Recent technologic advances have brought completely. Robotic Endoscopic Left Internal Mammary Artery Harvesting: What Have We Learned After 100 Cases?

Transcription

1 Robotic Endoscopic Left Internal Mammary Artery Harvesting: What Have We Learned After 100 Cases? Armin Oehlinger, MD, Nikolaos Bonaros, MD, Thomas Schachner, MD, Elisabeth Ruetzler, MD, Guy Friedrich, MD, Guenther Laufer, MD, and Johannes Bonatti, MD Departments of Cardiac Surgery and Cardiology, Innsbruck Medical University, Innsbruck, Austria Background. The development of robotic devices has recently offered the possibility of performing coronary artery bypass graft surgery (CABG) in a totally endoscopic way. An important step of this procedure is endoscopic harvesting of the left internal mammary artery (LIMA). It was the aim of our study to find factors influencing LIMA harvesting time and to describe the challenges associated with robotic endoscopic LIMA harvesting. Methods. From June 2001 to December 2005, a total of 100 patients underwent robotically assisted CABG. In all cases, the LIMA was harvested by using the robotic DaVinci device. Coronary artery bypass grafting procedures were completed through sternotomy, minithoracotomy, or in a totally endoscopic fashion. Results. The median LIMA harvesting time was 48 minutes (19 to 180). A significant learning curve was observed: y (min) ln (x), x LIMA takedown number, p less than Takedown time decreased from 140 minutes in the first 10 cases to 34 minutes in the last 10 cases. There was no independent demographic factor that significantly influenced the LIMA harvesting time. The LIMA takedown time also showed no significant correlation with thorax dimensions. Injury to the LIMA occurred in 3 patients (6%) during the first half of the experience and in 1 patient (2%) during the second half (p not significant). Conclusions. Robotic-enhanced LIMA takedown is a prerequisite for totally endoscopic CABG. After passing through a significant learning curve, IMA takedown can be performed safely and within an acceptable time frame. Demography and chest size do not seem to influence IMA harvesting time. The rate of LIMA injuries is within the limits of conventional thoracoscopic harvesting. (Ann Thorac Surg 2007;83:1030 4) 2007 by The Society of Thoracic Surgeons Recent technologic advances have brought completely new instruments into the heart surgery armamentarium and have enabled the development of minimally invasive operative techniques. Attempts to perform heart operations with thoracoscopic instruments were reported by different groups in the mid 1990s [1 3]. The use of voice-controlled or computer-assisted surgical systems and the development of advanced telemanipulation devices offered the heart surgeon the possibility of performing cardiac surgery in a totally endoscopic way. A limited number of totally endoscopic coronary artery bypass graft (TECAB) procedures, in most of the cases left internal mammary artery (LIMA) grafts to the left anterior descending artery, has so far been reported [4 9, 17]. Endoscopic LIMA harvesting is an important step in TECAB development [5]. We present the experience of robotic-enhanced LIMA harvesting in 100 cases using the DaVinci telemanipulator during implementation of arrested heart TECAB (AH-TECAB). It was the aim of our Accepted for publication Oct 23, Address correspondence to Dr Oehlinger, Department of Cardiac Surgery, Innsbruck Medical University, Anichstrasse 35, Innsbruck A-6020, Austria; armin_oehlinger@gmx.at. study to find factors influencing LIMA harvesting time, and to report on technical aspects of the procedure and the challenges encountered during the learning curve. Material and Methods From June 2001 to December 2005, a total of 100 patients underwent robotic-enhanced LIMA takedown at the Department of Cardiac Surgery at Innsbruck Medical University. During the same period, 1,432 conventional CABG operations were performed. Therefore, in 6.5%, the LIMA was harvested endoscopically. Demographic data are listed in Table 1. In all cases the LIMA was harvested using the robotic DaVinci device (Intuitive Surgical, Sunnyvale, California). Informed consent was obtained, and the performance of AH-TECAB was approved by the Institutional Ethics Committee. Operative Technique After induction of anesthesia, a double-lumen endotracheal tube for single-lung ventilation was used. Patients were positioned in a 30-degree right lateral position. After setting up the DaVinci system, the camera port was 2007 by The Society of Thoracic Surgeons /07/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg OEHLINGER ET AL 2007;83: ROBOTIC ENDOSCOPIC LIMA HARVESTING Table 1. Patient Demographics of 100 Patients Undergoing Endoscopic LIMA Harvesting (n 100) Age 59 (38 76) Male 83 (83%) Female 17 (17%) Height (cm) 172 ( ) Weight (kg) 79 (48 102) BMI 26 (19 36) Heart transverse diameter (cm) 14.6 ( ) Chest transverse diameter (cm) 31.4 ( ) Cardiothoracic ratio 0.46 ( ) Diaphragm-apex left diameter (cm) 24.5 ( ) Diaphragm-apex right diameter (cm) 22.9 ( ) St.p. myocardial infarction 33 (33%) St.p. PTCA/stent 31 (31%) Hypertension 77 (77%) Hypercholesterolemia 74 (74%) DM 9 (9%) Smoking 59 (59%) COPD 16 (16%) CVD 3 (3%) PVD 5 (5%) Chronic renal failure 4 (4%) Hemodialysis 0 (0%) Creatinine (mg/dl) 1.01 ( ) EuroSCORE 1 (0 7) BMI body mass index; COPD chronic obstructive pulmonary disease; CVD cerebrovascular disease; DM diabetes mellitus; EuroSCORE European System for Cardiac Operative Risk Evaluation; PTCA percutaneous transluminal coronary angioplasty; PVD peripheral vascular disease; St.p. status post. introduced by the patient-side surgeon into the left fifth intercostal space on the anterior axillary line. After CO 2 insufflation at target pressure between 8 and 10 mm Hg Table 2. Procedures Performed 1031 Total number of robotic enhanced LIMA 100 takedown LIMA takedown conventional CABG 6 (6%) through sternotomy LIMA takedown MIDCABG 9 (9%) LIMA takedown OPCABG 4 (4%) LIMA takedown robotically sutured 7 (7%) anastomosis through sternotomy AH-TECAB 1 (LIMA-LAD or Dg) 57 (57%) BH-TECAB 8 (8%) AH-TECAB 2 (LIMA-OM RIMA-LAD) 4 (4%) AH-TECAB arrested heart totally endoscopic coronary artery bypass; BH-TECAB beating heart totally endoscopic coronary artery bypass; CABG coronary artery bypass grafting; Dg diagonal branch; LAD left anterior descending artery; LIMA left internal mammary artery; MIDCABG minimally invasive direct coronary artery bypass; OM marginal branch; OPCABG off-pump coronary artery bypass graft; RIMA right internal mammary artery. and deflation of the left lung, the instrument ports were inserted through the third and the seventh intercostal spaces on the midclavicular line under thoracoscopic vision. The internal mammary artery (IMA) was exposed and harvested from the first to the fifth intercostal space using electrocautery at 20 W (Fig 1). Endoscopic clips were used selectively for division of side branches. After LIMA harvesting, heparin was given and a temporarily occluding bulldog clamp was placed. The graft was partly divided at its distal end and then incised. After free flow was checked, the distal part was prepared for the anastomosis. Coronary artery bypass grafting was completed in different manners, as listed in Table 2. The LIMA harvesting time was defined from completed port placement to completed IMA detachment from thoracic wall. Port placement was separately recorded; setup time was not available for the whole series. Definition of Thoracic Measurements For thoracic measurements, preoperative posteroanterior chest radiographs were taken. Heart transverse diameter was defined as the distance between the right and the left edge of cardiac silhouette. Chest transverse diameter was the external transverse diameter of the thorax. The distance between the left apex of the lung and the left diaphragm were taken for the left diaphragm-apex diameter. The diameter for the right side was measured equally. CARDIOVASCULAR Fig 1. The left internal mammary artery (LIMA) is harvested endoscopically using a left (L instr.) and a right (R instr.) robotic instrument which are controlled by the surgeon at the telemanipulator console. Statistics For statistical analysis and calculations, SPSS 11.0 statistical software package (SPSS, Chicago, Illinois) was used. Continuous variables are given as median and range, and categorical variables are presented as absolute numbers and percentages. For calculation of the learning curve, regression analysis with logarithmic curve fit was used. Correlations concerning influencing factors on LIMA harvesting time were calculated using nonparametric tests (Spearman correlation and Mann-Whitney U test). A p value less than 0.05 was regarded as significant.

