- Email: [email protected]
Thymoma: Results with complete resection and adjuvant postoperative irradiation in 141 consecutive patients
J
THoRAc CARDIOVASC SURG
1988;95:1041-7
Thymoma: Results with complete resection and adjuvant postoperative irradiation in 141 consecutive patients A series of 141 patients with thymoma was studied with regard to the way in which complete resection followed by postoperative radiation therapy influenced prognosis according to the stage and histologic type of thymoma. Postoperative radiation therapy (30 Gy in 3 weeks to 50 Gy in 6 weeks) was performed in 73.1 % of the patients. Thirty-five of 48 patients with thymoma invading the surrounding tissues (stage III) underwent complete resection, with survival rates of 100% at 5 years and 94.7% at 10 and 15 years. The prognosis in these patients was comparable with those in 45 patients with no invasion (stage I) and in 33 patients with capsular invasion (stage llI), all of whom underwent complete resection. Complete resection was done in 18 of 26 patients with epithelial cell thymoma, and there were no deaths by tumor until 15 years after the operation. There were no statistically significant differences in the survival rates of patients undergoing complete resection according to cell type (33 of 36 patients with lymphocyte predominant type and 61 of 77 patients with mixed cell type). The survival rate in six patients with epithelial cell type who underwent subtotal resection was not significantly different from that of 12 patients who underwent biopsy alone. Our findings indicate that complete resection of thymoma followed by postoperative radiation therapy results in a "benign" postoperative course, regardless of the stage and histologic type. Therefore, an aggressive surgical approach, such as resection of the superior vena caval system followed by reconstruction with a ringed polytetraftuoroethylene graft and/or complete pleuropneumonectomy, is justified for advanced thymoma, although the long-term results of such extended operations are stiU unclear. Kazuya Nakahara, MD, Kiyoshi Ohno, MD, Junpei Hashimoto, MD, Hajime Maeda, MD, Shinichiro Miyoshi, MD, Masami Sakurai; Yasumasa Monden, MD,b and Yasunaru Kawashima, MD, Osaka and Tokushima, Japan
Malignancy of thymoma has been discussed mainly with regard to the invasiveness of the tumor to the surrounding tissues at the time of the initial treatment':' and the histologic type.r" However, uncommon recurrence or dissemination of noninvasive thymomas, once removed completely, was reported by Fechner," who compiled eight other cases from the literature. The same experience in our institution led us to believe that thymoma is malignant, or at least potentially malignant. Therefore, we surmise that it is mandatory to resect the From the First Department of Surgery and Department of Pathology,' Osaka University Medical School, Osaka, Japan, and the Second Department of Surgery,' Tokushima University Medical School, Tokushima, Japan. Received for publication April 30. 1987. Accepted for publication July 29, 1987. Address for reprints: Kazuya Nakahara. MD, First Department of Surgery, Osaka University Medical School, Fukushima-ku, Osaka. 553, Japan.
tumor completely, including normal thymic tissue and invaded tissues, and to use postoperative radiation therapy. This therapeutic approach provided a "benign" long-term outcome in many patients with thymoma even thought the tumor had been invasive, especially when the tumor was removed completely. The purpose of this study is to determine how complete resection of the tumor affects the survival rate in terms of the stage of thymoma and histologic type, which have been thought to define the malignancy of thymoma. Patients and methods From 1957 to 1985, 142 patients with thymoma were treated. Seventy-nine patients were male and 63 were female. The ages ranged from 8 months to 73 yearssand the mean age was 44 ± 14 (standard deviation) years. Thymoma was defined as tumor originating from the epithelial cell of the thymus! Granulomatous thymoma, carcinoid, seminoma, malignant lymphoma, and mediastinal lung cancer were carefully excluded by a pathologist (M.S.). The associated diseases were myasthenia gravis in 81 patients, 1 n 41
The Journal of
1 042
Nakahara et al.
Thoracic and Cardiovascular Surgery
Table I. Surgical procedures. stage. and histologic type in patients with thymona, with or without myasthenia gravis (MG) Procedures Complete
Subtotal
Biopsy
Total
MG
NoMG
MG
NoMG
MG
NoMG
IS
IVa IVb Histologic type Lymphocyte Mixed Epithelial Unknown
30 23 20 0 0
10 15 0 0
0 0 2 3 0
0 0 3 5 3
0 0 I 2 0
0 0 7 2 0
30 23 23 5 0
24 39 9 I
9 22 9 0
2 3 0 0
0 5 6 0
I 1 1 0
0 7 I I
27 (25) 43 (32) 10 (6) I (1)
9 (4) 34 (26) 16 (8) I (1)
Total
73
40
5
9
81 (64)
60 (39)
MG
NoMG
Stage I II III
II
3
(23) (18) (19) (4)
15 (8) 10 (7) 25 (17) 7 (5) 3 (2)
Numbers of patients undergoing postoperative radiation therapy are shown in parentheses.
