Prognostic Factors in Patients Submitted to Surgical Treatment for Thymoma

Correlation of CT Scan Parameters With Histology, Stage and Prognosis in Surgically-treated Thymomas

Thymoma is the most common primary tumor of the anterior mediastinum. Complete surgical resection is the mainstay of treatment of these tumors.

The staging and histological classification of thymoma is still a matter of discussion. Preoperative computed tomography (CT) scan parameters that correlate with histology, stage and prognosis also still have to be completely assessed. The aim of this study is to evaluate the potential association between clinical, radiologic and pathologic characteristics in patients submitted to surgical treatment for thymoma, assessing their prognostic value.

Data of patients submitted to surgical resection for pathologically proven thymoma at our Department of Thoracic Surgery between January 2005 and December 2015 will be retrospectively reviewed.

The correlation of preoperative CT scan features, histological and pathological characteristics of thymomas will be evaluated, assessing the prognostic role of these factors.

• Rationale: Even though the World Health Organization (WHO) histological classification and Masaoka-Koga staging system are currently the most widely used, some studies have validated alternative pathologic classifications, as the one proposed by Suster and Moran. Following the proposals of a working group of the International Thymic Malignancies Interest Group (ITMIG) and the International Association for the Study of Lung Cancer (IASLC), a new staging system for thymic epithelial tumors has been included in the last Tumor Node Metastases (TNM) cancer staging system revision. TNM staging has been shown to have a stronger correlation with disease-free survival (DFS) analysis in comparison with the Masaoka-Koga staging system. Moreover, many studies sustained the role of preoperative CT-scan in predicting the pathologic features of thymic tumors. In this context, the potential association between clinical-radiological features and histo-pathological appearance may therefore provide a more accurate prognostic stratification in patients with thymoma.
• Targets: The aim of this study is to evaluate the potential association between clinical, radiologic and pathologic characteristics in patients submitted to surgical treatment for thymoma, and their prognostic value.
• Design of the study and parameters: Data about the patients submitted to surgical resection for pathologically proven thymoma at our Department of Thoracic Surgery between January 2005 and December 2015 will be retrospectively reviewed. Patients enrolled in the study will undergo a follow up until the date of death, and not later than March 31th, 2020. The following data will be entered in a prospective database: age, gender, presence of myasthenia gravis, neoadjuvant therapy, date of operation, surgical access (median sternotomy, thoracotomy or VATS), tumor histology according to WHO classification revised in 2015, Masaoka-Koga staging, adjuvant chemotherapy or radiotherapy, recurrence-free and overall survival, and cause of death. All histopathologic reports will be reviewed. Histology of each case will be reclassified according to the proposed Suster and Moran classification . Moreover, Masaoka stages will be reclassified according to the newly proposed IASLC/ITMIG TNM staging system. Preoperative CT scans will be reviewed. The following features will be evaluated using a dedicated software (Philips Intellispace): location, 2-axes diameters, volume, shape, presence of necrosis, calcifications, or lymphoadenopathy, presence and length of pleural contact, presence of pleural effusion or dissemination, pericardial effusion, invasion of mediastinal fat, great mediastinal vessels, pericardium or lung, and contrast-enhancement pattern. The association between clinical, radiological and pathologic features and survival will be analyzed to identify possible prognostic factors.
• Histopathological and staging review: all histopathological reports will be reviewed. Histology of each tumor will be classified according to the 2015 WHO classification and re-classified according to the classification proposed by Suster and Moran as follows: A, AB, B1 and B2 thymomas will be re-classified as well-differentiated thymic neoplasms (or typical thymoma); B3 tumors will be re-classified as moderately-differentiated thymic neoplasms (or atypical thymoma). Tumors will be staged according both to the Masaoka-Koga and the 8th edition of the IASLC/ITMIG TNM staging system.
• CT scan features and image interpretation: All patients will be sumbitted to chest CT scan before surgery. CT scans will be reviewed by two radiologists, expert in thoracic oncology. Differences will be resolved by consensus. The following features will be evaluated using a dedicated software (Philips Intellispace): location of the tumor, 2-axes diameters, volume, shape, presence of necrosis, calcifications, pathological lymph nodes, presence and length of pleural contact, presence of pleural effusion or dissemination, pericardial effusion, invasion of mediastinal fat, great vessels, pericardium or lung, and contrast-enhancement pattern. Location of the tumor will be classified as right, left, or median according to the site of the intersection of the two main diameters of the lesion. The size of the lesion will be defined as the length of the largest of the three lesion diameters. Tumor volume will be calculated with the dedicated software after the identification of the borders of the lesion. Shape and contours will be defined as regular or irregular. A contrast-enhancement pattern will be assessed in the images obtained 30 seconds after contrast medium injection, and described as homogeneous or heterogeneous. Necrosis will be defined as an area without contrast-enhancement or low-density value comparable to water. A short-axis diameter of ≥ 1 cm will be used as the threshold for pathological lymph-nodes. Length of pleural contact will be measured with multiple lines along all the borders of the lesion in contact with the mediastinal pleura. Invasion of the mediastinal fat will be established in case of disomogeneity, ill-definition and hypodensity. Contact with great vessels will be reported whenever clear distinction of margins was not possible. Lung invasion will be documented when the border between the tumor and the lung was markedly irregular and/or the lung was compressed and pinched by the lesion.
• Data management: The authorized staff of the Thoracic Surgery Department will acquire data from the institutional electronic dossier at IRCCS Ospedale San Raffaele. All data will be protected from unauthorized access. Data will be archived in an electronic database (Microsoft Excel) in pseudo-anonymous form, and patients enrolled in the study will be identified with a progressive number code.
• Sample size: We identified 50 patients who can potentially be enrolled in the study.
• Diagnostic and therapeutic procedures: The patients enrolled in the study gave a standard informed consent, and underwent routine preoperative workup (contrast enhanced chest CT-scan, FDG-PET scan, pulmonary function tests, blood tests) and surgical treatment for thymoma following international guidelines.
• Ethic evaluation: The study will be conducted following ethical principles of the Declaration of Helsinki and the current legislation on observational studies.
• Declaration on the management of patient's consent: All the patients who will be considered for this study have been staged and surgically treated following international guidelines. In addition, all patients provided standard informed consent to the planned therapy prior to surgery. In accordance with the Privacy Regulation, we wrote a specific informed consent form for the enrollment in this study, and all patients will be required to fill it out. However, considering the long follow-up (up to over 10 years), it is likely that some subjects have already died. The collection and analysis of data for this retrospective observational study does not involve physical, psychic and social risks towards enrolled patients. Nevertheless, reasonable efforts will be made to re-contact patients who are still alive.
• Internal organization and statistical analysis: This research will be carried out by the authorized staff of the Department of Thoracic Surgery involved in the study. The final analysis will be responsibility of the PI. Analysis was performed with the SPSS, v. 18 software (Chicago, IL, USA). In order to evaluate the association between clinical variables, histopathology and radiologic features, and their prognostic value, continuous variables will be dichotomized according to their median value. Comparisons of categorical variables among the groups of patients will be performed by means of either Chi-square test or Fisher Exact test as appropriate. Survival curves will be estimated by the Kaplan and Meier method. Cox regression analysis will be used to assess the risks of the variables. Survival rates of patients grouped according to selected variables will be compared by means of the log-rank test. On the univariate analysis basis, in order to evaluate the independent contribution of the variables on recurrence-free and overall survival, a multivariate analysis will be performed using the Cox regression method. The analysis of the retrospective data will be based on the consultation of the institutional electronic archive (Galileo). The visualization of the images (CT scan) will take place using using a dedicated software (Philips Intellispace) available on the workstations of the Radiology Department. A duration of approximately 30 days for the completion of the study is expected.
• Privacy: The PI and Ospedale San Raffaele own the data and results of the study, which will constitute the material for scientific publications. In these works data will be reported anonymously.

Inclusion Criteria:

• Patients with surgically-treated thymoma

Exclusion Criteria:

• Incomplete resection
• Thymic carcinoma