3 1032 OEHLINGER ET AL Ann Thorac Surg ROBOTIC ENDOSCOPIC LIMA HARVESTING 2007;83: Cumulative freedom from angina in this patient population was 97% at 5 years, and no target vessel reintervention was necessary in the immediate follow-up. Fig 2. The learning curve for left internal mammary artery (LIMA) harvesting followed the formula: y (min) x ln (x), x LIMA takedown number (Nr.). Results There were no major technical failures of the robotic system. The median for LIMA harvesting time was 48 minutes (19 to 180). A significant learning curve was observed: y (min) ln (x), x LIMA takedown number; p less than Robotic time decreased from 140 minutes (median) in the first 10 cases to 34 minutes (median) in the last 10 cases (Fig 2). No demographic variable significantly influenced harvesting time, and LIMA preparation showed no significant correlation to anatomic thoracic dimensions (Table 3). That was true for the whole series and for the phase after the most significant part of the learning curve (last 50 cases). Four cases of LIMA injury occurred, 3 (6%) during the first half of the experience and 1 (2%) during the second half (p 0.310). One patient had to be converted to median sternotomy owing to LIMA injury during harvesting that compromised graft flow. In 1 case of intended sternotomy, an end-to-end anastomosis of the LIMA was performed because of thermic damage and dissection in the middle third of the vessel caused by electrocautery. On-table revision after intraoperative angiography diagnosis was necessary in 2 cases: in the first case, a LIMA stenosis of the graft probably caused by electrocautery was detected, whereas in the second case, an intramural hematoma compromising graft flow additionally to a side branch was found. Both patients underwent on-table revision through median sternotomy and left the operating room with a patent LIMA graft, as confirmed in the angiography control. Overall clinical results are listed in Table 4. Comment The development of robotic surgical devices was a prerequisite for performing TECAB. However, TECAB is a highly complex procedure consisting of several not routinely performed surgical steps, which requires a modular and stepwise approach [5, 17]. An important step of this procedure is the endoscopic harvesting of the IMA. We harvested the LIMA in 100 cases using the DaVinci system and completed coronary artery bypass operations in different manners. In this study, we demonstrate that this important procedure module can reach a routine state and can be performed in an acceptable time after a small number of cases. Endoscopic harvesting of IMA using conventional thoracoscopic instruments was described by different groups [1, 2, 11]. Vassiliades and colleagues [11] demonstrated a LIMA harvesting time of 38 minutes using thoracoscopic instrumentation [11]. In another study, by Duhaylongsod and coworkers [3], this time was 42 minutes to 55 minutes. Our time frame required for the procedure in the clinical scenario falls well within this range, especially during the second half of the experience. We could demonstrate that thoracoscopic LIMA takedown using the DaVinci telemanipulator is feasible, but time consuming and demanding at the beginning. After 20 procedures, harvesting time already could be reduced to an acceptable duration between 30 and 60 minutes; and even after 70 cases, a further drop of harvesting time into the below-40-minute range was noted. Long preparation times at the beginning of our series may to be explained by the complexity of this new system and the lack of clinical experience of the operation team. Bolotin and associates [13] recently reported robotic LIMA harvesting speed in the 39-minute to 48-minute range when using the dog model. Learning curves and long preparation times at the beginning of implementation of robotic-enhanced CABG also had been described by Falk and associates [9] and Reuthebuch and coworkers [10]. The Falk group presented a LIMA takedown time of about 40 minutes after 50 cases. The Reuthebuch study noted that after the learning curve, a target LIMA harvesting speed in the 35-minute range can be achieved. It was interesting to see that except for an increasing case number, no other variable influenced the LIMA harvesting time. To correct for effects of the immediate learning phase, we also looked at variable influence during the second half of the series. Again, no influence of demographic and anatomic variables was seen. Vassiliades and coworkers [11] described increased thoracoscopic IMA harvesting time for patients with a higher body mass index. In our study, we could not find any significant influence of the body mass index on harvesting time. However, it has to be mentioned that a selected group was chosen for performance of robotic-enhanced