pure red cell aplasia in four patients and myokinesis, lung cancer, and malignant fibrous histiocytoma of the retroperitoneum in one patient each. One other patient had hypergamrnaglobulinemia, rheumatoid arthritis, and Sjogren disease. Table I presents the relationship between the surgical procedures and the stage of thymoma or histologic type in terms of the presence or absence of myasthenia gravis. Complete resection of the thymoma and thymic tissue, including invaded tissues such as the mediastinal pleura, pericardium, and lung, if tumor invasion was apparent or suspected, was performed in 73 patients (90.1 %) with and 40 patients (66.7%) without myasthenia gravis. Five myasthenic and II non myasthenic patients underwent subtotal resection. Three myasthenic and nine nonmyasthenic patients underwent biopsyonly. The staging system for thymoma that we] have proposed is as follows: Stage I: No capsular invasion. Stage II: Capsular or pleural invasion. Stage III: Invasion of surrounding tissues such as lung, pericardium, superior vena cava, and aorta Stage IV: Dissemination in the thoracic cavity (IVa) and distant metastasis (IVb) As Bernatz and associates" proposed, the thymomas were classified histologically by the degree of lymphocyte infiltration of the thymoma epithelial cells as lymphocyte predominant, mixed, and epithelial types. Before 1967 we treated four thymoma patients with no (stage I) or minimal (stage II) capsular invasion that had been resected completely but afterward recurred. Therefore, after 1968 our plan was to administer postoperative radiation therapy in a dosage of 30 Gy in 3 weeks or 40 Gy in 4 weeks for patients undergoing complete resection, and over 50 Gy in 6 weeks for patients undergoing incomplete resection or biopsy only. 10. II The number of patients in whom this strategy was accomplished is shown in parentheses in Table I. Since 1968 postoperative radiation therapy has been administered to 103 of 122 (84.2%)
patients. The main causes of failure to give the intended postoperative radiation therapy were severe myasthenic condition in five patients, far-advanced disease in five patients, pure red cell aplasia in two patients, Sjogren disease in one patient, and other causes in six patients. Postoperative chemotherapy was comtemplated when the disease was disseminated, with distant metastasis or recurrence. The regimens we applied were conventional: Combinations of cyclophosphamide. vincristine. and prednisone were used until 1978. Thereafter. doxorubicin was supplemented with cyclophosphamide and vincristine in some cases. I! At present, we are trying cisplatin and vindesine as the main regimen in these patients.') Death by tumor occurred in five of 81 patients (6.2%) with and in 18 of 61 patients (307<) without myasthenia gravis. Causes of death other than tumor were myasthenic crisis in 10 patients, pure red cell aplasia in three patients. and malignant fibrous histiocytoma, pneumonia. and hepatitis in one patient each. One patient died of respiratory failure soon after operation, and this patient was excluded from the study. Three patients were lost to follow-up I. 2. and 5 years after operation. Survival rates were computed by the Kaplan-Meier method." Patients dying of causes other than tumor recurrence and patients lost to follow-up were regarded as "drop-out cases" at the year of these events. Statistical significance was assessed by the Cox-Mantel method." A contingency table method was applied to analyze the statistical significance of frequencies of occurrence. 16 A probability value less than 5'/r was regarded as statistically significant.
Results Table II shows the survival rates in terms of the surgical procedures, stage of thymoma, and histologic type. The survival rates in patients undergoing complete resection were significantly higher than in those undergoing subtotal resection (p < 0.001) and biopsy
Volume 95 Number 6
Thymoma
June 1988
1 04 3
Table II. Survival rates in terms of surgical procedures, stage, and histologic type Survival rate (%) No. of patients
Procedures Complete Subtotal Biopsy Stage I II III IVa + IVb Histologic type§ Lymphocyte Mixed Epithelial
5 yr
10 yr
15 yr
8;') )1 t
113 16 12
97.6 68.2 25.0
94.2 68.2 0
45 33 48 15
100 91.5 87.8 46.6
100 84.4 77.2 46.6
36
96.3 89.5 76.3
87.9 86.3 65.5
77
26
o
t }
857) ) 70.4:j:
61.9
:j:
l1 •
t}.
70.3\ 68.5 • 65.4
'I' < 005. tl'
+1' < o.co I
§Twu patients were excluded because the histologic type was unknown.
only (p < 0.001). There was also a significant difference in the survival rates between patients undergoing subtotal resection and biopsy only (p < 0.01). The prognosis of thymoma was also stage dependent (Table II): The survival rate of patients with stage III disease was significantly different (p < 0.001) from that of patients with stage I disease. A slight but significant (p < 0.05) difference was also detected between stage I and II disease. Patients with stage IV disease had a poorer prognosis than those with stage I (p < 0.001), II (p < 0.01), and III disease (p < 0.05). An assessment of histologic types showed that the survival rate of patients with epithelial cell thymoma was significantly lower (p < 0.05) than that of patients with the lymphocyte predominant type. Fig. 1 shows the survival curves according to the surgical procedures and the stage of thymoma: All the patients with stage I disease (45 patients) underwent complete resection of the tumor. The survival rates in stage I disease were 100% at 5 and 10 years and 85.7% at 15 years. The causes of death in patients with thymoma are shown in Table III. One patient with stage I disease had tumor recurrence 4 years after complete resection, and reoperation revealed an intrapleural dissemination. This patient died of the tumor 15 years after the first operation. Five patients with stage I disease died of myasthenic crisis from 2 months to 4 years after thymectomy. Two other patients died of pure red cell aplasia 3 and 13 years after operation. Malignant fibrous histiocytoma and hepatitis were the causes of death in one patient each and 3 years after operation.
Complete resection was possible in all 33 patients with stage II disease. The survival rates were 91.5% at 5 years, 84.4% at 10 years, and 70.4% at 15 years (Fig. 1). Recurrence of the tumor was the cause of death from 2 to 11 years after operation in four patients; radiation therapy was administered only after the recurrence was found. Myasthenia gravis was the cause of death in three patients from 1 to 6 years after operation. One patient died of uncontrollable pure red cell aplasia 1 year after operation (Table III). In stage III disease, 35 patients (72.9%) underwent complete resection with survival rates of 100% at 5 years and 94.7% at 10 and 15 years (Fig. 1). Recurrence of the tumor was the cause of death in one patient 7 years after operation, and two patients died of myasthenic crisis 1 and 2 years after operation (Table III). Thus there was no significant differences in the survival rates of patients with stage III disease compared with patients with stage I or II disease undergoing complete resection (Fig. 1). On the other hand, the survival rates in five patients with stage III disease who underwent subtotal resection were 80% at 5 and 10 years and 0% at 15 years. Also, in 13 patients with stage IV disease who underwent subtotal resection, survival rates of 62.3% at 5 and 10 years were obtained (Fig. 1) The tumor was the cause of death in three patients with stage III and in four patients with stage IV disease (Table III). There was no statistically significant difference between the survival rates in patients with stage III and IV disease undergoing subtotal resection (Fig. 1).
The Journal of
1 044
Thoracic and Cardiovascular
Nakahara et al.
Surgery
%
100
:iii i1
+.
I.
~
N, subtotal
50 f-
m,
o
11
I
5
m,
complete
I. complete
r.
complete
m,
subtotal
------l....
N biopsy
15 years
10
Fig. 1. Survivalcurvesof patients with thymoma assessed by surgical procedures and stage in combination. There were no significant differences between stage 1 and III disease or stage II and III disease if completely resected. Survival rates in patients with stage III and IV disease who underwent subtotal resection were significantly different (p < 0.01 and p < 0.05) from those of patients undergoing biopsy only.
Table
m. Causes of death in patients with thymoma Procedures
Tumor
MG
PRCA
Others
(5) (3) (2)
2
2 (2)
Stage
I II
III III IV III + IV Histologic type Lymphocyte Mixed
Epithelial
Unknown Total
Complete Complete Complete Subtotal Subtotal Biopsy Complete Complete Subtotal Biopsy Complete Subtotal Biopsy Biopsy
1 4 (3) I
I
3 4 10 (2) 4 (2) 2 (I) 2 7 (I)
I (I)
2 (2)
(6)
2
1 (I)
3
3 (3)
(3) 5 2 (I) 1
23 (5)
(I)
(10)
The numbers of patients with thymoma and myasthenia gravis are shown in parentheses. MG. Myasthenia gravis; PRCA. pure red cell aplasia.
In 12 patients with stage III and IV disease, only biopsy for histoligic specimens was done. No patients so treated survived more than 8 years (Fig. 1). The tumor was the cause of death in all but one patient who died of pneumonia induced by chemotherapy (Table III). Six of 113 (5.3%) patients undergoing complete resection died of the tumor, 7 of 16 patients (43.8%) undergoing subtotal resection, and 10 of 12 patients (83.3%) undergoing biopsy only (Table III). The x' analysis" demonstrated the rates of death by tumor to be significantly (p < 0.01) different among these three groups.
The reasons that complete resection was impossible in stage III disease were infiltration of the superior vena cava in 10 patients, of the lung hilum in two patients, and of the trachea in one patient (Table IV). Three patients underwent replacement of the superior vena caval system with a polytetrafluoroethylene graft with rings so that the tumor could be removed en bloc (Fig. 2). One patient with stage IV disease died of tumor dissemination, but the other two patients were well 2 years after operation with sufficient blood flow through the graft. Fig. 3 shows the survival curves according to the
Volume 95 Number 6
Thymoma
June 1988
Table IV. Reasons for palliative operations in stage III disease Invasion of SVC Invasion of lung hilum or trachea
I 04 5
---it
---../!
Subtotal
Biopsy
3 2
7
) ::::>~. ~--
I
combined surgical procedures and histologic types. Thirty-three of 36 patients with lymphocyte predominant type underwent complete resection and had survival rates of 96.0% at 5 years, 86.9% at 10 years, and 69.5% at 15 years. Sixty-one of 77 patients with mixed cell type underwent complete resection and had the survival rates of 98.1 % at 5 and 10 years and 93.2% at 15 years. In epithelial cell thymoma, 18 of 26 patients underwent complete resection and there were no deaths until 15 years. Thus there were no significant differences in the survival rates according to histologic types among patients undergoing complete resection. No patients with epithelial cell thymoma who underwent complete resection died of tumor recurrence, whereas four patients with lymphocyte predominant type and two patients with mixed type died of the tumor after complete resection (Table III). On the other hand, eight patients with mixed type who underwent subtotal restriction had survival rates of 100% at 5 and 10 years but 0% at 15 years. Six patients with epithelial cell type who underwent subtotal resection had survival rates of 16.7% at 5 to 14 years. There was a significant difference (p < 0.001) in the survival rates between patients with epithelial type who underwent complete resection and subtotal resection. Also, a significant difference (p < 0.01) in the survival rates was detected between patients with mixed and epithelial types who underwent subtotal resection. There was no significant difference between patients with epithelial type who underwent subtotal resection and biopsy only.
Discussion As previously reported by Salyer and Egglestone," Bergh and associates,' and by US,3 the present results reconfirm that complete resection has a remarkable effect on the prognosis of patients with thymoma. Wilkins and Castleman? claimed that incomplete resection of thymoma provided little evidence to support the efficacy of more than biopsy. However, an exact comparison of survival rates in patients undergoing incomplete resection and biopsy only was not made. Our results clearly demonstrated that patients undergoing subtotal resection had a better survival rate than those
J svc.~ tumor
7// 1\
Fig. 2. Schematic representation of reconstruction of superior vena caval system with ringed polytetrafluoroethylene graft in three patients with invasive thymoma. PA, Pulmonary artery; RA. right atrium; SVC, superior vena cava.
undergoing biopsy only (Table 11), except for patients with epithelial cell type, in whom the two treatment modalities did not result in a significant difference in survival rates (Fig. 3). Controversy exists concerning the effect of the presence or absence of myasthenia gravis on prognosis. Batata and associates" and Wilkins and Castleman? claimed that the presence of myasthenia gravis was one factor causing a poor prognosis. However, the apparently low frequency of association with myasthenia gravis (4/54) in the series reported by Batata and associates" compared with that generally reported," and the high frequency of death caused by myasthenic crisis (18/37)
The Journal of
1 046
Nakahara et al.
Thoracic and Cardiovascular Surgery
%
100
epithelial complete
=
I
II
Q)
+'
"iii > .:; L
mixed complete lymphocytic complete
biopsy
50
mixed
:::J
subtotal
II)
I o
~ 5
epithelial subtotal
I
10
15
years
Fig. 3. Survival curves of patients with thymoma assessed by the surgical procedures and histologic type in combination. There were no significant differences between patients with lymphocyte predominant, mixed, and epithelial cell types if they underwent complete resection. In epithelial cell type, the survival curves showed a significant difference between patients undergoing complete resection and those having subtotal resection. Subtotal resection in patients with epithelial cell type seemed to have no better effect than biopsy alone.
in the series reported by Wilkins and Castleman,' seemed to lead to this conclusion. On the other hand, Bernatz," Shamji," and their colleagues showed that coexisting myasthenia gravis had no influence on the prognosis of thymoma. In our series, about 10% of myasthenic patients had incomplete resection, whereas about 35% of patients without myasthenia had incomplete resection (Table I). This difference was statistically significant (p < 0.01). However, in myasthenic patients the rate of complete resection was 88.8% to 90%, regardless of cell types. Although complete resection was performed in 100% of nonmyasthenic patients with lymphocytic cell type, 64.7% with mixed type, and 56.0% with epithelial cell type, these differences were statistically insignificant. Therefore, the cell type might be partly responsible, but not definitively so, for the difference in the complete resection rate between patients with thymoma with and without myasthenia gravis. The fact that a larger percentage of myasthenic (37.0%) than nonmyasthenic (25%) patients had stage I disease in our series suggested that the tumor might have been detected at a somewhat earlier stage in myasthenic patients because of their disease (Table I). However, in our previous report," we demonstrated that the recurrence rate for thymoma of the same stage was smaller in myasthenic than in nonmyasthenic patients, which suggested that coexisting myasthenia was actually a sign of a better prognosis than the absence of myasthenia. Although the mechanism by which myasthenia gravis affects the biologic behavior of thymoma is
still unknown, the survival rate in patients with thymoma and myasthenia gravis would have been better than the present result had the crisis been treated more effectivley." Concerning the histologic type, our result was consistent with those of other authors 4. 18 in that the survival rate for epithelial cell type was worse than for the lymphocyte predominant type. However, if we considered the surgical procedure and histologic type in combination, patients with epithelial type had the same survival rates as those with other histologic types when they underwent complete resection (Fig. 3). It is interesting that no patients with epithelial cell type died of tumor recurrence if they were treated by complete resection (Table III), whereas the survival rate after subtotal resection was almost the same as that after biopsy alone (Fig. 3). Verley and Hollman" emphasized that a radical operation may not always benefit patients, since death from tumor was as common in slightly invasive, completely excised tumors (3/21) as in largely invasive tumors for which only a biopsy had been done (8/31). However, our results showed that the survival rate in patients with stage III disease undergoing complete resection was comparable with that of patients with stage I and II diseae (Fig. 1), and that the rate of death by tumor was significantly (p < 0.01) different among patients undergoing complete resection, subtotal resection, and biopsy only (Table III). Therefore, we conclude that, even in invasive thymoma, the prognosis over the long term is comparable with
Volume 95 Number 6 June 1988
that of noninvasive or minimally invasive thymomas, regardless of histologic type, so long as the tumor is removed completely and postoperative radiation therapy (30 Gy in 3 weeks to 40 Gy in 4 weeks) is administered. In our series, the rate of death by tumor in patients with invasive (stage II and III), completely resected tumor was lower (5/68) than in Verley and Hollman's series"; in patients undergoing only a biopsy, it was higher (10/12) than in Verley and Hollman's series." Postoperative radiation therapy was contemplated in both their series" and ours, but it is not clear how many patients actually received the contemplated radiation therapy in their study." Thus the reason for this discrepancy is obscure. We applied postoperative radiation therapy even in patients with stage I diseaes, because before 1967 one of our patietns with stage I disease had had lymphocytic cell type at the time of the first complete resection; reoperation for recurrence 4 years later, however, revealed epithelial cell thymoma with intrapleural dissemination. Masaoka and associates," in our institution, showed that the epithelial component was more prevalent in recurrent thymoma than in the initial tumor. Therefore, we had believed that the complete treatment comprising a radical operation and postoperative radiation therapy would be mandatory initially. Currently, we are trying surgical treatment alone for stage I disease to determine whether radiation therapy is actually necessary in these patients. We did not resect the tumor completely in 10 patients because the tumor had invaded the superior vena cava (Table IV). However, recent advances in vascular surgery have enabled us to remove the superior vena caval system and reconstruct it adequately with a ringed polytetrafluoroethylene graft. We treated three patients in this manner. In the same sense, pneumonectomy or complete pleuropneumonectomy should have been performed in advanced thymoma, as Bergh and associates I did in three patients, since oncologically thymoma behaves in a much more benign manner than lung cancer. The long-term result of such an extended operation in thymoma is unresolved, but we believe that these procedures have the potential to improve the prognosis of invasive thymoma. REFERENCES 1. Bergh NP, Gatzinsky P, Larsson S, Lundin P, Rindell B. Tumor of the thymus and thymic region. I. Clinicopathologica studies on thymoma. Ann Thorac Surg 1978;25: 91-8. 2. Wilkins EW, Castleman B. Thymoma: a continuing survey at the Massachusetts General Hospital. Ann Thorac Surg 1979;28:252-6.
Thymoma
10 4 7
3. 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. 4. Bernatz PE, Khonsari S, Harrison EG, Taylor WF. Thymoma: factors influencing prognosis. Surg Clin North Am 1973;53:885-93. 5. Jain U, Frable WJ. Thymoma: analysis of benign and malignant criteria. J THORAC CARDIOVASC SURG 1974; 67:310-21. 6. Verley JM, Hollman KH. Thymoma: a comparative study of clinical stages, histological features and survival in 200 cases. Cancer 1985;55:1074-86. 7. 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: 130-8. 8. Fechner RE. Recurrence of noninvasive thymomas: report of four cases and review of literature. Cancer 1969; 23:1423-7. 9. Rosai J, LevineGO. Tumor of the thymus. Atlas of tumor pathology. 2nd series. Washington DC: Armed Forces Institution of Pathology, 1976:34-153. 10. Penn CRH, Hope-Stone HF. The role of radiotherapy in the management of malignant thymoma. Br J Surg 1972;59:533-9. 11. Marks RD Jr, Wallace KM, Pettit HS. Radiation therapy control of nine patients with malignant thymoma. Cancer 1978;41:117-9. 12. Lohrer PJ, Bonomi P, Goldman S, et al. Remission of invasive thymoma due to chemotherapy: two patients treated with cyclophosphamide, doxorubicin and vincristine. Chest 1985;87:377-80. 13. Hu E, Levine J. Chemotherapy of malignant thymoma: case report and review of the literature. Cancer 1986; 57:1101-4. 14. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:45781. 15. Cox DR. Regression models and life table. J R Stat Soc (B) 1972;34: 187-220. 16. Zar JH. Biostatistical analysis. Englewood Cliffs, New Jersey: Prentice-Hall, 1974:59-69. 17. Salyer WR, Egglestone .IC. Thymoma: a clinical and pathological study of 65 cases. Cancer 1976;37:229-49. 18. Batata MA, Martini N, Huvos AG, Aguilar RI, Beattie EJ. Thymoma: clinicopathologic features, therapy and prognosis. Cancer 1974;34:389-96. 19. Shamji F, Pearson FG, Todd TRJ, Ginsberg RJ, lives R, Cooper JD. Results of surgical treatment for thymoma. J THORAC CARDIOVASC SURG 1984;87:43-7. 20. Monden Y, Nakahara K, Iioka S, et al. Recurrence of thymoma: clinicopathological features, therapy and prognosis. Ann Thorac Surg 1985;39: 165-9. 21. Masaoka A, Nagaoka Y, Maeda M, Monden Y, Seike Y. Study on the ratio of lymphocytes to epithelial cells in thymoma. Cancer 1977;40:1222-8.
THoRAc CARDIOVASC SURG
1988;95:1041-7
Thymoma: Results with complete resection and adjuvant postoperative irradiation in 141 consecutive patients A series of 141 patients with thymoma was studied with regard to the way in which complete resection followed by postoperative radiation therapy influenced prognosis according to the stage and histologic type of thymoma. Postoperative radiation therapy (30 Gy in 3 weeks to 50 Gy in 6 weeks) was performed in 73.1 % of the patients. Thirty-five of 48 patients with thymoma invading the surrounding tissues (stage III) underwent complete resection, with survival rates of 100% at 5 years and 94.7% at 10 and 15 years. The prognosis in these patients was comparable with those in 45 patients with no invasion (stage I) and in 33 patients with capsular invasion (stage llI), all of whom underwent complete resection. Complete resection was done in 18 of 26 patients with epithelial cell thymoma, and there were no deaths by tumor until 15 years after the operation. There were no statistically significant differences in the survival rates of patients undergoing complete resection according to cell type (33 of 36 patients with lymphocyte predominant type and 61 of 77 patients with mixed cell type). The survival rate in six patients with epithelial cell type who underwent subtotal resection was not significantly different from that of 12 patients who underwent biopsy alone. Our findings indicate that complete resection of thymoma followed by postoperative radiation therapy results in a "benign" postoperative course, regardless of the stage and histologic type. Therefore, an aggressive surgical approach, such as resection of the superior vena caval system followed by reconstruction with a ringed polytetraftuoroethylene graft and/or complete pleuropneumonectomy, is justified for advanced thymoma, although the long-term results of such extended operations are stiU unclear. Kazuya Nakahara, MD, Kiyoshi Ohno, MD, Junpei Hashimoto, MD, Hajime Maeda, MD, Shinichiro Miyoshi, MD, Masami Sakurai; Yasumasa Monden, MD,b and Yasunaru Kawashima, MD, Osaka and Tokushima, Japan
Malignancy of thymoma has been discussed mainly with regard to the invasiveness of the tumor to the surrounding tissues at the time of the initial treatment':' and the histologic type.r" However, uncommon recurrence or dissemination of noninvasive thymomas, once removed completely, was reported by Fechner," who compiled eight other cases from the literature. The same experience in our institution led us to believe that thymoma is malignant, or at least potentially malignant. Therefore, we surmise that it is mandatory to resect the From the First Department of Surgery and Department of Pathology,' Osaka University Medical School, Osaka, Japan, and the Second Department of Surgery,' Tokushima University Medical School, Tokushima, Japan. Received for publication April 30. 1987. Accepted for publication July 29, 1987. Address for reprints: Kazuya Nakahara. MD, First Department of Surgery, Osaka University Medical School, Fukushima-ku, Osaka. 553, Japan.
tumor completely, including normal thymic tissue and invaded tissues, and to use postoperative radiation therapy. This therapeutic approach provided a "benign" long-term outcome in many patients with thymoma even thought the tumor had been invasive, especially when the tumor was removed completely. The purpose of this study is to determine how complete resection of the tumor affects the survival rate in terms of the stage of thymoma and histologic type, which have been thought to define the malignancy of thymoma. Patients and methods From 1957 to 1985, 142 patients with thymoma were treated. Seventy-nine patients were male and 63 were female. The ages ranged from 8 months to 73 yearssand the mean age was 44 ± 14 (standard deviation) years. Thymoma was defined as tumor originating from the epithelial cell of the thymus! Granulomatous thymoma, carcinoid, seminoma, malignant lymphoma, and mediastinal lung cancer were carefully excluded by a pathologist (M.S.). The associated diseases were myasthenia gravis in 81 patients, 1 n 41
The Journal of
1 042
Nakahara et al.
Thoracic and Cardiovascular Surgery
Table I. Surgical procedures. stage. and histologic type in patients with thymona, with or without myasthenia gravis (MG) Procedures Complete
Subtotal
Biopsy
Total
MG
NoMG
MG
NoMG
MG
NoMG
IS
IVa IVb Histologic type Lymphocyte Mixed Epithelial Unknown
30 23 20 0 0
10 15 0 0
0 0 2 3 0
0 0 3 5 3
0 0 I 2 0
0 0 7 2 0
30 23 23 5 0
24 39 9 I
9 22 9 0
2 3 0 0
0 5 6 0
I 1 1 0
0 7 I I
27 (25) 43 (32) 10 (6) I (1)
9 (4) 34 (26) 16 (8) I (1)
Total
73
40
5
9
81 (64)
60 (39)
MG
NoMG
Stage I II III
II
3
(23) (18) (19) (4)
15 (8) 10 (7) 25 (17) 7 (5) 3 (2)
Numbers of patients undergoing postoperative radiation therapy are shown in parentheses.
pure red cell aplasia in four patients and myokinesis, lung cancer, and malignant fibrous histiocytoma of the retroperitoneum in one patient each. One other patient had hypergamrnaglobulinemia, rheumatoid arthritis, and Sjogren disease. Table I presents the relationship between the surgical procedures and the stage of thymoma or histologic type in terms of the presence or absence of myasthenia gravis. Complete resection of the thymoma and thymic tissue, including invaded tissues such as the mediastinal pleura, pericardium, and lung, if tumor invasion was apparent or suspected, was performed in 73 patients (90.1 %) with and 40 patients (66.7%) without myasthenia gravis. Five myasthenic and II non myasthenic patients underwent subtotal resection. Three myasthenic and nine nonmyasthenic patients underwent biopsyonly. The staging system for thymoma that we] have proposed is as follows: Stage I: No capsular invasion. Stage II: Capsular or pleural invasion. Stage III: Invasion of surrounding tissues such as lung, pericardium, superior vena cava, and aorta Stage IV: Dissemination in the thoracic cavity (IVa) and distant metastasis (IVb) As Bernatz and associates" proposed, the thymomas were classified histologically by the degree of lymphocyte infiltration of the thymoma epithelial cells as lymphocyte predominant, mixed, and epithelial types. Before 1967 we treated four thymoma patients with no (stage I) or minimal (stage II) capsular invasion that had been resected completely but afterward recurred. Therefore, after 1968 our plan was to administer postoperative radiation therapy in a dosage of 30 Gy in 3 weeks or 40 Gy in 4 weeks for patients undergoing complete resection, and over 50 Gy in 6 weeks for patients undergoing incomplete resection or biopsy only. 10. II The number of patients in whom this strategy was accomplished is shown in parentheses in Table I. Since 1968 postoperative radiation therapy has been administered to 103 of 122 (84.2%)
patients. The main causes of failure to give the intended postoperative radiation therapy were severe myasthenic condition in five patients, far-advanced disease in five patients, pure red cell aplasia in two patients, Sjogren disease in one patient, and other causes in six patients. Postoperative chemotherapy was comtemplated when the disease was disseminated, with distant metastasis or recurrence. The regimens we applied were conventional: Combinations of cyclophosphamide. vincristine. and prednisone were used until 1978. Thereafter. doxorubicin was supplemented with cyclophosphamide and vincristine in some cases. I! At present, we are trying cisplatin and vindesine as the main regimen in these patients.') Death by tumor occurred in five of 81 patients (6.2%) with and in 18 of 61 patients (307<) without myasthenia gravis. Causes of death other than tumor were myasthenic crisis in 10 patients, pure red cell aplasia in three patients. and malignant fibrous histiocytoma, pneumonia. and hepatitis in one patient each. One patient died of respiratory failure soon after operation, and this patient was excluded from the study. Three patients were lost to follow-up I. 2. and 5 years after operation. Survival rates were computed by the Kaplan-Meier method." Patients dying of causes other than tumor recurrence and patients lost to follow-up were regarded as "drop-out cases" at the year of these events. Statistical significance was assessed by the Cox-Mantel method." A contingency table method was applied to analyze the statistical significance of frequencies of occurrence. 16 A probability value less than 5'/r was regarded as statistically significant.
Results Table II shows the survival rates in terms of the surgical procedures, stage of thymoma, and histologic type. The survival rates in patients undergoing complete resection were significantly higher than in those undergoing subtotal resection (p < 0.001) and biopsy
Volume 95 Number 6
Thymoma
June 1988
1 04 3
Table II. Survival rates in terms of surgical procedures, stage, and histologic type Survival rate (%) No. of patients
Procedures Complete Subtotal Biopsy Stage I II III IVa + IVb Histologic type§ Lymphocyte Mixed Epithelial
5 yr
10 yr
15 yr
8;') )1 t
113 16 12
97.6 68.2 25.0
94.2 68.2 0
45 33 48 15
100 91.5 87.8 46.6
100 84.4 77.2 46.6
36
96.3 89.5 76.3
87.9 86.3 65.5
77
26
o
t }
857) ) 70.4:j:
61.9
:j:
l1 •
t}.
70.3\ 68.5 • 65.4
'I' < 005. tl'
+1' < o.co I
§Twu patients were excluded because the histologic type was unknown.
only (p < 0.001). There was also a significant difference in the survival rates between patients undergoing subtotal resection and biopsy only (p < 0.01). The prognosis of thymoma was also stage dependent (Table II): The survival rate of patients with stage III disease was significantly different (p < 0.001) from that of patients with stage I disease. A slight but significant (p < 0.05) difference was also detected between stage I and II disease. Patients with stage IV disease had a poorer prognosis than those with stage I (p < 0.001), II (p < 0.01), and III disease (p < 0.05). An assessment of histologic types showed that the survival rate of patients with epithelial cell thymoma was significantly lower (p < 0.05) than that of patients with the lymphocyte predominant type. Fig. 1 shows the survival curves according to the surgical procedures and the stage of thymoma: All the patients with stage I disease (45 patients) underwent complete resection of the tumor. The survival rates in stage I disease were 100% at 5 and 10 years and 85.7% at 15 years. The causes of death in patients with thymoma are shown in Table III. One patient with stage I disease had tumor recurrence 4 years after complete resection, and reoperation revealed an intrapleural dissemination. This patient died of the tumor 15 years after the first operation. Five patients with stage I disease died of myasthenic crisis from 2 months to 4 years after thymectomy. Two other patients died of pure red cell aplasia 3 and 13 years after operation. Malignant fibrous histiocytoma and hepatitis were the causes of death in one patient each and 3 years after operation.
Complete resection was possible in all 33 patients with stage II disease. The survival rates were 91.5% at 5 years, 84.4% at 10 years, and 70.4% at 15 years (Fig. 1). Recurrence of the tumor was the cause of death from 2 to 11 years after operation in four patients; radiation therapy was administered only after the recurrence was found. Myasthenia gravis was the cause of death in three patients from 1 to 6 years after operation. One patient died of uncontrollable pure red cell aplasia 1 year after operation (Table III). In stage III disease, 35 patients (72.9%) underwent complete resection with survival rates of 100% at 5 years and 94.7% at 10 and 15 years (Fig. 1). Recurrence of the tumor was the cause of death in one patient 7 years after operation, and two patients died of myasthenic crisis 1 and 2 years after operation (Table III). Thus there was no significant differences in the survival rates of patients with stage III disease compared with patients with stage I or II disease undergoing complete resection (Fig. 1). On the other hand, the survival rates in five patients with stage III disease who underwent subtotal resection were 80% at 5 and 10 years and 0% at 15 years. Also, in 13 patients with stage IV disease who underwent subtotal resection, survival rates of 62.3% at 5 and 10 years were obtained (Fig. 1) The tumor was the cause of death in three patients with stage III and in four patients with stage IV disease (Table III). There was no statistically significant difference between the survival rates in patients with stage III and IV disease undergoing subtotal resection (Fig. 1).
The Journal of
1 044
Thoracic and Cardiovascular
Nakahara et al.
Surgery
%
100
:iii i1
+.
I.
~
N, subtotal
50 f-
m,
o
11
I
5
m,
complete
I. complete
r.
complete
m,
subtotal
------l....
N biopsy
15 years
10
Fig. 1. Survivalcurvesof patients with thymoma assessed by surgical procedures and stage in combination. There were no significant differences between stage 1 and III disease or stage II and III disease if completely resected. Survival rates in patients with stage III and IV disease who underwent subtotal resection were significantly different (p < 0.01 and p < 0.05) from those of patients undergoing biopsy only.
Table
m. Causes of death in patients with thymoma Procedures
Tumor
MG
PRCA
Others
(5) (3) (2)
2
2 (2)
Stage
I II
III III IV III + IV Histologic type Lymphocyte Mixed
Epithelial
Unknown Total
Complete Complete Complete Subtotal Subtotal Biopsy Complete Complete Subtotal Biopsy Complete Subtotal Biopsy Biopsy
1 4 (3) I
I
3 4 10 (2) 4 (2) 2 (I) 2 7 (I)
I (I)
2 (2)
(6)
2
1 (I)
3
3 (3)
(3) 5 2 (I) 1
23 (5)
(I)
(10)
The numbers of patients with thymoma and myasthenia gravis are shown in parentheses. MG. Myasthenia gravis; PRCA. pure red cell aplasia.
In 12 patients with stage III and IV disease, only biopsy for histoligic specimens was done. No patients so treated survived more than 8 years (Fig. 1). The tumor was the cause of death in all but one patient who died of pneumonia induced by chemotherapy (Table III). Six of 113 (5.3%) patients undergoing complete resection died of the tumor, 7 of 16 patients (43.8%) undergoing subtotal resection, and 10 of 12 patients (83.3%) undergoing biopsy only (Table III). The x' analysis" demonstrated the rates of death by tumor to be significantly (p < 0.01) different among these three groups.
The reasons that complete resection was impossible in stage III disease were infiltration of the superior vena cava in 10 patients, of the lung hilum in two patients, and of the trachea in one patient (Table IV). Three patients underwent replacement of the superior vena caval system with a polytetrafluoroethylene graft with rings so that the tumor could be removed en bloc (Fig. 2). One patient with stage IV disease died of tumor dissemination, but the other two patients were well 2 years after operation with sufficient blood flow through the graft. Fig. 3 shows the survival curves according to the
Volume 95 Number 6
Thymoma
June 1988
Table IV. Reasons for palliative operations in stage III disease Invasion of SVC Invasion of lung hilum or trachea
I 04 5
---it
---../!
Subtotal
Biopsy
3 2
7
) ::::>~. ~--
I
combined surgical procedures and histologic types. Thirty-three of 36 patients with lymphocyte predominant type underwent complete resection and had survival rates of 96.0% at 5 years, 86.9% at 10 years, and 69.5% at 15 years. Sixty-one of 77 patients with mixed cell type underwent complete resection and had the survival rates of 98.1 % at 5 and 10 years and 93.2% at 15 years. In epithelial cell thymoma, 18 of 26 patients underwent complete resection and there were no deaths until 15 years. Thus there were no significant differences in the survival rates according to histologic types among patients undergoing complete resection. No patients with epithelial cell thymoma who underwent complete resection died of tumor recurrence, whereas four patients with lymphocyte predominant type and two patients with mixed type died of the tumor after complete resection (Table III). On the other hand, eight patients with mixed type who underwent subtotal restriction had survival rates of 100% at 5 and 10 years but 0% at 15 years. Six patients with epithelial cell type who underwent subtotal resection had survival rates of 16.7% at 5 to 14 years. There was a significant difference (p < 0.001) in the survival rates between patients with epithelial type who underwent complete resection and subtotal resection. Also, a significant difference (p < 0.01) in the survival rates was detected between patients with mixed and epithelial types who underwent subtotal resection. There was no significant difference between patients with epithelial type who underwent subtotal resection and biopsy only.
Discussion As previously reported by Salyer and Egglestone," Bergh and associates,' and by US,3 the present results reconfirm that complete resection has a remarkable effect on the prognosis of patients with thymoma. Wilkins and Castleman? claimed that incomplete resection of thymoma provided little evidence to support the efficacy of more than biopsy. However, an exact comparison of survival rates in patients undergoing incomplete resection and biopsy only was not made. Our results clearly demonstrated that patients undergoing subtotal resection had a better survival rate than those
J svc.~ tumor
7// 1\
Fig. 2. Schematic representation of reconstruction of superior vena caval system with ringed polytetrafluoroethylene graft in three patients with invasive thymoma. PA, Pulmonary artery; RA. right atrium; SVC, superior vena cava.
undergoing biopsy only (Table 11), except for patients with epithelial cell type, in whom the two treatment modalities did not result in a significant difference in survival rates (Fig. 3). Controversy exists concerning the effect of the presence or absence of myasthenia gravis on prognosis. Batata and associates" and Wilkins and Castleman? claimed that the presence of myasthenia gravis was one factor causing a poor prognosis. However, the apparently low frequency of association with myasthenia gravis (4/54) in the series reported by Batata and associates" compared with that generally reported," and the high frequency of death caused by myasthenic crisis (18/37)
The Journal of
1 046
Nakahara et al.
Thoracic and Cardiovascular Surgery
%
100
epithelial complete
=
I
II
Q)
+'
"iii > .:; L
mixed complete lymphocytic complete
biopsy
50
mixed
:::J
subtotal
II)
I o
~ 5
epithelial subtotal
I
10
15
years
Fig. 3. Survival curves of patients with thymoma assessed by the surgical procedures and histologic type in combination. There were no significant differences between patients with lymphocyte predominant, mixed, and epithelial cell types if they underwent complete resection. In epithelial cell type, the survival curves showed a significant difference between patients undergoing complete resection and those having subtotal resection. Subtotal resection in patients with epithelial cell type seemed to have no better effect than biopsy alone.
