Objective:To investigate the short-term outcomes of Da Vinci robotic versus laparoscopic and open surgery for locally advanced Siewert type Ⅱ and Ⅲ adenocarcinoma of esophagogastric junction (AEG).Methods:The propensity score matching and retrospective cohort study was conducted. The clinicopathological data of 404 patients with locally advanced Siewert type Ⅱ and Ⅲ AEG who underwent radical gastrectomy in the First Hospital Affiliated to Army Medical University from January 2009 to April 2019 were collected. There were 331 males and 73 females, aged from 34 to 90 years, with a median age of 62 years. Of the 404 patients, 104 undergoing Da Vinci robotic radical gastrectomy were allocated into robotic group, 205 undergoing laparoscopic radical gastrectomy were allocated into laparoscopic group, and 95 undergoing open radical gastrectomy were allocated into open group. Observation indicators: (1) the propensity score matching conditions and comparison of general data among the three groups after propensity score matching; (2) surgical situations; (3) intraoperative lymph node dissection; (4) postoperative situations; (5) postoperative complications; (6) follow-up. Patients were followed up at postoperative 1 month by outpatient examination and telephone interview to detect survival and severe complications up to June 2019. The propensity score matching was used to perform 1∶2∶1 nearest neighbor matching by SPSS 23.0 and R software 3.6.1 Matchit among the robotic group, laparoscopic group and open group. Measurement data with normal distribution were represented as Mean± SD, and comparison among groups was done using one-way ANOVA analysis. Measurement data with skewed distribution were represented as M (range), and comparison was done using the Kruskal-Wallis H test. Comparison of ordinal data was analyzed using the Mann-Whitney U test. Count data were represented as absolute numbers or percentages, and comparison among groups was done using the chi-square test. Results:(1) The propensity score matching conditions and comparison of general data among the three groups after propensity score matching: 312 of 404 patients had successful matching, including 78 in the robotic group, 156 in the laparoscopic group, and 78 in the open group. The age, cases in G1, G2, G3 (histopathological classification) and cases with proximal gastrectomy or total gastrectomy (surgical resection range) before matching were (62.2±1.0)years, 0, 37, 67, 13, 91 in the robotic group, (60.9±8.1)years, 0, 98, 107, 31, 174 in the laparoscopic group, and (58.5±9.8)years, 1, 32, 62, 27, 68 in the open group, showing significant differences among the three groups ( F=4.269, 6.356, χ2=10.416, P<0.05). The above indicators after matching were (61.2±10.8)years, 0, 28, 50, 12, 66 in the robotic group, (60.7±8.0)years, 0, 56, 100, 25, 131 in the laparoscopic group, and (60.7±8.4)years, 0, 25, 53, 18, 60 in the open group, showing no significant difference among the three groups ( F=0.074, 0.379, χ2=2.141, P>0.05). (2) Surgical situations: the surgical time, volume of intraoperative blood loss, length of surgical incision, length of proximal margin after matching were 300.0 minutes(range, 188.0-420.0 minutes), 137.5 mL(range, 50.0-400.0 mL), 6.0 cm(range, 3.0-12.0 cm), 2.5 cm(range, 1.5-5.5 cm) in the robotic group, 276.0 minutes(range, 180.0-400.0 minutes), 150.0 mL(range, 40.0-800.0 mL), 6.0 cm(range, 3.0-12.0 cm), 3.0 cm(range, 1.0-5.0 cm) in the laparoscopic group, and 244.5 minutes(range, 125.0-461.0 minutes), 200.0 mL(range, 55.0-800.0 mL), 20.0 cm(range, 18.0-25.0 cm), 2.0 cm(range, 1.0-5.5 cm) in the open group, showing significant differences among the three groups ( χ2=27.619, 30.069, 179.367, 11.560, P<0.05). (3) Intraoperative lymph node dissection: the number of lymph node dissected, the number of lymph node dissected in the first station, the number of diaphragmatic and periesophageal lymph node dissected were 30.5(range, 10.0-70.0), 18.0(range, 6.0-42.0), 4.0(range, 0-13.0) in the robotic group, 29.0(range, 12.0-79.0), 19.0(range, 6.0-47.0), 5.0(range, 0-15.0) in the laparoscopic group, and 29.0(range, 18.0-58.0), 18.0(range, 12.0-38.0), 5.0(range, 0-8.0) in the open group, showing no significant difference among the three groups ( χ2=3.676, 1.014, 0.827, P>0.05). The number of lymph node dissected in the second station, the number of lymph node dissected in the superior pancreatic region, the number of No.110 lymph node dissected, the number of No.111 lymph node dissected after matching were 9.0(range, 2.0-30.0), 9.0(range, 2.0-30.0), 1.0(range, 0-4.0), 0(range, 0-3.0) in the robotic group, 6.5(range, 0-25.0), 7.0(range, 0-25.0), 0(range, 0-3.0), 0(range, 0-4.0) in the laparoscopic group, and 6.5(range, 0-19.0), 6.5(range, 0-19.0), 0(range, 0-1.0), 0(range, 0-1.0) in the open group, showing significant differences among the three groups ( χ2=19.027, 24.368, 19.236, 11.147, P<0.05). (4) Postoperative situations: the time to first flatus, time to initial out-of-bed activities, duration of postoperative hospital stay, treatment expenses after matching were 3 days(range, 2-5 days), 2 days(range, 1-4 days), 9 days(range, 5-20 days), 10.6×10 4 yuan [range, (5.4-18.0)×10 4 yuan] in the robotic group, 3 days(range, 2-8 days), 2 days(range, 1-7 days), 9 days(range, 6-56 days), 8.6×10 4 yuan[range, (5.7-40.8)×10 4 yuan] in the laparoscopic group, and 4 days(range, 2-10 days), 4 days(range, 2-10 days), 11 days(range, 8-41 days), 8.4×10 4 yuan[range, (5.8-15.2)×10 4 yuan] in the open group, showing significant differences among the three groups ( χ2=28.487, 95.069, 39.443, 83.899, P<0.05). (5) Postoperative complications: the incidence of overall complications, incidence of severe complications (Clavien-Dindo classification ≥grade 3), incidence of gastrointestinal complications, incidence of incisional complications, incidence of respiratory complications, incidence of infection were 21.8%(17/78), 5.1%(4/78), 10.3%(8/78), 1.3%(1/78), 7.7%(6/78), 2.6%(2/78) in the robotic group, 21.8%(34/156), 7.1%(11/156), 5.1%(8/156), 1.3%(2/156), 11.5%(18/156), 3.8%(6/156) in the laparoscopic group, and 29.5%(23/78), 6.4%(5/78), 9.0%(7/78), 2.6%(2/78), 14.1%(11/78), 2.6%(2/78) in the open group, showing no significant difference among the three groups ( χ2=1.913, 0.321, 2.394, 0.866, 1.641, 0.335, P>0.05). (6) Follow-up: 312 patients after propensity score matching were follow up at postoperative 1 month. During the follow-up, 2 cases with severe complications died after discharge. No severe complication such as obstruction of input or output loop, dumping syndrome was found in the other 310 patients. Conclusions:The Da Vinci robotic radical gastrectomy is safe and feasible for locally advanced Siewert type Ⅱ and Ⅲ AEG. Compared with laparoscopic and open radical gastrectomy, Da Vinci robotic radical gastrectomy has more advantages in the number of lymph node dissected in the second station (especially in the superior pancreatic region).