4 Ann Thorac Surg OEHLINGER ET AL 2007;83: ROBOTIC ENDOSCOPIC LIMA HARVESTING Table 3. Variables Tested for Influence on Endoscopic LIMA Harvesting Time Whole Series (n 100) After 50 Cases (n 50) LIMA harvesting time 48 (19 180) min 42 (19 66) min Male (min) 48 (19 180) 42 (19 66) Female (min) 47 (35 70) p (35 49) p Age r p r p Height r p r p Weight r p r p BMI r p r p Heart transverse r p r p diameter Chest transverse r p r p diameter Cardiothoracic ratio r p r p Diaphragm-apex r 0.19 p r p diameter left Diaphragm-apex r p r p diameter right Hypertension Yes (min) 49 (19 170) 45 (19 66) No (min) 42 (26 180) p (26 48) p Hypercholesterolemia Yes (min) 48 (26 180) 42 (26 59) No (min) 48 (19 80) p (19 66) p DM Yes (min) 48 (38 60) 47 (38 51) No (min) 48 (19 180) p (19 66) p Smoking Yes (min) 50 (19 180) 42 (19 66) No (min) 48 (27 140) p (27 60) p COPD Yes (min) 49 (29 180) 42 (29 59) No (min) 48 (19 180) p (19 66) p CVD Yes (min) 45 (35 66) 66 (66) No (min) 48 (19 180) p (19 60) p PVD Yes (min) 48 (19 80) 33 (19 48) No (min) 48 (26 180) p (26 66) p CARDIOVASCULAR BMI body mass index; COPD chronic obstructive pulmonary disease; CVD cerebrovascular disease; DM diabetes mellitus; LIMA left internal mammary artery; PVD peripheral vascular disease; r correlation coefficient. CABG because we wanted to restrict robotic surgery to a low-risk population. Hence, we chose patients with low comorbidities, but we did not have patients with severe obesity in our series. The idea of such a selection was to keep biological reserves for application of a new technology in CABG. According to our findings, thoracic dimensions on chest radiograph did not correlate with LIMA harvesting time. We had expected that anatomical factors and the chest size would have an influence. We admit that some important factors like intercostal space diameter, thickness of subcutaneous fat, or course of the LIMA could have an influence on IMA harvesting time. But these factors can only be measured with computed tomography or magnetic resonance tomography. Such examinations were not available for the whole cohort. With this study, we wanted to provide information whether common demographic variables and standard examinations such as chest radiography can predict robotic IMA harvesting time. Our group has previously seen that LIMA to left anterior descending artery distance on preoperative computed tomography scans has an impact on the LIMA to left anterior descending artery anastomotic time [14]. Reoperations or conversions because of a direct LIMA injury or a bleeding side branch have also been reported by other authors who applied robotic devices [8 10] or conventional thoracoscopic instruments [1, 3, 12]. Duhay-

5 1034 OEHLINGER ET AL Ann Thorac Surg ROBOTIC ENDOSCOPIC LIMA HARVESTING 2007;83: Table 4. Perioperative and Postoperative Outcomes of 100 CABG Patients After Endoscopic LIMA Harvesting Myocardial infarction 3 (3%) Atrial fibrillation 10 (10%) IABP 1 (1%) Ventilation time (hours) 10 (0 278) Pneumonia 3 (3%) Hemofiltration 2 (2%) Forced diuresis 9 (9%) TIA 0 (0%) Stroke 1 (1%) MSOF 1 (1%) Sepsis 1 (1%) Ventilation time (hours) 10 (0 278) ICU stay (hours) 20 (11 389) Hospital stay (days) 7 (5 25) 30-Day mortality 0 (0%) IABP intra-aortic balloon pump; ICU intensive care unit; LIMA left internal mammary artery; MSOF multisystem organ failure; TIA transitory ischemic attack. longsod and coworkers [3] noted a LIMA injury rate of 1.8% in a three-center experience with conventional thoracoscopic harvesting. Hard data on the frequency of these events in conventional CABG are difficult to