- Email: [email protected]
Results of a phase one study on robotically assisted myocardial revascularization on the beating heart
CARDIOVASCULAR
Results of a Phase One Study on Robotically Assisted Myocardial Revascularization on the Beating Heart J. W. Randolph Bolton, MD, and Jackie E. Connally, SFA/CST Carolina Heart Institute, Greenwood, South Carolina
Background. The objective of this study was to evaluate the feasibility of computer assisted myocardial revascularization on the beating heart. Methods. Ten patients underwent at least one robotically assisted internal mammary artery (IMA) anastomosis utilizing the da Vinci surgical system (Intuitive Surgical, Inc, Mountain View, CA) performed through an open incision as part of standard multivessel off pump revascularization. Following chest closure a selective IMA angiogram was performed to assess patency. Three month follow-up included a stress echocardiogram. Results. There were 12 anastomoses performed in 10 patients. The average age was 61 years with a mean ejection fraction of 56%. No patient required inotropic support. Eight of 10 patients were found to have fully patent IMA anastomoses by angiogram. One patient was noted to have an occluded left anterior descending coronary artery distal to the anastomosis and one had occlu-
sion at the anastomosis. Both patients had immediate manual revision of the anastomosis. One patient who required anastomotic revision experienced postoperative myocardial infarction and sternal wound infection requiring pectoralis flaps. Conclusions. Because robotic instrumentation is meant for closed chest procedures, there were major issues with positioning of the robotic arms in this study since the chest was open. Although two patients required anastomotic revision, there were no complications or technical failures related to the robotic system. Thus, based upon this study robotically assisted beating heart revascularization appears to be feasible, safe, and effective. Further evaluation will be necessary to determine the role of robotically assisted totally endoscopic coronary artery bypass on the beating heart in the United States. (Ann Thorac Surg 2004;78:154 – 8) © 2004 by The Society of Thoracic Surgeons
T
Material and Methods
otally endoscopic robotically assisted coronary artery bypass (TECAB) has been performed on the beating heart in some countries for a number of years. At the time of this study, however, there were no robotic devices approved for such use in the United States by the Federal Drug Administration (FDA). There were other concurrent studies investigating the use of robotics to facilitate TECAB on the arrested heart. The purpose of this study was to determine in a stepwise fashion the feasibility, applicability, and safety of robotically assisted coronary revascularization on the beating heart. This first phase study is intended to provide a first step in the logical progression in the future training program for those interested in pursuing the use of this technology. What follows are the results of this study and comments regarding the lessons learned during the development of the procedure.
Accepted for publication Jan 22, 2004. Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003. Address reprint requests to Dr Bolton, 303 W Alexander Ave, Ste E PO Box 3284, Greenwood, SC 29648; e-mail: rbolton@carolinaheartinstitute. com.
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
The safety and effectiveness of the da Vinci surgical system (Intuitive Surgical, Mountain View, CA) and endoscopic instruments was evaluated for use in myocardial revascularization on the beating heart following approval by the Institutional Review Committee and the FDA. In preparation for this phase-one investigational device exemption (IDE) study the authors underwent extensive laboratory training. To achieve the primary objective, ten subjects diagnosed with coronary artery disease underwent myocardial revascularization on the beating heart by sternotomy or minimally invasive direct coronary artery bypass (MIDCAB) in which the da Vinci surgical system was used to complete the distal internal mammary artery (IMA) anastomosis to the target coronary vessels. Potentially eligible subjects were screened for participation using prospectively established inclusion and exclusion criteria. Patients were considered suitable for inclusion if they were between the ages of 18 and 80 years and candidates for first time, nonemergent myocardial revascularization. Exclusion criteria included left ventricular ejection fraction (EF) less than 30%, severe peripheral vascular disease, acute stroke, renal failure, moderate to severe valve disease, or hemodynamic instability. Subjects meeting the appropriate criteria 0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.01.036
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
155
Table 1. Demographics Patient Number 1 2 3 4 5 6 7 8 9 10
Sex
Age Years
EF Percent
MI
Unstable Angina
Diabetes
PTCA Stent
Male Female Male Female Male Male Female Male Female Male
40 56 56 69 56 72 69 66 50 77
40 40 60 55 70 60 60 60 50 60
Yes Yes No No No No No No Yes No
Yes Yes Yes Yes Yes No Yes Yes Yes Yes
No No No No No Yes No Yes No No
Yes No Yes No No No No No No Yes
EF ⫽ left ventricular ejection fraction;
MI ⫽ myocardial infarction;
were invited to participate in the trial and required to give informed consent before enrollment (Table 1). Of the twelve patients invited to participate in the study, ten completed the study with twelve anastomoses being accomplished with robotic assistance. The robotic assistance was abandoned nonemergently in two patients. One patient had a planned bilateral MIDCAB with robotically assisted IMA harvest, which was abandoned during IMA takedown because of ongoing bleeding due to clopidogrel (Plavix). The other patient was abandoned following robotically assisted coronary arteriotomy due to inadequate visualization secondary to bleeding due to a large septal branch. Both patients had an uneventful recovery. The role of the surgical assistant was to change instruments, insure safety, and monitor observational real-time quality of the robotically assisted anastomosis. Successful use of the da Vinci surgical system was defined as a successful perioperative course and postsurgical, predischarge graft patency as determined using coronary angiography. Following completion of all other necessary bypass grafts, the previously prepared da Vinci patient side robot was positioned for the IMA anastomosis. The stabilizer was manually placed in the first eight patients and robotically placed in the final two patients. Following robotically assisted placement of silastic vessel occluders, the left anterior descending coronary artery (LAD) was opened and a running anastomosis was robotically performed using 6-0 monofilament polymide suture (Peters Cardionyl; Peters, Inc, Paris, France). At conclusion of the anastomosis the suture was robotically tied and the IMA bulldog occluder released. Hemostasis was ensured and the silastic LAD occluders removed. The stabilizer was taken away and the patient side robot withdrawn from the field. The chest was then closed in routine fashion and selective cardiac catheterization performed (by the radial arterial catheter site as often as possible) to assess patency while the patient remained intubated and anesthetized. Information was gathered to describe various operative characteristics including assisted ventilation time, intensive care unit, and total hospital stay duration, as
PTCA ⫽ percutaneous transluminal coronary angioplasty.
well as observed perioperative and postoperative (⬍30 days) morbidity and mortality. Additional clinical testing (eg, angiographic or echocardiographic studies) was conducted and documented at the following study phases: intraoperatively prior and subsequent to coronary artery bypass grafting, immediately postoperatively, at two weeks following discharge, and three months following surgery. Evaluation of the da Vinci surgical system in this study was specifically limited to completion of the anastomosis. Although the da Vinci surgical system, at times, was used for other portions of the procedure for which it is already FDA approved (eg, IMA mobilization) information characterizing the use of the device for such procedures was not collected in detail.
Results The study group consisted of ten patients (six male and four female) with twelve robotically assisted IMA anastomoses. (Table 2) There were nine IMA anastomoses to the LAD and three to the diagonal artery with two sequential bypasses to the diagonal artery and the LAD. Nine patients had multivessel off pump coronary artery bypass (OPCAB) by sternotomy with the robotically assisted anastomosis being performed as the final graft(s) and one patient had the robotically assisted anastomosis to the LAD by MIDCAB. There was an average of 2.5 grafts per patient (range, 1 to 4). The average age was 61.1 years (40 to 77 years) with a mean EF of 55.5% (40% to 70%). The average time for completion of the robotically assisted anastomoses, including setup and removal of stabilizing devices, was 40 minutes (31 to 60 min) with mean operating time of 200 minutes. Different stabilizing platforms were evaluated, including the Genzyme straight arm immobilizer (one), Genzyme flexible arm immobilizer (six), Genzyme straight arm MIDCAB retractor (one) (Genzyme Biosurgery, Cambridge, MA), and the Medtronic Cardioboy (Medtronic, Inc, Minneapolis, MN) endoscopic suction stabilizer with irrigator (two). The postoperative length of stay (LOS) averaged 2.8 days (1 to 9 days) with only one patient requiring a LOS of more than three days (patient 6). On immediate
CARDIOVASCULAR
Ann Thorac Surg 2004;78:154 – 8
CARDIOVASCULAR
156
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
Ann Thorac Surg 2004;78:154 – 8
Table 2. Results Patient Number
Total Anast
Robotic Anast
Anast Minutes
Postop LOS
1
2
2
2
Imp
None
2 3
3 2
1 1
39-Diag 31-LAD 40 37
2 3
Nml N/Aa
4 5
2 3
1 2
2 2
N/Aa Nml
6
3
1
40 60-Diag 29-LAD 53
None Occluded LAD (Anast revised) Transfusion Thorocentesis None IMA spasm
9
Apical akinesis
7 8 9 10
3 1 4 2
1 1 1 1
32 37 40 39
3 1 2 2
Nml Nml Nml Imp
a
Stress Echo
Complications
Occluded IMA (Anast revised) Myocardial infarction Atrial fibrillation Transfusion Thorocentesis Sternal infection None None None None
Refused echo.
Anast ⫽ anastomosis; Diag ⫽ diagonal artery; Imp ⫽ improved; artery; LOS ⫽ length of stay in days; Nml ⫽ normal.
LAD ⫽ left anterior descending artery;
IMA ⫽ internal mammary
postoperative angiography ten of the twelve anastomoses revealed no stenosis. However, although angiography revealed no anastomotic stenosis and good retrograde filling of the LAD and its diagonal branches, one patient (patient 3) had an LAD occlusion distal to the anastomosis. This was associated with minor electrocardiographic (ECG) changes, which resolved upon revision of the anastomosis. Cardiac enzymes were negative. Interestingly, postoperatively this patient was found to have deficiencies of proteins S and C. A second patient (patient 6) had total anastomotic occlusion of the left internal mammary artery (LIMA). Both of these patients were immediately returned to the operating room for revision of the anastomosis with a standard hand sewn technique. These two patients also represented the only patients requiring transfusion postoperatively (two units of packed red blood cells each). Nine patients were free of major complications; however, the patient with the occluded LIMA experienced a postoperative myocardial infarction (MI), atrial fibrillation, and eventual sternal wound infection requiring pectoralis flap closure. No patient experienced hemodynamic instability or required inotropic support and there was no mortality. All patients were extubated within one and one half hours after arrival to the single stay intensive care unit with eight patients being extubated in the operating room or cardiac catheterization laboratory following angiography. Stress echocardiogram performed at three month follow-up was normal or improved in seven of eight patients. The stress echocardiogram on patient 6 revealed a new finding of apical akinesis without the onset of angina following the postoperative MI. Two patients (patients 3 and 4)
refused the stress echocardiogram because they felt improved and physically unlimited since their surgery. There were three minor complications. Two patients (patients 3 and 6) had pleural effusions requiring thoracentesis. There was one episode of IMA spasm associated with ischemic chest pain and sudden ST segment elevation several hours postoperatively (patient 5) which responded to medical therapy with resolution of chest discomfort and ECG changes. Cardiac enzymes were negative. There were no complications or technical failures relative to the robotic system. One anastomosis was revised robotically after the suture broke while being tied.
Comment Aside from one episode of IMA spasm, all complications occurred in the two patients who required revision of the anastomosis. In regard to the first of these patients (patient 3) who had what appeared to be an LAD occlusion distal to the anastomosis, spasm of the LAD should also be considered since there was no clot found during revision of the anastomosis. This has been described by Bonatti and colleagues [1] to have occurred in eighteen of twenty-three patients following robotically assisted anastomoses. In addition, based upon the excellent retrograde filling of the LAD at angiogram, it is possible that had flow testing alone been performed to assess patency it may very likely have been adequate and the anastomosis would not have been revised [2, 3]. The second patient, a diabetic with small, calcified vessels (patient 6) accounted for six of the eight compli-
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
cations. Postoperative angiography revealed a totally occluded LIMA at the anastomosis. There were no ECG changes intraoperatively. However, it had been noted during the procedure that the anastomosis had been technically difficult. On review of the film of the anastomosis it was noted that the arteriotomy was somewhat small and that there was the suggestion of an intimal flap. Following hand sewn revision, the patient was extubated in the operating room hemodynamically stable without inotropic support and without ECG changes. Forty-eight hours postoperatively the patient had onset of chest pain with associated ECG changes and positive cardiac enzymes. On postoperative day four the patient experienced atrial fibrillation, which was treated medically and required transfusion of two units of packed red blood cells for a hemoglobin of 6.9 g. He was discharged to home on his ninth postoperative day, but returned with a deep sternal wound infection one week later (postoperative day 15) which required sternal debridement and pectoralis flap closure. Stress echocardiogram at three months postoperatively showed apical akinesis without associated angina. During the course of this study there were a number of technical issues and lessons learned which may assist in standardizing the procedure and a more efficient and reproducible operation. At the time of this study there was no published literature regarding the use of the da Vinci surgical system through a sternotomy. Since the instrumentation has been developed specifically for endoscopic work, there are a number of issues which must be considered when utilizing them through a sternotomy. For example, the most significant problem encountered in this study involved conflicts between the robotic arms and the sternal edge or stabilizing device. The angles were eventually resolved to minimize the conflicts and improve visualization. The camera arm had the remote center placed at the center of the left side of the sternal retractor while using a 30 degree endoscope. Likewise, the left and right arm remote centers were placed against the retractor cephalad and caudad with the “elbows” up and out in order to maximize range of motion. Another important consideration involved the type of stabilizer and blower or irrigator. Although the straight arm immobilizer offered superior stabilization and coronary occlusion, it was immediately obvious that the high profile caused conflict with the robotic instruments. Furthermore, the stabilizing window was somewhat limiting. The flexible arm stabilizer had fewer conflicts with the instruments; however, there remained problems regarding the “trailer hitch” attachment of the stabilizer with the supporting arm. Although the Cardioboy system with its integrated irrigator only became available for use on the final two patients, it was found to work best overall. Its low profile reduced conflicts and the irrigator allowed for adequate visualization without splatter onto the camera lens. A further advantage was the ability to place the stabilizer and adjust the irrigator robotically. There was, however, an issue with coronary occlusion since this is not integrated into the stabilizing system. There were no
157
commercially available occluders and the use of “home made” coronary occluders, although adequate, were less than ideal and difficult to adjust robotically. In order to stabilize and more closely approximate the setup for eventual robotically assisted beating heart TECAB, a 5-0 silk suture attached to the IMA pedicle distally was used to provide traction. This allowed for easier visualization of the IMA hood during the anastomosis. Also, the use of a silk tie attached to the occluding bulldog on the IMA pedicle allowed for easy retrieval following removal. Once the stabilizer was placed, it was important to completely clear the targeted coronary artery of epicardial fat so that the vessel is clearly delineated before arteriotomy. During the formation of the coronary arteriotomy there was a tendency, probably due to the increased magnification, to create a small opening. This can be avoided by a conscious effort to measure the arteriotomy using the graduations on the coronary scissors. Since a sutured anastomosis was performed in this study, there was care to ensure that the suture remained snug by utilizing the needle to tighten the suture before tying. Also, a crimp of the suture distal to the knot will allow the suture to break distal to the knot if too much tension is applied. From the assistant’s perspective, there was a need to be able to “duck and weave” while monitoring not only the endoscopic view, but also actions at the field and hemodyamics. This required full attention and peripheral vision. It was obvious, however, early in this experience that although the more traditional use of both a scrub member as well as surgical assistant is indeed possible, with proper training and set-up it was more efficient and less crowded to have the roles combined. This also allowed for less confusion. Due to the arm positioning, there was difficulty at times in the ability of the assistant to retrieve needles in the open sternotomy. The use of a sterile magnet on a telescoping handle solved the problem of needle retrieval. Finally, there was no issue with the ability to easily remove the robot from the field at the conclusion of the procedure or when the procedure was abandoned. This was accomplished quickly and efficiently before the surgeon arrived back to the operating table. The results of this study are similar to those reported by others [4 –7]. Although totally endoscopic robotically assisted beating heart coronary artery bypass (CAB) is being performed in centers outside the United States, the majority of studies to date in the United States have been by sternotomy or endoscopically on the arrested heart. Reports to date have concentrated primarily on the quality of the anastomosis and associated complications. The data demonstrating feasibility of robotically assisted CAB on the arrested heart as well as the beating heart is mounting as more surgeons evaluate this enabling technology. Currently, though feasible, it is recognized that robotically assisted CAB is a technically demanding procedure with a steep learning curve [5, 8, 9]. A logical stepwise progression has been outlined by Boyd and Stahl [10] which progresses from laboratory training to
CARDIOVASCULAR
Ann Thorac Surg 2004;78:154 – 8
CARDIOVASCULAR
158
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
endoscopic LAD harvest and identification of the target vessel to robotically assisted anastomosis on the beating heart by sternotomy and finally to TECAB. As technology and techniques involving the application of robotically assisted cardiac surgery are developed it will be important to discuss the process and lessons learned as the procedures are perfected. In conclusion, although two patients required anastomotic revision, there were no complications or technical failures related to the robotic system. Thus, based upon this study robotically assisted beating heart revascularization appears to be feasible, safe, and effective. Further evaluation will be necessary to determine the role of robotically assisted totally endoscopic coronary artery bypass on the beating heart in the United States.
Ann Thorac Surg 2004;78:154 – 8
4.
5.
6. 7. 8.
References 1. Bonatti J, Schachner T, Bonaros N, et al. Robotically sutured LIMA grafts—intraoperative quality control using graft angiography. Heart Surg Forum 2003;6(Suppl 1):S33. 2. Damiano RJ Jr, Ehrman WJ, Ducko CT, et al. Initial United States clinical trial of robotically assisted endoscopic coronary artery bypass grafting. J Thorac Cardiovasc Surg 2000; 119:77–82. 3. Damiano RJ Jr, Ducko CT, Stephenson ER Jr, Lawton JS,
9.
10.
Kuenzler RC, Chambers CE. Robotically assisted coronary artery bypass grafting: a prospective single center clinical trial. J Card Surg 2000;15:256 –65. Detter C, Boehm DH, Reichenspurner H, Deuse T, Arnold M, Reichart B. Robotically assisted coronary artery surgery with, and without cardiopulmonary bypass—from first clinical use to endoscopic operation. Med Sci Monit 2002;8:118 – 23. Bonatti J, Schachner T, Bonaros N, et al. Program development and learning curve issues in robotic totally endoscopic coronary artery bypass grafting. Heart Surg Forum 2003; 6(Suppl 1):S48. Damiano RJ Jr, Tabaie HA, Mack MJ, et al. Initial prospective multicenter clinical trial of robotically-assisted coronary artery bypass grafting. Ann Thorac Surg 2001;72:1263–8. Tabaie HA, Graper WP, Reinbolt JA. Clinical investigation: endoscopic coronary artery bypass grafting with robotic assistance. Heart Surg Forum 2002;5:1–6. Novick RJ, Fox SA, Kiaii BB, et al. Analysis of the learning curve in telerobotic, beating heart coronary artery bypass grafting: a 90 patient experience. Ann Thorac Surg 2003;76: 749 –53. Dogan S, Aybek T, Andressen E, et al. Totally endoscopic coronary artery bypass grafting on cardiopulmonary bypass with robotically enhanced telemanipulation: report of fortyfive cases. J Thorac Cardiovasc Surg 2002;123:1125–31. Boyd WD, Stahl KD. The Janus syndrome: a perspective on a new era of computer-enhanced robotic cardiac surgery. J Thorac Cardiovasc Surg 2003;126:625–30.
DISCUSSION DR D. GLENN PENNINGTON (Johnson City, TN): This is really courageous work and interesting, and I think we are all applauding you while we sit in the audience. I am interested about informed consent and how you went about that and what you told patients and why did they consent? DR BOLTON: They were fully informed, truly. Of course, the informed consent was approved by the Food and Drug Administration, and in addition, they saw a video of work that has been done in Germany so they had an idea of what we were doing. There were a couple of minimally invasive direct coronary artery bypasses in this series, and we were using the robotics to take down the mammary, but it is a Food and Drug Administration-approved instrument for that. These patients all knew that they would get an immediate postoperative angiogram. There were not only the risks of a normal off-pump coronary bypass but, in addition, with the robotics the patients were aware that it may be necessary, as it was in two patients, to redo their anastomosis. And, by the way, it did go before the bioethics committee and our Institutional Review Board to get approval for the consent. I feel very comfortable that they were as fully informed as anyone possibly could be, and I think they were just courageous patients. I believe we all have to remember that when we do any of this kind of work that it is the patients who are the heroes, not the guy driving the robotics or whatever the research might be. DR ROBERT POSTON (Baltimore, MD): You mentioned that you didn’t reverse heparinization during the case, but what other anticoagulation maneuvers did you use for these patients? DR BOLTON: It was just on that one patient I didn’t reverse it, and the reason was that I was expecting to find clot in the distal
left anterior descending coronary artery because the anastomosis was open. I didn’t feel comfortable ignoring the possibility of clot even though I didn’t see any whenever I revised the anastomosis. I just felt that I could leave the heparin as it was. My normal heparinization for an off-pump coronary is what we did, and that comes out to about a third of a heparinizing dose of heparin, and I do reverse the heparin at the end with protamine. DR POSTON: Did you just use aspirin postoperatively? DR BOLTON: Oh, yes. They get their first aspirin 6 hours postoperatively. All but one of these patients were extubated in the cath laboratory or in the OR. The other one was extubated on arrival to the intensive care unit. DR POSTON: Do you plan on more aggressively anticoagulating patients when you eventually reduce the risk for bleeding by doing coronary artery bypass totally endoscopic without a sternotomy or just maintain postoperative aspirin? DR BOLTON: No. There is nothing to guide us as far as anticoagulation. I know a number of people, I am sure in this audience, routinely give their patients Plavix after off-pump surgery. I have gone back and forth with that, and, in my opinion, we are being guided by data from our cardiologists where they are putting stents into vessels. I don’t think we have any real science. It will be interesting if someday we can query The Society for Thoracic Surgeons database and find out if there is a difference in long-term or intermediate-term outcome. It is an expensive drug, and actually I saw more patients come back to my practice with pleural effusions after using Plavix.
Results of a Phase One Study on Robotically Assisted Myocardial Revascularization on the Beating Heart J. W. Randolph Bolton, MD, and Jackie E. Connally, SFA/CST Carolina Heart Institute, Greenwood, South Carolina
Background. The objective of this study was to evaluate the feasibility of computer assisted myocardial revascularization on the beating heart. Methods. Ten patients underwent at least one robotically assisted internal mammary artery (IMA) anastomosis utilizing the da Vinci surgical system (Intuitive Surgical, Inc, Mountain View, CA) performed through an open incision as part of standard multivessel off pump revascularization. Following chest closure a selective IMA angiogram was performed to assess patency. Three month follow-up included a stress echocardiogram. Results. There were 12 anastomoses performed in 10 patients. The average age was 61 years with a mean ejection fraction of 56%. No patient required inotropic support. Eight of 10 patients were found to have fully patent IMA anastomoses by angiogram. One patient was noted to have an occluded left anterior descending coronary artery distal to the anastomosis and one had occlu-
sion at the anastomosis. Both patients had immediate manual revision of the anastomosis. One patient who required anastomotic revision experienced postoperative myocardial infarction and sternal wound infection requiring pectoralis flaps. Conclusions. Because robotic instrumentation is meant for closed chest procedures, there were major issues with positioning of the robotic arms in this study since the chest was open. Although two patients required anastomotic revision, there were no complications or technical failures related to the robotic system. Thus, based upon this study robotically assisted beating heart revascularization appears to be feasible, safe, and effective. Further evaluation will be necessary to determine the role of robotically assisted totally endoscopic coronary artery bypass on the beating heart in the United States. (Ann Thorac Surg 2004;78:154 – 8) © 2004 by The Society of Thoracic Surgeons
T
Material and Methods
otally endoscopic robotically assisted coronary artery bypass (TECAB) has been performed on the beating heart in some countries for a number of years. At the time of this study, however, there were no robotic devices approved for such use in the United States by the Federal Drug Administration (FDA). There were other concurrent studies investigating the use of robotics to facilitate TECAB on the arrested heart. The purpose of this study was to determine in a stepwise fashion the feasibility, applicability, and safety of robotically assisted coronary revascularization on the beating heart. This first phase study is intended to provide a first step in the logical progression in the future training program for those interested in pursuing the use of this technology. What follows are the results of this study and comments regarding the lessons learned during the development of the procedure.
Accepted for publication Jan 22, 2004. Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003. Address reprint requests to Dr Bolton, 303 W Alexander Ave, Ste E PO Box 3284, Greenwood, SC 29648; e-mail: rbolton@carolinaheartinstitute. com.
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
The safety and effectiveness of the da Vinci surgical system (Intuitive Surgical, Mountain View, CA) and endoscopic instruments was evaluated for use in myocardial revascularization on the beating heart following approval by the Institutional Review Committee and the FDA. In preparation for this phase-one investigational device exemption (IDE) study the authors underwent extensive laboratory training. To achieve the primary objective, ten subjects diagnosed with coronary artery disease underwent myocardial revascularization on the beating heart by sternotomy or minimally invasive direct coronary artery bypass (MIDCAB) in which the da Vinci surgical system was used to complete the distal internal mammary artery (IMA) anastomosis to the target coronary vessels. Potentially eligible subjects were screened for participation using prospectively established inclusion and exclusion criteria. Patients were considered suitable for inclusion if they were between the ages of 18 and 80 years and candidates for first time, nonemergent myocardial revascularization. Exclusion criteria included left ventricular ejection fraction (EF) less than 30%, severe peripheral vascular disease, acute stroke, renal failure, moderate to severe valve disease, or hemodynamic instability. Subjects meeting the appropriate criteria 0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.01.036
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
155
Table 1. Demographics Patient Number 1 2 3 4 5 6 7 8 9 10
Sex
Age Years
EF Percent
MI
Unstable Angina
Diabetes
PTCA Stent
Male Female Male Female Male Male Female Male Female Male
40 56 56 69 56 72 69 66 50 77
40 40 60 55 70 60 60 60 50 60
Yes Yes No No No No No No Yes No
Yes Yes Yes Yes Yes No Yes Yes Yes Yes
No No No No No Yes No Yes No No
Yes No Yes No No No No No No Yes
EF ⫽ left ventricular ejection fraction;
MI ⫽ myocardial infarction;
were invited to participate in the trial and required to give informed consent before enrollment (Table 1). Of the twelve patients invited to participate in the study, ten completed the study with twelve anastomoses being accomplished with robotic assistance. The robotic assistance was abandoned nonemergently in two patients. One patient had a planned bilateral MIDCAB with robotically assisted IMA harvest, which was abandoned during IMA takedown because of ongoing bleeding due to clopidogrel (Plavix). The other patient was abandoned following robotically assisted coronary arteriotomy due to inadequate visualization secondary to bleeding due to a large septal branch. Both patients had an uneventful recovery. The role of the surgical assistant was to change instruments, insure safety, and monitor observational real-time quality of the robotically assisted anastomosis. Successful use of the da Vinci surgical system was defined as a successful perioperative course and postsurgical, predischarge graft patency as determined using coronary angiography. Following completion of all other necessary bypass grafts, the previously prepared da Vinci patient side robot was positioned for the IMA anastomosis. The stabilizer was manually placed in the first eight patients and robotically placed in the final two patients. Following robotically assisted placement of silastic vessel occluders, the left anterior descending coronary artery (LAD) was opened and a running anastomosis was robotically performed using 6-0 monofilament polymide suture (Peters Cardionyl; Peters, Inc, Paris, France). At conclusion of the anastomosis the suture was robotically tied and the IMA bulldog occluder released. Hemostasis was ensured and the silastic LAD occluders removed. The stabilizer was taken away and the patient side robot withdrawn from the field. The chest was then closed in routine fashion and selective cardiac catheterization performed (by the radial arterial catheter site as often as possible) to assess patency while the patient remained intubated and anesthetized. Information was gathered to describe various operative characteristics including assisted ventilation time, intensive care unit, and total hospital stay duration, as
PTCA ⫽ percutaneous transluminal coronary angioplasty.
well as observed perioperative and postoperative (⬍30 days) morbidity and mortality. Additional clinical testing (eg, angiographic or echocardiographic studies) was conducted and documented at the following study phases: intraoperatively prior and subsequent to coronary artery bypass grafting, immediately postoperatively, at two weeks following discharge, and three months following surgery. Evaluation of the da Vinci surgical system in this study was specifically limited to completion of the anastomosis. Although the da Vinci surgical system, at times, was used for other portions of the procedure for which it is already FDA approved (eg, IMA mobilization) information characterizing the use of the device for such procedures was not collected in detail.
Results The study group consisted of ten patients (six male and four female) with twelve robotically assisted IMA anastomoses. (Table 2) There were nine IMA anastomoses to the LAD and three to the diagonal artery with two sequential bypasses to the diagonal artery and the LAD. Nine patients had multivessel off pump coronary artery bypass (OPCAB) by sternotomy with the robotically assisted anastomosis being performed as the final graft(s) and one patient had the robotically assisted anastomosis to the LAD by MIDCAB. There was an average of 2.5 grafts per patient (range, 1 to 4). The average age was 61.1 years (40 to 77 years) with a mean EF of 55.5% (40% to 70%). The average time for completion of the robotically assisted anastomoses, including setup and removal of stabilizing devices, was 40 minutes (31 to 60 min) with mean operating time of 200 minutes. Different stabilizing platforms were evaluated, including the Genzyme straight arm immobilizer (one), Genzyme flexible arm immobilizer (six), Genzyme straight arm MIDCAB retractor (one) (Genzyme Biosurgery, Cambridge, MA), and the Medtronic Cardioboy (Medtronic, Inc, Minneapolis, MN) endoscopic suction stabilizer with irrigator (two). The postoperative length of stay (LOS) averaged 2.8 days (1 to 9 days) with only one patient requiring a LOS of more than three days (patient 6). On immediate
CARDIOVASCULAR
Ann Thorac Surg 2004;78:154 – 8
CARDIOVASCULAR
156
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
Ann Thorac Surg 2004;78:154 – 8
Table 2. Results Patient Number
Total Anast
Robotic Anast
Anast Minutes
Postop LOS
1
2
2
2
Imp
None
2 3
3 2
1 1
39-Diag 31-LAD 40 37
2 3
Nml N/Aa
4 5
2 3
1 2
2 2
N/Aa Nml
6
3
1
40 60-Diag 29-LAD 53
None Occluded LAD (Anast revised) Transfusion Thorocentesis None IMA spasm
9
Apical akinesis
7 8 9 10
3 1 4 2
1 1 1 1
32 37 40 39
3 1 2 2
Nml Nml Nml Imp
a
Stress Echo
Complications
Occluded IMA (Anast revised) Myocardial infarction Atrial fibrillation Transfusion Thorocentesis Sternal infection None None None None
Refused echo.
Anast ⫽ anastomosis; Diag ⫽ diagonal artery; Imp ⫽ improved; artery; LOS ⫽ length of stay in days; Nml ⫽ normal.
LAD ⫽ left anterior descending artery;
IMA ⫽ internal mammary
postoperative angiography ten of the twelve anastomoses revealed no stenosis. However, although angiography revealed no anastomotic stenosis and good retrograde filling of the LAD and its diagonal branches, one patient (patient 3) had an LAD occlusion distal to the anastomosis. This was associated with minor electrocardiographic (ECG) changes, which resolved upon revision of the anastomosis. Cardiac enzymes were negative. Interestingly, postoperatively this patient was found to have deficiencies of proteins S and C. A second patient (patient 6) had total anastomotic occlusion of the left internal mammary artery (LIMA). Both of these patients were immediately returned to the operating room for revision of the anastomosis with a standard hand sewn technique. These two patients also represented the only patients requiring transfusion postoperatively (two units of packed red blood cells each). Nine patients were free of major complications; however, the patient with the occluded LIMA experienced a postoperative myocardial infarction (MI), atrial fibrillation, and eventual sternal wound infection requiring pectoralis flap closure. No patient experienced hemodynamic instability or required inotropic support and there was no mortality. All patients were extubated within one and one half hours after arrival to the single stay intensive care unit with eight patients being extubated in the operating room or cardiac catheterization laboratory following angiography. Stress echocardiogram performed at three month follow-up was normal or improved in seven of eight patients. The stress echocardiogram on patient 6 revealed a new finding of apical akinesis without the onset of angina following the postoperative MI. Two patients (patients 3 and 4)
refused the stress echocardiogram because they felt improved and physically unlimited since their surgery. There were three minor complications. Two patients (patients 3 and 6) had pleural effusions requiring thoracentesis. There was one episode of IMA spasm associated with ischemic chest pain and sudden ST segment elevation several hours postoperatively (patient 5) which responded to medical therapy with resolution of chest discomfort and ECG changes. Cardiac enzymes were negative. There were no complications or technical failures relative to the robotic system. One anastomosis was revised robotically after the suture broke while being tied.
Comment Aside from one episode of IMA spasm, all complications occurred in the two patients who required revision of the anastomosis. In regard to the first of these patients (patient 3) who had what appeared to be an LAD occlusion distal to the anastomosis, spasm of the LAD should also be considered since there was no clot found during revision of the anastomosis. This has been described by Bonatti and colleagues [1] to have occurred in eighteen of twenty-three patients following robotically assisted anastomoses. In addition, based upon the excellent retrograde filling of the LAD at angiogram, it is possible that had flow testing alone been performed to assess patency it may very likely have been adequate and the anastomosis would not have been revised [2, 3]. The second patient, a diabetic with small, calcified vessels (patient 6) accounted for six of the eight compli-
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
cations. Postoperative angiography revealed a totally occluded LIMA at the anastomosis. There were no ECG changes intraoperatively. However, it had been noted during the procedure that the anastomosis had been technically difficult. On review of the film of the anastomosis it was noted that the arteriotomy was somewhat small and that there was the suggestion of an intimal flap. Following hand sewn revision, the patient was extubated in the operating room hemodynamically stable without inotropic support and without ECG changes. Forty-eight hours postoperatively the patient had onset of chest pain with associated ECG changes and positive cardiac enzymes. On postoperative day four the patient experienced atrial fibrillation, which was treated medically and required transfusion of two units of packed red blood cells for a hemoglobin of 6.9 g. He was discharged to home on his ninth postoperative day, but returned with a deep sternal wound infection one week later (postoperative day 15) which required sternal debridement and pectoralis flap closure. Stress echocardiogram at three months postoperatively showed apical akinesis without associated angina. During the course of this study there were a number of technical issues and lessons learned which may assist in standardizing the procedure and a more efficient and reproducible operation. At the time of this study there was no published literature regarding the use of the da Vinci surgical system through a sternotomy. Since the instrumentation has been developed specifically for endoscopic work, there are a number of issues which must be considered when utilizing them through a sternotomy. For example, the most significant problem encountered in this study involved conflicts between the robotic arms and the sternal edge or stabilizing device. The angles were eventually resolved to minimize the conflicts and improve visualization. The camera arm had the remote center placed at the center of the left side of the sternal retractor while using a 30 degree endoscope. Likewise, the left and right arm remote centers were placed against the retractor cephalad and caudad with the “elbows” up and out in order to maximize range of motion. Another important consideration involved the type of stabilizer and blower or irrigator. Although the straight arm immobilizer offered superior stabilization and coronary occlusion, it was immediately obvious that the high profile caused conflict with the robotic instruments. Furthermore, the stabilizing window was somewhat limiting. The flexible arm stabilizer had fewer conflicts with the instruments; however, there remained problems regarding the “trailer hitch” attachment of the stabilizer with the supporting arm. Although the Cardioboy system with its integrated irrigator only became available for use on the final two patients, it was found to work best overall. Its low profile reduced conflicts and the irrigator allowed for adequate visualization without splatter onto the camera lens. A further advantage was the ability to place the stabilizer and adjust the irrigator robotically. There was, however, an issue with coronary occlusion since this is not integrated into the stabilizing system. There were no
157
commercially available occluders and the use of “home made” coronary occluders, although adequate, were less than ideal and difficult to adjust robotically. In order to stabilize and more closely approximate the setup for eventual robotically assisted beating heart TECAB, a 5-0 silk suture attached to the IMA pedicle distally was used to provide traction. This allowed for easier visualization of the IMA hood during the anastomosis. Also, the use of a silk tie attached to the occluding bulldog on the IMA pedicle allowed for easy retrieval following removal. Once the stabilizer was placed, it was important to completely clear the targeted coronary artery of epicardial fat so that the vessel is clearly delineated before arteriotomy. During the formation of the coronary arteriotomy there was a tendency, probably due to the increased magnification, to create a small opening. This can be avoided by a conscious effort to measure the arteriotomy using the graduations on the coronary scissors. Since a sutured anastomosis was performed in this study, there was care to ensure that the suture remained snug by utilizing the needle to tighten the suture before tying. Also, a crimp of the suture distal to the knot will allow the suture to break distal to the knot if too much tension is applied. From the assistant’s perspective, there was a need to be able to “duck and weave” while monitoring not only the endoscopic view, but also actions at the field and hemodyamics. This required full attention and peripheral vision. It was obvious, however, early in this experience that although the more traditional use of both a scrub member as well as surgical assistant is indeed possible, with proper training and set-up it was more efficient and less crowded to have the roles combined. This also allowed for less confusion. Due to the arm positioning, there was difficulty at times in the ability of the assistant to retrieve needles in the open sternotomy. The use of a sterile magnet on a telescoping handle solved the problem of needle retrieval. Finally, there was no issue with the ability to easily remove the robot from the field at the conclusion of the procedure or when the procedure was abandoned. This was accomplished quickly and efficiently before the surgeon arrived back to the operating table. The results of this study are similar to those reported by others [4 –7]. Although totally endoscopic robotically assisted beating heart coronary artery bypass (CAB) is being performed in centers outside the United States, the majority of studies to date in the United States have been by sternotomy or endoscopically on the arrested heart. Reports to date have concentrated primarily on the quality of the anastomosis and associated complications. The data demonstrating feasibility of robotically assisted CAB on the arrested heart as well as the beating heart is mounting as more surgeons evaluate this enabling technology. Currently, though feasible, it is recognized that robotically assisted CAB is a technically demanding procedure with a steep learning curve [5, 8, 9]. A logical stepwise progression has been outlined by Boyd and Stahl [10] which progresses from laboratory training to
CARDIOVASCULAR
Ann Thorac Surg 2004;78:154 – 8
CARDIOVASCULAR
158
BOLTON AND CONNALLY ROBOTICALLY ASSISTED MYOCARDIAL REVASCULARIZATION
endoscopic LAD harvest and identification of the target vessel to robotically assisted anastomosis on the beating heart by sternotomy and finally to TECAB. As technology and techniques involving the application of robotically assisted cardiac surgery are developed it will be important to discuss the process and lessons learned as the procedures are perfected. In conclusion, although two patients required anastomotic revision, there were no complications or technical failures related to the robotic system. Thus, based upon this study robotically assisted beating heart revascularization appears to be feasible, safe, and effective. Further evaluation will be necessary to determine the role of robotically assisted totally endoscopic coronary artery bypass on the beating heart in the United States.
Ann Thorac Surg 2004;78:154 – 8
4.
5.
6. 7. 8.
References 1. Bonatti J, Schachner T, Bonaros N, et al. Robotically sutured LIMA grafts—intraoperative quality control using graft angiography. Heart Surg Forum 2003;6(Suppl 1):S33. 2. Damiano RJ Jr, Ehrman WJ, Ducko CT, et al. Initial United States clinical trial of robotically assisted endoscopic coronary artery bypass grafting. J Thorac Cardiovasc Surg 2000; 119:77–82. 3. Damiano RJ Jr, Ducko CT, Stephenson ER Jr, Lawton JS,
9.
10.
Kuenzler RC, Chambers CE. Robotically assisted coronary artery bypass grafting: a prospective single center clinical trial. J Card Surg 2000;15:256 –65. Detter C, Boehm DH, Reichenspurner H, Deuse T, Arnold M, Reichart B. Robotically assisted coronary artery surgery with, and without cardiopulmonary bypass—from first clinical use to endoscopic operation. Med Sci Monit 2002;8:118 – 23. Bonatti J, Schachner T, Bonaros N, et al. Program development and learning curve issues in robotic totally endoscopic coronary artery bypass grafting. Heart Surg Forum 2003; 6(Suppl 1):S48. Damiano RJ Jr, Tabaie HA, Mack MJ, et al. Initial prospective multicenter clinical trial of robotically-assisted coronary artery bypass grafting. Ann Thorac Surg 2001;72:1263–8. Tabaie HA, Graper WP, Reinbolt JA. Clinical investigation: endoscopic coronary artery bypass grafting with robotic assistance. Heart Surg Forum 2002;5:1–6. Novick RJ, Fox SA, Kiaii BB, et al. Analysis of the learning curve in telerobotic, beating heart coronary artery bypass grafting: a 90 patient experience. Ann Thorac Surg 2003;76: 749 –53. Dogan S, Aybek T, Andressen E, et al. Totally endoscopic coronary artery bypass grafting on cardiopulmonary bypass with robotically enhanced telemanipulation: report of fortyfive cases. J Thorac Cardiovasc Surg 2002;123:1125–31. Boyd WD, Stahl KD. The Janus syndrome: a perspective on a new era of computer-enhanced robotic cardiac surgery. J Thorac Cardiovasc Surg 2003;126:625–30.
DISCUSSION DR D. GLENN PENNINGTON (Johnson City, TN): This is really courageous work and interesting, and I think we are all applauding you while we sit in the audience. I am interested about informed consent and how you went about that and what you told patients and why did they consent? DR BOLTON: They were fully informed, truly. Of course, the informed consent was approved by the Food and Drug Administration, and in addition, they saw a video of work that has been done in Germany so they had an idea of what we were doing. There were a couple of minimally invasive direct coronary artery bypasses in this series, and we were using the robotics to take down the mammary, but it is a Food and Drug Administration-approved instrument for that. These patients all knew that they would get an immediate postoperative angiogram. There were not only the risks of a normal off-pump coronary bypass but, in addition, with the robotics the patients were aware that it may be necessary, as it was in two patients, to redo their anastomosis. And, by the way, it did go before the bioethics committee and our Institutional Review Board to get approval for the consent. I feel very comfortable that they were as fully informed as anyone possibly could be, and I think they were just courageous patients. I believe we all have to remember that when we do any of this kind of work that it is the patients who are the heroes, not the guy driving the robotics or whatever the research might be. DR ROBERT POSTON (Baltimore, MD): You mentioned that you didn’t reverse heparinization during the case, but what other anticoagulation maneuvers did you use for these patients? DR BOLTON: It was just on that one patient I didn’t reverse it, and the reason was that I was expecting to find clot in the distal
left anterior descending coronary artery because the anastomosis was open. I didn’t feel comfortable ignoring the possibility of clot even though I didn’t see any whenever I revised the anastomosis. I just felt that I could leave the heparin as it was. My normal heparinization for an off-pump coronary is what we did, and that comes out to about a third of a heparinizing dose of heparin, and I do reverse the heparin at the end with protamine. DR POSTON: Did you just use aspirin postoperatively? DR BOLTON: Oh, yes. They get their first aspirin 6 hours postoperatively. All but one of these patients were extubated in the cath laboratory or in the OR. The other one was extubated on arrival to the intensive care unit. DR POSTON: Do you plan on more aggressively anticoagulating patients when you eventually reduce the risk for bleeding by doing coronary artery bypass totally endoscopic without a sternotomy or just maintain postoperative aspirin? DR BOLTON: No. There is nothing to guide us as far as anticoagulation. I know a number of people, I am sure in this audience, routinely give their patients Plavix after off-pump surgery. I have gone back and forth with that, and, in my opinion, we are being guided by data from our cardiologists where they are putting stents into vessels. I don’t think we have any real science. It will be interesting if someday we can query The Society for Thoracic Surgeons database and find out if there is a difference in long-term or intermediate-term outcome. It is an expensive drug, and actually I saw more patients come back to my practice with pleural effusions after using Plavix.