Current status of minimally invasive esophagectomy for esophageal cancer: Is it truly less invasive?

Abstract Esophagectomy with extended lymphadenectomy remains the mainstay of treatment for localized esophageal cancer. However, it is one of the most invasive procedures with high morbidity. To reduce invasiveness, minimally invasive esophagectomy (MIE), which includes thoracoscopic, laparoscopic, mediastinoscopic, and robotic surgery, is becoming popular worldwide. Thoracoscopic esophagectomy in the prone position is ergonomic for the surgeon and has better perioperative arterial oxygen pressure/fraction of inspired oxygen (P/F) ratio. Thoracoscopic esophagectomy in the left decubitus position is easy to introduce because it is similar to standard right posterolateral open esophagectomy (OE) in position. It has a relatively short operative time. Laparoscopic approach could potentially have a substantial effect on pneumonia prevention under the condition of thoracotomy. Mediastinoscopic surgery has the potential to reduce pulmonary complications because it can avoid a transthoracic procedure. In robotic surgery, in the future, less recurrent laryngeal nerve palsy will be expected as a result of polyarticular fine maneuvering without human tremors. In studies comparing MIE with OE, mediastinoscopic surgery and robotic surgery are usually not included; these studies show that MIE has a longer operative time and less blood loss than OE. MIE is particularly beneficial in reducing postoperative respiratory complications such as atelectasis, despite no dramatic decrease in pneumonia. Reoperation might occur more frequently with MIE. There is no significant difference in mortality rate between MIE and OE. It is important to recognize that the advantages of MIE, particularly “less invasiveness”, can be of benefit at facilities with experienced medical personnel.


| INTRODUC TI ON
Torek reported the world's first case of transthoracic esophagectomy in 1913. 1 It was carried out for a 67-year-old woman through the left thoracic cavity, with the proximal ends of the fourth through seventh ribs transected near their tubercles. 1,2 Food passed from a stoma in the proximal esophagus through an external tube to the gastrostomy. 1 Since this first case, esophagectomy with extended lymphadenectomy remains the mainstay of treatment for localized esophageal cancer. [3][4][5] However, it is one of the most invasive procedures and is associated with high morbidity. 6,7 One hundred years after the world's first case, minimally invasive esophagectomy (MIE) in the form of thoracoscopic and/or laparoscopic surgery is spreading around the world. [8][9][10] Furthermore, mediastinoscopic and robotic surgery are being introduced as new MIE for esophageal cancer. 11,12 Some investigators have reported advantages of faster recovery and lower morbidity with MIE compared with open esophagectomy (OE). 13 However, whether or not MIE is truly less invasive remains controversial. In the present review, MIE is defined as esophagectomy carried out using endoscopy, which provides a magnified view with decreased body wall destruction. MIE comprises several procedures, namely, thoracoscopic, laparoscopic, mediastinoscopic, and robotic esophagectomy. Moreover, "less invasive" is comprehended as fewer respiratory complications. The purpose of this review is to clarify the advantages and challenges of MIE.

| Genesis of thoracoscopic esophagectomy
Cuschieri et al 9 first reported on thoracoscopic esophagectomy in the left decubitus position (TELD) in 1992. It has attracted attention as a potentially less invasive procedure. In the first case, an electronic pressure-controlled carbon dioxide (CO 2 ) insufflator was already being used with 8 mmHg of pressure to achieve lung collapse. 9 As the positioning and approach for TELD and standard right posterolateral OE are almost identical, TELD became popular and spread worldwide quickly. 14 Thoracoscopic esophagectomy in the prone position (TEP) was also reported first by Cuschieri et al in 1994. 10 They carried out TEP for six patients in the full prone jackknife position. They described achieving excellent access to the mediastinum and the entire intrathoracic esophagus as well as good visual exposure because the right lung fell away from the operative field by gravity. 10

| Learning curve for the thoracic procedure of MIE
There is a learning curve for both TELD and TEP. Osugi et al 15 reported that there were significant differences between the first 36 cases and the 41 subsequent cases of TELD in terms of length of thoracic procedure, amount of blood loss, and pneumonia. Guo et al 16 reported that at least 30 cases are needed to   reach a plateau for TELD. After more than 60 TELD, lower morbidity could be achieved. Concerning TEP, Oshikiri et al reported   on an individual surgeon's learning curve over the course of 100 procedures. They concluded that approximately 30-60 cases are needed to reach a plateau for TEP and a lower morbidity rate. 17 For both procedures, 30-60 cases are needed to reach a plateau in the learning curve.

| Outcomes of TELD versus TEP
Many studies have compared TELD and TEP. Shen et al 18 evaluated the surgeon's physical and mental stress during both procedures in a randomized control study. The drop in the eye-blink rate of the surgeon at the end of the thoracic procedure from baseline was higher in the TELD group than in the TEP group (3.0 ± 1.4 blinks/ min vs 1.2 ± 0.9 blinks/min, P < 0.001). The surgeon's symptom scale score was worse after TELD compared with TEP. The authors concluded that TEP is associated with a lighter workload and better ergonomic results than TELD. 18 Noshiro et al 19 also reported that TEP is associated with better surgeon ergonomics and better operative exposure than TELD because it is easier to explore the operative field around the left recurrent laryngeal nerve (RLN) during TEP. Mean duration of TEP was 307 minutes, which was significantly longer than the mean duration of TELD. However, the total estimated blood loss with TEP was significantly lower. There was no significant difference in the incidence of postoperative complications for the two procedures. 19 Concerning postoperative oxygenation, some investigators reported that the TEP group had a significantly higher arterial oxygen pressure/fraction of inspired oxygen (P/F) ratio after surgery than the TELD group. 20,21 In contrast, no significant differences were observed in the frequency of respiratory complications. 20,21 In the TEP group, blood loss was significantly lower (P < 0.001), and the number of dissected intrathoracic lymph nodes was significantly higher (P = 0.03) than in the TELD group. 20 In the TEP group, length of thoracic procedure was significantly longer and there was less blood loss. 21 Comparison of short-term outcomes between 54 cases of TEP and 33 cases of TELD showed that total and thoracic estimated blood loss, incidence of postoperative pulmonary complications, duration of intensive care unit (ICU) stay, and duration of hospital stay were significantly lower in the TEP group. 22 Consequently, the advantages of TEP include better ergonomics, less blood loss, more dissected mediastinum lymph nodes, and a better P/F ratio. The advantages of TELD are similarity in positioning and approach as standard right posterolateral OE and shorter operative time than TEP. Concerning pneumonia, the advantage of TEP is controversial even though the perioperative P/F ratio is better with TEP. However, there were no significant differences between the TEP and TELD groups in the incidence of adverse events other than pneumonia.  26 They compared hybrid minimally invasive laparoscopic-thoracotomic esophagectomy (HMIE) with OE. 26 Their analysis showed that HMIE is associated with a reduction in postoperative pulmonary morbidity, less perioperative blood loss, and shorter duration of hospital stay. This was essentially a comparison between laparoscopic and open laparotomy; the findings suggested that laparoscopic surgery might be advantageous. 26 In conclusion, laparoscopic approach could potentially have a substantial effect on pneumonia prevention at least under the condition of thoracotomy.

| Mediastinoscopic surgery
In this section, mediastinoscopic surgery indicates THE with mediastinoscope and/or laparoscope assistance.
As the conventional mediastinoscope has a specialized design for procedures involving a narrow operative field around the tip, it is unsuitable for radical esophagectomy with en bloc lymphadenectomy.
In fact, use of a conventional mediastinoscope has been limited to esophageal mobilization with or without lymph node sampling in mediastinoscope-assisted THE (MATHE). [27][28][29] Fujiwara et al 30 developed hand-assisted laparoscopic THE with a systematic procedure for en bloc infracarinal lymph node dissection. In their procedure, transhiatal mobilization of the esophagus with lymph node dissection is done using a standardized method with hand-assisted laparoscopic techniques. The cervical esophagus is mobilized using a left cervical approach. A total of 57 patients underwent esophagectomy, of whom 34 underwent the transthoracic procedure for upper mediastinal lymphadenectomy following esophagectomy and gastric tube reconstruction via the retrosternal route. Total operative time was significantly shorter in the laparoscopic THE group than in the laparoscopic THE with transthoracic procedure group (216 and 370 minutes, P < 0.001). Blood loss in the laparoscopic THE group was less than that in the laparoscopic THE with transthoracic procedure group (139 and 238 mL, respectively), even though this difference was not statistically significant. However, fewer lymph nodes were retrieved in the laparoscopic THE group than in the laparoscopic THE with thoracic procedure group (24 and 39, P < 0.001).
The no residual tumor (R0) resection rate in both groups was similar.
The incidence of RLN palsy was significantly higher in the laparoscopic THE with transthoracic procedure group, whereas there were no significant differences in the incidence of pneumonia between the two groups. The authors concluded that hand-assisted laparoscopic THE, which includes a systematic mediastinal lymph node dissection, is safe and feasible as a type of MIE. 30 Lymphadenectomy in the upper mediastinum is insufficient with laparoscopic THE alone, because the number of retrieved nodes was significantly lower than with laparoscopic THE followed by a thoracoscopic procedure. 30   that transcervical mediastinoscopic lymphadenectomy is a safe and feasible procedure that enabled total non-transthoracic radical esophagectomy in combination with a transhiatal approach. 34 Consequently, a summary of Takeuchi, Nozaki, and Seesing's reports concluded that MIE is a safe procedure with similar mortality to OE and is particularly beneficial in reducing postoperative respiratory complications, but might be associated with higher reoperation rates (Table 1). [37][38][39] Concerning long-term health-related quality of life, Barbour et al 40 compared 110 OE with 377 MIE. Mean symptom scores for pain were significantly higher in the OE group than in the MIE group for 2 years after surgery (P = 0.036). In addition, mean constipation scores were significantly better for the MIE group at 3 months after surgery (P = 0.037). They concluded that OE is associated with more pain and constipation than MIE. Characteristics compared between OE and MIE are summarized (Table 2).

| MINIMALLY INVA S IVE E SOPHAG EC TOMY VER SUS OPEN E SOPHAG EC TOMY
Practically, some difficulties are encountered in the application of the satisfactory data from the high-volume centers to low-volume institutions. The reported data were results obtained at limited outstanding high-volume facilities. Moreover, Nishigori et al 41 reported that high-volume hospitals had lower risk-adjusted 30-day and operative mortality rates compared with low-volume hospitals. Data from the Japanese nationwide web-based database included data on outcomes not only from high-volume, but also from low-volume hospitals, so that data from the Japanese nationwide web-based database may not reflect the actual reduced invasiveness of MIE. 6,37 For example, lymphadenectomy around the RLN in MIE requires advanced skills to prevent nerve palsy, which may lead to pneumonia.
In the Japanese nationwide web-based database, RLN palsy rate was significantly higher for MIE than for OE. 37 High RLN palsy rates in low-volume centers may have an effect on the non-significant decrease in the incidence of pneumonia in MIE. High reoperation rates in MIE were also similarly accounted for. Nozaki   Registry: NTR3291). 47 The study started in January 2012. Follow up will be 5 years. Short-term results will be analyzed and published after discharge of the last randomized patient. 47 In conclusion, it is not clear whether RAMIE is better for reducing complication rates. A decreased incidence of RLN palsy leading to pneumonia will be expected in the future as a result of fine polyarticular maneuvering without human tremors. If RAMIE cannot reduce the incidence of RLN palsy and pneumonia, the significance of RAMIE will be controversial.

| CON CLUS ION
Minimally invasive esophagectomy is particularly beneficial in reducing the incidence of postoperative respiratory complications.
Reoperation or reintervention might occur more often with MIE than with OE. The mortality rates for MIE and OE are similar. It is important to recognize that the advantages of MIE, particularly "less invasiveness", can be availed at facilities with experienced medical personnel.

CO N FLI C T S O F I NTE R E S T
Authors declare no conflicts of interest for this article.