D pects of the pathophysiology and treatment of thymomas

CURRENT REVIEW Thymoma: Results of 241 Operated Cases Giuliano Maggi, MD, Caterina Casadio, MD, Antonio Cavallo, MD, Roberto Cianci, MD, Massimo Molinatti, MD, and Enrico Ruffini, MD Department of Thoracic Surgery, University of Torino, Torino, Italy Clinical and histopathological aspects of 241 thymomas were reviewed. One hundred sixty of the patients with thymoma had myasthenia gravis and 15 had other autoimmune diseases; 55% of the thymomas were encapsulated and 45% invasive. Operation was radical resection in 87.5% of the patients, subtotal resection with residual tumor in 8.7%, and simple biopsy in 3.7%. A tumor relapse was observed in 24 patients (10%): 2 (1.5%) of 133 with encapsulated thymomas and 22 (20.4%) of 108 with invasive thymomas; among these patients, a relapse was found in 20.6% of the patients who received radiotherapy postoperatively and in 24.6% who did not. Adverse prognostic factors were clinical stage IVa (multiple pleural nodes), not feasible resection (for technical reasons), inoperable tumor relapse, and association with one of the following autoimmune diseases: pure red cell aplasia, hypogammaglobulinemia, and lupus erythematosus. Conversely, myasthenia gravis is now a curable disease; it contributes to early discovery of associated thymoma, thus allowing a better survival for patients with thymoma who have myasthenia gravis compared with patients with thymoma but without myasthenia gravis (p < 0.05). Postoperative radiotherapy does not seem necessary after removal of encapsulated thymomas, but it is advisable in case of invasive thymomas, regardless of the extent of the resection. (Ann Thorac Surg 1992;52:252-6) espite an impressive number of studies, many as- D pects of the pathophysiology and treatment of thymomas have not been clearly defined yet. In addition, the unique association with specific autoimmune diseases, notably myasthenia gravis (MG), represents a further source of difficulty in the interpretation of the results. Malignancy of thymomas has been definitely demonstrated and has been discussed with regard to invasiveness of surrounding tissues and to histology [l, 21. Therefore, most centers advocate the most complete resection possible, including the tumor, the thymus gland, and the surrounding fatty tissue as well [3]; the appropriate use of neoadjuvant or adjuvant therapy (particularly radiotherapy and/or chemotherapy) is still controversial [4]. Simple transsternal versus a more extended operative approach, the treatment of recurrences [5, 61, and the use of adjuvant plasmapheresis constitute additional items of controversy. The present retrospective study reports our experience with 241 consecutive operated thymomas to assess the most significant prognostic factors and the most appropriate therapeutical strategy. Material and Methods From 1959 to 1988, 241 consecutive patients with thymoma underwent operation at the Department of Thoracic Surgery of the University of Torino; there were 107 male and 134 female patients, ranging in age from 16 to 76 years with a mean age of 47 years. Address reprint requests to Dr Maggi, 39/1, Via Millefonti, 10126 Torino, Italy. Thymoma was defined as a tumor originating from the epithelial cells of the thymus [7]. Granulomatous masses (such as Hodgkin s disease), carcinoids, seminomas, lymphomas, and thymolipomas were excluded from the study. Associated diseases were myasthenia gravis (MG) in 160 patients (half of whom had a roentgenographically detectable tumor), pure red cell aplasia in 7, hypogammaglobulinemia in 2, and lupus erythematosus in 6. The tumor was asymptomatic and discovered after a routine chest roentgenogram in 31 patients; in 6 patients a superior vena cava syndrome was present due to the extent of the neoplasm. The surgical approach used was different for the patients with and without MG. In the former group we preferred a median longitudinal sternotomy (partial or total depending on the extent of the tumor), which allows us to remove both the thymoma and the thymus gland. In the patients without MG sternotomy was less used (44% of patients), and we used thoracotomy when the mass was omolateral. Complete resection of the thymoma and the thymic gland, including perithymic fat tissue if tumor invasion was apparent or suspected, was carried out in 152 patients (95%) with and 59 patients (72.8%) without MG. Five patients with and 16 without MG underwent subtotal resection. A simple biopsy was the only surgical procedure in 3 myasthenic and 6 nonmyasthenic patients. The method of surgical-pathological staging proposed by Masaoka and associates [8] was used to indicate the extent of tumor invasion at operation. Five stages were defined as follows: stage I: no capsular invasion; stage I1 capsular or pleural invasion; stage 111: invasion of surrounding tissues such as lung, pericardium, superior 0 1991 by The Society of Thoracic Surgeons 0003-4975/91/$3.50

