Assessment of Robotic Thymectomy Using the Myasthenia Gravis Foundation of America Guidelines

Transcription

1 Assessment of Robotic Thymectomy Using the Myasthenia Gravis Foundation of America Guidelines Seth D. Goldstein, MD, and Stephen C. Yang, MD Division of Thoracic Surgery, Department of Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland Background. Robotic thymectomy is an emerging treatment for myasthenia gravis. However, the Myasthenia Gravis Foundation of America clinical research standards have been infrequently adopted in the surgical literature. Methods. Twenty-six patients underwent robotic thymectomy for myasthenia gravis between 2003 and 2008, performed by a single surgeon using the da Vinci system (Intuitive Surgical; Sunnyvale, CA) through a four-port right-sided approach. Results. Mean operative times were minutes of robotic system activation and minutes from incision to closure. There were no intraoperative or postoperative mortalities; the most common intraoperative complication was desaturation after single-lung ventilation, for which four procedures were converted to open. On histologic examination, there were five thymomas. The average follow-up after surgery was 26 months. Median preoperative and postoperative Myasthenia Gravis Foundation of America disease classifications were 2 and 0, respectively, reflecting a statistically significant decrease in symptoms (p < 0.01). Additionally, the average daily dose of cholinesterase inhibitor decreased by 63% postoperatively. Overall, 82% of patients improved and 18% were unchanged; no worsening disease was observed. Conclusions. Robotic thymectomy is a safe and efficacious treatment option for myasthenia gravis. There were no notable differences in patient demographics compared with previously published reports of open thymectomies. Furthermore, surgical and neurologic outcomes in this series compare favorably with conventional approaches in the literature. Of those with follow-up greater than 6 months, 82% of patients undergoing robotic thymectomy demonstrated significant clinical improvement postoperatively, indicating that this approach in concert with optimized medical management is an effective treatment for myasthenia gravis. (Ann Thorac Surg 2010;89:1080 6) 2010 by The Society of Thoracic Surgeons Accepted for publication Jan 7, Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5 8, Address correspondence to Dr Yang, The Johns Hopkins Hospital, Division of Thoracic Surgery, 600 N Wolfe St, Blalock 240, Baltimore, MD 21287; syang@jhmi.edu. Thymectomy as a treatment for myasthenia gravis (MG) was introduced by Alfred Blalock in 1939 [1] after he cared for a patient whose MG went into remission after removal of a cystic thymoma. Although thymectomy is currently an accepted treatment option for MG [2 4], there are long-standing debates regarding the ideal surgical approach. Over the past decade, the development of video-assisted thoracic surgery has led to less frequent use of the classic transsternal approach. The most recent innovation in minimally invasive thymectomy has been the use of robotic systems. Threedimensional binocular vision and multiarticulated instruments have helped to overcome some of the technical limitations of video-assisted thoracic surgery while retaining the improved cosmesis and decreased morbidity of minimally invasive surgery [5]. There are a number of published studies of thymectomy for MG employing a range of open and minimallyinvasive approaches [6 16]. However, these investigations are all retrospective reviews, and thus a consensus of best practice has never been reached. This is in part due to the fact that the incidence of MG is relatively low, at 1 in 10,000 to 15,000 [17]. Furthermore, the fluctuating extent and severity of the disease have historically made classification difficult. Studies have traditionally used a modified version of Osserman s criteria (Perlo and colleagues [18]), which relies on subjective assessments of mild, moderate, and severe disease. The lack of uniformity has led to variability between investigations that preclude meta-analysis [3]. Other classification systems have been proposed but were never adopted in the literature [6]. Recognizing this, in 2000 the Myasthenia Gravis Foundation of America (MGFA) released recommendations for MG clinical research standards [19], with a disease classification system reproduced in Table 1. Despite the distribution of the MGFA recommendations to the surgical and neurologic communities, subsequent studies have continued to use the Osserman and defilippi classification systems that the guidelines were meant to replace, particularly in the surgical literature [20]. Therefore, the purpose of this study is to report the efficacy and safety of robotic thymectomy for MG using the MGFA guidelines, to compare the data to published 2010 by The Society of Thoracic Surgeons /10/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg GOLDSTEIN AND YANG 2010;89: ROBOTIC THYMECTOMY USING MGFA GUIDELINES Table 1. Myasthenia Gravis Foundation of America Clinical Classification Class I Class II IIa IIb Class III IIIa IIIb Class IV IVa IVb Class V Any ocular muscle weakness May have weakness of eye closure All other muscle strength is normal Mild weakness affecting other than ocular May also have ocular muscle weakness of any severity Predominantly affecting limb, axial, or both May also have lesser involvement of oropharyngeal Predominantly affecting oropharyngeal, respiratory, or both May also have lesser or equal involvement of limb, axial, or both Moderate weakness affecting other than ocular May also have ocular muscle weakness of any severity Predominantly affecting limb, axial, or both May also have lesser involvement of oropharyngeal Predominantly affecting oropharyngeal, respiratory, or both May also have lesser or equal involvement of limb, axial, or both Severe weakness affecting other than ocular May also have ocular muscle weakness of any severity Predominantly affecting limb, axial, or both May also have lesser involvement of oropharyngeal Predominantly affecting oropharyngeal, respiratory, or both May also have lesser or equal involvement of limb, axial, or both Defined by intubation, with or without mechanical ventilation, except when employed during routine postoperative management. The use of a feeding tube without intubation places the patient in class IVb Reprinted with permission from Jaretzki III A, Barohn RJ, Ernstoff RM, et al. Myasthenia gravis: recommendations for clinical research standards. Ann Thorac Surg 2000;70: [19]. literature addressing traditional approaches, and to consider factors that may positively or negatively influence outcomes. Patients and Methods This study was a retrospective analysis of a prospectively kept database of 26 patients undergoing robotic thymectomy between 2003 and Approval was given by the Johns Hopkins Medical Institutions Investigational Review Board; a waiver was granted for patient consent. The surgeries were performed using the da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA) through a four-port right-sided approach, class T-2c in the MGFA thymectomy classification (Table 2). All patients were given a clinical diagnosis of MG by a neurologist and received preoperative thoracic computed tomographic (CT) scans and pulmonary function tests. The CT scans were carefully reviewed preoperative to assess for thymoma and to optimize approach. Quantification of preoperative and postoperative MG disease status was based on hospital patient records and direct questionnaires. Per the MGFA clinical research recommendations, preoperative disease classification was recorded as the most severe pretreatment status. Similarly, postintervention status follows criteria also set forth in the guidelines; most notably, pharmacologic remission is defined as symptom free after surgery following continued use of steroid or other immunosuppression without the need for a cholinesterase inhibitor. Patient records typically did not include sufficient information to compute a total quantitative MG score as outlined in the MGFA recommendations; however, fewer examination findings were necessary to sort into MGFA category. Typically this was done on the basis of time to ptosis on upward gaze, arm abduction time, and grip strength. Postoperatively, patients in pharmacologic remission were analyzed as a number zero with respect to numeric calculations. All patients were optimized medically by their referring neurologist prior to surgery. If intravenous immunoglobulin was part of the patient s pharmacologic regimen, it was given within two weeks of surgery. An epidural catheter was placed prior to induction of anesthesia. General anesthesia was administered avoiding the use of nondepolarizing neuromuscular blocking Table 2. Myasthenia Gravis Foundation of America Thymectomy Classification (Modified) T-1 Transcervical thymectomy a- Basic b- Extended g- Extended with partial sternal split d- Extended with videoscopic technology T-2 Videoscopic thymectomy a- Classic VATS (unilateral) b- VATET (bilateral with neck dissection) c- Videoscopic with robotic technology (unilateral) d- Videoscopic with robotic technology (bilateral) T-3 Transstemal thymectomy a- Standard b- Extended T-4 Transcervical & transsternal thymectomy T-5 Infra-sternal thymectomy a- Combined transcervical subxiphoid b- Videoscopic subxiphoid c- Videoscopic with robotic technology subxiphoid c- Infrastemal mediastinoscopy 1081 Reprinted with permission from Sonett JR, Jaretzki III A. Thymectomy for nonthymomatous myasthenia gravis: a critical analysis. Ann N Y Acad Sci 2008;1132: [21]. VATET video-assisted thoracoscopic extended thymectomy; VATS video-assisted thoracic surgery. GENERAL THORACIC

3 1082 GOLDSTEIN AND YANG Ann Thorac Surg ROBOTIC THYMECTOMY USING MGFA GUIDELINES 2010;89: agents, and a double-lumen tube or bronchial blocker was placed for single lung ventilation. In all patients, the approach was through the right side. The patient was placed in a supine position with arms tucked and a bump under the right chest. The table was then turned so the shoulder axis was at 45 degrees to the floor. The patient was carefully padded to avoid any pressure points on the face, and left arms and legs. The robotic equipment was brought in from the left side. Four port incisions were made as depicted in Figure 1. Typically, the left and right robotic arm ports were placed in the third and eighth intercostal spaces, respectively, along the midclavicular line; the camera (30 degree angled lens pointing posteriorly) through the fifth intercostal space along the anterior axillary line; and the working port through the fifth or sixth intercostal space along the mid-axillary line. After engaging the robotic system with left lung ventilation through the double-lumen endotracheal tube, the thymus was first carefully dissected off of the right phrenic nerve carefully preserving the neurovascular bundle. The innominate vein was then dissected inferiorly, clipping the venous drainage. After dividing the attachments to the posterior shelf of the sternum and to the pericardium, the left thymus was mobilized beginning at the inferior aspect and retracted toward the right pleural space. Using downward traction on the pericardium through the working port, this allowed adequate visualization of the thymus working cephalad, often visualizing the left phrenic particularly well with the angled camera. The pyramidal lobes were brought down from the neck and dissected off; the small branches from the inferior thyroid were cauterized if bleeding. The specimen was then placed into a bagged retrieval instrument and carefully brought out through the working port site. Lastly, a single 19F Blake drain was inserted through the most inferior port site (where the right robotic arm was placed), and guided to the apex along the posterior sulcus and looped to the anterior mediastinum. Fig 1. Robotic port site placement through a right-sided approach. (L left robotic arm port site; C camera port site; R right robotic arm port site; W working port site.) Fig 2. Frequency histogram of cohort Myasthenia Gravis Foundation of America (MGFA) classification preoperatively (brown) and postoperatively (green). Statistical analysis was performed using the log-rank test when appropriate. A cutoff of p equal to or less than 0.05 was considered as significant. Results The patient population was comprised of 26 patients (20 females and 6 males; mean age 39 years; range, 21 to 61) undergoing robotic thymectomy for MG between 2003 and 2008, performed by a single surgeon at an academic medical center. Mean duration of symptoms prior to surgery was 24 months, the average length of postoperative follow-up was 26 months. All patients had a confirmed diagnosis of MG by a neurologist; 75% were known seropositive for acetylcholinesterase antibodies. Mean operative times were minutes of robotic system activation and minutes total from incision to closure. There was no operative mortality; the most common intraoperative complication was desaturation during single-lung ventilation, for which four procedures were converted to conventional thymectomy through an upper sternal split; these patients were not included in the outcomes analysis. Other complications included bleeding from the internal mammary artery (n 2), postoperative flair of myasthenic symptoms (n 1), and an infected suture (n 1). The average hospital stay was 2 days (range, 1 to 4). Though only 4 patients had preoperative CT scans suspicious for small thymomas, a total of 5 were found on histologic examination of the resected specimens. Four patients were lost to long-term follow-up. Median (range) preoperative MGFA disease classification was 2 (1 to 5) (6% class 1, 59% class 2b, 12% class 3a, 18% class 3b, 6% class 5). The median postoperative MGFA classification was 0 (0 to 2) (62% symptom free, 15% class 1, 8% class 2a, 15% class 2b) (Figs 2; 3). Overall, the postoperative disease classification was significantly lower than preoperative (p 0.001, paired t test). The symptom-free group included 38% reaching pharmacologic remission (MGFA postintervention status PR), and 23% that were clinically asymptomatic on a cholinesterase inhibitor as well as some form of immunosuppres-

4 Ann Thorac Surg GOLDSTEIN AND YANG 2010;89: ROBOTIC THYMECTOMY USING MGFA GUIDELINES Fig 3. Individual patient preoperative and postoperative Myasthenia Gravis Foundation of America (MGFA) classifications. sion. None of the patients were in complete stable remission. Patients were typically managed on a regimen of pyridostigmine, prednisone, and mycophenolate mofetil. The average daily dose of cholinesterase inhibitor decreased by 63% postoperatively, from 225 mg per day to 83 mg per day. Overall, 82% of patients improved and 18% were unchanged; no worsening disease was observed. Comment 1083 The MGFA clinical classification retains the spirit of Osserman s criteria by separating pure ocular weakness from generalized symptoms. In addition, the MGFA guidelines further recognize manifestations of disease as predominately bulbar or predominately limb weakness. Categoric variables such as moderate remain; however, also provided is a rubric for objectively calculating a quantitative MG score, which is based on pulmonary function tests and 12 physical exam findings. The postintervention status permits documentation of continued immunosuppression and cholinesterase inhibitor use in addition to improvement or worsening of symptoms. This is an improvement on the DeFilippi classification of remission, which lacks the complexity to account for the severity of postoperative disease interpreted in the context of various degrees of medical management. Lastly, a framework regarding primary approach to thymectomy was offered by the MGFA. As the range of surgical options proliferate, it is important to clearly state the extent of the resection and the method by which it was achieved. Jaretski and Sonett [21] have offered modifications to this system, which subdivide the major categories based on method and degree of thymus removal in addition to including techniques reported since 2000 such as the one described in this series (Table 2). Neurologic outcomes in this series of robotic thymectomies compare favorably with conventional approaches in the literature [8]. Of those with follow-up greater than 6 months, 82% of patients receiving a robotic thymectomy demonstrated significant clinical improvement postoperatively, indicating that this approach in concert with optimized medical management is an effective treatment for MG. There were no cases of worsening disease long term nor were there any thymoma recurrences. The difference in MGFA preoperative and postoperative disease severity was statistically significant; however, the role of surgery in this improvement is not precisely known. Immunosuppression with prednisone and mycophenolate mofetil was continued in most cases postoperatively, and for 5 patients a regimen of mycophenolate mofetil was begun after recovery from surgery. However, as the length of average follow-up time increased, many patients in this cohort were noted to have a decreasing need for medical management, including immunosuppression and cholinesterase inhibitors. Thymectomy remains a widely accepted treatment option for MG, although there is a lack of rigorous evidence regarding optimal disease treatment. The controversy surrounding thymectomy is twofold. First, is thymectomy superior to optimized medical management with immunosuppression? Second, if thymectomy is indicated, which surgical approach will result in the best patient outcomes? This investigation falls under the heading of the latter, adding to literature demonstrating the safety and efficacy of a robotic minimally invasive approach. The first question is the subject of an ongoing international multicenter randomized clinical trial comparing thymectomy and prednisone therapy with prednisone alone [22]. Without evidence demonstrating the superiority of any one technique, the choice of surgical approach for thymectomy becomes a multifaceted decision, including consideration of factors such as anatomic access, operative complications, and available resources. One of the perceived drawbacks of robotic thymectomy is anatomic access to retrosternal tissue. It is thought by some groups that the robotic thymectomy may remove less thymus than the open approaches and will result in poorer outcomes [21, 23]. However, the extent to which that is clinically relevant has not been rigorously demonstrated. Incomplete gland resections have been demonstrated to be inferior to complete resections [14, 24], but it is unknown whether areas of microscopic or ectopic thymus that may remain after a minimally invasive procedure affect neurologic outcome. A single right-sided approach was used in this series, providing an adequate window into the anterior mediastinum that allowed resection of the thymus. Although criticism of our report avoided an additional left-sided approach, we felt there was adequate visualization of the left thymus. A bilateral approach could easily be implemented and forewarned to the patient preoperatively based on the CT scan if further exploration and dissection are desired. Decreased surgical morbidity is the driving force behind the trend toward minimally invasive thoracic surgical techniques. Robotic thymectomy compared with open theoretically offers faster recovery of pulmonary function, fewer wound infections, and a shorter average hospital stay. The average hospital stay in this series was 2 days, compared with 8 days in a recent report of transsternal outcomes [8]. There were no major postoperative surgical complications and only one patient received antibiotics GENERAL THORACIC

5 1084 GOLDSTEIN AND YANG Ann Thorac Surg ROBOTIC THYMECTOMY USING MGFA GUIDELINES 2010;89: for an infected suture. As with other minimally invasive procedures, high body mass index may adversely affect tolerance of single-lung ventilation and should be considered when offering surgical treatment options. This may be due in part to the fact that paralytic agents are not used for the anesthetic portion. Though this study was not designed to detect this, it was noted that the four procedures that were converted to open transsternal thymectomies involved patients with a mean body mass index of 36.1, compared with 27.0 in the robotic group. Other variables such as age, disease severity, and duration of symptoms preoperatively were not significantly different between those two groups. With these considerations in mind, the need for consistent reporting of outcomes should be highlighted. Myasthenia gravis is a protean disease with numerous manifestations and a fluctuating natural progression, making it intrinsically difficult to gauge the effectiveness of an intervention. This makes it all the more important to report data using the MGFA clinical research standards which will allow data to be pooled or approaches compared head-to-head in a consistent manner. To the best of our knowledge, there is currently only one paper using the recommended guidelines for each of the accepted approaches: transsternal [8], transcervical [16], thoracoscopic [12], and robotic [10]. As this body of literature increases, the consistency offered by the MGFA standards should allow for future meta-analysis comparison in anticipation of designing a prospective surgical approach trial. Ideally, as follow-up time postoperatively increases and patients are added to this cohort by receiving the same surgery, a Kaplan-Meier curve or other disease progression model will be used to track outcomes. This may allow discrimination of the benefit due to surgery to be differentiated from that of immunosuppression. It could additionally be revealing to administer a qualityof-life survey, as patient perception of disease burden may not scale directly with MGFA severity [25]. The small number of patients and the retrospective analysis of our own database are the main limitations of this study. Although a prospective, randomized study would be optimal, this would require wide participation; in particular, acceptance by referring neurologists. Bias is already present at the opposite ends of the surgical spectrum, as some neurologists still believe an aggressive resection through sternotomy remains the gold standard, while at the same time patients have become well-informed by word of mouth or the internet and actively seek out these minimally invasive techniques, whether indicated or not. Robotic thymectomy for MG appears to be a safe and efficacious surgery for the treatment of MG. It appears to offer similar treatment efficacy to traditional open approaches. This cohort seems adequately representative of MG patients presenting for thymectomy as there were no notable differences in the patient population compared with previously published reports. Robotic thymectomy should continue to be considered as a treatment option for MG in light of accumulating evidence confirming desirable outcomes and safety profile, although the need remains for controlled studies comparing surgical approaches. Irrespective of the technique, these procedures should still be performed at institutions with experienced thoracic surgeons well versed in all potential approaches in addition to neurologists familiar with MG, as these patients can quickly become unstable during the perioperative period. This study was supported by the Katz Foundation in Thoracic Surgery, at Johns Hopkins School of Medicine. References 1. Blalock A, Mason MF, Morgan HJ, Riven SS. Myasthenia gravis and tumors of the thymic region: Report of a case in which the tumor was removed. Ann Surg 1939;110: Jaretzki A, Steinglass KM, Sonett JR. Thymectomy in the management of myasthenia gravis. Semin Neurol 2004;24: Gronseth GS, Barohn RJ. Practice parameter: thymectomy for autoimmune myasthenia gravis (an evidence-based review): report of the quality standards subcommittee of the American Academy of Neurology. Neurology 2000;55: Saperstein DS, Barohn RJ. Management of myasthenia gravis. Semin Neurol 2004;24: Hashizume M, Konishi K, Tsutsumi N, Yamaguchi S, Shimabukuro R. A new era of robotic surgery assisted by a computer-enhanced surgical system. Surgery 2002;131(1 suppl):s Bulkley GB, Bass KN, Robert Stephenson G, et al. Extended cervicomediastinal thymectomy in the integrated management of myasthenia gravis. Ann Surg 1997;226: Cakar F, Werner P, Augustin F, et al. A comparison of outcomes after robotic open extended thymectomy for myasthenia gravis. Eur J Cardiothorac Surg 2007;31: Kattach H, Anastasiadis K, Cleuziou J, et al. Transsternal thymectomy for myasthenia gravis: surgical outcome. Ann Thorac Surg 2006;81: Rea F, Marulli G, Bortolotti L, Feltracco P, Zuin A, Sartori F. Experience with the da vinci robotic system for thymectomy in patients with myasthenia gravis: report of 33 cases. Ann Thorac Surg 2006;81: Ruc kert JC, Ismail M, Swierzy M, et al. Thoracoscopic thymectomy with the da Vinci robotic system for myasthenia gravis. Ann N Y Acad Sci 2008;1132: Shrager JB, Nathan D, Brinster CJ, et al. Outcomes after 151 extended transcervical thymectomies for myasthenia gravis. Ann Thorac Surg 2006;82: Tomulescu V, Ion V, Kosa A, Sgarbura O, Popescu I. Thoracoscopic thymectomy mid-term results. Ann Thorac Surg 2006;82: Wagner AJ, Cortes RA, Strober J, et al. Long-term follow-up after thymectomy for myasthenia gravis: Thoracoscopic vs open. J Pediatr Surg 2006;41: Zieliníski M, Kuzdzał J, Szlubowski A, Soja J. Comparison of late results of basic transsternal and extended transsternal thymectomies in the treatment of myasthenia gravis. Ann Thorac Surg 2004;78: Calhoun RF, Ritter JH, Guthrie TJ, et al. Results of transcervical thymectomy for myasthenia gravis in 100 consecutive patients. Ann Surg 1999;230: Yu L, Shan M, Jiang J, et al. Combined transcervical and unilateral-thoracoscopic thymectomy for myasthenia gravis: 2 years of follow-up. Surg Laparosc Endosc Percutan Tech 2008;18: Nicolle MW. Myasthenia gravis. Neurologist 2002;8:2 21.

6 Ann Thorac Surg GOLDSTEIN AND YANG 2010;89: ROBOTIC THYMECTOMY USING MGFA GUIDELINES 18. Perlo VP, Poskanzer DC, Schwab RS, Viets HR, Osserman KE, Genkins G. Myasthenia gravis: evaluation of treatment in 1,355 patients. Neurology 1966;16: Jaretzki A III, Barohn RJ, Ernstoff RM, et al. Myasthenia gravis: recommendations for clinical research standards. Ann Thorac Surg 2000;70: Jaretzki A III, Sonett JR. Evaluation of results of thymectomy for MG requires accepted standards. Ann Thorac Surg 2007;84: Sonett JR, Jaretzki A III. Thymectomy for nonthymomatous myasthenia gravis: a critical analysis. Ann N Y Acad Sci 2008;1132: Newsom-Davis J, Cutter G, Wolfe GI, et al. Status of the thymectomy trial for nonthymomatous myasthenia gravis 1085 patients receiving prednisone. Ann N Y Acad Sci 2008; 1132: Jaretzki A III. Thymectomy for myasthenia gravis: analysis of controversies - patient management. Neurologist 2003; 9: Zieliníski M, Kuzdzał J, Staniec B, et al. Extended rethymectomy in the treatment of refractory myasthenia gravis: Original video-assisted technique of resternotomy and results of the treatment in 21 patients. Interact Cardiovasc Thorac Surg 2004;3: Burns TM, Conaway MR, Cutter GR, et al. Less is more, or almost as much: a 15-item quality-of-life instrument for myasthenia gravis. Muscle Nerve 2008;38: GENERAL THORACIC DISCUSSION DR STEPHEN CASSIVI (Rochester, MN): I wanted to, first of all, thank the Association for the privilege of discussing this paper, and I want to start off by congratulating Seth on a superb presentation. I am going to open the discussion by noting that Seth is a medical student, and he worked with Dr. Steve Yang to do the research, analyze it, and present it here in a large room full of eminent cardiothoracic surgeons. So congratulations. I would also like to say that I think Dr. Yang s continued willingness to work with young promising prospects like Mr. Goldstein is not only commendable but in the true spirit of the STSA Inspiration Award of the Southern Thoracic Surgery Association. It is also courageous, Seth, to start your bourgeoning academic pursuits with the subject of myasthenia gravis. I am reminded of what was written by one of our Mayo Clinic icons in thoracic surgery, Jim Clagett, when he said, It is of considerable interest to find a relationship between a mysterious disease, myasthenia gravis, and a mysterious gland, the thymus. Your study is important as it underlines two important issues in the investigation of the treatment options for myasthenia gravis. Number one, and this is most interesting to us surgeons here, is what are the available approaches and can we define which are best in which circumstance? So my question on this is, what is the role of the robot? What does it add apart from the 50-minute setup time that you showed us in your presentation today? The robot is primarily a constructive instrument. In prostate surgery it is used to anastomose the urethra. It could be used to repair the mitral valve. I have seen my colleagues, Dr. Suri and Dr. Burkhart, at the Mayo Clinic do some very intricate mitral valve repairs using the robot. But thymectomy is primarily, if not completely, a dissection operation. So what does the robot add? And the second point is less interesting from a technical surgical standpoint but probably more important from a scientific perspective is how do we best measure treatment effect with this mysterious disease? You have mentioned in your presentation and in your manuscript that you measured the results after an adequate time of follow-up. What is that? I think what we have learned from some other eminent surgeons in this field, especially Larry Kaiser, is that maybe the best way to do this is with long-term follow-up using Kaplan-Meier remission graphs. So I would like to hear your comments on that. And once again, congratulations on the start of a great career. DR GOLDSTEIN: Thank you for your comments and for your insight into our manuscript. Your first question was regarding the actual utility of the robot to perform thymectomy. The quest for minimally invasive thymectomy is predicated on the assumption that pulmonary outcomes are improved when avoiding a median sternotomy, and to some extent, an upper sternal split or a transcervical incision. As I understand it, VATS (video-assisted thoracic surgery) thymectomy is a somewhat difficult procedure, because maneuvering around anatomic structures on a two-dimensional screen using straight graspers and dissectors can present a challenge. And so, really what the robot adds is a sense of three-dimensional perception as you sit in the console, as well as multiarticulated instruments that give extra degrees of freedom. And while you mention that it is most often used as an anastomotic tool, we believe that as a dissection device it facilitates removal of the thymus just as efficiently as our experiences with open approaches via an upper sternal split. The second question, how should treatment outcomes can be measured, I believe is the outstanding question in all of the myasthenia gravis literature, both neurologic and surgical. We gauged adequate follow-up in this cohort to be six months. Interestingly, there are well-done studies in the literature suggesting that improvement after thymectomy continues as long as three to five years postoperatively. Certainly in our series, with a median follow-up of 19 months, we expect subsequent improvement, and so I think this database is not yet complete. Drs Sonett and Jaretzki recently published a paper in the Annals of the New York Academy of Sciences suggesting, as you mentioned, that Kaplan-Meier survival may be the best way to track outcomes. Our patient cohort was too small to use that approach. And I would like to add to that that Kaplan-Meier in and of itself is not sufficient. One must remain faithful to the guidelines of the MGFA (Myasthenia Gravis Foundation of America) classification, because previous scales like the De Filippi classification of remission really do not contain the complexity to interpret myasthenia symptoms in the context of medical management and immunosuppression, which we all know has become increasingly sophisticated. And so I think the MGFA classifications within life table analysis using Kaplan- Meier plots is likely optimal for outcomes analysis. Again, thank you for your questions. DR KEITH S. NAUNHEIM (St. Louis, MO): Seth, great presentation and congratulations on a great research effort. I would say that the da Vinci robot reminds me a little bit of the laser. It is one of those instruments that you keep in the corner of the OR for use when it s appropriate. Patients really like the concept and often ask Can you use the robot and/or the laser? and you reply that, yes, if it is appropriate, I will utilize them. The robot is one of those items that seems like it is a solution in search of a problem. We have been working very hard to find appropriate uses for the da Vinci, whether it be mitral valve repair or coronary bypass surgery. There is even a paper on later in the meeting about its use in a Belsey antireflux

7 1086 GOLDSTEIN AND YANG Ann Thorac Surg ROBOTIC THYMECTOMY USING MGFA GUIDELINES 2010;89: procedure. But I think we have a duty to be intellectually honest and responsible when we address these questions as well. We all now realize that we have limited health care resources and the da Vinci is not exactly a cheap item. It has been estimated to add at least $1,500 of cost to every procedure by the time you add up the cost of the procedure, the disposables, the maintenance, et cetera. And so I think it is going to be very important for you and for other investigators using the da Vinci or similar robots to achieve not just similar bur significantly superior results. This is one area where the use of a noninferiority testing strategem will not be appropriate. If you use a da Vinci robot and add an additional $1,500 dollars of cost, you have to show $1,500 extra benefit. My question is whether within this preliminary trial, you building in cost estimates and collecting quality of life assessments. If we are really going to determine whether or not a da Vinci robot is worth that $1,500, we need to know about long term about quality of life issues, survival, and we really have to have accurate costs. Right now it is hard to really identify a great use for the da Vinci other than prostatectomy. DR GOLDSTEIN: Thank you. That point is well taken, and I understand that the evolution of the da Vinci system is not complete. I would argue, however, that it is a natural solution to the progression of limitations that have been encountered regarding thymectomy approach. It is hard to argue that a sternotomy is not to be avoided, particularly a complete median sternotomy, and thus many surgeons have yearned for a minimally invasive approach. As I mentioned, VATS has technical limitations that restrict its utility, and so, in a sense, robotic thymectomy is the answer to at least some of those questions. You are right, though, that the onus is on us to prove superiority given the extra cost. $1,500 is comparable to our estimates of what it adds, although I might argue that depending on how one looks at it, that money may be saved at the back end. Our mean hospital stay was two days, range one to four. That is far superior to the classic transsternal literature, which reported mean hospital stays of anywhere from six to ten days. So I think costbenefit should be a question in and of itself, certainly plausible to study. As far as quality of life, you are right, disease progression is not the only outcome, particularly considering that patients who seek out the robot perhaps are more satisfied because they got what they wanted. Myasthenia is a hard disease to assess, and I think there is certainly a role for quality of life surveys in these patients. DR M. BLAIR MARSHALL (Washington, DC): I echo Dr Naunheim s comments, excellent presentation. I would like to caution you, though, on equating a transcervical approach as an open approach, because actually that is done as an out-patient procedure, and so I think one has to consider those costs in as well. I agree that the exposure isn t as good as a VATS approach, but we have taken to adding a 5 mm camera with a 30-degree angle through the neck incision, and that gives you that AP [anteriorposterior] window exposure, which I think is the most difficult part of the transcervical approach. DR GOLDSTEIN: That is a great point, and I would just like to mention that if the ongoing randomized NIH (National Institutes of Health) trial does demonstrate a benefit of thymectomy plus prednisone over prednisone alone, these matters of optimal surgical approach are exactly the types of questions that we will be prospectively randomizing in order to study. Thank you.