The TandemHeart as a Bridge to a Long-Term Axial-Flow Left Ventricular Assist Device (Bridge to Bridge)

Abstract

End-stage heart-failure patients in acute refractory cardiogenic shock with multi-organ dysfunction require aggressive medical therapy that includes inotropic support. Historically, the intra-aortic balloon pump was the last option for patients who were dying of acute cardiogenic shock. Short-term extracorporeal pulsatile or nonpulsatile cardiac assist devices or extracorporeal membrane oxygenation offered further treatment options; however, these therapies required invasive surgical procedures. Patients in this high-risk group had increased mortality rates from major procedures that required cardiopulmonary bypass. We used the TandemHeart®, a percutaneously implanted device for short-term cardiac assistance, to lower the risk of death and improve hemodynamic performance and end-organ perfusion before implanting long-term assist devices in selected patients with signs of profound cardiogenic shock.

Nine end-stage heart-failure patients (mean age, 37.7 yr) in acute refractory hemodynamic decompensation received a percutaneously implanted TandemHeart pump as a bridge to an implantable axial-flow pump. To determine the relative risk for these patients, prognostic scores were calculated before and after insertion of the TandemHeart.

Percutaneous support times ranged from 1 to 22 days (mean, 5.9 d). The mean cardiac index before support, 1.02 L/(min·m²) (range, 0.0–1.8 L/[min·m²]) (0.0 L/[min·m²] implies active cardiopulmonary resuscitation), improved to 2.97 L/(min·m²) (range, 2.2–4.0 L/[min·m²]) during support. Three patients underwent successful cardiac transplantation; 5 are currently supported by axial-flow pumps; and 1 died of complications unrelated to the axial-flow pump, after 587 days.

End-organ function and overall condition improved uniformly in our patients, thus decreasing the preoperative risk factors for implantation of the long-term device.

Despite improvements, medical therapy for chronic heart failure is still of limited efficacy in patients who have advanced disease, especially in those who present with acute cardiac decompensation. Inotropic support and intra-aortic balloon pumps (IABPs) were historically the only tools available to clinicians for the treatment of patients who were dying of acute cardiogenic shock. The advent of temporary support offered by short-term extracorporeal pulsatile or nonpulsatile cardiac assist devices or by extracorporeal membrane oxygenation (ECMO) in pre-terminal patients enabled the introduction of further treatment options. However, this therapy required invasive surgical procedures. Patients in this high-risk group were more likely to die when they were subjected to major surgical procedures that required cardiopulmonary bypass. Patients in cardiogenic shock with signs of multi-organ system failure had unfavorable outcomes when treated emergently with surgically implanted long-term assist devices.^1,2

In recent years, the use of temporary assist devices has increased as a means of stabilizing patients before long-term assist device implantation or cardiac transplantation. However, even with these rescue therapies, mortality rates remain high. The overall survival rate with the AB5000® (ABIOMED, Inc.; Danvers, Mass) as a bridge to left ventricular assist device (LVAD) support or transplantation has been reported at 62% by Morgan and colleagues.³

To lower the risk of death that is associated with long-term device support or with the temporary assist devices mentioned above, we have used the TandemHeart® (CardiacAssist, Inc.; Pittsburgh, Pa), a percutaneous ventricular assist device (pVAD) for short-term cardiac assistance, in selected patients who were experiencing profound cardiogenic shock.⁴ This device improves hemodynamic performance and end-organ perfusion before long-term assist device implantation.⁵ Herein, we report the use of the TandemHeart for temporary mechanical support as a bridge to implantation of a long-term axial-flow LVAD. The aim of this approach is to reduce the risks involved in subsequent surgery and implantation of a long-term assist device.

Patients and Methods

From February 2005 through February 2007, we treated 9 end-stage heart-failure patients (5 men, 4 women; mean age, 37.7 yr; range, 17–65 yr) who presented in acute refractory hemodynamic decompensation and received the TandemHeart percutaneous pump as a bridge to an implantable axial-flow pump. Patients were considered to be in refractory acute cardiogenic shock if they had a low cardiac index (<2.0) and signs of end-organ dysfunction (hepatic, renal, or both) despite maximal inotropic therapy and IABP support. If cardiac catheterization showed no evidence of significant coronary artery disease, a diagnosis of nonischemic cardiogenic shock was established. Prognostic scores, as described by Oz and colleagues,⁶ were calculated before and after TandemHeart insertion to monitor the relative risk for patients before subsequent implantation of a long-term axial-flow device. This investigation was approved by our center's institutional review board, and the patients provided written informed consent.

