In the current study, we sought to evaluate the impact of surgical methods on short-and long-term outcomes after neoadjuvant therapy in patients with ESCC. We only included patients underwent neoadjuvant treatment followed by Mckeown esophagectomy to guarantee the negative margins and systematic lymphadenectomy. MIE resulted in a shorter operative time, and less bleeding during operation, and after surgery, the incidence of anastomotic leakage was lower than open esophagectomy. For long-term survival, the OS and DFS between the two procedures was comparable in patients underwent preoperative chemotherapy. Nevertheless, MIE was significantly associated with better OS in patients underwent neoadjuvant chemoradiation. Minimal invasive surgery was further demonstrated as an independently prognostic factor for favorable OS.
Traditionally, surgery is considered as the primary treatment for resectable esophageal cancer patients, but the surgical outcomes for the locally advanced disease seems hard to improve. Since previous studies have demonstrated the positive role of neoadjuvant therapy on the prognosis of esophageal cancer patients [3, 6, 15], the use of preoperative therapy followed by surgery has become common practice for the locally advanced ESCC patients in the clinical application. Neoadjuvant CT and CRT contribute to the clearance of micrometastatic disease and tumor down staging, which benefits a more radical surgical resection and a better survival. But at the same time, both chemotherapy and radiotherapy may result in tissue edema and adhesion, and increase the difficulty of the surgical procedure. Furthermore, chemoradiotherapy followed by surgery is associated with significant side effects, including radiation pneumonitis, postoperative pulmonary complications, and pericarditis [16]. Recent years, two primary methods for esophagectomy are traditional thoracotomy and minimally invasive surgery combined thoracoscopy and laparoscopy. Although a number of institutions have investigated the benefits and disadvantages of the two procedures, there is very limited data on the short-and long-term clinical outcomes of the open esophagectomy and MIE concerning ESCC patients underwent neoadjuvant therapy [10, 17–19]. Considering the treatment-related adverse effects, therefore, it is essential to clarify the appropriate surgical method for esophagectomy after induced therapy.
In terms of operative outcomes, our study indicated that shorter operation time and less intraoperative blood loss were noted in the MIE group, which was consistent with the previous reports [19, 20]. However, we found the lymph node yield in open surgery group was higher than in MIE group (mean: 30 vs 23). There is reason to suspect the traditional open surgery is beneficial to perform the systematic lymphadenectomy under direct vision, while the operation field observed by the monitor in MIE was lacking in partial depth perception due to its two dimensions. Although the number of lymph node harvested was fewer in MIE than open surgery in our study, the lymph node yield was in excess of the recommended threshold of 11–15 nodes required for accurate staging [21]. It is well established that dissected lymph node comprehensively provides useful prognostic information after surgery. However, it is controversial whether excise all involved lymph nodes actually improves long-term prognosis [22]. Particularly, the number of lymph nodes removed has relatively low sensitivity for N staging after neoadjuvant therapy. The radical mediastinal lymphadenectomy would prolong the surgery time and the duration of collapsed lung, and increase the risk of damage in nerve and lymphatics. In addition, during surgery, the dissection of the lymph nodes located deep in the mediastinum is inevitable to avoid stretching or compressing of the lung parenchyma, which is related to the pulmonary complications after esophagectomy. Furthermore, lung toxicities are the most common adverse effect of neoadjuvant chemoradiotherapy, and the incidence of postoperative pulmonary complications was demonstrated a significant increase in patients received neoadjuvant treatment [16, 23]. Thus, the significance of extensive lymphadenectomy after induction needs to be further discussed.
In most previous studies, the leakage rate was reported as similar for both procedures [10, 19, 24]. Nevertheless, in this study, the major complications between two groups were comparable except the anastomotic leakage. Open esophagectomy was suggested to be associated with a higher risk of cervical anastomotic leakage than MIE. According to our clinical experience, postoperative pain in patients underwent open esophagectomy is still more severe than those had MIE, even if the high doses of analgesic has been taken in the first week after operation. The intense pain would lead to the fragile cough and patients are unwilling to expectorate spontaneously. Meanwhile, the intense pain prevents patients from getting out of bed and do activities as soon as possible. These disadvantages may cause pulmonary infection and hypoxemia, which are the major risk factors for anastomotic leakage.
The comparisons of long-term outcomes between open esophagectomy and MIE have been widely discussed [17, 19, 25, 26]. However, the effect of different procedures on prognosis is still controversial. Most studies demonstrated that MIE appeared to produce better survival compared with open surgery [19, 26], while similar outcomes between two procedures were also reported [17, 25, 27]. In our study, Cox model demonstrated that MIE was an independently prognostic factor (HR = 0.606; 95% CI, 0.384–0.958; P = 0.032). Nevertheless, in the subgroup analysis, we found the prognostic effect of surgical method was various in patients received different neoadjuvant treatment. For patients underwent nCT, OS and DFS were comparable between patients in open and MIE group. On the contrary, significant difference in OS between the two surgical methods was observed in patients underwent nCRT and MIE related to a better prognosis. The reason may be concluded as follows: firstly, neoadjuvant chemoradiation improves the pathologic complete response (pCR) rate compared with chemotherapy alone, which shrinks the tumors dramatically [28]. In particular, MIE with the thoracoscope and laparoscope could amplify the surgical filed, which is beneficial for the dissection of tumor in relatively small size and decreases the surgery duration. Thus, patients with nCRT followed by minimal invasive surgery demonstrated improved OS. Subsequently, some ESCC patients could not tolerate the concurrent preoperative chemoradiotherapy due to the serious side effects and only underwent neoadjuvant chemotherapy. Under the circumstances, pCR is hardly achieved and the size of primary tumor would not change as obviously as after nCRT. According to our experience, MIE did not prevail over traditional open surgery in dissecting complex structure or oversize tumor surrounded by denser tissue. Open esophagectomy with hand-assisted seems more precise in operating complicated surgery than MIE that can only rely on the surgical apparatus. Number of harvested nodes was less in the MIE group, suggesting that MIE may involve a steep learning curve. Therefore, comparing MIE with the open procedure to esophageal resection showed comparable long-term survival in patients received nCT.
In this study, we only enrolled ESCC patients who underwent neoadjuvant therapy followed by three incisions esophagectomy with traditional open surgery and total MIE. Other approaches such as Ivor Lewis esophagectomy or hybrid MIE [26] were excluded, which guaranteed the en bloc resection and normalized lymphadenectomy. Nevertheless, it is important to recognize the limitations of our study. This is a retrospective analysis, and the selection bias is inevitable. Patients were not randomly assigned to open esophagectomy or MIE group but were treated based on surgeon evaluation and patient deliberation. Therefore, patients in advanced disease would be inclined in the open esophagectomy group, which may lead to the worse outcomes than the other group intrinsically. In addition, the sample size is limited because of the rigorous enrollment criteria, and large population study is needed to validate the association between surgical procedures and prognosis for ESCC patients underwent neoadjuvant treatment.