Robotic minimally invasive cell transplantation for heart failure Harald C. Ott, MD, a Johannes Brechtken, MD, b Cory Swingen, PhD, c Tanya M. Feldberg, BS, a Thomas S. Matthiesen, BS, a Samuel A. Barnes, BS, a Wendy Nelson, PhD, a and Doris A. Taylor, PhD, a Minneapolis, Minn C ardiac cell transplantation offers new opportunities as a potent therapeutic tool to improve left ventricular (LV) function and reverse postinfarction remodeling in isch- emic heart disease. Skeletal myoblasts (SKMBs) en- graft within infarcted myocardium, form myotubes, induce angio- genesis, and improve both diastolic and systolic LV function. 1 Bone marrow– derived mononuclear cells (BM-MNCs) likewise engraft, increase angiogenesis, and improve myocardial perfu- sion. 2 Both cell types have moved to clinical testing, and preclin- ical studies suggest that they could have synergistic functional benefits that argue for combined transplantation. 3,4 Intramyocar- dial injections are currently performed either percutaneously through an endoventricular or transvenous approach or surgically through a thoracotomy or sternotomy. We recently reported a video-assisted thoracoscopic technique to reduce invasiveness and perioperative risk of surgical cell delivery that was tested in uninjured swine hearts. 5 In the setting of heart failure (HF), me- chanical manipulation of the left ventricle both by means of stabilization and cell injection must be minimized to prevent hemodynamic compromise, arrhythmia, and ventricular perfora- tion. Robotically assisted cardiac surgery combines the advantages of minimal invasiveness and thoracoscopic access but adds a 3-dimensional view and 7 degrees of freedom that requires less cardiac manipulation than with the 2-dimensional view and limited freedom of motion of video-assisted thoracoscopic surgery. 6 We therefore propose a robot-assisted, beating-heart cell transplanta- tion technique for use in severe HF to increase safety, optimize targeting, and reduce procedural time. Procedure Description Eleven injured swine in which HF was previously induced by means of coronary occlusion and coronary embolism (left anterior descending coronary ar- tery, n = 9; circumflex artery, n = 2) under- went robot-assisted cell (n = 7) or vehicle (n = 4) injection by using the daVinci robotic system (Intuitive Surgical, Sunnyvale, Calif). During right single-lung ventilation and antiarrhythmic prophylaxis (amiodarone, 3 mg/kg; lidocaine, 1 mg/kg), we in- serted the camera port, 2 instrument ports, and an auxiliary port (Figure 1, A). After removal of the pericardial fat pad, we incised the pericardium along the sternal border, dissected pericardial adhesions, and created a triangular pericardial flap. We inserted the prefilled injection needle (27-gauge needle attached to 12-inch tubing; Saf-T E-Z Set, BD, Sandy, Utah) through the auxiliary port and injected a 7-mL cell suspension containing a combination of 2.9  10 8  5.9  10 7 autologous SKMBs and 1.1  10 8  6.8  10 6 autologous BM-MNCs (Figure 1, B) at 6 to 10 sites. SKMBs were iron oxide labeled, as previously described. 5 Viabil- ity at the time of injection was greater than 85%, and CD56 expression was greater than 80%. BM-MNCs were acutely isolated from bone marrow aspirate through Ficoll density gradient cen- trifugation. Injections were performed tangentially to minimize perforation risk and injectate backflow, covering a target area of 15 to 20 cm 2 . After cell delivery, the pericardial flap was readapted, all instrument ports were removed, and the left lung was expanded under visual control. We inserted a chest tube through the inferior instrument port, closed the port sites in 3 layers, and flushed the left thoracic cavity with 0.9% saline (200 mL). After removal of the chest tube, animals were extubated and recovered according to postoperative standards. Baseline magnetic resonance imaging (MRI) was performed 5 weeks after myocardial injury. Follow-up MRI was repeated at 4 and 7 weeks after cell/vehicle transplantation. Results Cell transplantation was completed successfully in 6 of 7 cases. Intractable ventricular fibrillation occurred in one animal during cell injection. No conversion to open chest surgery was necessary, and no other procedure-related complications occurred. Over the course of the study, single-lung ventilation time was reduced to a minimum of 23 minutes, and total anesthesia time was reduced to a minimum of 44 minutes. Cells were successfully transplanted into the apical, anterior, and lateral target regions of the left ventricle, including into thinned sections of the scar (target region wall thickness, 3-14 mm), without ventricular perforation. Postop- erative MRI studies confirmed retention of iron oxide–labeled cells in the apex (Figure 2, A) and lateral wall (Figure 2, B) up to 7 weeks after injection. Prussian blue staining of tissue sections showed engraftment of iron oxide–labeled myotubes in treated areas (Figure 2, C). Immunofluorescent staining for slow skeletal From the Center for Cardiovascular Repair, a the Division of Cardiology, b and the Department of Radiology, c University of Minnesota, Minneapolis, Minn. This work was supported in part by National Heart, Lung, and Blood Institute/National Institutes of Health awards to Dr Taylor (R-01 HL-63346, HL-63703). Received for publication Nov 19, 2005; revisions received Feb 1, 2006; accepted for publication Feb 21, 2006. Address for reprints: Doris A. Taylor, PhD, Center for Cardiovascular Repair, University of Minnesota, 312 Church St SE, BSBE 7, Minneapolis, MN 55455 (E-mail: dataylor@umn.edu). J Thorac Cardiovasc Surg 2006;132:170-3 0022-5223/$32.00 Copyright © 2006 by The American Association for Thoracic Surgery doi:10.1016/j.jtcvs.2006.02.017 Drs Swingen, Matthiesen, Nelson, Ott, Taylor, and Brechten (left to right) Brief Communications 170 The Journal of Thoracic and Cardiovascular Surgery ● July 2006