0.005, respectively; Table 3). A Cox regression analysis was performed on the cohort using age, body surface area, mean systolic blood pressure, and duration of time in fixed mode of device operation to assess for significant variables associated with device malfunction in a time-dependent fashion. No significant variables were identified using stepwise forward regression. The freedom from all major device malfunctions was significantly greater for the XVE cohort as compared with the VE group (p ; Fig 3). The freedom from all major device malfunctions was 76% 6% and 97% 2% at 1 year, and 36% 7% and 76% 11% at 18 months for the VE and XVE groups, respectively. Freedom from death attributable to major device malfunction at 1 year was 97% 2% for the HeartMate XVE device and 98% Fig 3. Freedom from all major device malfunctions for the VE (thin line) and XVE (thick line) groups. Fig 4. Freedom from death caused by major device malfunction for the VE (thin line) and XVE (thick line) groups. 2% for the HeartMate VE device, and was not significantly different between the groups (Fig 4). Comment Data from this study demonstrate that important design modifications to the HeartMate XVE LVAD have resulted in a significantly decreased incidence of major device malfunctions. Importantly, significantly fewer major device malfunctions occurred with similar duration of device support between the groups. This decrease in incidence of major device malfunctions was primarily attributable to the significant reduction in incidence of both inflow valve dysfunction and bearing wear (either main bearing or cam follower bearing). Inflow valve dysfunction was the most common major device malfunction and accounted for approximately 53% of all major device malfunctions in the Heartmate VE device in this study; this was similar to the experience observed in the REMATCH trial [10]. With the addition of the reinforcing stitch to the valve graft assembly and later total redesign of the inlet valve housing, the number of major device malfunctions secondary to inflow valve dysfunction was significantly reduced with the HeartMate XVE device experience in this study. Previous reports of the durability of the HeartMate XVE device did not include these most recent changes to the inlet valve assembly and housing [11, 12]. The second major area of improvement was the significant reduction in number of major device malfunctions attributable to bearing wear, either main bearing or cam follower bearing. The major design enhancement to the HeartMate XVE device to reduce bearing wear has been developed through an improved understanding of the

5 Ann Thorac Surg PAGANI ET AL 2006;82: MECHANICAL RELIABILITY OF THE HEARTMATE XVE 1417 effect that high pump-chamber pressure has on both subsequent bearing and inlet valve durability. The incorporation of the Opti-Fill software within the system controller of the HeartMate XVE device was designed to decrease internal pump-chamber pressures by slowing the beat rate of the device to maximize device filling and reduce the initial surge in internal pump-chamber pressures associated with early device systole. The effect of this modification has been to reduce bearing load, improve bearing durability, and reduce load on the inlet valve graft assembly resulting in reduced graft distortion. The association of pump-chamber pressures and durability of the HeartMate XVE device has also led to an improved understanding of the impact of physiologic and surgical factors on device durability. Increased afterload secondary to hypertension, kinking of the outflow graft, or pressure gradient at the anastomotic site of the aorta and outflow graft secondary to improper sizing can also contribute to high pump-chamber pressures independent of device-related factors. It is unknown how improvement in patient management with control of hypertension and improved surgical technique reducing the incidence of outflow graft distortion or anastomotic narrowing impacted the results of the current study. The design enhancements to the HeartMate XVE device and better understanding of the clinical factors that contribute to device failure may be responsible for the improvement in outcomes observed with destination therapy beyond what was achieved with the REMATCH experience [13]. Diaphragm rupture occurred in two VE pumps. This catastrophic malfunction was not reported in the Heart- Mate XVE device that has incorporated improved diaphragm support and positioning. There was a small increase in number of other types of major device malfunctions that were associated with the HeartMate XVE device and not observed in the earlier VE iteration. A commutator, which is an electrical control circuit for the internal electric motor within the LVAD, failed in one XVE device. This failure mode was traced to a defective component within the assembly. This is a rare event in that the commutator is tested completely in all of its operating modes before incorporation into the Heart- Mate XVE device. The particular component that failed prematurely had never failed before this event. The other failure mode not observed with the VE device was related to restricted airflow at the internal orifice of the driveline within the pump resulting in an inability of the air to move into and out of the motor housing. Further design modifications and changes in pump manufacturing have since been performed to prevent this problem. The mortality associated with serious device malfunction was reduced in the XVE cohort versus the VE cohort (one event versus three events at 24 months). Although this is not statistically significant (p 0.87), it does demonstrate that the HeartMate system with its failure warning systems and backup control system can reduce the effects of system failure and reduce the mortality associated with device failure. As shown in Figure 4, a patient with an XVE would have a 98% freedom from death at 24 months caused by a major device failure. Comparisons of serious mechanical failures before and after HeartMate LVAD modifications have been encouraging in multicenter and large single-center experiences [11, 12]. These reports, however, did not examine the impact of the two most recent design modifications of the HeartMate XVE device that included the stitch reinforcement at the inlet valve graft assembly and new inflow valve housing design. This current study and the previous reports importantly highlight the need for continued surveillance of device performance after introduction of a device for clinical use and the importance of clinical experience in assisting in design modifications. The limitations of long-term mechanical support as a result of issues of device reliability is, perhaps, one of the most important factors facing the use of LVAD therapy, as highlighted in the outcomes observed in the REMATCH trial. Thus, limitations in device durability have significant impact on patient survival and outcomes, quality of life, and cost of care. It is important to emphasize that the current generation of HeartMate XVE device is not the same device used in the REMATCH trial (HeartMate VE), and significantly improved device reliability has now been demonstrated with the latest Heart- Mate XVE design. Further, although newer generations of rotary pumps with continuous flow characteristics are anticipated to have significantly improved durability, displacement pumps with pulsatile flow are still thought to be necessary for some patients with severe shock and multiorgan dysfunction. Additionally, the absence of a requirement for warfarin or heparin anticoagulation for the HeartMate XVE system makes this device an attractive alternative for some patients with contraindications to anticoagulation. These reasons make relevant the need for continued device enhancements and improvements to durability with the HeartMate XVE device. The major limitations of this study are attributable to (1) the retrospective nature of the study design, and, importantly, (2) the fact that pump implantations of the VE and XVE designs were sequential in time, with the overwhelming majority of XVE pumps being implanted at a later date than the VE design. Thus, center experience in areas of surgical technique and patient care (control of infection and hypertension) may have had important influences on pump durability. Indeed, significant changes in patient management and center experience led to important improvements in survival for patients treated with LVAD therapy during the course of the REMATCH trial [14]. However, we believe that not all changes in patient management and center experience can account for the reductions in major device malfunctions for the HeartMate XVE group. The recent XVE design enhancements are thought to be a major contributor to improved device reliability and durability. Further, results were limited to 18 months as an insufficient number of patients achieved the 24-month period to allow statistically meaningful data to be presented.

6 1418 PAGANI ET AL Ann Thorac Surg MECHANICAL RELIABILITY OF THE HEARTMATE XVE 2006;82: In conclusion, major device malfunctions represent an important obstacle to successful long-term LVAD therapy and result in significant morbidity and mortality. However, technologic advances incorporated into the new iteration of the HeartMate XVE have resulted in significantly fewer device malfunctions within the first 18 months of follow-up, specifically related to a reduction in inflow valve dysfunction and bearing wear. Further follow-up of device performance will be necessary to establish the long-term reliability of this new generation of HeartMate XVE pump. With the recent design enhancements and changes in patient management, we expect to see continued improvement in the outcomes of patients supported with the HeartMate XVE. References 1. Long JW. Advanced mechanical circulatory support with the HeartMate left ventricular assist device in the year Ann Thorac Surg 2001;71(3 Suppl):S Poirier VL. Worldwide experience with the TCI HeartMate system: issues and future perspective. Thorac Cardiovasc Surg 1999;47(Suppl 2): Frazier OH, Rose EA, McCarthy P, et al. Improved mortality and rehabilitation of transplant candidates treated with a long-term implantable left ventricular assist system. Ann Surg 1995;222: Hosenpud JD, Bennett LE, Deck BM, et al. The registry of the International Society for Heart and Lung Transplantation; seventeenth official report J Heart Lung Transplant 2000;12: Heart Disease and Stroke Statistics 2004 update. Dallas: American Heart Association, Task Force for Diagnosis and Treatment of Chronic Heart Failure, European Society of Cardiology. Guidelines for diagnosis and treatment of chronic heart failure. Eur Heart J 2001;22: Expert Panel Review of the NHLBI Artificial Heart (TAH) Program. June 1998 November Available at (accessed Dec 2005). 8. Rose EA, Gellijns AC, Moskowitz AJ, et al. Long-term use of the left ventricular assist device for end-stage heart failure. N Engl J Med 2001;345: Frazier OH, Myers TJ, Radovancevice B. The HeartMate left ventricular assist system overview and 12-year experience. Tex Heart Inst J 1998;25: Dembitsky WP, Tector AJ, Park S, et al. Left ventricular assist device performance with long-term circulatory support: lessons from the REMATCH trial. Ann Thorac Surg 2004;78: Pagani FD, Patel HJ, Wright S, Dyke BD, Aaronson KD. Significant reduction in major LVAD device failures: comparison of the HeartMate VE and XVE LVAS. J Heart Lung Transplant 2003;22(1 Suppl):S Dowling RD, Park SJ, Pagani FD, et al. HeartMate VE LVAS design enhancements and its impact on device reliability. Eur J Cardiothorac Surg 2004;25: Long JW, Kfoury AG, Slaughter MS, et al. Long-term destination therapy with the HeartMate XVE left ventricular assist device: improved outcomes since the REMATCH study. Cong Heart Fail 2005;11: Park SJ, Tector A, Piccioni W, et al. Left ventricular assist devices as destination therapy: a new look at survival. J Thorac Cardiovasc Surg 2005;129:9 17. INVITED COMMENTARY The Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial investigating lifetime use of left ventricular assist devices (LVAD) in non-transplant-eligible patients with end-stage heart disease was performed between 1998 and 2001 using the HeartMate vented electric (VE) device (Thoratec Corp, Pleasanton, CA). This seminal study rigorously and prospectively established new standards for survival, quality of life, and adverse events in this severely ill patient population. Despite the favorable results for LVAD versus optimal medical management, REMATCH is often criticized by the medical community: few patients survived more than 2 years secondary to high rates of adverse events (infection, bleeding, neurologic) and device failure (35% at 2 years). As reported by other investigators, thorough infection prophylaxis, careful anticoagulation protocols, and increased surgical experience have markedly reduced adverse events and have improved survival. The Achilles heel of lifetime LVAD therapy ultimately rests on the latter point of device durability. The current article reports the impact of changes in technology on device malfunction. Pagani and colleagues [1], from four active LVAD centers, compared their results using the HeartMate VE, which was used in REMATCH, versus the two latest perturbations in the HeartMate XVE pump. These changes include a new inflow valve housing (this accounted for more than 50% of REMATCH device failures), new software for the controller to optimize filling pressure volume relationships, a slicker percutaneous lead, outflow graft bend relief, and locking screw rings to secure the pump, valves, and grafts. The results show the marked decrease in major device malfunctions with the XVE, including bearing failures (likely from the software improvements), inflow and outflow valve dysfunction, and diaphragm rupture. Although this report clearly demonstrates a decrease in mechanical problems in the latest HeartMate pump, the data need to be viewed with some circumspection. Most of the devices (82%) were used for bridging to transplantation. As such, long-term reliability remains unknown. In fact, fewer than 30% of the devices were implanted for more than 1 year and fewer than 5% of XVE pumps were still implanted at 18 months. It is not clear how many pumps, if any, were replaced secondary to their malfunctions (with the attendant morbidity and mortality) or if pump failure became the indication for transplantation. Because of the limitations of numbers with extended follow-up, it is difficult to extrapolate how the latest structural and software changes can influence long-term survival. With the increasing number of pulsatile and continuous pumps either in use or in development, it is clear that we have good solutions for bridging patients with mechanical support. The challenge of an efficient, safe, and durable LVAD for lifetime use still exists. Although there 2006 by The Society of Thoracic Surgeons /06/$32.00 Published by Elsevier Inc doi: /j.athoracsur