During the past 20 years, the trend has been to make

Transcription

1 Early Experience With Robotic Lung Resection Results in Similar Operative Outcomes and Morbidity When Compared With Matched Video- Assisted Thoracoscopic Surgery Cases Brian E. Louie, MD, Alexander S. Farivar, MD, Ralph W. Aye, MD, and Eric Vallières, MD Division of Thoracic Surgery, Swedish Cancer Institute, Seattle, Washington Background. Robotic lung resection is gaining popularity despite limited published evidence. Comparative studies are needed to provide information about the safety and effectiveness of robotic resection. Therefore, we compared our initial experience with robotic anatomic resection to our most recent video-assisted thoracoscopic surgery (VATS) cases. Methods. A case-control analysis of consecutive anatomic lung resections by robot or VATS from 2009 through 2011 was performed. Results. In the robotic group, 52 resections were attempted. Three conversions and 3 wedges were excluded, leaving 40 lobectomies, 5 segments, and 1 conversion to VATS. In the VATS group, 35 resections were attempted with 1 conversion. The distribution of resected lobes or segments and demographics was similar. Clinical outcomes between robotics and VATS were similar in tumor size (2.8 versus 2.3 cm), operative time (213 versus 208 minutes), blood loss (153 versus 134 ml), intensive care unit stay (0.9 versus 0.6 days), and length of stay (4.0 versus 4.5 days). There was no operative mortality. Major (n 8; 17%) and minor morbidity (n 12; 26%) with robotics was similar to VATS. The percentage of expected nodal stations sampled was similar. The duration of narcotic use after discharge (p 0.039) and the time to return to usual activities (p 0.001) was shorter in the robotic group. Conclusions. Early experience with robotic resection resulted in similar outcomes compared with mature VATS cases. A potential benefit of robotics may relate to postoperative pain reduction and earlier return to usual activities. Robotic lung resection should be studied further in selected centers and compared with VATS in a randomized fashion to better define its potential advantages and disadvantages. (Ann Thorac Surg 2012;93: ) 2012 by The Society of Thoracic Surgeons During the past 20 years, the trend has been to make surgery less invasive. Cholecystectomy, fundoplication, prostatectomy, hysterectomy, and Roux-en-Y bypass are routinely performed using minimally invasive techniques. But video-assisted thoracoscopic surgery (VATS) lobectomy is not performed routinely, with estimates as high as 32% of lobectomies performed using VATS in the highly selected Society for Thoracic Surgeons database [1] and as low as 6% in the Nationwide Inpatient Sample database [2]. Robotic-assisted lung resection (uses an access incision [ 5 cm] without a port and communicates with ambient air) and completely portal robotic lung resection (uses 1-cm incisions with Accepted for publication Jan 16, Presented at the Fifty-eighth Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 9 12, Address correspondence to Dr Louie, Division of Thoracic Surgery, Swedish Cancer Institute, Madison St, Seattle, WA 98105; brian.louie@swedish.org. sealed ports, carbon dioxide insufflation, and extraction through an enlarged incision after resection) are gaining popularity and are being promoted as an alternative to VATS despite very limited published evidence beyond feasibility [3 5] and two large series from highly specialized centers and surgeons [6, 7]. Whether these outcomes can be generalized to most thoracic surgeons remains to be determined as thoracic surgeons of all skills are beginning to contemplate robotics for lung surgery, either to avoid learning the VATS techniques, to see whether they can improve on their VATS outcomes, or simply as a marketing strategy. Because little comparative analysis has been published on robotic lung resection, more studies are needed to provide information about the safety and effectiveness compared with standard techniques, and to delineate the nuances of robotic surgery relative to VATS. Dr Louie and Dr Farivar disclose that they have a financial relationship with Intuitive Surgical by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg LOUIE ET AL 2012;93: ROBOT AND VIDEO-ASSISTED LUNG RESECTION 1599 Material and Methods We performed a case-control analysis of consecutive completely portal anatomic (lobectomy or segmentectomy) lung resections (CPR) and VATS anatomic lung resections. Data were collected between May 2009 and October 2011 using a prospective database of robotic cases and a retrospective review of patients charts for VATS cases. Consecutive patients were included if they underwent an anatomic lung resection, defined as a lobectomy or segmentectomy in which the individual branches of the pulmonary arteries and veins and the lobar or segmental bronchus were divided separately, during the study period. Resection was primarily for clinical stage (I or II) non small cell lung cancer, but also solitary pulmonary metastases and benign lung conditions such as bronchiectasis, congenital malformations, and localized fungal infections. We excluded patients requiring induction therapy, those with large, central lesions, and those with multiple positron emission tomography positive hilar or mediastinal nodes to facilitate our learning experience on the robot. Statistical comparisons were done with Student s t test, Pearson s 2, and the Mann-Whitney U test. The Swedish Medical Center Institutional Review Board approved the study. The four surgeons had a wide range of laparoscopic and advanced VATS training and experience. Surgeon 1 performed two VATS lobes during training, but has since gained significant experience in advanced laparoscopy and VATS lobectomy. Surgeon 2 trained in advanced laparoscopy and VATS lobectomy and started on the robot when he commenced practice. Surgeon 3 does not have any formal laparoscopic or VATS training, nor does he perform advanced laparoscopy or VATS lobectomy. Surgeon 4 pioneered the laparoscopic Hill repair, performed the first VATS lobectomy in Washington State, and has performed more than 250 VATS lobectomies. Preoperatively, all patients underwent computed tomography and pulmonary function testing. Cardiac assessment was individualized. Patients with a malignant diagnosis or suspicion of malignancy underwent computed tomography positron emission tomography imaging as appropriate and a preoperative diagnosis was established with a transthoracic needle biopsy or electromagnetic navigational bronchoscopy guided biopsy to avoid the need for diagnostic wedge biopsy at the time of resection. Staging cervical video mediastinoscopy was used in virtually all non small cell lung cancer cases before resection. Postoperatively, patients were managed on the thoracic surgery ward. Drainage tubes were removed when the output was less than 450 ml for a 24-hour period and there was no air leak. Multimodal analgesia with patientcontrolled intravenous narcotics, oral acetaminophen, or ketorolac, or a combination of these, was used to achieve a patient-reported pain level of less than 3 of 10 before discharge. Oral narcotics were initiated on the first postoperative day. After discharge, no activity restrictions or precautions were used for either group. Two dedicated advanced registered nurse practitioners, not blinded to the procedure and independent of the surgeon, determined the amount of narcotic patients required, and with the patient, made the decisions about returning to work or usual activities. They were taught in advance of our group, prospectively collecting robotic data to record these data at predetermined postoperative time intervals. They instructed patients about pain management strategies on discharge. At the first postsurgery visit (7 to 10 days), patients were instructed about weaning from narcotics if this had not already been initiated. Patients were allowed to return to usual activities if pain control was adequate and narcotic weaning had been initiated. Morbidity and mortality was graded and classified using the definitions outlined by Seely and colleagues [8]. This classification system has been validated specifically for thoracic surgery and lobectomy. To compare nodal evaluation between techniques, we used a predetermined lobe-specific nodal sampling as defined in the American College of Surgery Oncology Group Z0030 trial [9] and others [15]. For each lobe, if the station was sampled during the resection or had been sampled at mediastinoscopy, then the expected station was marked as successfully sampled. If pathologic examination did not confirm that a lymph node was sampled from a station determined by the surgeon, then the station was not considered successfully sampled. Operative Techniques VIDEO-ASSISTED THORACOSCOPIC SURGERY ANATOMIC RESECTION. Video-assisted thoracoscopic surgery resections were performed in the lateral decubitus position with the table flexed to bring the hip level with the chest wall. Pleural blocks were placed at incision and closure of the incisions using bupivacaine 0.25% with epinephrine. Access was established using 1-cm incisions at the sixth intercostal space anterior axillary line, fourth or fifth intercostal space behind the pectoralis major, and sixth interspace infrascapular, and a 5-mm camera incision was made in the ninth interspace along the scapular line. Extraction of the specimen in a bag was accomplished by extending the ninth interspace incision to approximately 5 cm or by extending the fifth interspace incision. Standard open and laparoscopic instruments were used. The interlobar fissures were taken with staplers and only dissected when a segmental resection was performed. ROBOTIC ANATOMIC RESECTION. A CPR-3 arm robotic resection was performed as previously described [10]. The first 30 resections used 2 attending surgeons: one on the console and one at the bedside. After that, a thoracic surgery fellow was used as the bedside assistant and given approximately 30 to 60 minutes on the console. Positioning was similar to VATS lobectomy with one exception; the operating room table was reversed to put the patient s head at the foot to allow for positioning of the robot. Anesthesia was positioned to the face side of the patient to facilitate access to the double-lumen tube and positioning of the robot over the head. Port

3 1600 LOUIE ET AL Ann Thorac Surg ROBOT AND VIDEO-ASSISTED LUNG RESECTION 2012;93: placement was also similar with the initial 8-mm incision placed in the sixth interspace anterior axillary line. Using the robotic 8-mm camera, the remaining ports were placed under direct visualization based on internal anatomy and external position in the following order: 12-mm camera port in the posterior axillary line in the ninth or tenth interspace; 8-mm robotic arm 2 inferior and posterior to the scapular tip at the level of the superior segment; and a 10-mm port in the fourth interspace for suctioning or retraction. Carbon dioxide insufflation was administered to a pressure of 6 mm Hg with a flow of 6 ml/min until the lung was deflated and then stopped. Four robotic instruments are used: Cadiere forceps, fenestrated bipolar, curved bipolar dissector, and L hook. Through the 10-mm port either a curved ring forceps for retraction or a suction device was introduced. The specimen was placed in an extraction bag and removed by enlarging the camera port to 5 cm, which we adopted from our historic VATS experience. Results Between May 2009 and October 2011, 52 CPRs were attempted. There were 3 conversions to thoracotomy: 1 was urgent for bleeding from a truncus arterial branch and 2 were not urgent for unexpected chest wall invasion and a positive parenchymal margin in the major fissure on frozen section. The pulmonary artery branch was controlled with a sponge gauze inserted through the 10-mm port, the robot was undocked, and the lobectomy was completed through a standard posterolateral thoracotomy without the need for transfusion. In total, 40 robotic lobectomies and 5 segmentectomies were completed. There was 1 conversion that was not urgent caused by failure to progress after a reasonable time limit that was completed by VATS. This case is analyzed in the robotic arm with intention-to-treat. In the VATS group, the most recent 35 resections (total of 407) performed during the study period were included with 1 conversion to open thoracotomy for poor visualization. In total, there were 27 VATS lobectomies and 7 segmentectomies. The groups have similar characteristics (Table 1) with the exception of the Eastern Cooperative Oncology Group performance status that favored the CPR group despite a similar American Society of Anesthesiology status. The majority of patients had non small cell lung cancer (Table 2). The distribution of resected lobes and segments was similar (Table 2). Key clinical and operative outcomes were similar, including tumor or lesions size, operative time, estimated blood loss, intensive care unit days, and length of stay (Table 3). There were no operative or 30-day mortalities in either group. The CPR group experienced 8 (17%) major morbidities versus 5 (15%) in the VATS group (Table 4). There were two post robotic hemorrhages. One was a presumed port site bleed that was taken immediately back to the operating room using VATS and no obvious source was identified. The second was presumed to be a staple line bleed after lobar resection of an infected and Table 1. Patient Demographics and Characteristics Characteristic Robotic VATS p Value Sex (male/female) 13/33 13/ Age (mean, y) Comorbidities (%) Coronary artery disease 21 (46) 11 (32) 0.23 COPD 16 (35) 7 (21) 0.17 Renal disease 2 (4) 0 (0) 0.22 Diabetes mellitus 6 (13) 6 (18) 0.57 Pulmonary function % FEV 1 (mean) % DLCO-V A (mean) Performance status ECOG (0/1) 35/9 19/ ASA (2/3) 20/23 11/ ASA American Society of Anesthesiology status; COPD chronic obstructive pulmonary disease; DLCO-V A diffusing capacity corrected for alveolar volume; ECOG Eastern Cooperative Oncology Group status; FEV 1 forced expiratory volume in 1 second; VATS video-assisted thoracoscopic surgery. bleeding pneumatocele. This was managed with instillation of lytic agents. Neither case required a transfusion. One bronchopleural fistula occurred after superior segmentectomy and completion lobectomy for a close margin 1 week later. The patient was treated for a dense pneumonia and had a pigtail catheter in place. Once her pneumonia resolved, the bronchopleural fistula healed without further intervention. There were 12 (26%) minor morbidities in the CPR group compared with 7 (21%) in the VATS group (Table 4). Prolonged air leaks were similar between the groups, but all of the robotic air leaks occurred early in the experience while becoming familiar with robotic touch. All 4 surgeons contributed cases to the robotic arm of the study. Surgeons with more VATS or minimally invasive surgery experience (1, 2, and 4) had shorter operative times compared with surgeon 3 with limited VATS or minimally invasive surgery experience (Table 5). However, surgeon 3 was still able to successfully complete each resection without complication and shorten his operative times with each successive case. The percentage of expected nodal stations sampled was similar between both groups (Fig 1). Patients undergoing CPR demonstrated a shorter duration of narcotic use after discharge (p 0.039) and an earlier return to work or usual activities (p 0.003) compared with the VATS group (Fig 2). Comment In this initial experience with completely portal robotic lung resection, we have shown that robotic resection can be performed safely with similar clinical outcomes when compared with a similar group of patients undergoing resection with a mature VATS technique. Although these groups are at opposite ends of the learning curve, this comparison is more appropriate than thoracotomy [6]

4 Ann Thorac Surg LOUIE ET AL 2012;93: ROBOT AND VIDEO-ASSISTED LUNG RESECTION 1601 Table 2. Diagnosis and Distribution of Lobes and Segments Characteristic Robotic VATS Diagnoses NSCLC Metastasis 3 2 Sequestration 1 0 Carcinoid 2 0 Pneumatocele 1 0 p 0.53 Lobes/segments Right upper Right middle 4 2 Right lower 5 5 Left upper 8 4 Left lower 5 5 Lingula 0 1 Superior segment 4 1 Upper lobe segment 0 3 Basilar 2 2 p 0.45 Table 4. Major and Minor Morbidity a Grade Complication Robotic VATS Major 8 (17%) 5 (15%) Grade IVa Acute renal failure 0 1 Respiratory failure 1 0 Grade IIIb Postoperative hemorrhage 2 0 Pleural effusion 1 0 Bronchopleural fistula 1 0 Bimalleolar ankle fracture 1 0 Grade IIIa Prolonged air leak 2 4 ( 5 days) Minor 12 (26%) 7 (21%) Grade II Prolonged air leak 4 2 (3 5 days) Atrial fibrillation 3 2 Pneumonia 2 1 Ileus 1 2 Urinary tract infection 1 0 Grade I Lobar collapse 1 0 a p VATS video-assisted thora- NSCLC non small cell lung cancer; coscopic surgery. VATS video-assisted thoracoscopic surgery. Table 3. Key Clinical Outcomes Characteristic Robotic VATS p Value Tumor or lesion size (cm), 2.8 ( ) 2.3 ( ) 0.07 median (range) Operative time (min, incision to close) Length of stay (days), 4.0 (2 21) 4.5 (2 22) 0.63 median (range) ICU stay (days) Estimated blood loss (ml) ICU intensive care unit; surgery. VATS video-assisted thoracoscopic and with the similar outcomes at this stage may favor robotic surgery after 400 resections. As a result, robotic and VATS lung resection can be considered for minimally invasive lung resection and currently should be seen as complementary minimally invasive techniques that may increase the number of minimally invasive lung resections performed. We consider this to be an analysis of a thoracic surgical group s learning curve for robotic surgery. Our experience differs from that of others [6, 7] in that 4 surgeons with widely varying levels of training and experience contributed robotic cases as opposed to reports of a single surgeon with no comparative analysis or comparison to thoracotomy. Despite these differences, the conversion rate and outcomes were similar to the singlesurgeon series [6, 7]. For experienced VATS surgeons, our data suggest that after six robotic cases operative times and length of stay could resemble their VATS experience based on surgeon 4 s experience. However, for traditional open surgeons with limited VATS experience, initial operative times may be 90 minutes longer than a VATS case in our experience and certainly longer than thoracotomy until greater experience and comfort is achieved [6]. We are encouraged by the fact that surgeons with varied minimally invasive experience can readily move to robotic surgery for lung resection, but the preference between VATS and CPR is surgeon dependent. Surgeons 1 and 2 favor robotic resection because of the enhanced vision, stability and maneuverability of instruments, and the difference in postoperative pain and return to usual activities. Surgeon 4 prefers VATS lobectomy despite being facile with the robotic technique. His rationale is that not every trainee will have access to a robot and trainees should be exposed to and be facile in both approaches to allow for conversion from CPR to VATS or continuation of a VATS wedge resection to VATS lobectomy because ad hoc conversion to a robotic lobe during a VATS case is laborious. Surgeon 3 remains more comfortable with open surgery. He offers robotic resection to early stage candidates; however, if robotic access requires rescheduling of other booked patients, will Table 5. Robot and Video-Assisted Thoracoscopic Surgery Cases and Robotic Operative Time by Surgeon Surgeon Robotic VATS Robotic OR Time (min) a a a p OR operating room; VATS video-assisted thoracoscopic surgery.