in the series reported by Wilkins and Castleman,' seemed to lead to this conclusion. On the other hand, Bernatz," Shamji," and their colleagues showed that coexisting myasthenia gravis had no influence on the prognosis of thymoma. In our series, about 10% of myasthenic patients had incomplete resection, whereas about 35% of patients without myasthenia had incomplete resection (Table I). This difference was statistically significant (p < 0.01). However, in myasthenic patients the rate of complete resection was 88.8% to 90%, regardless of cell types. Although complete resection was performed in 100% of nonmyasthenic patients with lymphocytic cell type, 64.7% with mixed type, and 56.0% with epithelial cell type, these differences were statistically insignificant. Therefore, the cell type might be partly responsible, but not definitively so, for the difference in the complete resection rate between patients with thymoma with and without myasthenia gravis. The fact that a larger percentage of myasthenic (37.0%) than nonmyasthenic (25%) patients had stage I disease in our series suggested that the tumor might have been detected at a somewhat earlier stage in myasthenic patients because of their disease (Table I). However, in our previous report," we demonstrated that the recurrence rate for thymoma of the same stage was smaller in myasthenic than in nonmyasthenic patients, which suggested that coexisting myasthenia was actually a sign of a better prognosis than the absence of myasthenia. Although the mechanism by which myasthenia gravis affects the biologic behavior of thymoma is
still unknown, the survival rate in patients with thymoma and myasthenia gravis would have been better than the present result had the crisis been treated more effectivley." Concerning the histologic type, our result was consistent with those of other authors 4. 18 in that the survival rate for epithelial cell type was worse than for the lymphocyte predominant type. However, if we considered the surgical procedure and histologic type in combination, patients with epithelial type had the same survival rates as those with other histologic types when they underwent complete resection (Fig. 3). It is interesting that no patients with epithelial cell type died of tumor recurrence if they were treated by complete resection (Table III), whereas the survival rate after subtotal resection was almost the same as that after biopsy alone (Fig. 3). Verley and Hollman" emphasized that a radical operation may not always benefit patients, since death from tumor was as common in slightly invasive, completely excised tumors (3/21) as in largely invasive tumors for which only a biopsy had been done (8/31). However, our results showed that the survival rate in patients with stage III disease undergoing complete resection was comparable with that of patients with stage I and II diseae (Fig. 1), and that the rate of death by tumor was significantly (p < 0.01) different among patients undergoing complete resection, subtotal resection, and biopsy only (Table III). Therefore, we conclude that, even in invasive thymoma, the prognosis over the long term is comparable with
Volume 95 Number 6 June 1988
that of noninvasive or minimally invasive thymomas, regardless of histologic type, so long as the tumor is removed completely and postoperative radiation therapy (30 Gy in 3 weeks to 40 Gy in 4 weeks) is administered. In our series, the rate of death by tumor in patients with invasive (stage II and III), completely resected tumor was lower (5/68) than in Verley and Hollman's series"; in patients undergoing only a biopsy, it was higher (10/12) than in Verley and Hollman's series." Postoperative radiation therapy was contemplated in both their series" and ours, but it is not clear how many patients actually received the contemplated radiation therapy in their study." Thus the reason for this discrepancy is obscure. We applied postoperative radiation therapy even in patients with stage I diseaes, because before 1967 one of our patietns with stage I disease had had lymphocytic cell type at the time of the first complete resection; reoperation for recurrence 4 years later, however, revealed epithelial cell thymoma with intrapleural dissemination. Masaoka and associates," in our institution, showed that the epithelial component was more prevalent in recurrent thymoma than in the initial tumor. Therefore, we had believed that the complete treatment comprising a radical operation and postoperative radiation therapy would be mandatory initially. Currently, we are trying surgical treatment alone for stage I disease to determine whether radiation therapy is actually necessary in these patients. We did not resect the tumor completely in 10 patients because the tumor had invaded the superior vena cava (Table IV). However, recent advances in vascular surgery have enabled us to remove the superior vena caval system and reconstruct it adequately with a ringed polytetrafluoroethylene graft. We treated three patients in this manner. In the same sense, pneumonectomy or complete pleuropneumonectomy should have been performed in advanced thymoma, as Bergh and associates I did in three patients, since oncologically thymoma behaves in a much more benign manner than lung cancer. The long-term result of such an extended operation in thymoma is unresolved, but we believe that these procedures have the potential to improve the prognosis of invasive thymoma. REFERENCES 1. Bergh NP, Gatzinsky P, Larsson S, Lundin P, Rindell B. Tumor of the thymus and thymic region. I. Clinicopathologica studies on thymoma. Ann Thorac Surg 1978;25: 91-8. 2. Wilkins EW, Castleman B. Thymoma: a continuing survey at the Massachusetts General Hospital. Ann Thorac Surg 1979;28:252-6.
Thymoma
10 4 7
3. 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. 4. Bernatz PE, Khonsari S, Harrison EG, Taylor WF. Thymoma: factors influencing prognosis. Surg Clin North Am 1973;53:885-93. 5. Jain U, Frable WJ. Thymoma: analysis of benign and malignant criteria. J THORAC CARDIOVASC SURG 1974; 67:310-21. 6. Verley JM, Hollman KH. Thymoma: a comparative study of clinical stages, histological features and survival in 200 cases. Cancer 1985;55:1074-86. 7. 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: 130-8. 8. Fechner RE. Recurrence of noninvasive thymomas: report of four cases and review of literature. Cancer 1969; 23:1423-7. 9. Rosai J, LevineGO. Tumor of the thymus. Atlas of tumor pathology. 2nd series. Washington DC: Armed Forces Institution of Pathology, 1976:34-153. 10. Penn CRH, Hope-Stone HF. The role of radiotherapy in the management of malignant thymoma. Br J Surg 1972;59:533-9. 11. Marks RD Jr, Wallace KM, Pettit HS. Radiation therapy control of nine patients with malignant thymoma. Cancer 1978;41:117-9. 12. Lohrer PJ, Bonomi P, Goldman S, et al. Remission of invasive thymoma due to chemotherapy: two patients treated with cyclophosphamide, doxorubicin and vincristine. Chest 1985;87:377-80. 13. Hu E, Levine J. Chemotherapy of malignant thymoma: case report and review of the literature. Cancer 1986; 57:1101-4. 14. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:45781. 15. Cox DR. Regression models and life table. J R Stat Soc (B) 1972;34: 187-220. 16. Zar JH. Biostatistical analysis. Englewood Cliffs, New Jersey: Prentice-Hall, 1974:59-69. 17. Salyer WR, Egglestone .IC. Thymoma: a clinical and pathological study of 65 cases. Cancer 1976;37:229-49. 18. Batata MA, Martini N, Huvos AG, Aguilar RI, Beattie EJ. Thymoma: clinicopathologic features, therapy and prognosis. Cancer 1974;34:389-96. 19. Shamji F, Pearson FG, Todd TRJ, Ginsberg RJ, lives R, Cooper JD. Results of surgical treatment for thymoma. J THORAC CARDIOVASC SURG 1984;87:43-7. 20. Monden Y, Nakahara K, Iioka S, et al. Recurrence of thymoma: clinicopathological features, therapy and prognosis. Ann Thorac Surg 1985;39: 165-9. 21. Masaoka A, Nagaoka Y, Maeda M, Monden Y, Seike Y. Study on the ratio of lymphocytes to epithelial cells in thymoma. Cancer 1977;40:1222-8.