find in the literature and are rarely recorded in CABG data bases. At our institution, IMA damages are rarely noted by experienced surgeons but do occur during the phase when residents start harvesting the graft. It was satisfying to see that LIMA injury happened only once during the second half of the experience. We would like to point out that all LIMA injuries were detected during the procedure and repaired immediately. We can state that all patients left the operating room with a functioning IMA graft. For avoidance of these problems, we recommend visualization of the whole pedicle by detaching the endothoracic facia before starting conduit harvest. Special care should be taken to set the electrocautery at 20 W. We think that in 1 case in the beginning of endoscopic harvesting, LIMA injury was caused by high electrocautery energy. In light of the fact that multiple learning curves were present in this series, we think that a 0% mortality rate and the low rate of perioperative complications can be regarded as very positive. Other authors also reported safe introduction of robotic technology into their coronary surgery programs [4, 8, 9]. In conclusion, robotic-enhanced IMA takedown is prerequisite for TECAB operations and can be safely implemented. After passing through a learning curve, IMA takedown can be done in an acceptable time. Demographics and chest size do not seem to influence IMA harvesting time. The rate of LIMA injuries is comparable with the rates reported for conventional thoracoscopic harvesting.[15, 16] The authors received research grants from Intuitive Surgical (Sunnyvale, California). References 1. Nataf P, Lima L, Regan M, et al. Thoracoscopic internal mammary artery harvesting: technical considerations. Ann Thorac Surg 1997;63(Suppl): Mack M, Acuff T, Yong P, Jett GK, Carter D. Minimally invasive thoracoscopically assisted coronary artery bypass surgery. Eur J Cardiothorac Surg 1997;12: Duhaylongsod F, Mayfield W, Wolf RK. Thoracoscopic harvest of the internal thoracic artery: a multicenter experience in 218 cases. Ann Thorac Surg 1998;66: 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: Bonatti J, Schachner T, Bernecker O, et al. Robotic totally endoscopic coronary artery bypass: program development and learning curve issues. J Thorac Cardiovasc Surg 2004; 127: Cichon R, Kappert U, Schneider J, et al. Robotic-enhanced arterial revascularization for multivessel coronary artery disease. Ann Thorac Surg 2000;70: Kappert U, Schneider J, Cichon R, et al. Wrist-enhanced instrumentation: moving toward totally endoscopic coronary artery bypass grafting. Ann Thorac Surg 2000;70: Kappert U, Schneider J, Cichon R, et al. Development of robotic enhanced endoscopic surgery for treatment of coronary artery disease. Circulation 2001;104(12 Suppl 1):I Falk V, Diegeler A, Walther T, et al. Total endoscopic computer enhanced coronary artery bypass grafting. Eur J Cardiothoracic Surg 2000;17: Reuthebuch O, Comber M, Gruenenfelder J, Zund G, Turina M. Experiences in robotically enhanced IMA preparation as initial step towards totally endoscopic coronary artery bypass grafting. J Cardiovasc Surg 2003;11: Vassiliades TA Jr, Nielsen JL, Lonquist JL. Effects of obesity on outcomes in endoscopically assisted coronary artery bypass operations. Heart Surg Forum 2003;6: Nataf P, Al-Attar N, Ramadan R, et al. Thoracoscopic IMA takedown. J Card Surg 2000;15: Bolotin G, Scott W, Austin T, et al. Robotic skeletonizing of the internal thoracic artery: is it safe? Ann Thorac Surg 2004;77: Schachner T, Feuchtner G, Bonaros N, et al. Does preoperative multisclice computed tomography predict operative times in total endoscopic coronary artery bypass grafting? Heart Surg Forum 2005;8:E Mohr F, Falk V, Diegeler A, et al. Computer-enhanced robotic cardiac surgery: experience in 148 patients. J Thorac Cardiovasc Surg 2001;121: Loulmet D, Carpentier A, d Attelis N, et al. Endoscopic coronary artery bypass grafting with the aid of robotic assisted instruments. J Thorac Cardiovasc Surg 1999;118: Bonatti J, Schachner T, Bonaros N, et al. Technical challenges in totally endoscopic robotic coronary artery bypass grafting. J Thorac Cardiovasc Surg 2006;131: