Clinical situations demanding weaning from long-term ventricular assist devices

Transcription

1 European Journal of Cardio-thoracic Surgery 26 (2004) Clinical situations demanding weaning from long-term ventricular assist devices Abstract C. Schmid a, *, C. Etz a, H. Welp a, M. Rothenburger a, H. Reinecke b, M. Schäfers c, Ch. Schmidt d, H.H. Scheld a a Department of Thoracic and Cardiovascular Surgery, University Hospital Münster, Albert-Schweitzer-Str. 33, Münster, Germany b Department of Internal Medicine C Cardiology, University Hospital Münster, Münster, Germany c Department of Nuclear Medicine, University Hospital Münster, Münster, Germany d Department of Anesthesia and Operative Intensive Care Medicine, University Hospital Münster, Münster, Germany Received 17 May 2004; received in revised form 7 July 2004; accepted 8 July 2004; Available online 20 August 2004 Objective: Ventricular assist devices are increasingly used to treat patients with acute or chronic end-stage heart failure. We report on circumstances, exemplified on four cases, where a surprisingly favorable clinical course of the patients ultimately demanded early explantation of the device, which was not anticipated prior to its implantation. Methods: The four patients were provided with implantable (Micromed BeBakeye, Incore) and external pneumatically driven (Thoratece, Excore) devices under emergency conditions and were listed for heart transplantation. Results: All four patients had an unexpected recovery of myocardial pump function. After careful diagnostic evaluation, all device components were completely removed without extracorporeal circulation. No stepwise weaning protocol was employed. Conclusions: Weaning patients from ventricular assist devices after recovery of myocardial pump function can become necessary. Diagnostic evaluation and the implementation of a weaning protocol is still a matter of debate, while complete surgical removal of all device components without extracorporeal circulation is possible with a low risk. q 2004 Elsevier B.V. All rights reserved. Keywords: Mechanical circulatory support; VAD; Heart failure; Heart transplantation 1. Introduction Ventricular assist devices (VADs) are increasingly used to treat patients with acute or chronic end-stage heart failure. The predominant indication is to bridge these patients until an appropriate donor organ for heart transplantation is available. The second intention is to provide patients not amenable to heart transplantation a therapeutic alternative [1,2]. Recently, the REMATCH study has proven, that permanent device implantation may allow patients with end-stage heart disease superior survival and quality of life as compared to medical therapy alone [3]. A third option, which is currently not yet fully approved by the FDA but repeatedly used outside the US is the bridge-torecovery setting. Frequently, the intention to wean patients later from their long-term ventricular assist device is expressed already prior to implantation of the device. Accordingly, VAD patients undergo repetitive hemodynamic studies and are included in special weaning protocols. In some institutions, VAD patients are even trained for recovery, i.e. they are treated with drugs to strengthen their myocardial pump activity. We report on circumstances, exemplified on four cases, where the clinical course of the patients awaiting heart transplantation demanded early explantation of the device, which was not anticipated prior to its implantation (Table 1). 2. Case 1 delayed recovery after severe acute rejection in a heart transplant patient * Corresponding author. Tel.: C ; fax: C address: schmid@uni-muenster.de (C. Schmid). A 15-year-old patient who had undergone heart transplantation for dilative cardiomyopathy was referred to our hospital 4 months after surgery with severe cellular rejection /$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi: /j.ejcts

2 C. Schmid et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Table 1 Patient data Case Age Gender Cause of heart failure Implanted device Complication Explantation at (years) 1 15 M Acute rejection (ISHLT grade IV) Thoatec BiVAD POD M Acute myocarditis Micromed DeBakey Cerebral bleeding POD 133 LVAD 3 42 M Dilative cardiomyopathy Incor LVAD Hypertension POD F Acute myocardial infarction Excor LVAD POD 125 (ISHLT grade IV) triggered by an acute gastrointestinal infection (Fig. 1). Within 12 h, the patient developed a hemodynamical compromising pericardial effusion and progressive cardiac failure despite corticosteroid bolus treatment. Being dependent on high dose catecholamines the patient developed a cardiovascular collapse and had to be connected to extracorporeal membrane oxigenation (ECMO) during ongoing external heart massage. After hemodynamic stabilization, a polycloncal anti-human-t cell antibody therapy was started and continued for 4 days according to the institution s immunosuppressive treatment protocol. However, the patient s condition did not get better, and daily transesophageal echocardiography studies did not show any sign of improvement of myocardial pump function. Audit members of Eurotransplant rejected a high urgency request for retransplantation. Due to the young age of our patient, we decided to implant a biventricular assist device (Thoratec BiVAD, Thoratec Inc., Berkeley, CA, USA). The postoperative course was prolonged but uneventful. Weekly controls of transesophageal echocardiography revealed a steadily improving cardiac function; complete myocardial recovery of the graft was noted 9 weeks after the implant of the BiVAD. The decision to wean the patient from the device was based on normal findings during echocardiography, and an endomyocardial biopsy indicating a vanished rejection (ISHLT grade I). No special weaning program to downgrade the pump parameters was initiated, except that the pump had been reduce by 30% prior to explantation, which was performed on the beating heart after 70 days of mechanical support. Postoperatively the patient did not need catecholamines, and could be extubated the same day. Meanwhile, 2 years after BiVAD explantation, the young patient is doing well with an excellent quality of life. the operating room under cardiac resuscitation conditions. Intraoperatively a cherry-size thrombus was removed from the left ventricular cavum. The postoperative course was uneventful. And the patient rapidly recovered. On day 130 after LVAD implantation, the patient developed intracerebral bleeding, became comatous, and had to undergo neurosurgery (occipital drainage) to lower intracranial pressure (Fig. 2). The patient survived with a mild neurological impairment, but was fully orientated right after surgery. Echocardiography demonstrated an almost normal LV function. To reduce the risk of another intracranial bleeding and also because of evolving signs of pocket infection, the decision was made to explant the device. The pump rate was stepwise lowered to minimum flow, which was well tolerated by the patient. Explantation took place 133 days after placement. The LVAD was completely removed via median sternotomy without extracorporeal circulation with an uneventful postoperative course. Control echocardiography showed a moderately decreased ventricular function as compared to pre-explant studies. After neurological rehabilitation, the patient could be discharged home. 4. Case 3 recovery after LVAD implantation for dilative cardiomyopathy A 42-year-old patient presenting with dilative cardiomyopathy was admitted to our hospital with fatal congestive 3. Case 2 cerebral bleeding during LVAD support A 29-year-old patient suffering from an excessively dilated heart with poor contractility after acute myocarditis was submitted to our hospital for emergency left ventricular assist device (LVAD) implantation. An axial flow pump (DeBakey-NASA-LVAD, Micromed Inc., Houston, USA) was implanted on an emergency basis for mechanical circulatory support when the asystolic patient entered Fig. 1. Endomyocardial biopsy demonstrating severe cellular rejection (case 1).

3 732 C. Schmid et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Fig. 2. Cranial computed tomography exhibiting severe cerebral bleeding (case 2). heart failure (NYHA IV) for implantation of a LVAD. The patient was intubated and mechanically ventilated, and inotrop dependent. Echocardiography demonstrated extremely poor left ventricular function with an ejection fraction of 15%, end-organ function was already impaired. An axial flow LVAD (INCOR, BerlinHeart AG, Berlin, Germany) was implanted on an emergency basis. The inflow cannula was fixed at the left ventricular apex, the outflow conduit was anastomosed to the proximal ascending aorta. The pathologic examination of the intraoperatively removed apical myocardial tissue ruled out active myocarditis, but demonstrated DNA of Parvo B19 virus as well as a CD3 positive lymphocyte infiltration. The patient could be extubated the following day and made an uneventful and fast recovery, apart from transient neurologic symptoms on day 58 and 68 (normal cranial computed tomography). Two months following device implantation, the patient became progressively hypertensive and finally required quadruple antihypertensive treatment. Echocardiography demonstrated complete recovery of myocardial pump function. Cardiac catheterization confirmed ventricular contractility (left ventricular ejection fraction 75%) and showed a normal coronary pattern. To assess Fig. 3. Improvement of left ventricular pump function (case 3). endothelial-dependent perfusion capacity, H 2 O-PET scans were performed after 90 days of mechanical circulatory support, revealing normal myocardial perfusion under non-stress conditions and a reduced endothelium-dependent perfusion reserve (0.68) (with increased perfusion during reduced pump action) [4]. SPECT scans (Tc-99m-Tetrofosmin) demonstrated a left ventricular ejection fraction of 80%. At least, a left ventricular biopsy was taken, which demonstrated normal findings. Based on the aforementioned findings, the device was explanted on day 97, without stepwise reduction of pump parameters. All components of the device were explanted on the beating heart and without extracorporeal circulation; no remnants of the inflow or outflow conduit were left in place. Follow-up controls days and 10 weeks after LVAD explantation demonstrated preserved myocardial pump function. Systolic left ventricular function remained normal, as were the results of the H 2 O-PET. Now, 6 months later, the patient is doing well and still has his remarkable recovery of myocardial function enabling excellent quality of life (Figs. 3 and 4). Fig. 4. Pre-explant angiography, demonstrating excellent systolic contraction properties (marked inflow cannula at left ventricular apex) (case 3).

4 C. Schmid et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Case 4 recovery after LVAD implantation for acute myocardial infarction A 39-year-old female patient suffered cardiovascular collapse with ventricular fibrillation and external heart massage. After resuscitation, she was transferred to a local hospital, where an acute myocardial infarction with occlusion of the left anterior descending artery (LAD) was diagnosed. The vessel was reopened via PTCA and provided with a stent. Despite this procedure, she remained in critically low out put and was referred to our institution, where an intraaortic balloon pump (IABP) was inserted as first measure. As the IABP did not stabilize the patient, it was decided to place her on extracorporeal membrane oxygenation via the femoral vessels, since the neurological status was unclear. Within the following 2 days, sedation was stopped to allow her to wake up and to prove neurological integrity, i.e. to exclude significant neurological damage. Repeat echocardiography studies consistently demonstrated severely impaired pump function with a left ventricular ejection fraction!20%. Kidney and liver function steadily worsened. On day 3, the ECMO was replaced by the pneumatically driven EXCOR system, which was inserted as an LVAD. A large cannula was inserted in the left ventricular apex of the relatively small heart; the outflow conduit was anastomosed to the ascending aorta. During the following weeks, the patient continuously improved with regard to end-organ function. On postoperative day 16, the patient was transferred to the intermediate care unit. One week later, puncture of a pleural effusion led to a hematothorax, which had to be operated upon. Nevertheless, the patient s physical condition further improved, so that she could spend Christmas and New Years Eve at home. Discharge with ambulatory care of the patient was denied for medical and psychological reasons. Routine control of myocardial pump function at 3 months demonstrated a left ventricular ejection fraction of 68%, which was confirmed by left heart catheterization. PET scans depicted vital myocardium except in the left ventricular distal anterior wall and apex (prior infarction). With this remarkable recovery, the LVAD was explanted on day 125. All components of the device were removed without the use of a heart-lung machine. The pump output of the device was set to minimum action with the onset of surgery and a catecholamine therapy was initiated. After explantation, dobutamine and epinephrine were carefully weaning over a 10-day period. Postoperative echocardiography consistently demonstrated preserved left ventricular function (ejection fraction 50 60%), as well as a normal endothelial perfusion reserve in the H 2 O-PET scan (!1.37). 6. Discussion Weaning from long-term ventricular assist devices is not new, but is gaining more and more consideration in heart failure programs due to the extreme lack of appropriate donor organs for heart transplantation and due to the favorable results reported in the few expert centers. Recovery of myocardial pump function, however, is not uniform. There are patients with chronic heart failure such as dilative cardiomyopathy, and there are patients who underwent VAD placement after acute deterioration following acute myocardial infarction or myocarditis. These patients may behave rather different with their potential to recover including the time frame, histological changes and hemodynamic improvement. However, in most heart failure centers, all LVAD patients with severely impaired ventricular contractility are considered bridge-to-transplant patients regardless of underlying heart disease. Moreover, myocardial recovery and weaning protocols are more exemption than routine as many questions including patient selection, diagnosis of adequate myocardial recovery, timing of explant surgery, surgical techniques, and longterm results are still unresolved. The first large series of successful weaning from a longterm device was reported by Mueller et al. from the Berlin group [5]. They reported on 17 patients with non-ischemic idiopathic dilated cardiomyopathy in NYHA class IV. Apart from severely impaired cardiac function, they all tested positive for anti-beta1-adrenoreceptor autoantibodies and had also histological evidence of myocardial fibrosis. Five of these patients with significant recovery of pump function could be weaned after a support interval days. Interestingly, anti-beta1-adrenoreceptor autoantibodies disappeared thereafter. Yacoub et al. have recently developed a strategy of combining LVAD support with pharmacologic therapies to produce maximal reverse remodeling followed by the induction of physiologic cardiac hypertrophy using clenbuterol, a selective beta2-adrenergic receptor agonist (the Harefield protocol) [6]. In contrast to the mentioned experiences in Berlin and Harefield, where patients were at least in part electively included in the respective protocols to remove ventricular support, we present four cases, where we were urged to remove the devices again for an unexpected recovery of myocardial pump function. The first patient had received the BiVAD for intractable heart failure associated with a severe rejection episode. According to the biopsy findings, recovery seemed so unlikely that we immediately decided to implant the BiVAD. In case of less myocardial damage, one may argue that a short-term device might have been preferred. However, the hemodynamic follow-up demonstrated recovery of the patient and his ventricular pump function only after several weeks, which proves in our mind that the decision was right. With the astonishing late recovery, we could no longer justify aiming at retransplantation and had to remove the device. The second patient suffered from cerebral bleeding, a known and quite unfavorable complication after VAD implantation. Being confronted with the risk of further bleeding complications,

5 734 C. Schmid et al. / European Journal of Cardio-thoracic Surgery 26 (2004) the only way to have the patient survive was seen in an urgent explantation of the device. The third patient was referred for device implantation from another university hospital for decompensated dilative cardiomyopathy. The patient fulfilled all criteria for implantation and therefore underwent the procedure. To our surprise, the patient recovered, an active myocarditis could never be proven. When the patient became hypertensive we started to intensively investigate recovery of myocardial pump function, which led to removal of the device. The fourth patient had been connected to ECMO with no other option that to die. Due to her young age, she underwent placement of a pneumatically driven LVAD and fortunately recovery. When hemodynamic control studies revealed sufficient left ventricular function, the anticipated heart transplantation could be no longer justified. The clinically most important question in these situations was the status of myocardial recovery. According to the aforementioned, predicting myocardial recovery is difficult. Usually, serial echocardiograms are used to assess left ventricular function and dimensions over time. However, it is well known, that predictions based on echocardiography can fail as it can be difficult to differentiate between the effects of unloading and recovery [7]. Dobutamine stress echocardiography has been found to demonstrate higher cardiac index, ejection fraction and smaller left ventricular diameters in patients which could be successfully weaned from the device [8]. In our patients we did not expect recovery. Accordingly, no early serial hemodynamic studies were undertaken. No patient was included in a weaning protocol of a step-by-step reduction of pump activity to follow gradually improving myocardial pump function [9]. Instead, we were surprised by a severe complication (case 2) or an obvious satisfactory recovery (cases 1,3,4). The time point of recovery was later than in respective reports by other centers. Therefore, these patients underwent immediate left heart catheterization (and coronary angiography) with the device at full and minimum pump flow, as well as right heart catheterization, apart from repeat echocardiography. The decision to explant a device was based only on the findings of left heart catheterization and echocardiography, i.e. whether the heart demonstrated good overall myocardial contractility, and was able to maintain cardiac output during minimum pump flow (Fig. 5). Nuclear imaging of endothelial function was included as a new investigational tool in the last two patients and it seems that this new diagnostic tool may provide helpful information in the near future [4]. A related prospective study is ongoing. The surgical technique had been frequently debated. In most cases reported, the inflow cannula has been left in situ, being transected and occluded [10]. Is has been argued to avoid the use of extra-corporeal circulation and to avoid resternotomy. In our experience, it is technically no problem to perform redo sternotomy and to remove all device components on the beating heart. Following sternotomy, Fig. 5. LV ejection fraction before and after LVAD therapy. one can follow the cannulas to their insertion and free the left ventricular apex and the ascending aorta from adhesions (as well as the pulmonary artery in case of RVAD implantation). The right atrial wall is always prepared to be able to institute cardiopulmonary bypass if necessary. After mobilization of the left ventricular apex, two deep sutures are placed around the inflow cannula. The knots of the inflow cannula are cut and the latter is pulled out, controlling bleeding digitally while tying the two sutures. Additional sutures help to optimize hemostasis. Extracaorporeal circulation has never been necessary, and no complications occurred in our hands. Since no stepwise reduction of pump flow and consecutive adaptation of the heart were performed prior to LVAD explantation, patients were kept on catecholamines after surgery. Replacing the intraoperatively administered nitric oxide by inhaled iloprost and administering dobutamine or low dose epinephrine, the patients myocardial pump function was allowed to gradually accommodate to the new hemodynamic situation. Catecholamines were replacement by ACE inhibitors and b-blockers within one to two weeks after hemodynamic stabilization. In conclusion, weaning patients from ventricular assist devices after recovery of myocardial pump function can become necessary in selected cases. Diagnostic evaluation and the implementation of a weaning protocol is a frequently considered routine but not possible or necessary in all cases. Complete surgical removal of all device components without extracorporeal circulation is possible with a low risk. References [1] Jurmann MJ, Weng Y, Drews T, Pasic M, Hennig E, Hetzer R. Permanent mechanical circulatory support in patients of advanced age. Eur J Cardiothorac Surg 2004;25: [2] Pae Jr WE, Koerfer R, El-Banayosy A, Arusoglu L, Hetzer R, Weng Y, Jurmann M, Vigano M, Rinaldi M, Pavie A, LaPrince P, Wolner E, Wieselthaler G, von Segesser L, Wahlers T, Franke U, Ruzevich-Scholl S, Swartz MT, Fey O, Reeders M, Lewis J. European Lion Heart Clinical Utility Baseline Study (CUBS). Circulation 2003; 108:IV-365. [3] Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W, Long JW, Ascheim DD, Tiemey AR, Levitan RG, Watson JT, Meier P, Ronan NS, Shapiro PA, Lazar RM, Miller LW, Gupta L, Frazier OH, Desvigne-Nickens P, Oz MC, Poirier VL.

6 C. Schmid et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) Study group long-term mechanical left ventricular assistance for end-stage heart failure. N Engl J Med 2001;345: [4] Schäfers KP, Spinks TJ, Camici PG, Bloomfield PM, Rhodes CG, Law MP, Baker CS, Rimoldi O. Absolute quantification of myocardial blood flow with H(2)(15)O and 3-dimensional PET: an experimental validation. J Nucl Med 2002;43: [5] Muller J, Wallukat G, Weng YG, Dandel M, Spiegelsberger S, Semrau S, Brandes K, Theodoridis V, Loebe M, Meyer R, Hetzer R. Weaning from mechanical cardiac support in patients with idiopathic dilated cardiomyopathy. Circulation 1997;96: [6] Hon JK, Yacoub MH. Bridge to recovery with the use of left ventricular assist device and clenbuterol. Ann Thorac Surg 2003;75: S36 S41. [7] Houel R, Vermes E, Tixier DB, Le Benhaiem-Sigaux N, Loisance DY. Myocardial recovery after mechanical support for acute myocarditis: is sustained recovery predictable? Ann Thorac Surg 1999;68: [8] Khan T, Delgado RM, Radovancevic B, Torre-Amione G, Abrams J, Miller K, Myers T, Okerberg K, Stetson SJ, Gregoric I, Hernandez A, Frazier OH. Dobutamine stress echocardiography predicts myocardial improvement in patients supported by left ventricular assist devices (LVADs): hemodynamic and histologic evidence before LVAD explantation. J Heart Lung Transplant 2003;22: [9] Hetzer R, Müller J, Weng Y, Wallukat G, Spiegelsberger S, Loebe M. Cardiac recovery in dilated cardiomyopathy by unloading with a left ventricular assist device. Ann Thorac Surg 1999;68: [10] Hetzer R, Müller JH, Weng Y, Meyer R. Dandel M Bridging-torecovery. Ann Thorac Surg 2001;71:S109 S113.