An 11 year (1995-2005) experience with thymic epithelial tumors was previously reported . The 5-year survival rate for a Masaoka stage IV (including IVA and IVB) thymoma was only 52%, which is significantly lower compared to stages I (96%), II (100%) and III (71%) tumors. The complete resection rate was also low (26.3%) for stage IV thymomas compared to stages I (100%), II (100%) and III (85.3%) tumors. The reason for the poor prognosis of stage IV thymomas was incomplete resections. Despite the fact that complete resection has been identified as a prognostic factor associated with long term survival of patients [9, 10], few studies have been carried out in patients with stage IVA thymomas due to the difficulty of achieving a total resection, its rare occurrence and its indolent natural history.
Stage IVA patients can be treated with primary chemotherapy, radiotherapy  or chemoradiotherapy  without surgical resection, as well as debulking surgery with adjuvant radiation . Although several investigators have attempted to improve the survival of patients with locally advanced thymoma, the data has been inconsistent with regard to the response rate and survival. Our approach is with aggressive surgery to eradicate the tumor. The problem lies in how a thymoma with extensive pleural dissemination, should be treated; in these cases, complete resection is not possible by simple parietal pleurectomy and lung preserving surgery. The most reliable resection method for stage IVA thymomas might be the EPP; because this procedure can remove invisible tumor cells as well as all gross implants.
Recent studies have reported favorable experiences with the EPP for cure of stage IVA thymomas [14–16]. The EPP for stage IVA thymomas has been performed at this hospital since 2005. This series had no cardiorespiratory morbidities, no perioperative death, and a reasonable hospital stay. These favorable results may be due to relatively young patients, with a good performance status and no underlying pulmonary disease in addition to the meticulous postoperative care.
Wright and colleagues  carried out EPP in five stage IVA patients. The five-year survival rate was reported to be 75% and was 50% for 10-years, which was fairly good compared to previous reports. Ishikawa and colleagues  reported 11 patients with invasive thymomas disseminated into the pleural cavity that underwent multimodality therapy. The patients that underwent EPP (n = 4) had better local recurrence free survival compared to the patients that did not have an EPP (n = 7) (5-year: 75% vs. 16%, 10-year: 75% vs. 0%). Huang and colleagues  reported on multimodality therapy in 18 patients with stage IVA thymomas. Complete resections were performed in 12 patients including nine pleurectomies and three EPP procedures. Among three out of the nine patients with pleurectomies, recurrences developed in the pleura. By contrast, three patients that underwent an EPP were alive without disease recurrence at 4, 32, and 112 months. These results suggest that the EPP achieves a higher complete local control rate than other surgical procedures. Among the completely resected patients in this series, distant metastases developed in one patient, who was the first EPP case at this hospital, and had no systemic therapy. After this experience, systemic therapy for stage IVA thymomas became an important part of treatment.
Many studies suggest a multimodality approach may lower the recurrence rate and increase the resectability of advanced thymomas and this approach is currently widely accepted [6, 7, 15, 17]. However, how to combine these modalities remains controversial in patients with stage IVA thymomas. We concluded that an optimal treatment sequence for stage IVA thymomas might include induction chemotherapy, en bloc extended total thymectomy with EPP, and adjuvant chemotherapy. The three most recent consecutive patients (patients 5, 6 and 7) were treated by the protocol discussed above. They received the same chemotherapy agents (doxetaxel and cisplatin) for induction chemotherapy and they all had a partial response. Complete resection was performed by an en bloc extended total thymectomy and EPP. Adjuvant chemotherapy included CAP (cyclophosphamide, doxorubicin and cisplatin), and all the patients are currently alive without disease recurrence. The patients that were planned to have EPP did not receive preoperative radiotherapy. Preoperative radiotherapy may have adverse effects on the postoperative outcomes because it can damage the heart and lungs with the wide extent of the radiation field needed for treatment. Radiotherapy was performed in only one patient with macroscopic residual tumor around the innominate vein. Hemithoracic radiation was not carried out in this series. All of the patients that received adjuvant chemotherapy alone (n = 4) had no recurrence. Huang and colleagues  treated four patients with induction chemotherapy followed by EPP and then adjuvant hemithoracic radiation for stage IVA thymomas. Wright  suggested induction chemotherapy followed by EPP, and then adjuvant chemoradiation. However, further study of the role of adjuvant chemotherapy or radiotherapy is needed.
En bloc extended total thymectomy and EPP was performed with extended incision of a posterolateral thoracotomy. During the surgery, meticulous attention was needed to prevent droplet metastasis. In order to prevent tumor cell spillage, black silk 3-0 sutures were used immediately when tearing of the parietal pleura occurred during extrapleural dissection. In two cases (patients 5 and 7) requiring resection of the innominate vein, a median sternotomy followed the posterolateral thoracotomy. Performing a posterolateral thoracotomy followed by a median sternotomy will lessen the risk for pleural droplet metastasis in the opposite thoracic cavity. If EPP is performed after opening the opposite pleura, gravity may enhance the possibility of droplet metastasis on the opposite side.
Selection criteria for the EPP must be considered. First, the patient's functional status has to be good enough to tolerate the pneumonectomy. Second, there should be no metastatic disease in the opposite thorax or the extrathoracic cavity. In this series, the PET-CT was used to rule out distant metastases. Third, complete or nearly complete resection should be expected when performing an EPP. Otherwise, a palliative approach should be considered in inoperable cases. The CT has been used to examine tumor invasion of neighboring organs, such as the heart, great vessels, and chest wall. Usually, the innominate vein and the superior vena cava can be safely resected. For patient 7, preparations were made for cardiopulmonary bypass because of concerns about aortic arch invasion. Fortunately, there was no need for aortic arch replacement; aortic invasion was not considered an absolute contraindication to the procedure. Forth, if a limited resection can achieve a complete resection, in individual pleural and pulmonary lesions, the EPP should be reserved for where it is most effective.
In this series, the outcome of extrapleural pneumonectomy was favorable with low morbidity and no mortality. However, the follow-up duration for this study was comparatively short for assessment of late recurrence and long term survival. In the future, multicenter trials are needed to establish standard treatment using a multimodality therapy including surgical procedures.