Ann Thorac Surg 1991;51;1524 REVIEW MAGGI ET AL 153 Table 1. Distribution of Surgical-Pathological Stages Stage Myasthenia Gravis I I1 Ill IVa Present 88 28 36 8 Absent 45 6 17 13 100 f ----- NO MG 81 66.7 vena cava, and aorta; stage IVa: dissemination in the thoracic cavity; and stage 1%: distant metastases. Histologically, we adopted the classification of Verley and Hollman [9] into four types: spindle cell, lymphocyte-rich, epithelial-rich, and undifferentiated, according to the ratio of lymphocytes to epithelial cells and the cell shape. Table 1 summarizes the surgical-pathological staging in the presence or absence of MG. A greater percentage of patients without MG had stage IVa tumors, as compared with MG patients. Possibly, the tumor may have been detected earlier in these patients because of the clinical surveillance of MG. Table 2 indicates the histological classification of the tumor in relation to the presence or absence of MG. Survival rates were computed by the actuarial life-table method and the curve differences were tested using the log-rank test. A probability value less than 5% was regarded as significant. Results Fifty-two of 241 patients (21.6%) died. Seven patients died in the postoperative period, 5 of whom died of MG crises. Death caused by tumor occurred in 2 of 160 patients (1.2%) with and in 11 of 81 patients (13.6%) without MG. Other causes of death were MG crises in 14 patients, associated autoimmune diseases in 10, and unrelated causes in 9. The results were analyzed in terms of the presence or absence of MG. Follow-up was conducted in 234 patients, with a minimum follow-up time of 2 years (mean, 10 years). Overall Survival and Survival With and Without Myasthenia Gravis The overall survival curve shows that 80.8% of the patients are alive at 5 years and 75.6% at 10 years. The comparison between the patients with and without MG establishes that survival rates are higher in the presence of MG (p < 0.05) (Fig 1). 50. I 1 2 3 4 5 6 1 8 9 10year 4.4995 pe 0.05 Fig I. Overall 5-year and 10-year survival and survival according to the presence or absence of myasthenia gravis (MG). Survival by Patient Age Survival rates were calculated by grouping the patients in four age classes: 15 to 30 years, 31 to 45 years, 46 to 60 years, and more than 60 years. Although survival gradually decreases with age, as expected in a normal population distribution, the differences among the four groups are not significant. It is unlikely, therefore, that thymomas become more aggressive with age. Survival by Histological Stage We found no survival rate differences among the three differentiated forms (spindle cell, lymphocytic, epithelial), whereas the rarer undifferentiated tumors carried a poorer prognosis. Survival by Surgical-Pathological Staging Our results confirm the validity of Masaoka and associates' classification. Encapsulated thymomas (stage I) have the best prognosis. Thymomas with capsular or pleural invasion (stage 11) behave like those invading the adjacent structures (stage 111), whereas an intrathoracic dissemination (stage IVa) predicts a bad outcome. The differences are significant ( p < 0.001) (Fig 2). Survival by Extent of Resection When considering the three possibilities of interventioncomplete resection, subtotal resection, and biopsy only- 100 1 Table 2. Distribution of Histological Types Type With MG Without MG Total 1 (Spindle cell) 6 11 17 2 (Lymphocyte-rich) 49 19 68 3 (Epithelial-rich) 102 48 150 4 (Undifferentiated) 3 3 6 MC = myasthenia gravis. I L 39.6 IVa 21 - I,,,,,,,,,, 1 2 3 4 s 6 7 8 9 loyear Log rank (1, II+III. [Val X2 = 21.8445 p 4 0.001 Fig 2. Survival according to surgical-pathological staging.

154 REVIEW MAGGI ET AL ';( 100........'.. *-.... \... '\.--. \ 50 _. 26.7 J.,,,,,,,,, 1 2 3 4 5 6 7 0 9 10year Log rank X2 = 21.7556 p 0.001 Ann Thorac Surg 1991;51;152-6 We analyzed tumor recurrence according to the surgical-pathological staging. Encapsulated thymomas (stage I) 03.0 very rarely recurred (2 of 132 patients, 1.5%); in the... 5......... 'L -- -- -- 00.6.. present series, only 5 patients with encapsulated thymo- 71.7 71.7 mas received adjuvant radiotherapy and none had a recurrence.... rad. 211 Invasive thymomas recurred more frequently (22 of 108 patients, 20.4%; stage 11, 12.5%; stage 111, 29.7%, and ----- Subt. 21 stage IVa, 25%). Of the 108 patients with invasive thy- Fig 3. Survival according to the extent of resection (radical, subtotal, biopsy only). the survival rate curve differences are significant (p < 0.001) (Fig 3). It seems worthwhile, therefore, to strive to obtain the most complete resection possible during the removal of a thymoma; nonetheless, patients undergoing subtotal resection still have a survival rate of 70% at 7 years. Survival by Associated Autoimmune Diseases (Other Than Myasthenia Gravis) Survival curves from the 15 patients with thymomas in whom pure red cell aplasia, hypogammaglobulinemia, or lupus erythematosus were diagnosed before or after operation indicate that these autoimmune diseases carry a poor prognosis (p < 0.001). Their presence is therefore to be regarded as an important negative prognostic factor (Fig 4). Tumor Recurrence Tumor recurrence was observed in 14 patients with and 10 patients without MG. The time lapse between operation and the recurrence ranged from 1 to 13 years (mean, 5 years). Ten of the 14 MG patients with relapse had a second operation, 2 were irradiated, and 2 received chemotherapy. Of the 10 patients with tumor relapse without MG, 2 had a second operation, 6 had radiotherapy, and 2 had chemotherapy. -biopsy 9 moma, 34 had adjuvant therapy (18, radiotherapy; 6, chemotherapy; and 10, both) and 7 had recurrence (20.6%); in these patients, adjuvant therapy was used in 21 radical resections and in 13 subtotal resections, and a reduction in tumor recurrence was observed (3 and 4 recurrences, respectively) as compared with 55 radical and 6 subtotal resections with no adjuvant treatment (11 and 4 recurrences, respectively). Although the numbers are too small to be statistically significant, our results suggest that postoperative radiotherapy and chemotherapy may be useful in these patients, not only after subtotal resection, but after radical resection as well. Since 1986 we have started treating patients in stage I11 and stage IVa with preoperative radiotherapy (30 Gy) when there is clinical evidence of invasiveness. In 3 patients the mass greatly shrank and we could perform a subtotal resection after 3 weeks. Comment Thymoma is a histologically well-differentiated tumor in more than 95% of cases, and the carcinomatous variant is rare [9]. Histological classifications based on morphological criteria, such that of Verley and Hollman (91, indicate that spindle cell and predominantly lymphocytic forms are less aggressive than predominantly epithelial forms, but in the present series we found that the difference was not significant. The classification proposed in 1985 by Marino and Muller-Hermelink [ 101 into medullary, cortical, and mixed tumors has been adopted by several authors [ll, 121 and the results are promising. They demonstrate that medullary thymoma is a somewhat benign disease, whereas cortical thymoma must be considered a malignant tumor, and that a multidisciplinary approach (ie, postoperative radiotherapy in all patients and chemotherapy in select Fig 4. Survival in patients with pure red cell aplasia (PRCA), hypogammglobulinemia flpoy), and lupus eythematosus (LE) as compared with the remaining patients with thymoma. - 06.2 4- - 226 cases '. 25 ---- PRCA, IPOX L.E. - no L - -_ 15 cases 1 2 3 4 5 6 7 0 9 10year Log rank X2 = 30.1570 p 0.001

Ann Thorac Surg 1991;51;152-6 REVIEW MAGGI ET AL 155 cases) is warranted. Most tumors have a mixed pattern with an intermediate behavior. Studies are in progress at our department to reclassify our series according to this classification. Macroscopically, in the present series more than 50% of thymomas were encapsulated at the time of operation; prognosis is excellent for these patients and we observed only two recurrences (1.5%). When neoplasm spreads beyond the capsule, the classification of Masaoka and associates based on the extent of tumor invasion is valuable: survival curves show that locally invasive thymomas have a less favorable outcome compared with encapsulated forms. We found no survival rate difference between invasion of perithymic fat and pleura (stage 11) and involvement of adjacent lung or pericardium (stage 111); conversely, when multiple pleural or pericardial nodes are found (stage IVa), survival rates decrease significantly. The extent of the resection is another important prognostic factor: in our series radical resection was feasible for all patients in stage I and stage I1 and for 60% of patients in stage 111, but hardly ever for patients in stage IVa; nonetheless, a survival rate of 71% at 7 years was obtained in case of partial resection, while far less survival rates (26% at 8 years) occurred when the tumor was deemed to be unresectable. Thymoma is a low-grade malignancy and our data confirm that a "debulking" operation followed by postoperative radiotherapy achieves high survival rates; some authors [13] reported good results with radiotherapy alone. In the present study invasive thymomas recurred in 20.4% of the cases: 18% after radical resection and 42% after subtotal resection. Relapses were less frequent when the resection had been followed by radiotherapy or chemotherapy (20.6%) than when it had not (24.6%). This is more evident for subtotal resections, but it has been observed after radical resections as well. The role of reoperation in the management of recurrence of thymoma has not been clearly defined. In a recent report, Kirschner [ 141 demonstrated that reoperation is a safe procedure and that the presence of MG has no influence on reintervention: moreover, reoperation is a valid treatment after thymomectomy or incomplete thymectomy. Finally, in cases of initially unresectable tumor treated with adjuvant radiotherapy or chemotherapy, reoperation may be considered as a valid completion of a radical therapy. Our data confirm that survival is better when tumor recurrence was surgically treated than when alternative therapies alone (radiotherapy or chemotherapy) were used, even if the difference was not found to be significant (Fig 5). There is no general agreement whether radiotherapy should follow the removal of encapsulated thymomas [6]: for these tumors recurrence rarely occurs, and, on the other hand, the risk of pulmonary fibrosis should be kept in mind. Therefore, we think that postoperative radiotherapy should not be routinely performed after removal of encapsulated thymomas. In our series 5 of 133 encapsulated thymomas received postoperative radiotherapy, 100 4 - second surg 12 \54.5 ---_ RT + CHEM. 11 \ 40.9 40.9 50 1 L------- 1 2 3 4 5 6 years Log rank K2 = 0 2788 p N.S Fig 5. Survival of patients with tumor relapse comparing second operation uersus adjuvant therapy alone (RT + CHEM.). (N.S. = not significant.) and only 2 of 127 nonirradiated patients had recurrence: both underwent successful reoperation. Prognosis is poor when thymoma is found to be unresectable, or when an association with the autoimmune disorders pure red cell aplasia, hypogammaglobulinemia, or lupus erythematosus is present, as indicated by previous reports [15], but fortunately both these events are rare (3.7% and 4.1%, respectively, in our series). In the present study, MG was the most frequent cause of death among patients with thymoma and MG, whereas in the group of patients without MG death was most frequently caused by local progression of the neoplasm. Concerning the effect of MG on survival, our data are in accordance with those of other authors [16, 171, indicating that the presence of MG does not adversely affect the prognosis. Indeed, we found a more favorable prognosis in patients with thymoma and MG, compared with patients who had thymoma without MG (p < 0.05). This may be due to either the refinement in techniques for long-term management of patients with MG or the earlier detection of the thymoma, owing to the careful surveillance of the myasthenic patients. Several conclusions can be drawn from this study: (1) The finding of a mediastinal mass in a patient with MG is highly suggestive of thymoma. If the patient has no MG, we think that a preoperative diagnosis should be established because the treatment will be different according to the histological results. We generally use a parasternal mediastinotomy, with removal of the second costal cartilage. (2) Subtotal resection of invasive thymomas may yield surprisingly high survival rates, particularly if followed by postoperative radiotherapy. Our results demonstrate that partial resection is effective, and we think that the surgeon must strive to remove most of the tumor, even if some tumor fragments will be left behind. (3) Controversy persists over the optimal surgical approach for patients with thymoma. Our standard surgical approach is a median longitudinal sternotomy and complete resection of the thymus and perithymic fat from the sternal notch down to the pericardium. We think that this

156 REVIEW MAGGIETAL Ann Thorac Surg 1991;51;1524 access permits a thorough view of the great vessels, not limiting a possible exposure of the pleural space. No major complications occurred with this procedure; a focus of sternal osteomyelitis developed in 1 patient and eventually healed after the removal of a sternal stitch a few months after the operation. (4) The first line of therapy for thymoma is operation: radiotherapy or chemotherapy, alone or together, can yield long-term survival, but hardly ever complete remission. Neoadjuvant radiotherapy or chemotherapy, if clinical evidence of invasiveness exists, permits easier tumor excision; adjuvant therapies are indicated for all patients with invasive thymoma, regardless of the extent of the resection. With such a multidisciplinary approach, in the present series 86.9% of the patients with encapsulated thymoma, 62% with local invasion, and 40% with pleural metastatic seeding are alive up to 10 years. These results are extremely impressive for a potentially malignant and invasive neoplasm. References Bergh NP, Gatzinsky P, Larsson S, Lundin P, Rindell B. Tumors of the thymus and thymic region. I. Clinicopathological studies on thymoma. Ann Thorac Surg 1978;25:91-8. Monden Y, Tanioka T, Maeda M, Masaoka A, Nakahara K. Kawashima Y. Malignancy and differentiation of neoplastic epithelial cells of thymoma. J Surg Oncol 1986;31:1304. Nakahara K, Ohno K, Hashimoto J, et al. Thymoma: results with complete resection and adjuvant postoperative irradiation in 141 consecutive patients. J Thorac Cardiovasc Surg 1988;95:1041-7. Nakahara K, Fujii Y, Ikeda H, Kawashima Y. Surgery and irradiation in the management of thymoma. La Chir Tor 1989;42:211-6. 5. Fechner RE. Recurrence of non-invasive thymomas. Report of four cases and review of literature. Cancer 1969;23:142%7. 6. Monden Y, Nakahara K, Iioka S, et al. Recurrence of thymoma: clinicopathological features, therapy, and prognosis. Ann Thorac Surg 1985;39:165-9. 7. Rosai J, Levine GD. Tumor of the thymus. Atlas of tumor pathology. 2nd series. Washington DC: Armed Forces Institution of Pathology, 1976. 8. Masaoka A, Monden Y, Nakahara K, Tanioka T. Follow-up study of thymoma with special reference to their clinical stages. Cancer 1981;48:2485-92. 9. Verley JM, Hollmann KH. Thymoma. A comparative study of clinical stages, histologic features and survival in 200 cases. Cancer 1985;551074-86. 10. Marino M, Muller-Hermelink HK. Thymoma and thymic cardnoma. Relation of thymoma epithelial cells to the cortical and medullary differentiation of the thymus. Vichows Arch 1985;4O7119-49. 11. Elert 0, Buchwald J, Wolf K. Epithelial thymus tumors: therapy and prognosis. Thorac Cardiovasc Surg 1988;36: 109-13. 12. Ricci C, Rendina EA, Pescarmona EO, et al. Correlations between histological type, clinical behaviour and prognosis in thymoma. Thorax 1989;44:455-60. 13. Cohen DJ, Ronningen LD, Graeber GM, et al. Management of patients with malignant thymoma. J Thorac Cardiovasc Surg 1984;87301-7. 14. Kirschner PA. Reoperation for thymoma: report of 23 cases. Ann Thorac Surg 1990;49:550-5. 15. Batata MA, Martini N, Huvos AG, Aguilar RI, Beattie EJ Jr. Thymomas: clinicopathologic features, therapy and prognosis. Cancer 1974;34:389-96. 16. Shamji F, Pearson FG, Todd TRJ, Ginsberg RJ, Ilves R, Cooper JD. Results of surgical treatment for thymoma. J Thorac Cardiovasc Surg 1984;8743-7. 17. Lewis JE, Wick MR, Scheithauer BW, Bernatz PE, Taylor WF. Thymoma: a clinicopathologic review. Cancer 1987;60: 272743.