The 140 years' journey of gastric cancer surgery: From the two hands of Billroth to the multiple hands of the robot

Abstract After the initial achievement by Billroth in 1881, surgery for gastric cancer has become increasingly extended. However, it turned out to be limited in Western countries after the publication that denied the role of extended surgery in the 1960s. While surgeons in Japan were still enthusiastic about extended surgery, the Japan Clinical Oncology Group (JCOG) conducted clinical trials to validate the role of extended surgery. Contrary to expectations, the efficacy of extended surgery was not demonstrated. In gastric cancer surgery, postoperative complications were reported to be associated with poor survival. A survival benefit could not be obtained by extended surgery, with high morbidity. Therefore, the paradigm had been changed from extended surgery to minimally invasive surgery (MIS). As an MIS for gastric cancer, laparoscopic surgery has been considered a practical method. Initial laparoscopic gastrectomy (LG) was first performed by Kitano in 1991. Thereafter, LG became increasingly common. Several clinical trials demonstrated the noninferiority of LG to open gastrectomy. LG is now regarded as the standard for cStage I gastric cancer, and the indication is expanding to advanced cancer. However, LG has some drawbacks owing to the restriction of movement caused by straight‐shaped forceps. Robotic gastrectomy (RG) is considered a major breakthrough to circumvent the drawbacks in LG using articulated devices. However, the solid evidence demonstrating the advantage of RG has not been proved yet. The JCOG is now conducting a randomized controlled trial to evaluate the superiority of RG to LG in terms of reducing morbidity.


| 271
TERASHIMA the standard treatment for early gastric cancer. Furthermore, in recent years the effectiveness of robotic surgery has been confirmed.
Likewise, gastric cancer surgery has changed on the basis of the results of clinical trials, and establishing a standard treatment based on scientific evidence is extremely important for the future. The history of gastric cancer surgery is reviewed in this article.

| THE DAWN OF G A S TRI C C AN CER SU RG ERY
The major history of gastric cancer is shown in Table 1. Since the initial achievement by Billroth in 1881, 1

| D IREC TI ON TOWARD E X TENDED SU RG ERY
In 1910, Groves, of the Bristol Hospital in the UK, proposed that resection of the omentum (partially the omental bursa) was necessary for a secure lymph node dissection, 6 after which omentectomy and bursectomy became widespread worldwide. Brunschwig, 7 of the Memorial Hospital for Cancer and Allied Diseases, New York, reported the results of combined resection of the pancreas and spleen for the purpose of lymph node dissection for gastric cancer.
Total gastrectomy with pancreato-splenectomy was performed in 14 patients. Only two patients died from surgery, and the safety was confirmed. 7 Lahey, of the Lahey Clinic, Boston, recommended total gastrectomy even for the lower part of gastric cancer because of the concept that gastric cancer progresses through lymphatic vessels in the stomach wall. 8 The surgical mortality rate was reduced from 34.6% to 9.4% in the latter stage, and the 5-year survival rate was as high as 12.5%, which was a sufficient result at that time. Appleby, from Vancouver, Canada, proposed the so-called Appleby operation, in which the celiac artery is ligated at the root. 9 The operation was performed in 13 patients, and operative mortality occurred in only one patient, which suggests the safety of the procedure. Thereby, extended surgery has been actively performed for gastric cancer, mainly in the US. Later, from the theory that lymph node metastasis is an indicator of systemic disease 11 and the results of a randomized controlled trial that proved that lymph node dissection did not improve survival, 12 systematic lymph node dissection has not been used in breast cancer. Gastric cancer is thought to have similar biological characteristics, and systematic lymph node dissection has gradually ceased in Europe and the US.

| TURNING P OINT
In Japan, lymph node metastasis was considered a local disease, and surgical removal of the regional lymph node was believed to lead to improvement of treatment results. In 1962, the Japanese Research Society for Gastric Cancer was established, and the Japanese Classification for Gastric Cancer was published. Lymph nodes were numbered, 13 and the national registration was also started. As data have shown the importance of lymph node dissection and the treatment results have demonstrated the effectiveness of extended dissection, surgery has become increasingly expanded.
Even para-aortic lymph nodes are considered the target of dissection, and some specialized institutions have reported better survival due to para-aortic lymph node dissection. 14 In these periods, a huge discrepancy existed between Japan and the Western countries.

| CLINI C AL TRIAL S FOR G A S TRIC C AN CER SURG ERY AND E S TAB LIS HMENT OF A S TANDARD TRE ATMENT
The major clinical trials conducted for gastric cancer are shown in Table 2. In the 1980s, two randomized controlled trials were conducted in Europe to verify the effectiveness of Japanese-style D2 lymph node dissection. One was the Medical Research Council ST01 trial conducted in the UK, 15 and the other was the Dutch trial conducted in the Netherlands. 16 Although large differences in quality assurance exist between the trials, the results were generally com- deaths in the D2 group, and no additional survival effect of D2.
Therefore, in the latter half of the 1990s, D1 dissection was positioned as the standard treatment in Europe and the US.
On the contrary, local control with lymph node dissection was believed to increase the curability of cancer in Japan, so clinical trials were planned to establish the evidence of extended surgery. The JCOG9501 study, planned by the Japan Clinical Oncology Group (JCOG), was designed to validate the superiority of prophylactic para-aortic lymph node dissection to conventional D2 dissection in advanced gastric cancer. As a result, the superiority of para-aortic lymph node dissection was not proved, and the significance of prophylactic para-aortic lymph node dissection was denied. 17 This result parallels that of a recently published ovarian cancer trial investigating the role of prophylactic para-aortic lymph node dissection. 18 The survival benefit of prophylactic pelvic and para-aortic lymph node dissection has not been proven yet. It has been speculated that the high incidence of morbidity and reoperation rate may have contributed to the negative results obtained in that study.
Since then, D2 dissection has been regarded as the standard with greater curvature invasion. 21,22 Additionally, the effectiveness of splenic hilar lymph node dissection with spleen preservation has been reported. 23 This procedure appears to be attractive not only from a safety perspective but also because of the oncological benefits. JCOG is now conducting a phase II trial evaluating the safety of laparoscopic splenic hilar lymph node dissection (UMIN000037580), and it is expected that this issue will be resolved in the near future.

| CLINICAL TRIALS CONDUCTED BY THE JCOG
In addition to the above-mentioned trials, the JCOG conducted Considering these two trials, as well as the aforementioned JCOG9501 and JCOG0110 trials, the superiority of extended surgery has not been confirmed. Although mortality was extremely low in either trial, an increase in the postoperative complication rate was observed in the extended surgery group in all the studies.
Postoperative complications have been reported to be an independent prognostic factor in gastric cancer. 26,27 We speculated that a survival benefit cannot be obtained by highly invasive surgery with high morbidity. By a curious coincidence, the contradiction of extended surgery reported by the US in the 1960s was proved more scientifically after 50 years. Since then, the paradigm had shifted from extended surgery to MIS, the establishment of a standard surgery, and the development of perioperative chemotherapy (Figure 1).

| L APAROSCOPI C SURG ERY
Laparoscopic gastrectomy (LG) has been developed as a practical method for MIS for gastric cancer. The world's first LG was reported by Kitano in 1991. 28 Since then, LG has become common worldwide with the advancement in surgical instruments such as the video system, forceps, and energy devices. However, owing to the technically demanding procedure using long linear forceps, confirming the safety and oncological tolerability took time. At first, the evidence was established through clinical trials only for early gastric cancer, for which the degree of lymph node dissection is limited and handling the stomach is easy.
The JCOG conducted a phase II clinical trial in patients with cStage I gastric cancer who underwent distal gastrectomy (DG) to investigate the safety of LG in JCOG0703. 29 The primary endpoints were the incidence of anastomotic leakage and pancreatic fistula. A total of 176 patients were enrolled, and the incidence of anastomotic leakage and pancreatic fistula were both extremely low, at 1. On the other hand, LG for advanced gastric cancer has been considered to require further examination from the viewpoints of tumor handling, accuracy of D2 lymph node dissection, and safety. Phase III clinical trials were conducted in Japan, South Korea, and China to examine the noninferiority of LDG to ODG for advanced gastric cancer. [34][35][36] The result of the primary analysis was first reported in the CLASS01 trial performed in China, which proved the noninferiority of LDG to ODG. Noninferiority was confirmed using the point estimation of 3-year relapse-free survival. 34 Next, the results of the KLASS02 study in Korea were reported, and the noninferiority of LDG was also proved. 35 The results of the JLSSG study in Japan will be reported in 2021, 36 and these results may allow laparoscopic surgery for advanced gastric cancer as the standard treatment. The incidence of pancreatic fistula has also been reported to be significantly higher in LDG than in ODG in both retrospective and prospective studies. 39,40 This is probably owing partly to the compression of the pancreas during suprapancreatic lymph node dissection using straight-shaped forceps. Circumvention of the restriction of movement is a major issue in laparoscopic surgery. Another important issue with LG is the learning curve. Even among the technically certified physicians of the Japanese Society of Endoscopic Surgery, which is the only credited system for laparoscopic surgery in the world, the pass rate is in the 20% range, and the acquisition of LG techniques is considerably difficult, which cannot be denied. These points are the problems that must be solved by laparoscopic surgery.