The insertion technique and management of the TandemHeart pVAD have been described previously.⁷ The inflow cannula is inserted into the femoral vein and is advanced across the interatrial septum into the left atrium; the outflow cannula returns oxygenated blood to the femoral artery. This low-prime-volume (10-cc) centrifugal pump can deliver blood flows of up to 5 L/min.

The TandemHeart was surgically explanted in all 9 patients at the time of LVAD insertion, and repair of the femoral vessels was surgically performed. Continuity of the atrial septum (the access site for the TandemHeart insertion) was restored surgically in 4 patients and percutaneously in 1 patient. In the other 4 patients, the atrial septum did not need repair.

Results

Percutaneous support times ranged from 1 to 22 days (mean, 5.9 d) (Table I). Seven patients had nonischemic cardiomyopathy, 1 had familial cardiomyopathy, and 1 had postpartum cardiomyopathy.

TABLE I. Patient Characteristics and Outcomes

All 9 patients were in refractory cardiogenic shock and were receiving maximal inotropic therapy, and 8 of the 9 were on IABP support before TandemHeart insertion. One patient had a pVAD inserted without prior IABP insertion. The mean left ventricular ejection fraction was 0.13 (range, 0.0–0.25 [0.0 implies a patient in active cardiopulmonary resuscitation]). At the time of TandemHeart insertion, two thirds of the patients were on mechanical ventilatory support. In addition, 3 patients were receiving cardiopulmonary resuscitation. The mean cardiac index—which before TandemHeart implantation was 1.02 L/(min·m²) with a range of 0.0–1.8 L/(min·m²) (0.0 L/[min·m²] implies patient in active cardiopulmonary resuscitation)—improved to 2.97 L/(min·m²) with a range of 2.2–4.0 L/(min·m²) during TandemHeart support. Mean TandemHeart flow was 3.3 L/min (range, 2.55–3.67 L/min). The patients' renal function improved significantly after TandemHeart placement. However, liver function showed only nonsignificant improvement (Table II). All Tandem Heart insertion procedures were performed in the cardiac catheterization laboratory by the same team. All clinic patients were evaluated for cardiac transplantation and were placed on the transplant list. If donor hearts did not become available in a timely fashion, long-term LVADs were implanted.

TABLE II. Mean Renal and Liver Function Test Results Before and After TandemHeart Placement

Six patients received a HeartMate® II left ventricular assist system (Thoratec Corporation; Pleasanton, Calif), 2 patients a Jarvik 2000® Flowmaker (Jarvik Heart, Inc.; New York, NY), and 1 patient a DeBakey VAD® (MicroMed Cardiovascular, Inc.; Houston, Tex). Suitable donor hearts became available for 3 patients, who successfully underwent cardiac transplantation at 14, 258, and 463 days. Five patients are still being supported by axial-flow pumps (mean duration, 556.0 d; range, 366–958 d). One patient died of complications unrelated to the axial-flow device after 587 days of support. Follow-up (Table I) was completed for all patients (mean duration, 18.4 mo; range, 12.0–31.4 mo).

Mean results of the risk factor selection scale (RFSS, developed by Oz and colleagues⁶) were 6.7 on hospital admission and 2.6 after percutaneous support but before LVAD insertion. The scale distribution of all our patients indicated that 67% would have died, had they undergone immediate implantation of a long-term device.

Discussion

We used the TandemHeart percutaneous device as a rescue treatment for these 9 heart-failure patients, who were in refractory cardiogenic shock upon arrival at the hospital. The TandemHeart is the only percutaneous rescue therapy available for patients in refractory cardiogenic shock who are too unstable to receive long-term implantable assist devices immediately. The TandemHeart can provide up to 5.0 L/min of continuous cardiac output support during and after intervention at the catheterization laboratory. Blood flows provided by the TandemHeart are sufficient to maintain both coronary and peripheral perfusion.

Attempts have been made to use other devices, such as ECMO, the AB5000, and the CentriMag® (Levitronix® LLC; Waltham, Mass) to provide such immediate short-term cardiac support. However, these devices require invasive surgical implantation. The TandemHeart device can provide adequate circulatory support, regardless of a patient's native heart rhythm. Patients have been supported with a pVAD for up to 2 weeks, and reported complications have been minimal.⁷

The immediate use of implantable long-term mechanical pumps in patients in refractory cardiogenic shock inevitably increases the risks of morbidity and death. Implantation of intracorporeal assist devices for long-term support in such ill patients usually requires the use of cardiopulmonary bypass (CPB), which has negative effects. The length of time that a patient spends on CPB depends on a multitude of factors: primarily on whether the patient has had a previous sternotomy and on which type of device is being implanted. Although the HeartMate II usually also requires CPB for implantation,⁸ the time on CPB is shorter. Frequently, the Jarvik 2000 can be implanted “off-pump”^9,10 or with the aid of CPB only for insertion of the device into the ventricle, thus minimizing CPB trauma.

Various means of temporary mechanical support have been used to improve the condition of patients in refractory cardiogenic shock, so that they can be bridged to an LVAD or to heart transplantation. However, patient outcomes with these different temporary circulatory support devices have varied.^3,11

Pagani and colleagues¹¹ reported on a series of 60 patients who were evaluated for ECMO support for cardiac failure. Of the 25 patients who were placed on ECMO, only 8 survived to undergo LVAD insertion. Of the 30 patients who underwent LVAD implantation without ECMO support, 23 underwent successful cardiac transplantation. Three patients who were placed on extracorporeal ventricular assistance were also successfully bridged to transplantation. The 1-year actuarial survival rate from the initiation of extracorporeal support was 36% in the ECMO group. In another study performed by Pagani and associates,² the results showed marginal improvement in the RFSS (risk factor selection scale) scores of patients who received ECMO support (4.1 ± 1.2 to 3.5 ± 1.6).

In a recent series reported by Hoefer and coworkers,¹² ECMO support was used to treat cardiogenic shock in 131 patients. Among the 24 (21%) of these patients who were bridged to an LVAD, the overall mortality rate was 42%; only 14 patients in this subgroup survived, and 11 underwent successful transplantation.

More recently, Anderson and co-authors¹³ reported the use of the short-term extracorporeal BVS®5000 (ABIOMED) in patients who later received the surgically implanted AB5000. Only 42% of patients survived to recovery or to their next therapy: 61% of the survivors were weaned from support, 33% underwent transplantation, and 6% received a device for destination therapy.

Percutaneously inserted mechanical pumps were initially used during high-risk percutaneous coronary interventions, but their use has since expanded to different patient populations. The surgically inserted Impella® circulatory support system (ABIOMED) has been used successfully to treat postcardiotomy low-output syndrome, to sustain patients with acute myocarditis, and as a bridge to heart transplantation.^14,15 (A newer system configuration, in use only in Europe, enables percutaneous insertion of a larger version of the Impella that is capable of providing flows of up to 5 L/min.) The TandemHeart has been used in the treatment of cardiogenic shock and acute myocarditis and as a bridge to cardiac transplantation.^4,16–18

In the current series of 9 patients, we used the TandemHeart as a bridge to an axial-flow pump for long-term support in patients who were experiencing refractory cardiogenic shock and could not be weaned from TandemHeart support. None of the patients described here experienced right ventricular failure, and there was no incidence of limb ischemia. During standard LVAD implantation, native atrial septal defects (ASDs) need to be closed. In our series, the ASD was surgically closed from the right atrium when patients underwent median sternotomy for LVAD insertion (n=4). In non–median-sternotomy approaches to LVAD implantation, the iatrogenic ASD created by the TandemHeart is usually not closed unless evidence of right-to-left shunting is present. In these cases, the ASD can be closed percutaneously in the cardiac catheterization laboratory by interventional cardiologists. We used this technique in 1 patient. The remaining 4 patients did not require repair of iatrogenic ASD.

TandemHeart insertion is a safe and exceptional life-saving procedure. In all of our patients, successful cannulation was achieved by the transseptal approach. The blood flows provided by the TandemHeart are high (up to 5 L/min) and, therefore, uniformly improve end-organ function and overall condition in this severely ill patient population. This device allows time for proper long-term decision-making regarding patient management while minimizing preoperative risk factors in patients who are in refractory cardiogenic shock if implantation of a long-term device is indicated. In all 9 patients described here, the use of a temporary percutaneous pump as a bridge enabled the safe, successful implantation of a long-term axial-flow device, and the patients were all alive at 1-year follow up.