| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of General Thoracic Surgery, Catholic University, Rome, Italy
e Institute of Histopathology, Catholic University, Rome, Italy
f Institute of Neurology, Catholic University, Rome, Italy
b CDC, San Raffaele Velletri, Rome, Italy
c Department of Radiotherapy, Campus Biomedico University, Rome, Italy
d Unit of Molecular and Clinical Epidemiology, IRCCS San Raffaele, Rome, Italy
Accepted for publication August 31, 2009.
* Address correspondence to Dr Cusumano, Catholic University, Policlinico Gemelli Rome, Largo F. Vito 1, Rome, 000168, Italy (Email: giacomare55{at}hotmail.com).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: We reviewed a surgical series of 317 (1972 to 2007) patients with thymoma: clinical and pathologic features were analyzed as prognostic factors matched against the short- and long-term survival and recurrence rates.
Results: Male to female ratio was 153:164; median age, 49 years. Myasthenia gravis coexisted in 276 patients (87.1%). Thymomas were classified according to the Masaoka (42.0% stage I, 32.2% stage II, 21.5% stage III, and 4.4% stage IV) and the World Health Organization (3.5% type A, 9.5% type AB, 19.2% type B1, 57.7% type B2, 8.2% type B3, and 1.9% thymic carcinoma) staging systems. The resection was complete in 295 patients (93.1%). Operative mortality and morbidity were respectively 1.6% and 7.6%. No differences were recorded in postoperative outcome stratifying for myasthenia gravis or comorbidities. Mean follow-up was 144.7 ± 104.4 months. The overall 5-, 10-, 20-, and 30-year survival rates were 89.9%, 84.1%, 73%, and 58.6%, respectively. The completeness of resection (p < 0.001), the Masaoka staging (p = 0.010), and the World Health Organization classification (p < 0.001) all significantly influenced the long-term survival (univariate analysis). Only completeness of resection was significantly correlated with a better prognosis (p < 0.001) in multivariate analysis. Masaoka staging (p < 0.001) and World Health Organization classification (p < 0.001) significantly correlated with the disease-free survival in the univariate and multivariate analyses as significant prognostic factors (Masaoka, p < 0.001; World Health Organization, p = 0.011). Myasthenia gravis patients showed a better prognosis in terms of long-term survival (p = 0.046) and disease-free survival (p = 0.012) in the univariate analysis.
Conclusions: We confirm the evidence that the clinical staging and the histologic classification influence long-term survival. The presence of myasthenia gravis was not significantly related to operative outcome, but prolongs both long-term survival and disease-free survival.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Thymomas are often associated with paraneoplastic syndromes, with myasthenia gravis (MG) being the most frequent of these (30% of patients). Myasthenia gravis is an autoimmune disease characterized by the production of antibodies against the acetylcholine receptors at the neuromuscular junction. Muscular weakness and fatigability of skeletal muscles are the hallmarks of this disease [10]. Despite publication of extensive review papers on the treatment of thymoma, the effect of thymoma-associated features like MG on morbidity, mortality, overall survival, and disease-free survival is still a matter of discussion. The lack of prospective data, the rarity of this neoplasm, and the presence of confounding factors such as the frequent incidence of paraneoplastic syndromes delayed the definition of the guidelines for thymoma classification and treatment. In particular, contrasting results have been published about the possible role of MG on the prognosis of thymomatous patients [5, 11, 12]. Indeed, the presence of MG was considered for a long time as an adverse prognostic factor; quite surprisingly and unexpectedly, in recent studies, it was found to confer a survival advantage [13–18].
The aim of this retrospective study is to analyze the short- and long-term outcome of the surgical treatment of thymic tumors, taking into account the role of clinical, oncologic, and surgical features. Furthermore, we exploited the opportunity of a 35-year follow-up in a cohort of thymoma patients affected by MG who were observed and treated in a single institution to investigate the impact of MG as a predictor of postoperative and long-term outcomes.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Based on the information available from the clinical records, the definitive retrospective clinical staging was possible in 317 patients diagnosed with thymoma and operated on during a 35-year period (from January 1972 to December 2007) at the Catholic University Hospital in Rome, Italy. Demographic and clinical features including age, gender, comorbidity, and surgical reports were collected and taken into consideration in the statistical analysis. We also analyzed the behavior, in terms of postoperative mortality and long-term survival (LTS), of patients who had undergone surgery in the first 15 years compared with those observed and treated in the most recent 15 years.
Paraffin-embedded tissue blocks were retrieved for all patients from the hospital tissue archive, and new sections were cut for histologic review. Patients were reclassified according to the new diagnoses using the WHO classification [3].
The tumor stage was determined according to the modified Masaoka classification [2]. The extended transsternal thymectomy, as defined by Jaretzki and Wolff [19], was the surgical approach of choice including the complete resection of the thymoma, along with the thymic gland and perithymic fat tissue. Whenever appropriate, the resection of the pleura, pericardium, lung, phrenic nerve, and pleural implants was performed.
The indications to adjuvant radiotherapy were not uniform because of the long time span concerned and because some of the patients were observed and treated (received adjuvant treatment) in other centers. However, the main indications to adjuvant therapy were the confirmed infiltration of the thymic capsule or the pathologic evidence of thymic carcinoma. Radiotherapy was administered to the tumor bed (clinical target volume) plus a 1.5-cm margin (planning target volume) to a total dose of 45 to 55.5 Gy, with 1.5- to 1.8-Gy fractionation. Between 1974 and 1995, patients were treated with an anteroposterior–posteroanterior treatment field technique; thereafter, a mixed approach was adopted, with the anteroposterior–posteroanterior treatment field technique up to 30 Gy followed by an angled field technique for the remaining 15 to 25.5 Gy. Cisplatin-based chemotherapy was administered using doxorubicin, cyclophosphamide, and prednisone or etoposide. This treatment was offered to patients with stage IV disease and, more recently (from 2000 onward), to those with stage III disease.
Data of patients' outcome were obtained from the hospital records, telephone interviews with patients or their family members, and the general practitioners.
Percentage comparisons were made by the continuity-corrected 
2 test. Overall and disease-free survival were calculated by the actuarial method and compared by the log-rank test [20, 21]. Multivariate analysis was performed by the Cox proportional hazard model using a step forward selection progress. The following variables were evaluated as the starting set of covariates: sex, age, comorbidity, MG, Masaoka staging, WHO histologic classification, and completeness of resection (CR). A new categorical variable taking into account the time period (first 15 and last 15 years) was included and matched with postoperative outcomes and LTS. Results were considered significant if the probability value was less than 0.05. All analyses have been performed by the statistical software SPSS 13.0 (SPSS Inc, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
Twenty-four patients (7.6%) had postoperative complications: 6 (with MG) experienced respiratory failure, 5 pneumonia, 2 pulmonary embolism, 1 thrombosis of the subclavian artery, 4 hemothorax, 4 surgical wound infection, and 2 persistent air leak.
Perioperative mortality occurred in 5 patients (1.6%), and the causes of death were respiratory failure in 3 patients and cardiovascular complications in 2 patients. Among these patients, four events were recorded in the first 15 years while only one death occurred in the more recent 15 years, but this observed difference was not statistically significant (p = 0.13; hazard ratio, 0.18; 95% confidence interval, 0.02 to 1.65). As well, no statistical differences were recorded in the perioperative mortality and morbidity among the patients when these were stratified by the presence of MG, age, sex, stage, type of resection and presence of comorbidities.
Overall, postoperative radiotherapy was administered in 7 patients (5.3%) with stage I tumors, 40 patients (39.6%) with stage II tumors, and 58 patients (89.3%) with stage III tumors. All patients with metastatic thymoma (14 patients) received postoperative radiotherapy, chemotherapy, or a combination of both.
Survival
At the time of completion of the data gathering for the purposes of the analysis (December 2007), 65 patients (20.5%) were dead. Thirty-five deaths were attributable to intercurrent causes (cardiovascular events, n = 22; neoplasm other than thymoma, n = 9; other miscellaneous causes, n = 4); 9 deaths were attributable to MG, 6 deaths to complication of immunosuppressive therapies (gastroenteric bleeding, tracheoesophageal fistula, pneumonia, and sepsis) and the remaining 15 deaths (4.8%) were attributable to recurrent disease.
For the entire population 5-, 10-, 20-, 30-, and 35-year overall LTS was 89.9%, 84.1%, 73%, 58.6%, and 58.6%, respectively (Fig 2). No differences in terms of LTS were observed with respect to gender (p = 0.28), age (p = 0.54) and comorbidity, excluding MG (p = 0.08; Table 2). Indeed, no significant differences were found between patients treated in the more recent 15 years and those treated in the first 15 years in terms of LTS (p = 0.12), even if we observed a better survival in the first group (92.1%, 88.4%, and 88.4% at 5, 10, and 15 years, respectively, in the first group, and 90.2%, 83.3%, and 74.8% at 5, 10, and 15 years, respectively, in the second group). The actuarial 5- and 10-year survival rates were respectively 90% and 86% for patients with MG and 85% and 75% for patients without MG respectively (p = 0.046; Fig 3).
|
|
|
The WHO classification has been confirmed to be a powerful prognostic predictor (p < 0.001; Table 2). The best survival rates were recorded for those patients with type A and AB thymomas and the worst for B3 and thymic carcinomas, as expected. Overall 10- and 20-year survival rates were 100% for patients with subtype A tumors; 92% and 77%, respectively, for subtype AB; 82% and 73%, respectively, for subtype B1; 87% and 75%, respectively, for subtype B2; 64% and 42%, respectively, for subtype B3; and 27% for thymic carcinoma (Fig 3).
Regarding the CR, the actuarial 5-, 10-, and 20-year survival for patients with CR and with incomplete resection were 92%, 87%, and 77%, respectively, and 73%, 32%, and 21%, respectively (p < 0.001; Fig 3).
The multivariate analysis clearly showed that only the CR was an independent factor of worse outcome (hazard ratio, 5.40; 95% confidence interval, 2.97 to 9.85; p < 0.001). Among the 312 patients, 25 patients (8%) had a recurrence. The ratio was clearly higher (10 of 22; 45%) in patients with incomplete resection. In the population of CR patients (291 patients) only 15 recurrences occurred (5.2%).
Eleven patients had a mediastinal recurrence, 12 patients had a pleural recurrence, and 2 patients had pleural and pericardial recurrences. These recurrences were managed by surgery whenever possible and with radiochemotherapy in other cases. Five patients with mediastinal recurrence had a second pleural recurrence.
The univariate and multivariate analyses were performed considering first the whole population and then the CR one.
For the entire population the 5-, 10-, 20-, and 30-year disease-free survival rates were 97%, 93%, 86%, and 86%, respectively. The analysis of the survival curves demonstrates a statistical correlation in terms of the Masaoka staging (p < 0.000), the WHO histological classification (p < 0.000) and the presence of MG (p = 0.012; Table 2, Fig 4). Only the Masaoka staging was an independent prognostic factor (p < 0.000) in the multivariate analysis.
|
Finally, the analysis of the impact of recurrences on survival in the CR population showed no statistically significant differences in terms of survival among the patients with and without recurrence (p = 0.21).
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
It is widely recognized that surgery is the best treatment in patients with early stage thymic tumors with a good LTS [14, 22, 24]. Revisiting our more than 30 years' experience, we found that the most important independent positive prognostic factor was the CR. These results were not surprising; in fact Regnard and colleagues [22], in a study conducted on a large number of patients, confirmed the results of other previous reports [16, 25–28] and clearly identified the CR as the only significant independent prognostic factor in this class of patients. This evidence was so strong that they proposed to include the CR in a modified clinicopathologic staging system of thymomas, a proposal that we can support in light of our own experience.
Because of the importance of CR, the role of debulking surgery can be revised. Relatively few reports have been published [29, 30], and this shows a certain degree of distrust that puts doubts on this surgical strategy. Our experience shows that among patients treated with incomplete resection, only 10 of 22 (45%) had a clinically significant relapse of the disease. Nevertheless, in the absence of a comparative surgery versus exclusive radiochemotherapy analysis, we believe that the debulking surgery should be further investigated in patients with local advanced disease when this is associated with adjuvant therapy.
Unfortunately, in our experience, adjuvant therapy was administered adopting several different methodologies owing to the continuous upgrade of radiation therapy techniques and availability of novel drugs, and no definitive conclusions can be drawn as these would be supported by inhomogeneous data. Regarding the clinical determinants of the decision-making process in the treatment of thymomatous patients, many clinical [2, 31, 32] and histologic classifications [3, 33, 34] have been proposed and widely adopted in the recent past. However, the application of the clinical Masaoka classification and WHO histopathologic classification remains debated, and new classification instruments have been recently proposed by Chen and associates [35], as well as by ourselves [36]. The results of our experience confirm the importance of the level of extracapsular invasiveness and histologic classification as powerful factors determining the mortality and the incidence of tumor relapse. The Masaoka staging system and the WHO histologic classification showed a significant correlation with survival and disease-free survival; however, we are confident that the predictive power of the CR is stronger in terms of power to predict survival than the pathologic and histologic features (as clearly proven in the multivariate analysis). In terms of LTS, the concurrent presence of MG in patients with thymoma was considered a negative prognostic feature in many reported experiences prevalently published before the 1980s [11, 12]. From 1980 onward some reports have shown that the presence of MG should not be considered as a negative factor [13, 14]. Consistently though, the presence of MG in thymomatous patients was related with a better prognosis [15–18].
This survival improvement rate was directly linked with improvement of the postoperative respiratory support so as to improve the long-term pharmacologic treatment of MG. Maggi and colleagues [14] have shown that the 5- and 10-year survival rates of patients affected by thymoma and MG (85% and 82%, respectively) were significantly higher in patients with thymoma in the absence of MG (78% and 67%, respectively); similarly, Monden and coworkers [15] have shown that the survival curve, when the deaths for MG in patients with MG and thymoma were not considered, showed a significantly better behavior than that plotted with data from patients without MG. Additional evidence was provided by Okumura and associates [17]; in fact they reported that patients with thymoma and MG had a more favorable prognosis than patients with thymoma without MG (p < 0.07), but this association was not a prognostic independent factor in multivariate analysis. The reports from Maggi and coworkers [14] and Okumura and associates [17] suggest that the improvement in terms of survival in MG thymomatous patients could be directly correlated with an earlier diagnosis of thymoma obtained thanks to the appearance of MG symptoms. In this setting it is worth mentioning the experience of Kondo and colleagues [18], in which it was shown that thymomas in patients with MG in the earliest stages (I and II) are more frequently diagnosed than thymomas in patients without MG; on the contrary, when the LTS of patients with the same stage are matched no significant differences between the two groups have been shown, thus clearly suggesting the same biologic behavior of thymomas with and without MG. Our experience has shown that patients with thymoma and MG have an advantage compared with patients without MG, but differently from the experience of Kondo and colleagues [18], we did not outline any particular correlation among MG and the clinical stage of the disease (p = 0.9). In the past experience of our group, as reported by Evoli and associates [37], MG was associated with some particular histotypes (B1 and B2), whereas B3 and thymic carcinomas were more rarely associated with MG (p < 0.001; Fig 1). Indeed, other studies support the hypothesis that invasive thymomas are diagnosed more frequently in non-MG patients than in MG patients and suggest that thymomas in non-MG patients behave more malignantly [38]. This particular issue needs, as well, further research to be definitively assessed.
In agreement with our experience, other reports showed that a thymoma concurrent with MG has a lower rate of recurrence than in the absence of MG [22, 35, 39]. This evidence suggests a different biologic behavior between the two patterns.
In conclusion, we can affirm that the results of our work are not dissimilar to those reported in other series. The merits of this experience are associated with the relatively large series and the long-lasting follow-up. Moreover, the unusually high incidence of MG strengthens the hypothesis that MG, per se, does not represent an increased risk in the brief and long-term outcomes, but a protective factor in terms of LTS and disease-free survival in this subset of patients. This last aspect remains unclear and could be related to a closer medical follow-up that possibly facilitates the earlier detection and better management of other medical problems, therefore prolonging survival. However, on the other hand, a biologically more aggressive pattern of growth of thymomas in patients without MG could be responsible for worse prognosis. Further research is needed.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Based on your experience, is there any surveillance that we should be doing after doing a resection for thymoma, and, if so, what, and, if so, for how long should we be doing it? It looked to me like your curves flattened out at some point and I didn't quite catch where that flat point was. Is there a point when we can maybe stop getting CT (computed tomography) scans, or should we be doing that at all?
DR MARGARITORA: In our institution the neurologists take care of the patients' follow-up since more than 90% are myasthenic. They have set the currently scheduled protocol that foresees an entire set of exams, including CT scan of the thorax once a year for the first 10 years, and after that, every 2 years. As you correctly outlined it is probable that a careful analysis of the survival and disease-free curves could induce a rethinking of the follow-up protocol in the near future.
DR BRYAN F. MEYERS (St. Louis, MO): The fraction of your patients who have myasthenia is quite high based on the comparative literature search that you reported to us and I wonder if you can account for that. Also, the good results that you report must have a lot to do with the way you manage your myasthenic patients perioperatively. Would you just review that in some detail for us about how you prepare a myasthenic patient for an operation?
DR MARGARITORA: Thank you for the question. The reason why we have so many myasthenic patients in our series is because they are referred to us by one of the most important and big centers for myasthenia gravis in Italy. As well, the clinical management of these patients is left entirely to the neurologists. I cannot therefore enter into the details of the preoperative pharmacologic therapy used in our center by our neurologists. Of course I assume that the optimization of the preoperative preparation along with a careful intra[operative] and immediately postoperative management reduces the early morbidity and sets the basis for a long-term good result in surgical and neurological terms.
DR DONALD E. LOW (Seattle, WA): As Dr Meyers pointed out, your percentage of myasthenic patients is very high, but you also identify that in that same population you had a high percentage of Masaoka I type tumors. These tumors, when unrelated to myasthenic symptoms, can get quite large before they become symptomatic. Is it your assessment that it's the myasthenic symptoms that led to the improvement in outcome because you were finding these tumors much earlier?
DR MARGARITORA: Thank you for the question. Of course we assume that the clinical management of myasthenia induces a high rate of early detection of concurrent thymomas, especially due to the fact that chest imaging is routinely administered to every patient being admitted with a suspect diagnosis of myasthenia.
The second part of the question?
DR LOW: That was the question.
DR MARGARITORA: Okay. We acknowledge that the clinical response to thymectomy in patients with thymoma is slightly worse that that of patients with a simple thymus hyperplasia. This is a well-known information.
DR ERIC VALLIERES (Seattle, WA): Stefano, that is a very good paper, with superb follow-up; congratulations.
This question is not related to the data that you just presented, but I would like to have your opinion and the opinion of your colleagues about VATS (video-assisted thoracoscopic surgery) thymectomy for myasthenia gravis. What is the role of a VATS approach to MG (myasthenia gravis) in Rome?
DR MARGARITORA: We presented our data on hyperplasia last fall in Bologna at the ESTS (European Society of Thoracic Surgeons) meeting. Without any question, our policy is to perform a radical thymectomy in every case of thymoma in the classical way with a median sternotomy. Because of the high percentage of myasthenic patients, we have to perform the thymectomy and we have to remove all of the mediastinal fat to cure the myasthenia: this is why we adopt the transsternal approach. With hyperplasia, it is different. I think that you can cure the myasthenia by performing thymectomy in any way that you want, by monolateral or bilateral VATS. In this setting we use a mini-invasive access with a minimal skin incision at the level of the xiphoid; the operation is therefore carried out, always through a sternotomy with the aid of the thoracoscope. We have described this approach and we are about to publish our long-term results on a large cohort of patients. These are in line with those related to the classical open transsternal approaches. So I think that you can adopt the technique in which you are confident. You can use the robot if you have a robot in the hospital. In Italy, Dr Federico Rea has a good experience with robotic-assisted thymectomy. You can do it, absolutely. I think that the most important goal is to have a good, complete remission rate of the disease, and it is not important the way that you opt for to obtain it.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. J. Kelly, I. Petrini, A. Rajan, Y. Wang, and G. Giaccone Thymic Malignancies: From Clinical Management to Targeted Therapies J. Clin. Oncol., December 20, 2011; 29(36): 4820 - 4827. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sakamoto, T. Murakawa, C. Konoeda, Y. Inoue, K. Kitano, A. Sano, M. Fukayama, and J. Nakajima Survival after extended thymectomy for thymoma Eur J Cardiothorac Surg, October 18, 2011; (2011) ezr026v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Margaritora, A. Cesario, G. Cusumano, F. Lococo, V. Porziella, E. Meacci, A. Evoli, and P. Granone Single-centre 40-year results of redo operation for recurrent thymomas Eur J Cardiothorac Surg, October 1, 2011; 40(4): 894 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Li Myasthenia Gravis and Recurrent Thymoma Ann. Thorac. Surg., July 1, 2011; 92(1): 406 - 406. [Full Text] [PDF]
|
|
|
|
|
|
S. Margaritora, G. Cusumano, F. Lococo, and P. Granone Reply Ann. Thorac. Surg., July 1, 2011; 92(1): 406 - 407. [Full Text] [PDF]
|
|
|
|
|
|
E. Ruffini, P. L. Filosso, C. Mossetti, M. C. Bruna, D. Novero, P. Lista, C. Casadio, and A. Oliaro Thymoma: inter-relationships among World Health Organization histology, Masaoka staging and myasthenia gravis and their independent prognostic significan a single-centre experience Eur J Cardiothorac Surg, July 1, 2011; 40(1): 146 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Fabre, E. Fadel, S. Mussot, O. Mercier, B. Petkova, B. Besse, J. Huang, and P. G. Dartevelle Long-term outcome of pleuropneumonectomy for Masaoka stage IVa thymoma Eur J Cardiothorac Surg, May 1, 2011; 39(5): e133 - e138. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Wick Histopathologic Prognosis of Thymomas: Another Example of Medical Surrogacy Am J Clin Pathol, November 1, 2010; 134(5): 703 - 705. [Full Text] [PDF]
|
|
|
|
 |
|
 R. T. Cheney The Biologic Spectrum of Thymic Epithelial Neoplasms: Current Status and Future Prospects J Natl Compr Canc Netw, November 1, 2010; 8(11): 1322 - 1328. [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 |
