|
|
||||||||
Division of Thoracic Surgery, Department of Cardiothoracic Surgery, Stanford University School of Medicine, Stanford, and Veterans Administration Palo Alto Health Care System, Palo Alto, California
* Address correspondence to Dr Shrager, Division of Thoracic Surgery, 2nd Floor Falk Building, Stanford Medical Center, 300 Pasteur Dr, Stanford, CA 94305 (Email: shrager{at}stanford.edu).
Presented at the 2nd International Bi-Annual Minimally Invasive Thoracic Surgery Summit, Boston, MA, October 9–10, 2009.
| Dr Shrager has no conflicts of interest to disclose.
|
| Abstract |
|---|
|
|
|---|
The autoimmune disease myasthenia gravis (MG) is one of the most common disorders of neuromuscular transmission, and its pathogenetic basis, natural history, diagnostic modalities, and means of pharmacologic therapy are well defined. Significant debate continues, however, over the role of surgery in the management of the disease. At the most basic level, the very effectiveness of thymectomy in MG remains a subject of debate. If one believes that thymectomy has a role, then one must entertain the next controversial issue: What is the ideal surgical approach to thymectomy? This review will focus upon results with one technique of thymectomy, the minimally invasive approach termed "extended transcervical thymectomy" (TCT).
The Myasthenia Gravis Foundation of America (MGFA) has classified the various surgical approaches to thymectomy for MG into four main categories and several subcategories according to the presumed extent of thymic resection performed [1]. These categories are the following: T1, transcervical thymectomy (including subdivisions T1a ["basic" TCT] and T1b ["extended" TCT], the latter being the subject of this paper); T2, videoscopic thymectomy (T2a ["classic"] and T2b ["video-assisted thoracoscopic extended thymectomy"]); T3, transsternal thymectomy (T3a ["standard"] transsternal and T3b ["extended"] transsternal); and T4, transcervical and transsternal thymectomy. Each of these procedures (with the exception, perhaps, of "basic" TCT) typically allows extracapsular resection of the thymus gland. The operations vary, however, in the extent to which extracapsular mediastinal and cervical fat (tissues that may contain variable amounts of ectopic thymic tissue) are excised. The MGFA Taskforce appears to have created this thymectomy classification with the intention that as one moved from T1 toward T4, there was a progressively greater amount of tissue resection being described. We would suggest, however, that when extended TCT is performed by an experienced operator, the extent of resection certainly equals that which can be performed by any videoscopic approach and may well equal many of the "extended" transsternal operations currently performed.
Outcomes after the various approaches to thymectomy have been reported across many studies in the form of complete and partial remission rates, generally using crude remission rates only. However, as also delineated in the recent MGFA Recommendations for Clinical Research Standards [1], response to thymectomy is most reliably judged by the rate of drug-free, complete stable remission (CSR) reported by Kaplan-Meier analysis. We will demonstrate that the CSR rate by Kaplan-Meier analysis after TCT is similar to results obtained by video-assisted thoracic surgery (VATS) and even to those obtained by the markedly more invasive, "maximal" transsternal-transcervical approach. Further, TCT has substantially decreased morbidity, length of stay, and presumably decreased costs versus any of the other available approaches.
| History |
|---|
|
|
|---|
Interestingly, surgical therapy actually developed before medical therapy. Also of interest is the fact that the first thymectomy for MG was in fact done transcervically. In 1912, Sauerbach removed a 49 gram hyperplastic thymus from a 19-year-old woman using the transcervical approach. This consisted apparently of only an enucleation of the thymus from within its capsule, but the patient's symptoms abated. In 1939, Blalock and associates removed a thymus gland by sternotomy for resection of a mass in the gland. This patient was also noted to experience improvement in her myasthenic symptoms, and by 1944 Blalock [6] had accumulated a total of 20 patients in whom he had performed thymectomy through median sternotomy for MG.