5 1602 LOUIE ET AL Ann Thorac Surg ROBOT AND VIDEO-ASSISTED LUNG RESECTION 2012;93: A B C RUL Robot VATS p 4R 94% 82% RML Robot VATS p 4R 100% 100% RLL Robot VATS p 4R 100% 100% LUL Robot VATS p 4L 75% 50% / LLL Robot VATS p 4L 100% 75% Fig 1. Percentage of expected nodal stations sampled by lobe. The percentage of expected nodal stations sampled is shown for patients with non small cell carcinoma. For example, the expected nodal stations to be sampled for a right upper lobectomy (RUL) during a video-assisted thoracoscopic surgery (VATS) or robotic procedure would be 4R, 7, 9, and 10R. (A) Right upper lobe; (B) right middle lobe (RML); (C) right lower lobe (RLL); (D) left upper lobe (LUL); (E) left lower lobe (LLL). delay the patient s operation, or the patient prefers not to wait until the next opening on the robot, open thoracotomy is completed. Two robotic techniques have been defined: 3-arm completely portal robotic lobectomy (CPR-3) and the 4-arm CPR-4 [6]. We believe our approach is consistent with a CPR-3 even though we use a 10-mm port for occasional suctioning and retraction. The operation is completed using incisions that are 10 mm or smaller, and the specimen is extracted by enlarging the lowermost incision only after the lobe or segment has been detached. In the transition from VATS to robotics, our D E CPR-3 ports may have several benefits over other described port placements. First, they may be more intuitive for VATS surgeons who are familiar approaching a resection primarily from anterior to posterior. Second, they allow the console surgeon time to get used to controlling the camera and two arms with the bedside surgeon assistant retracting through the fourth port. Third, traditional minimally invasive triangulation concepts may be more familiar for a novice robotic surgeon to visualize the anatomy. Finally, it facilitates conversion to VATS if the surgeon is frustrated, fails to progress, or has adopted the recommendation to set a time limit during the learning curve [6]. One of the purported benefits of robotic lobectomy is that the superior vision and stability will allow surgeons to perform an extensive lymphadenectomy [14]. Although our initial impression was that robotics would be an advantage over VATS because of the robotic platform, our data suggest that the ability to perform a mediastinal lymph node dissection is similar between VATS and robotic cases. This lack of difference may reflect the importance and significant effort placed on a thorough lymph node dissection regardless of the technique. Previous studies comparing mediastinal lymph node dissection have demonstrated equivalence between VATS and open, [11 13] and robotic and open approaches [6, 14]. But it should be recognized that these data are from centers at which the importance of lymph nodes is stressed, and the experience may not be generalizable to all surgeons using these different techniques. There are two other observations related to lymph node dissection. First, robotics has given us greater confidence in dissecting N1 lymph nodes adjacent to the pulmonary artery. This may ultimately have an impact on oncologic outcomes in the long term, but in the immediacy of the operation it allows easier and safer passage of the stapler. Second, assessing the adequacy of lymphadenectomy using the number of nodes resected was difficult to interpret particularly if the surgeon did not specify the number of nodes resected and the count was left to the Fig 2. (A) Duration of narcotic use is shown at set intervals by the number of patients who have ceased using narcotics since discharge during that time. (B) The number of patients who returned to usual activities or work is shown set intervals. (VATS video-assisted thoracoscopic surgery.)

6 Ann Thorac Surg LOUIE ET AL 2012;93: ROBOT AND VIDEO-ASSISTED LUNG RESECTION 1603 pathologist. Without a process to ensure accurate counts, we performed our analysis using methods from the American College of Surgery Oncology Group Z0030 study [9] and others [15]. Since then, we have instituted a new reporting process in which each complete node resected at lobectomy is counted by the surgeon and reported directly on a jointly managed pathologic reporting form for more accurate assessment and staging. One of the unexpected benefits from robotic surgery was the duration of narcotic use and return to usual activities. Initially, we had concerns about the larger ports causing more intercostal nerve irritation and pain. However, our pain evaluation suggests that robotic patients required fewer narcotics compared with VATS patients and this translated into an earlier return to usual activities. The increased pain after VATS surgery may be a result of more torque on the intercostal bundle during levering of the instruments. The EndoWrist movements of the robotic instruments probably lessen the torque by allowing articulation within the hemithorax rather than on the on the chest wall and ribs. Clearly, this evaluation is not satisfactory without a proper and blinded pain assessment including a visual pain scale and more accurate narcotic measurement. Definitive conclusions about pain will require more comprehensive evaluation and investigation. Nonetheless, we were encouraged by these findings. This issue of cost cannot be ignored when discussing the differences between VATS and robotic resection. The hospital s financial reporting system was used to gather casespecific cost for all included patients. There was no statistical difference between robotic and VATS resection, but there was an average of $1,400 difference in labor cost and $340 difference in supply costs favoring VATS. We decided to exclude the cost data after meeting with financial representatives because the methods for allocating the capital cost of the robot, the minimally invasive surgery suites, and other items were not consistent, and the methods of allocation were not consistent with known cost analysis methodologies and arbitrarily decided. This reinforces that cost analysis outside of a clinical trial or not specifically designed to answer a hypothesis may be subject to bias and misinterpretation. There remain a number of unanswered questions and controversies around robotic surgery that no doubt will be answered by further research as more surgeons gain experience in robotic surgery. We disagree with the assertion that a clinical trial is not possible because there is a lack of equipoise [6]. Our comparative series compares VATS and robotic surgery, and even though there are intangible benefits to robotic surgery, the outcomes we evaluated suggest that there is enough equivalence to consider a trial. Moreover, the issues of oncologic benefit, pain control, and especially cost-effectiveness can only truly be answered within a randomized, controlled trial. We do believe that robotic lobectomy may benefit from study in certain higher volume centers to determine relative benefits compared with standard thoracotomy and VATS before being more widely implemented. Early experience with complete portal robotic lobectomy and segmentectomy resulted in similar outcomes compared with a mature VATS technique. Surgeons with a wide range of training and experience in advanced VATS and laparoscopy can perform robotic surgery safely. A potential benefit of robotic lung resection may be less postoperative pain as measured by the duration of narcotic use and an earlier return to usual activities. Robotic lung resection should be studied further in selected centers and compared with VATS in a randomized fashion to better define its potential advantages and disadvantages in terms of oncologic efficacy, pain control, and cost. References 1. Boffa DJ, Allen MS, Grab JD, Gaissert HA, Harpole DH, Wright CD. Data from The Society of Thoracic Surgeons General Thoracic Surgery database: the surgical management of primary lung tumors. J Thorac Cardiovasc Surg 2008;135: Gopaldas RR, Bakaeen FG, Dao TK, Walsh GL, Swisher SG, Chu D. Video-assisted thoracoscopic versus open thoracotomy lobectomy in a cohort of 13,619 patients. Ann Thorac Surg 2010;89: Ashton RC Jr, Connery CP, Swistel DG, DeRose JJ Jr. Robot-assisted lobectomy. J Thorac Cardiovasc Surg 2003; 126: Bodner J, Wykypiel H, Wetscher G, Schmid T. First experiences with the da Vinci operating robot in thoracic surgery. Eur J Cardiothorac Surg 2004;25: Park BJ, Flores RM, Rusch VW. Robotic assistance for videoassisted thoracic surgical lobectomy: technique and initial results. J Thorac Cardiovasc Surg 2006;131: Cerfolio RJ, Bryant AS, Skylizard L, Minnich DJ. Initial consecutive experience of completely portal robotic pulmonary resection with 4 arms. J Thorac Cardiovasc Surg 2011; 142: Dylewski MR, Ohaeto AC, Pereira JF. Pulmonary resection using a total endoscopic robotic video-assisted approach. Semin Thorac Cardiovasc Surg 2011;23: Seely AJ, Ivanovic J, Threader J, et al. Systematic classification of morbidity and mortality after thoracic surgery. Ann Thorac Surg 2010;90: Darling GE, Allen MS, Decker PA, et al. Randomized trial of mediastinal lymph node sampling versus complete lymphadenectomy during pulmonary resection in the patient with N0 or N1 (less than hilar) non-small cell carcinoma: results of the American College of Surgery Oncology Group Z0030 Trial. J Thorac Cardiovasc Surg 2011;141: Farivar AS, Wagner O, Vallières E, Aye RW, Louie BE. Robotic right upper lobectomy. CTSnet, Available at expert_tech-46.html. Accessed January 31, D Amico TA, Niland J, Mamet R, Zornosa C, Dexter EU, Onaitis MW. Efficacy of mediastinal lymph node dissection during lobectomy for lung cancer by thoracoscopy and thoracotomy. Ann Thorac Surg 2011;92: Kondo T, Sagawa M, Tanita T, et al. Is complete systematic nodal dissection by thoracoscopic surgery possible? A prospective trial of video-assisted lobectomy for cancer of the right lung. J Thorac Cardiovasc Surg 1998;116: Scott WJ, Allen MS, Darling G, et al. Video-assisted thoracic surgery versus open lobectomy for lung cancer: a secondary analysis of data from the American College of Surgeons Oncology Group Z0030 randomized clinical trial. J Thorac Cardiovasc Surg 2010;139: Veronesi G, Galetta D, Maisonneuve P, et al. Four-arm robotic lobectomy for the treatment of early-stage lung cancer. J Thorac Cardiovasc Surg 2010;140:19 25.

7 1604 LOUIE ET AL Ann Thorac Surg ROBOT AND VIDEO-ASSISTED LUNG RESECTION 2012;93: Ishiguro F, Matsuo K, Fukui T, Mori S, Hatooka S, Mitsudomi T. Effect of selective lymph node dissection based on patterns of lobe-specific lymph node metastases on patient outcome in patients with resectable non-small cell lung cancer: a large-scale retrospective cohort study applying a propensity score. J Thorac Cardiovasc Surg 2010;139: DISCUSSION DR ROBERT J. CERFOLIO (Birmingham, AL): I am a little surprised at some of your findings and not by others I have not seen that much difference in pain between VATS (video-assisted thoracoscopic surgery) and the robot when I use the 30-degree camera but with the 0 I think it is less, but not much. You showed a difference. Maybe it s your port placement. Technically, our port placement is very different. With your camera down low and with arm 1 and 2 up high, do you have a problem getting the inferior pulmonary ligament or do you need a bedside assistant to pull the lung up and do you use the 0-degree camera most of the time, that tweaks the intercostal nerve less I think? DR LOUIE: In the majority of the cases we haven t had trouble, and as you and I have talked before, we push the arms up and we come straight down into the cul-de-sac. DR CERFOLIO: But you don t have case reports? DR LOUIE: No. DR CERFOLIO: That s number one. Number two, I think the biggest advantage of the robot is the lymph nodes, and although you showed a slight advantage to the nodes, it just wasn t statistically significant. Can you tell me why? Don t you think lymph nodes is often how tired the surgeon is? Are you doing the lymph nodes before you do the lobe or after? DR LOUIE: Well, in the group we tend to assess the lymph nodes, either open and VATS, before the lobectomy. So we do it up front, because we know that we are tired, and you and I have had this conversation at another meeting. Similarly, with the robotics we think taking out the nodes early is better because we get tired, and, number two, it facilitates the operation. We only looked at the N2 nodes, and I think our N2 node assessment, VATS and robots, is very similar, as has been shown by others. Where I think the robot has a significant advantage is if you are in the N1 stations, because our confidence on the pulmonary artery and taking lymph nodes off in station 11, station 10, as we get into the hilum is dramatically better with the robot, and that may actually facilitate segmentectomy, because I think we are better at segmentectomy on the robot than we are VATS. But I obviously don t have data to prove that. DR CERFOLIO: I think the biggest advantage of the robot outside the experience of true VATS expert is the robot lowers the bar and it allows more surgeons to be able to get more N2 nodes. DR LOUIE: And in the manuscript, if we had shared that with you, we made that same observation. I think all the series that compare VATS to open in terms of lymph node sampling are done by experienced and outstanding VATS surgeons. That may not be generalizable to the average thoracic surgeon, and the robot may be the game changer that you have been talking about for surgeons who aren t as adept at VATS or have concerns about the limitations of VATS, but who really want to get into minimally invasive surgery. DR CERFOLIO: Another finding that surprises me a bit was the operative times. Your VATS times were a little longer than most. I think the robot takes more time than VATS, especially at first, do you agree? DR LOUIE: I think it is probably a little unfair comparison. We have the fellow operating on both sides and it takes more time with a fellow, particularly on the robot. There is no question. DR CERFOLIO: A final question, a key question. Why did surgeon 2, who was doing VATS, now stop doing VATS and only does robotics? What are his or her real reasons why he has chosen the robot over VATS? DR LOUIE: Because when he got into practice after all the VATS lobes he did as a fellow, he came in, he saw the robot, and he was convinced, like you, that this is a way better operation. It is more stable. He finds it is unstable with the VATS platform. He has greater confidence taking the lymph nodes. We keep talking about a trial comparing VATS to robot at our organization. He has said no, I am not going to randomize patients. I am doing a robot. So we have had this discussion over and over. DR TODD L. DEMMY (Buffalo, NY): Two quick questions. As far as the scope that you use for the VATS, I noticed you use flexible tip scopes. Did you notice if some of the pain was from the scope port? Another quick question relates to the better performance status for the robot group. And we know that patients with reduced activity levels tend to be more vulnerable to chronic pain. Does that favor the robot? DR LOUIE: On the first question with the straight 30-degree scope, we haven t found that the pain issues are related to the camera incision. They seem to be higher up on our VATS cases. We have tried the flexible scope, and I have to agree, I think the flexible scope doesn t provide as much torque. And then your second point was? DR DEMMY: The performance status and pain measures. DR LOUIE: We were surprised that the ECOG (Eastern Cooperative Oncology Group) performance status favored the robot, because the ASA (American Society of Anesthesiology) status doesn t. We determined what the ECOG performance status was. Maybe we are not as correct. It may have something to do with that. We certainly put through some pretty sick people on the robot. The bimalleolar fracture I had was on a morbidly obese lady we did a segment on. Postop[erative] day one, she stood up and shattered her ankle because she weighs so much and spent a week in the ICU (intensive care unit) after her bimalleolar fracture, got a DVT (deep venous thrombosis) and the whole 9 yards. DR FRANK A. BACIEWICZ (Detroit, MI): A couple of questions and a couple of comments. First, I think you are fortunate to have 2 attendings doing a case with 1 at the operating table. In

8 Ann Thorac Surg LOUIE ET AL 2012;93: ROBOT AND VIDEO-ASSISTED LUNG RESECTION 1605 the real world, most surgeons have a general surgery resident or physician assistant at the operating table. That certainly will affect your time. My questions are, one of the few things I ve noticed after doing the robotic cases, and I have only done about seven of these with the robot, is that patients have significantly more secretions afterwards which hinders their recovery. I thought that the secretions were due to the length of the procedure and the time between stapling the vascular branches. The other thing is I have had two chyle fistulas to this point. The final question, I found that surgery on large patients or barrel chest patients is much easier. I tend to do it on that patient population much more than the average size patient. Thank you. DR LOUIE: With respect to your first question, even early in our experience when our operative times were longer, we didn t find the secretions to be a problem or have the need to bronchoscope people. Actually, I think they clear their secretions very well after the robot. We have not had to rebronchoscope any of the 46 we put through. The third comment, I think it s much easier to do a robotic lobectomy on a morbidly obese patient than a VATS. I would have to agree with that. And then your second point, the chylothorax, we have not seen that despite the fact that we harvest everything out of station 7, most things out of station 4. We are down taking out all the stuff around station 9. So even in the subcarinal space, we haven t seen coming across the chyle leak. Cerf, do you have a comment? DR CERFOLIO: We are so blessed with the nodes both open and robotically and do such an aggressive lymph node dissection with both with the robot you see so well so we have had trouble with both getting chylothoraces. We have done 400 robots now and 160 lobes and have had at least 2 chylothoraces, maybe 3. I don t know if it is high. But I would say that esophagectomy, there has been a couple of groups that have come to watch us and who have gotten too close to the left main stem bronchus because they see so well that they are trying to take the lymph nodes out, they are overdissecting the nodes. So I think that is a problem, because everything is 10 magnified and 3D (three-dimensional) and you see every node so much better than before. You can get a little obsessive and compulsive. So I do think you are going to see a higher incidence of chylothoraces. DR LOUIE: Well, until we recognized how close we are, that magnification, things always appear closer. It is kind of like the mirror says in your car. DR STEPHEN R. HAZELRIGG (Springfield, IL): I think with regard to the pain, I have to agree with Dr Cerfolio, I don t believe that there is any difference. I do not think the amount of narcotics used is an adequate way to compare pain. We have had a number of delayed chylothoraces after lung resection thoracoscopically. I suggest we have become so focused on getting so much nodal tissue, much more than we ever did before, and this is the reason for more chyle leaks. I have had 2 patients that developed a chylothorax almost 3 weeks after surgery. They did not have any chyle drainage when we pulled the chest tube. They went home and came back with a pleural effusion, which proved to be a chylothorax. So I am not sure that it has anything to do with the robot. I just think as we have become more and more focused on getting as many nodes as possible, we have increased the incidence of chylous leaks, which can happen in a delayed fashion. DR LOUIE: No, I would agree. We get focused on the nodes and I think that s the case. We readily admit that one of the limitations in this paper is the pain assessment, and it is very rudimentary, which is why we have softened our conclusions, and we have written that if we need to do a pain assessment, it should be a formalized one with pain scores, narcotics, counting, and all that stuff. DR DANIEL L. MILLER (Atlanta, GA): Over the last 2 years we have been using energy devices to perform our lymph node dissections. We use the Enseal device, which has the least dissipation of heat of all the devices. It seams to me that our postoperative pleural drainage is less and the incidence of chyle leaks is lower than with electrocautery alone. DR LOUIE: For our esophagectomies we have learned that the Harmonic is too hot, and we have had a bunch of airway injuries and we have had chyle leaks despite ligating the duct. DR JOSHUA ROBERT SONETT (New York, NY): I am not convinced that the LigaSure takes care of the lymph leaks, and if you are worried about the lymph leak, you are going to have to clip at the base at the sort of juicy area of tissue that you are leaving behind. Are you convinced that the LigaSure will seal a lymph leak? Would you go back into it and ligate a thoracic duct with a LigaSure? You are talking about lymphatic tissue now. You are not talking about a pulmonary artery or vascular vessel. I have assumed that it doesn t work, but I may be wrong.