| ROBOTI C G A S TREC TOMY
In recent years, robot-assisted or robotic gastrectomy (RG) has been highlighted as an MIS for gastric cancer. RG overcomes the Several reports of meta-analyses have compared RG and LG. [45][46][47] RG has been reported to show a prolonged operation time and a slightly smaller amount of blood loss; however, no reports have suggested the absolute advantage of RG. In the report from Japan, the safety of RG was evaluated in single-arm clinical phase II studies, 48,49 and one report indicated that RG had fewer complications by comparing RG and LG. 50 No such report has been made by countries other than Japan. A recent meta-analysis from Italy has reported that robotic surgery excels in short-term results, but all the underlying articles are from Japan. 51 In a multicenter prospective cohort study conducted in Japan, the postoperative complication rates were reported to be reduced to less than half in RG as compared with laparoscopic surgery as the historical control. 52 While in Korea a multicenter prospective nonrandomized control study was conducted with 434 patients who underwent RG or LG. 53 No significant difference was observed between RG (n = 223) and LG (n = 211) in the incidence of postoperative complications (11.9% vs 10.3%), and the mortality rate was 0% in both groups. However, the operative time was approximately 40 minutes longer and the cost of surgery was approximately 5000 USD higher in RG than in LG. Thus, the authors concluded that RG has no advantage that counterbalances the time and cost disadvantages.
A single-center randomized controlled trial recently reported in China has not shown the usefulness of RG over LG in short-term results. 54 The slight differences in the content of surgery between Japan and other countries are undeniable, but it is significant to show the reduction of postoperative complication rates as proof of the usefulness of RG. The JCOG is currently conducting a multicenter prospective randomized phase III trial to validate the superiority of RG in terms of reducing the morbidity in JCOG1907 (UMIN000039825). The primary endpoint is the incidence of intraabdominal infectious complications, and the sample size is 1040 cases. This trial will reveal the real benefit of RG.

| FUTURE PROS PEC TS
In the future, endoscopic surgery is expected to be used for more complicated surgical procedures and to improve prognosis by minimizing surgical invasiveness, even in highly advanced stages of gastric cancer. Furthermore, the introduction of artificial intelligence is expected to help in the development of new surgical procedures. It can be said with some certainty that endoscopic surgery will play a leading role in gastric cancer therapy in the near future. Evidence based on clinical studies must be urgently established to facilitate such advances in endoscopic surgery.
In addition, the efficacy of perioperative chemotherapy has remarkably progressed these days. 55,56 Immuno-checkpoint inhibitors were also introduced in the perioperative treatment for gastric cancer. With the progress of systemic chemotherapy, conversion therapy is becoming more and more common. 57 It has become sometimes possible to curatively resect a tumor that was thought to be unresectable before treatment. The indication and procedure of surgical treatment for advanced gastric cancer may drastically change in the future.

ACK N OWLED G M ENTS
The author thanks the main office and Stomach Cancer Study Group of JCOG for their support in conducting clinical trials for gastric cancer. This work was supported in part by the National Cancer Center