Sternotomy became the generally accepted approach to thymectomy for MG from the 1950s to the 1980s. This was despite (or perhaps partially as a result of) the work of Kirshner and colleagues (1969) [7] and Papatestas and colleagues (1987) [8], who championed a "basic" transcervical thymectomy that, without the benefit of a sternal-lifting retractor and thus with compromised exposure of the mediastinum, likely led to many cases of incomplete resection of the gland. Jaretzki and colleagues [9] in the 1970s established the presence of relatively high rates of extracapsular foci of thymic tissue in the adipose tissues of the mediastinum and neck, and on the basis of this finding they proposed and strongly advocated over many years an aggressive operation, "maximal transcervical-transsternal thymectomy," designed to remove all of these ectopic thymic foci [10].
In 1988, Cooper and colleagues [11] reported a modified approach to transcervical thymectomy which they termed "extended" TCT to emphasize the contrast with "basic" TCT reported by Kirshner and colleagues [7], Papatestas and colleagues [8], and others. Extended TCT involved use of a newly developed, sternum-lifting, self-retaining retractor which dramatically improved mediastinal exposure and allowed more complete removal of mediastinal thymic tissue and extrathymic fat. Since that time, this technique of extended TCT has reemerged as a suitable alternative to thymectomy through larger and more morbid incisions.
Despite the publication of multiple retrospective reports delineating outcomes after the various approaches to thymectomy, it remains highly controversial today which approach is optimal. It remains unknown whether the wider excision of extrathymic fat that is probably afforded by sternotomy or "maximal" approaches actually translates into improved response rates, and if so, whether the degree of difference in response rates is sufficient to overcome the clearly reduced morbidity of the transcervical operation.
| Indications |
|---|
|
|
|---|
Individual neurologists maintain widely divergent opinions regarding the role of thymectomy, but for those who believe that surgery has a role, the available literature generally supports thymectomy in the following patients: Osserman class I or II myasthenics whose symptoms, despite medication, interfere with their lives substantially; and Osserman class III patients between the ages of puberty and 60 years regardless of the effectiveness of medical therapy. Osserman class IV patients, although amenable to thymectomy, generally require aggressive medical management to control symptoms prior to receiving surgery. There appears to be a benefit to referring for surgery early in the course of the disease, as response rates are generally higher in patients who have had a shorter preoperative duration of disease.
For TCT specifically, there are certain criteria that must generally be met. Most importantly, the patient must be able to extend his or her neck to a reasonable extent, as the access to the mediastinum from the head of the operating table is impossible in a patient who cannot extend. This issue makes the operation difficult in some but not all elderly patients. Morbid obesity is also a relative contraindication, as neck extension and also the sternum-lifting essential to the operation may be compromised in such patients. Prior mediastinal surgery is a strong contraindication, and most feel that TCT is also contraindicated in the presence of known or suspected thymoma. It should be noted, however, that as experience with the operation grows, complete resection of small (<3 cm), clearly noninvasive thymomas can be performed with a low risk of spillage, as long as one has no hesitation to convert to sternotomy when needed [13].
| Operative Technique |
|---|
|
|
|---|
The superior poles of the thymus gland are located behind the strap muscles; generally the left superior pole is identified first as it tends to be larger, and this tubular structure is dissected cephalad to where it trails off into the thyrothymic ligament, where it is ligated and divided. A 0-silk ligature is left attached to the pole and is critical in allowing retraction during the subsequent dissection. This pole is then followed caudally to the origin of the opposite pole, which is subsequently dissected up into the right side of the neck in the same manner, and a long silk is left on this pole as well. Typically, both superior poles join above the sternal notch, and the thymus then passes anterior to the innominate vein into the mediastinum.
The cleido-cleido ligament is divided next, and the retrosternal plane is developed by digital dissection. The self-retaining, sternal-elevating retractor is then placed. With deflation of the inflatable bag beneath the shoulders, the sternum is maximally elevated, providing open exposure of the anterior mediastinum. Army-Navy retractors are usually placed at the two corners of the incision and tied to the retractor using Penrose drains to pull the skin edges out of the surgeon's view.
With the surgeon now seated at the head of the table working with a headlight, the thymus is next retracted forward and upward (using the aforementioned ligatures left on each of the upper poles), and the thymic veins draining into the innominate are identified. These are sequentially doubly ligated and divided. There are at least 2 and often as many as 4 of these veins. It is advisable to ligate rather than place clips upon these vessels as the space anterior to the vein will serve as an avenue for the remainder of the dissection, and clips are likely to be disrupted during this dissection.
The freeing of the gland and surrounding mediastinal fat is next carried progressively deeper into the mediastinum, largely by blunt dissection carried out under direct vision. The gland and surrounding fat is separated gradually from the anterior and lateral chest wall, the pericardium posteriorly, the pleurae laterally, and ultimately from the diaphragm. The dissection is taken as far as the phrenic nerves laterally. Ventilation is held intermittently to facilitate the view and the dissection off of the pleurae. The tissue to be resected is for the most part bluntly swept off of these structures until it is delivered through the incision. Vascular attachments encountered (for example, small branches of the internal mammary vessels) are clipped and divided as encountered. Dense attachment to the pleural membrane or pericardium can be easily managed with en bloc resection of portions of these when necessary.
Careful inspection of the specimen and the mediastinum for evidence of any residual, unresected thymic tissue is then performed to ensure that complete, extracapsular resection of the gland has been achieved. In rare cases where there is a question, frozen sections may be obtained to assure no residual thymic tissue. If there is a question of having performed an incomplete resection that cannot be reliably completed transcervically, there should be no hesitation to convert to partial or full sternotomy. In our most recent cases (since publication of our series), we have begun to place a 5 mm video camera along the lateral aspect of the incision as described by de Perrot and colleagues [14] to improve exposure in difficult cases. This provides both light and magnification, and perhaps most importantly, it facilitates teaching the operation to trainees.
Prior to wound closure, the dissection planes are inspected for hemostasis. When either pleural space has been entered, air can be evacuated with a red rubber catheter while holding a sustained positive pressure breath prior to completion of the closure. No drains are routinely left in place. Patients are nearly always extubated in the operating room and discharged home a few hours later after assuring a normal postoperative chest radiogram.
| Results |
|---|
|
|
|---|
Our recent report of 151 patients undergoing TCT for MG represents the largest published experience with the operation [15] and incorporates our previous report on the first approximately one-half of the patients in this series [16]. The mean age of patients was 42.5 years and 60.3% were female. Mean preoperative Osserman class was 2.3; 21.2% class I, 39.1% class II, 27.6% class III, and 12.2% class IV. Duration of symptoms was greater than two years in 31.4% of patients. At the time of operation, 75 patients were undergoing single drug therapy (pyridostigmine), and 51 patients were undergoing two drug or modality therapy (pyridostigmine and steroids or pyridostigmine and plasmapheresis). The remainder were receiving three or four drug therapy with or without plasmapheresis. On pathologic examination, the gland was normal in 38% of patients, showed follicular hyperplasia in 36%, thymoma in 8.3%, and other pathology in 11%. Mean postoperative follow-up was 53 months, and 97.4% of patients had complete follow-up.
The conversion rate from planned TCT to more extended incision, virtually always a partial upper sternotomy, was 7.9%. Seventy-three out of the last 74 TCTs were performed on an out-patient basis. The total complication rate was 7.3%, with 6.7% of these classified as minor (wound infections, seromas, atrial fibrillation, pneumothorax). The only major complication was a single case of unilateral recurrent laryngeal nerve injury.
Although we agree with the view that response rates after thymectomy are best reported as complete stable responses (CSR) in the form of Kaplan-Meier analysis, for the sake of comparison with previous studies we also reported our results in the form of crude cumulative response rates. We defined a "response" as a decrease of at least one Osserman class with either less or the same amount of medication, or maintaining the same Osserman class with a decrease in number or dose of medications. By this definition, 80.8% of patients responded to thymectomy. Mean Osserman class fell from a mean of 2.3 preoperatively to 1.0 postoperatively. Crude CSR rate defined as asymptomatic off medication for at least 6 months was 28.8%. When we broaden the definition of CSR to include patients who are asymptomatic but on low dose (< 10 mg prednisone or 150 mg azathioprine) single-drug immunosuppression, the rate rises to 37.1%.
We have felt that this latter, broader definition of CSR is perhaps the more appropriate one given the way our neurologists tend to manage these patients. Despite absence of symptoms for as long as several years after thymectomy, our neurologists are hesitant to completely stop immunosuppressive agents because of reports in the literature of relapses occurring in this setting [17]. Given the minimal morbidity of maintaining a patient on a single-drug regimen consisting of 5 to 10 mg prednisone every day or 100 to 150 mg azathioprine every day, several of the neurologists we have worked with will maintain patients on one of these drugs indefinitely after thymectomy even with prolonged absence of symptoms. Because many patients remain in this category of "asymptomatic-single low dose immunosuppressive" for prolonged periods while additional operated patients will be added to this group over time, the proportion of the entire cohort that this group represents will increase over time. Thus, with increased duration of follow-up, there will be a larger proportion of patients in the asymptomatic-low dose immunosuppressive group and a lower proportion in drug-free complete remission. If one considers only patients in drug-free CSR as true "CRs," then as time of follow-up accrues the CR rate will falsely appear to fall. As we believe it is likely that many of the patients in the asymptomatic-low dose immunosuppression group would remain asymptomatic off of those medications, we believe that they are appropriately included within the definition of a CSR.
More important than the crude response rates reviewed above are the CR rates by Kaplan-Meier analysis after TCT. By Kaplan-Meier analysis, CSR rates are 43% and 45% at 3 and 6 years using our broader definition of a CSR (Fig 1), and 33% and 35% at 3 and 6 years excluding the asymptomatic-low dose immunosuppression group (Fig 2). It is also important to note that no patient suffered a relapse after an initial CSR.
|
|
| Controversy |
|---|
|
|
|---|
Some may argue that VATS or robotic approaches are also "minimally invasive" and demonstrate morbidity and costs as low as TCT, but we believe that this is not the case. Transcervical thymectomy does not require even an overnight hospital stay, and patients typically return to their normal activities within a few days. Thoracoscopic and robotic thymectomy, as a result of the need for postoperative chest drainage and pain control for intercostal incisions, typically requires at least a 2 or 3 day hospital stay. All thoracic surgeons have had patients with chronic intercostal neuralgia after VATS incisions, whereas prolonged pain is unheard of after a central neck incision. Beyond the hospital stay, costs for VATS and robotic approaches will clearly be greater than for TCT if reusable ports and instruments are employed.
The real controversy that remains concerns whether the CR rates after TCT versus sternotomy approaches, and TCT versus VATS-robotic approaches, are the same or similar. If one could establish without a doubt that response rates after TCT are identical to those after thymectomy through sternotomy (and presumably, also as good as those after VATS-robotics), then it would be appropriate for TCT to be adopted as the preferred operation. All of the sizeable published studies reporting postoperative response rates, however, are retrospective and without control groups. We are thus left trying to compare results across studies which have enrolled patients widely differing across a variety of important preoperative factors, many of which have been shown to have an impact on response rates. Additionally, as we have discussed above, only very few studies report their results using Kaplan-Meier analysis.
Our opinion is that when the results obtained with TCT are compared with those obtained after transsternal thymectomy the complete remission rates are very similar. Table 1 shows selected results of some of the larger studies of thymectomy published within the last 25 years that used reasonable data reporting and analysis. This table gives a sense of the complete response rate ranges reported within the literature [10, 14, 15, 18–31]. In summary of this data it can be stated that crude CR rates range from 19 to 58% after sternotomy and 29 to 44% after TCT. Clearly these ranges are overlapping. Kaplan-Meier CR rates which, however, have been rarely reported (as far as we know they are reported exclusively in the qualifying publications that show a value in the last column in Table 1) also appear similar. It is worth noting that Kaplan-Meier estimates for CR at extended time points (not shown in table) are even higher (91% at 10 years) after one version of TCT [14] than after "maximal" thymectomy [10] (81% at 7.5 years).
|
Comparing CSR rates after TCT to those after VATS and robotic approaches can be difficult because there are so few VATS and robotic studies with both a reasonable number of patients enrolled (we chose at least n = 48 to list a study in Table 1) and full reporting of the important parameters at this early stage in the development of the techniques. Additionally, the VATS approach has been described with so many different technical variations (eg, right versus left versus bilateral approaches, addition of a cervical incision, addition of a sternum-lifting retractor) that it is hard to know which variation one should be evaluating. Overall, we believe that the literature shows an advantage of TCT over VATS in response rates, and there is only one carefully evaluated robotic series with which to compare. Other than the single study by Tomulescu and colleagues [29] listed in Table 1, crude CSR rates after VATS thymectomy cover a range from 14 to 40% (versus 29 to 44% for TCT), and Kaplan-Meier CSR rates after VATS thymectomy have either not been reported, or are so high (in small studies) as to be impossible to believe given all other published data. The largest report of robotic thymectomy [31] describes 106 patients and represents nearly one-quarter of the world's published experience. The 5-year Kaplan-Meier CSR rate in this study was approximately 40% (drawn from a figure in the publication), which does suggest that its results are at least equal those of TCT. The need for the robot, however, creates costs and other hurdles that currently prevent wide dissemination of this technique.
Only with a randomized, prospective study comparing TCT to sternotomy (and possibly VATS-robotic approaches) or at a minimum a carefully designed prospective registry, will these controversies be resolved. The latter is far more likely to be successful given the lack of equipoise that many surgeons have on this issue and thus the unlikelihood of accruing well to a randomized study. Neither study would be appropriate, however, until the primary question is resolved; ie, whether any form of thymectomy is more effective than medical therapy in MG or some subset of MG patients. This question should be answered by the ongoing multicenter trial that has been organized by the MGFA. If the results of that study show a benefit of thymectomy over prednisone, the next important step would be to organize a means of effectively evaluating TCT versus thymectomy through sternotomy (and possibly VATS-robotics). Such a study would find wide support within the surgical community.
As Jaretzki and colleagues have written [32], the optimal surgical technique for thymectomy in MG will be the one that "... balances extent of resection, morbidity, patient acceptance, and results." In the current environment, where when carries out a thymectomy for MG one is performing a potentially morbid operation that has no level I data to support it, minimizing morbidity and mortality becomes paramount. As it is clear that TCT is less morbid and costly than all other approaches to thymectomy, it is a very attractive option in this environment. We believe that CSR rates with TCT are as good or better than those after VATS approaches. We believe further that although the data are admittedly inconclusive, CSR rates with TCT are sufficiently similar to those after the extended transsternal and "maximal" thymectomy procedures to allow one to argue cogently that whatever small difference may exist in remission rates is likely outweighed by the larger difference in morbidity between the approaches.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. Marulli, M. Schiavon, E. Perissinotto, A. Bugana, F. Di Chiara, A. Rebusso, and F. Rea Surgical and neurologic outcomes after robotic thymectomy in 100 consecutive patients with myasthenia gravis J. Thorac. Cardiovasc. Surg., March 1, 2013; 145(3): 730 - 736. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. K. Freeman, A. J. Ascioti, J. M. Van Woerkom, A. Vyverberg, and R. J. Robison Long-Term Follow-Up After Robotic Thymectomy for Nonthymomatous Myasthenia Gravis Ann. Thorac. Surg., September 1, 2011; 92(3): 1018 - 1023. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |