Thymomas are tumors of epithelial cell origin arising. Preoperative CT Findings of Thymoma are Correlated with Postoperative Masaoka Clinical Stage

Transcription

1 Preoperative CT Findings of Thymoma are Correlated with Postoperative Masaoka Clinical Stage Yan-juan Qu, MS, Guo-bing Liu, MS, He-shui Shi, MS, PhD, Mei-yan Liao, MS, PhD, Gui-fang Yang, MS, PhD, Zhi-xiong Tian, MS Rationale and Objectives: Both preoperative computed tomography (CT) staging and postoperative surgical Masaoka clinical staging are of great clinical importance for diagnosing thymomas. Our study aimed to investigate the relationships between these two staging systems. Materials and Methods: This was a retrospective review of 129 patients who had undergone thymoma surgery. Helical CT and 16-slice CT were performed preoperatively. Surgical findings were evaluated according to the Masaoka clinical staging system. Results: A significant association was shown between Masaoka clinical staging and CT staging, especially of features including tumor size (P =.004), tumor shape (P <.001), tumor density (P <.001), capsule completeness (P <.001), and involvement of surrounding tissues (P <.001). Based on the CT findings, there were 35.09% of Masaoka stage I patients who had a tumor size <5 cm as compared to 14.81% of stage IV patients. Only 8.77% of Masaoka stage I patients had a tumor size $10 cm as compared to 40.74% of stage IV patients. In stages III and IV, most tumors were irregularly shaped with an uneven density and incomplete capsule. Invasive tumors were more frequently found in stages III (81.48%) and IV (88.89%) than in stages I (0%) and II (38.89%). The incidence of myasthenia gravis was comparable in different stages. Consistency between CT and Masaoka clinical stages was higher in stage I (37.98%) than other stages (approximately 10%). Conclusion: This study documented a close relationship between preoperative CT thymoma staging and postoperative Masaoka clinical staging. Thus, preoperative CT findings can be beneficial for determining the proper management and prognosis of thymoma patients. Key Words: Clinical stage; computed tomography; Masaoka staging system; thymic epithelial tumors; thymoma. ªAUR, 2013 Thymomas are tumors of epithelial cell origin arising from the thymus gland. Although thymomas are rare, occurring in only about 0.15 per 100,000 population (1), they account for 15% to 21.7% of tumors in the mediastinum and 47% of tumors in the anterior mediastinum (2). Clinical presentation of thymomas may vary considerably, but they are usually accompanied by either an immune or nonimmune mediated paraneoplastic syndrome, of which myasthenia gravis (MG) is the most common presentation (3). Acad Radiol 2013; 20:66 72 From the Departments of Radiology (Y-j.Q., G-b.L., M.-y.L., Z.-x.T.) and Pathology (G.-f.Y.), Zhongnan Hospital of Wuhan University, 169 East Lake Road, Wuhan , PR China; Department of Radiology, Union Hospital affiliated to Tongji Medical College of Huazhong University of Science and Technology, Wuhan , PR China (H.-s.S.). Received April 27, 2012; accepted August 1, Address correspondence to: Y-j.Q. quyanjuan7676@163.com ªAUR, The clinical stage of thymoma is analyzed based on the Masaoka staging system and is usually determined according to surgical findings (4). The histologic classification system, which was introduced by the World Health Organization (WHO) in 1999 (5) and updated in 2004 (6), is based on the pathological findings after collecting the specimen by either biopsy or during surgery. These two systems are applied at the postoperative stage and are considered independent prognostic indicators (7). However, there is no unequivocal correlation between WHO classification and Masaoka stages (8,9). The preoperative distinction of thymoma is also of great clinical importance; it is critical for the determination of a proper therapeutic regimen, which will necessarily vary for different stages of thymomas (9). Regarding the role of radiographic imaging modalities in thymoma assessment, a recent study showed that computed tomography (CT) was currently the first choice for characterizing mediastinal masses in terms of anatomic features and invasiveness of surrounding structures at the preoperative staging of thymoma (10). The standard way to assess patient responses to therapy is tumor 66

2 Academic Radiology, Vol 20, No 1, January 2013 CT VS. MASAOKA CLINICAL STAGE FOR THYMOMA PATIENT measurement by CT according to the Response Evaluation Criteria in Solid Tumors guidelines (11). CT is helpful in differentiating invasive from noninvasive thymoma; the presence of lobulated or irregular contours, areas of low attenuation, and multifocal calcifications is suggestive of invasive thymoma (12). The primary tumor CT imaging features can differentiate between stage I/II and stage III/IV disease, helping to identify patients more likely to benefit from neoadjuvant therapy (8). However, correlation between CT scan results and the WHO classification system remains controversial (8,13,14). Studies on the correlation between preoperative CT findings and Masaoka clinical stage of thymomas are limited (8). The purpose of this study was to investigate the correlation between preoperative CT findings and the Masaoka clinical stage of thymomas. Our findings may deepen the understanding of the value of CT in preoperatively determining thymoma stages. MATERIALS AND METHODS Participants and CT Scanning This study recruited a total of 129 patients with pathologically proven thymoma who underwent surgery in our institution between March 1999 and December Demographic data (age, gender) and clinical symptoms (chest tightness and chest pain, cough, fever, edema) were collected. Helical CT (PQ6000, PICKER, Cleveland, OH) (120 Kv, 100 ma, thickness 5 10 mm, index 5 10 mm, pitch 1.5) and 16-slice CT (Somatom Sensation, Siemens, Erlangen, Bavaria, Germany) (120 Kv, 120 mas, mm, 0.5 second/rotation, Recon increment 5 10 mm, pitch 1.125) were performed in all patients preoperatively. Scanning was performed from the inlet of the thoracic cage to the level of the diaphragm with a slice thickness and interslice interval of 5 to 10 mm. Plain scanning was performed in all 129 patients and enhanced scanning was performed in 65 patients. The contrast medium used was iopromide injection (Ultravist, Schering Pharmaceutical Ltd, Guangzhou, 300 mg/ml); a total of 80 to 100 ml of medium was injected at a rate of 3.0 ml/s. Features Analyzed by CT Two radiologists were chosen with expertise in diagnosing thymomas to review the CT findings and radiologically stage the tumor. However, the radiologists were blinded to the respective pathological results and clinical pathology staging. Consensus on the results was achieved by negotiation. The following features were analyzed by anatomic and invasive characteristics. 1. Tumor size: the maximal cross-section was used to measure the long and short diameters. Tables 1 3 showed this measurement using the longest axis diameter. 2. Tumor shape: oval, lobulated, and irregular (the ratio of long and short diameters <1.5 as round, as oval, and >3.0 as irregular) (12). TABLE 1. Characteristics and Clinical Features of 129 Patients Variable n = 129 Characteristics Age (y) Gender (male) 64 (49.61%) Symptoms Chest tightness and 46 (35.66%) chest pain Cough 24 (18.60%) Fever 4 (3.10%) Edema 4 (3.10%) CT features Tumor size (cm) <5 33 (25.58%) (54.26%) $10 26 (20.16%) Irregular shape 85 (65.89%) Uneven density 70 (54.26%) Calcification 27 (20.93%) Necrotic or cystic 32 (24.81%) components Incomplete capsule 63 (48.84%) Involvement of 53 (41.09%) surrounding tissues Lymph node 31 (24.03%) enlargement CT stage I 55 (42.64%) II 31 (24.03%) III 25 (19.38%) IV 18 (13.95%) Clinical status Myasthenia gravis 42 (32.56%) Resectable tumor 118 (91.47%) Masaoka clinical stage I 57 (44.19%) II 18 (13.95%) III 27 (20.93%) IV 27 (20.93%) CT, computed tomography. Data were expressed as mean SD for age; data were represented as counts (percentages) for categorical variables. 3. Tumor edge: smooth or coarse texture. 4. Tumor density: homogeneous, heterogeneous, calcification, necrotic, or cystic components. The heterogeneous density included both the gross unevenness and the difference of CT value of 20 Hounsfield units. The nonenhanced region with water density was regarded as either cystic or necrotic. 5. Mediastinal fat layer: clear, blurred, or disappeared. 6. Pleura: irregular interface with an absent space between tumor and pleura; pleural thickening; pleural effusion. One of the documented findings suggested pleural involvement. 7. Pericardium: irregular interface with an absent space between tumor and pericardium; pericardial thickening; 67

3 QU ET AL Academic Radiology, Vol 20, No 1, January 2013 TABLE 2. The Association between CT Features and Masaoka Clinical Stage Masaoka Clinical Stage CT Features I II III IV P Value Tumor size (cm).004* <5 20 (35.09%) 5 (27.78%) 4 (14.81%) 4 (14.81%) (56.14%) 12 (66.67%) 14 (51.85%) 12 (44.44%) $10 5 (8.77%) 1 (5.56%) 9 (33.33%) 11 (40.74%) Shape <.001* Oval 37 (64.91%) 6 (33.33%) 1 (3.7%) 0 (0%) Irregular 20 (35.09%) 12 (66.67%) 26 (96.3%) 27 (100%) Density <.001* Even 37 (64.91%) 8 (44.44%) 8 (29.63%) 6 (22.22%) Uneven 20 (35.09%) 10 (55.56%) 19 (70.37%) 21 (77.78%) Capsule <.001* Complete 56 (98.25%) 5 (27.78%) 5 (18.52%) 0 (0%) Incomplete 1 (1.75%) 13 (72.22%) 22 (81.48%) 27 (100%) Involvement of <.001* surrounding tissues Yes 0 (0%) 7 (38.89%) 22 (81.48%) 24 (88.89%) No 57 (100%) 11 (61.11%) 5 (18.52%) 3 (11.11%) Lymph node enlargement.272 Yes 47 (82.46%) 14 (77.78%) 20 (74.07%) 17 (62.96%) No 10 (17.54%) 4 (22.22%) 7 (25.93%) 10 (37.04%) CT, computed tomography. *P <.05 indicated a significant association between CT feature and clinical stage. TABLE 3. The Association between Masaoka Clinical Stage and Myasthenia Gravis Masaoka Clinical Stage Myasthenia Gravis I II III IV P Value No 39 (30.23%) 11 (8.53%) 15 (11.63%) 22 (17.05%).211 Yes 18 (13.95%) 7 (5.43%) 12 (9.3%) 5 (3.88%) Data were expressed as counts (percentages). The association was conducted by a chi-square test. pericardial effusion. One of the documented findings suggested pericardial involvement. 8. Blood vessels: absent space between tumor and blood vessels; oppression, deformation and occlusion of vessels; tumor thrombus in the vessels; tumor capsuling >1/4 circumference of vessel. One of the documented findings suggested vascular involvement. 9. Lymph node enlargement, distant metastasis. A measurement of the shortest lymph node diameter >1.0 cm was defined as lymph node enlargement. 10. Enhancement of tumor (even or uneven enhancement): the uneven enhancement included gross unevenness and a difference of CT value of >20 Hounsfield units. CT Tumor Staging CT tumor staging was assessed according to the following previously described methods (12): stage I: the capsule is intact and the space between fat tissues is clear; stage II: the capsule is blurred, with tumor involving mediastinal fat or mediastinal pleura; stage III: tumor invades pericardium, major blood vessels, and lung tissues; stage IV: tumor invades distant pleura and pericardium and lymph node enlargement-distant metastasis may be present. Masaoka Staging of Surgical Findings Based on surgical findings, the thymomas were staged according to the following criteria developed by Masaoka et al (4) stage I: the capsule is intact and microscopic pleural involvement is absent; stage II: the tumor involves surrounding pleura or mediastinal fat, or microscopic pleural involvement is present; stage III: the tumor involves surrounding tissues (pericardium, major blood vessels, or lung); stage IV: the tumor diffuses into the pleura or pericardium (stage IVa) or lymphatic or hematogenous metastasis (stage IVb). Statistical Analysis Data were expressed as mean standard deviation (SD) for continuous variables and counts with percentages for categorical variables. The associations between CT features and MG 68

4 Academic Radiology, Vol 20, No 1, January 2013 CT VS. MASAOKA CLINICAL STAGE FOR THYMOMA PATIENT with clinical staging were analyzed by the chi-square test. The consistency between CT stage and clinical staging was estimated by weighted kappa coefficient. The level of significance was set at Statistical analyses were performed by SAS 9.1 (SAS Institute Inc., Cary, NC). RESULTS Patient Demographic and Clinical Data A total of 129 patients including 64 (49.61%) males and 65 (50.39%) females with mean age of years old were enrolled in the study. The patients ranged from 18 to 75 years of age. Among these patients, 46 (35.66%) had symptoms of chest tightness and chest pain; 24 (18.60%) had cough; 4 (3.10%) had fever; 4 (3.10%) had edema; and 42 (32.56%) were diagnosed with MG. The tumors of most subjects (118; 91.47%) were completely resected. According to the surgical findings, Masaoka stages included 57 (44.19%) patients in clinical stage I; 18 (13.95%) patients in clinical stage II; 27 (20.93%) patients in stage III; and 27 (20.93%) patients in stage IV (Table 1). TABLE 4. Consistency between CT Stage and Masaoka Clinical Stage CT Stage Masaoka Clinical Stage I II III IV I 49 (37.98%) 3 (2.33%) 3 (2.33%) 0 (0%) II 8 (6.20%) 13 (10.08%) 7 (5.43%) 3 (2.33%) III 0 (0%) 1 (0.78%) 13 (10.08%) 11 (8.53%) IV 0 (0%) 1 (0.78%) 4 (3.10%) 13 (10.08%) CT, computed tomography. Weighted kappa coefficient = Distribution of CT Features among Clinical Stages To demonstrate the association between CT scan results and Masaoka clinical stage, we analyzed the distribution of CT features among four clinical stages. A significant association was shown between tumor size and clinical stage (P =.004). In each stage, approximately half of the subjects had a tumor size ranging from 5 to 10 cm. Additionally, the shape, density, capsules, and involvement of surrounding tissues had a significant association with the clinical stage (all P <.001). Most cases had an irregularly shaped, unevenly dense tumor and an incomplete capsule in grade II to grade IV; more patients whose surrounding tissues were invaded were in stages III and IV than in stages I and II (Table 2). As shown in Table 3, the clinical stage was not associated with the presence of MG (P =.211). Consistency between CT Stage and Clinical Stage. To assess the consistency between tumor stage by CT scan and clinical stage (Masaoka stage), we applied the weighted kappa coefficient, which was (P =.029). This represented a strong consistency between CT stage and clinical stage (Table 4). Figures 1 5 show representative CT thymoma images in different stages. DISCUSSION In the present study, we retrospectively analyzed the clinical information of a series of patients with thymoma. Preoperative CT findings were analyzed to stage the thymoma and results were compared with clinical stages obtained according to Masaoka et al (4). Limited studies have compared preoperative CT findings and Masaoka clinical staging of thymoma (8,13,14). Hence, our results demonstrating a strong consistency between preoperative CT stages of thymoma Figure 1. Computed tomography showed oval nodules in the anterior mediastinum, which had clear and smooth edges, even densities, and complete capsules (stage I in multi-slice spiral computed tomography and Masaoka staging system). and Masaoka clinical tumor staging (weighted kappa coefficient of 0.819; P =.029). This correlation remained unchanged in thymoma patients with and without MG. The Masaoka staging system has been widely applied towards the clinical staging of thymoma. In a study by Rios et al (15) using the Masaoka staging system, 5-year survival was 93.7% in stage I patients, 79.2% in stage II patients, 51.4% in stage III patients, and 0% in stage IV patients; this showed significant differences between thymoma patients with different tumor stages. In another study, multivariate analysis showed that the clinical stage was an independent factor influencing prognosis (16). Previous studies revealed that tumor size was related to thymoma type: that is, the larger the tumor size, the more malignant the tumor (17). Although tumor size (the longest axis diameters) can be readily obtained from CT images, it is not easily measured in surgical cases. The tumor mass is often located near the root of the aorta and pulmonary arteries and may be partially or completely outlined by fat. Thus, it is hard to completely remove the tumor intact and accurately measure the tumor size. Our findings suggested that tumor size can be measured by CT imaging and that size correlated with clinical stage (8,18). CT findings revealed that stage I or 69

5 QU ET AL Academic Radiology, Vol 20, No 1, January 2013 Figure 2. Computed tomography showed a soft-tissue mass with irregular shape in the upper anterior mediastinum that had an even density, unclear edge, and absent fat line (stage II in multi-slice spiral computed tomography and Masaoka staging system). Figure 4. Computed tomography showed an irregular soft-tissue mass in the mediastinum that had even density and absent fat space. The tumor encapsulated the major vessels, resulting in enlargement of internal mammary lymph nodes present behind the sternum. Figure 3. Computed tomography showed a soft-tissue mass with oval shape in the upper anterior mediastinum that had an even density as well as superior vena cava and pulmonary trunk (filling defect) oppression (stage III in multi-slice spiral computed tomography and Masaoka staging system). II thymomas were oval or round, had smooth edges and even density, whereas stage III or IV thymomas were usually irregularly shaped or lobulated. These findings also are consistent with those from recent studies (7,18). Preoperative differentiation of invasive and noninvasive thymomas can help determine prognosis and plan treatment. Tomiyama et al found that the presence of lobulated or irregular contours, areas of low attenuation, and also multifocal calcifications were most indicative of invasive thymoma (12). Preoperative CT features such as larger tumor size, lobulated tumor, and infiltration of fat surrounding the tumor suggest a high probability of Masaoka stage III (macroscopic invasion into neighboring organs, such as pericardium, Figure 5. From the same patient as Figure 4. In the lung window, multiple metastatic tumors were shown (stage IV in multi-slice spiral computed tomography and Masaoka staging system). great vessels, or lung) or IV (lymphogenous or hematogenous metastasis) disease; presentation of normal appearing pleura, tumor heterogeneity and infiltration of surrounding fat are most often associated with stage III disease (8). In the present study, most tumors were incompletely encapsulated, both macroscopically and microscopically, consistent with Masaoka stages IIa and IIb thymoma. Furthermore, the incidence of incomplete encapsulation was observed along with increases in clinical stage, such as microscopic invasion into the capsule or macroscopic invasion into mediastinal fat. Our results are consistent with other CT imaging studies (8,18). In the present study, the absence of mediastinal fat made it difficult to accurately identify the stage of thymoma. When space between the thymoma and mediastinal fat or mediastinal pleura is absent, the presence of mediastinal pleural thickening 70

6 Academic Radiology, Vol 20, No 1, January 2013 CT VS. MASAOKA CLINICAL STAGE FOR THYMOMA PATIENT may increase the accuracy of CT in thymoma staging when compared with partial absence of mediastinal fat space (12,19). Some investigators have also proposed that the presence of small nodules or cusp-like processes at the tumor edge indicate that the thymoma has invaded the capsule, serving as a marker for diagnosis of malignant thymoma (20). Although CT has a low specificity in assessing the infiltration of adjacent mediastinal structures, pleura, lung, and the chest wall (12), it was recently shown to have an increased sensitivity in identifying mediastinal masses and multiplanar reformatting. Such scans can be performed easily with multidetector CT scanners and can aid in surgical planning (8,18). The characteristic appearance and spread of thymoma on CT scans is especially useful preoperatively; for example, thymomas rarely present with metastatic lymphadenopathy of metastatic pulmonary nodules that may be present in other anterior mediastinal lesions (eg, lung cancer) (18). In this study, macroscopic invasion into the lungs, pericardium, or larger vessels was observed by CT, consistent with clinical stage III. The involvement of the pericardium was characterized by close adhesion of the thymoma to the pericardium along with pericardial thickening. With invasive thymomas, lung involvement is characterized by adhesion between the thymoma and adjacent lung tissues. This oppresses the lung, resulting in poor inflation or patchy inflammation of lung tissue, which is clear in the lung window. In the mediastinal window, lung involvement is difficult to identify or is characterized by cusp-like processes or cord-like shadows (21). Differentiating MG is critical for the diagnosis and treatment of thymoma because there has been a previously documented correlation between these medical conditions (22 24). For example, Whooley et al noted 33% to 75% of thymoma patients had MG in their study (22). De Perrot et al reported that more than two-thirds of MG patients had stage I or II thymomas, and proposed that thymoma patients with MG were more likely to be diagnosed than those without MG; deterioration during follow-up was not related to thymoma recurrence (23). Zhang et al reported that 49.2% of thymoma patients (n = 185) in their study had MG; most had stage I or II thymoma and 21 patients developed a myasthenic crisis (24). However, in our study, the correlation between stages indicated in CT scans and clinical stages remained unchanged between thymoma patients with and without MG. However, thymoma in combination with MG was not related to the Masaoka clinical stage. The findings of irregular contour, necrotic, or cystic component, heterogeneous enhancement, lymphadenopathy, and great vessel invasion were helpful. Today, preoperative radiographic imaging, including CT, magnetic resonance imaging, and positron emission tomography, is considered essential for determining the treatment strategy for thymoma (7,25). As our study was retrospective, it allowed us to only include information from 65 preoperative patients with a thymoma diagnosis made from enhanced CT. Although enhanced CT images may be more accurate, unenhanced CT images can also evaluate tumor density. Large tumors in unenhanced CT images had a central region that was also identified as a water-density lesion. However, we could not define whether these were cystic or necrotic in character because cystic or necrotic regions were also nonenhanced regions with water density. Necrotic regions usually had a slightly higher density in unenhanced CT images. Cystic and necrotic lesions can only be confirmed by pathological analysis after surgery (9,25). CT provides an accurate preoperative assessment by demonstrating the full extent of abnormalities, which is especially useful in surgical planning, monitoring therapy and detecting recurrence (21). Although CT is limited in making WHO histologic subtype differentiation (26), CT findings are useful for both differentiating low- or high-risk thymomas from thymic carcinomas (25) and predicting metastasis and postoperative recurrence (7). CTalso effectively distinguishes thymomas from benign mediastinal lesions of lymphoma when multiple mediastinal abnormalities are present (27). An increasing number of asymptomatic patients have been discovered to have thymoma because of the increasing application of CT for screening and investigation of other diseases (18). In the present study, as clinical stages increased from I to IV, the proportion of patients with stage I or II thymoma decreased, but that of stage III or IV increased; significant correlation was shown between stages in CT and clinical stages determined using the Masaoka staging system. Inconsistencies between CT and clinical stages were attributed to the difficulty in identifying the involvement of fat and blood vessels. Most important, the correlation of CT stages with Masaoka clinical stages could be applied to determine the invasiveness of thymoma, provide evidence of prognosis and to help plan the therapeutic regimen. Additionally, the CT finding of a cast-like relationship between malignant thymoma and major vessels of the heart suggests that surgical intervention is not feasible. Limitations Because this was a retrospective study and surgical findings were not evaluated prospectively, precise comparisons were limited. Although all study patients underwent surgical resection, patients who did not undergo surgical intervention were not included in the results, thereby causing a study bias. Further research should include all participants included in the study. CONCLUSION The thymoma tumor stages identified by CT are consistent with the clinical stages determined by Masaoka staging. CT can determine tumor size, shape, density, and involvement of the capsule and surrounding tissues, which are all related to the clinical stages of thymoma. Although thymoma is usually concomitant with MG, the incidence of MG is comparable among thymoma of different stages. 71

7 QU ET AL Academic Radiology, Vol 20, No 1, January 2013 Our results provide important information for both the accurate staging of thymoma and demonstrate the value of preoperative CT in determining a proper therapeutic regimen and prediction of prognosis in patients with thymoma. REFERENCES 1. Engels EA, Pfeiffer RM. Malignant thymoma in the United Stated: demographic patterns in incidence and associations with subsequent malignancies. Int J Cancer 2003; 105: Chen G, Marx A, Chen WH, et al. New WHO histologic classification predicts prognosis of thymic epithelial tumors: a clinicopathologic study of 200 thymoma cases from China. Cancer 2002; 95: Tormoehlen LM, Pascuzzi RM. Thymoma, myasthenia gravis and other paraneoplastic syndromes. Hematol Oncol Clin N Am 2008; 22: Masaoka A, Monden Y, Nakahara K, et al. Follow-up study of thymomas with special reference to their clinical stages. Cancer 1981; 48: Rosai J, Sobin LH. Histological typing of tumours of the thymus. In: World Health Organization, International histological classification of tumours. 2nd ed. New York, NY: Springer-Verlag, Okumura M, Shiono H, Minami M, et al. Clinical and pathological aspects of thymic epithelial tumors. Gen Thorac Cardiovasc Surg 2008; 56: Yanagawa M, Tomiyama N. Prediction of thymoma histology and stage by radiographic criteria. Thorac Clin Surg 2011; 21: Marom EM, Milito MA, Moran CA, et al. Computed tomography findings predicting invasiveness of thymoma. J Thorac Oncol 2011; 6: Priola AM, Priola SM, Di Franco M, et al. Computed tomography and thymoma: distinctive findings in invasive and noninvasive thymoma and predictive features of recurrence. Radiol Med 2010; 115: Tomaszek S, Wigle DA, Keshavjee S, et al. Thymomas: review of current clinical practice. Ann Thorac Surg 2009; 87: Huang J, Detterbeck FC, Wang Z, et al. Standard outcome measures for thymic malignancies. J Thorac Oncol 2011; 6:S1691 S Tomiyama N, M uller NL, Ellis SJ, et al. Invasive and noninvasive thymoma: distinctive CT features. J Comput Assist Tomogr 2001; 25: Han J, Lee KS, Yi CA, et al. Thymic epithelial tumors classified according to a newly established WHO scheme: CT and MR findings. Korean J Radiol 2003; 4: Jeong YJ, Lee KS, Kim J, et al. Does CT of thymic epithelial tumors enable us to differentiate histologic subtypes and predict prognosis? Am J Roentgenol 2004; 183: Rıos A, Torres J, Galindo PJ, et al. Prognostic factors in thymic epithelial neoplasms. Eur J Cardiothorac Surg 2002; 21: Okumura M, Ohta M, Tateyama H, et al. The World Health Organization histologic classification system reflects the oncologic behavior of thymoma: a clinical study of 273 patients. Cancer 2002; 94: Wang AW, Ji WB, Yang H, et al. Characteristics of CT and clinical pathology of thymoma. J Pract Med 2008; 24: Marom EM. Imaging thymoma. J Thorac Oncol 2010; 5:S296 S Sun ZH, Yu H, Liu HS. Diagnosis and evaluation of invasive thymoma by using CT. Chin J Radiol 1998; 32: Ge JL, Wang PJ. Diagnostic value of CT in the diagnosis of malignant thymoma. J Pract Radiol 1997; 13: Yang WT, Lei KI, Metreweli C. Plain radiography and computed tomography of invasive thymomas: clinico-radiologic-pathologic correlation. Australas Radiol 1997; 41: Whooley BP, Urschel JD, Antkowiak JG, et al. A 25-year thymoma treatment review. J Exp Clin Cancer Res 2000; 19: de Perrot M, Liu J, Bril V, et al. Prognostic significance of thymomas in patients with myasthenia gravis. Ann Thorac Surg 2002; 74: Zhang XF, Zhang QG. Clinical characteristics of 185 cases of thymoma. Chin J Clin Thorac Cardiovasc Surg 2007; 14: Sadohara J, Fujimoto K, M uller NL, et al. Thymic epithelial tumors: comparison of CT and MR imaging findings of low-risk thymomas, high-risk thymomas, and thymic carcinomas. Eur J Radiol 2006; 60: Nakagawa K, Asamura H, Matsuno Y, et al. Thymoma: a clinicopathologic study based on the new World Health Organization classification. J Thorac Cardiovasc Surg 2003; 126: Maher MM, Shepard JA. Imaging of thymoma. Semin Thorac Cardiovasc Surg 2005; 17: