Chapter 1 – Introduction Gastric cancer is conventionally treated by means of open distal or total gastrectomy. The open surgical approach is associated with high morbity and long postoperative hospital stay. Minimally invasive surgery is upcoming for gastric cancer, since promosing results in the treatment of other abdominal diseases were booked. This thesis consists of 2 parts. In the first part the value of laparoscopic gastrectomy is described. In the second part the surgical treatment of tumors located at the gastro-esophageal junction is evaluated. These tumors are located at the transistional area between the esophagus and stomach and are therefore difficult to treat. PART 1 – GASTRIC CANCER SURGERY Chapter 2 In this systematic meta-analysis of comparative cohort studies the short-term outcomes of laparoscopic total gastrectomy versus open total gastrectomy in the treatment of gastric cancer were evaluated. Laparoscopic total gastrectomy was associated with reduced blood loss, lower short-term complication rates, and quicker functional recovery, at the price of a longer operating time. The included studies did not elaborate on oncologic outcome, long-term survival, or quality of life. Chapter 3 This international cross-sectional survey revealed the current preferences in gastric cancer surgery. Members of the International Gastric Cancer Association filled in a web-based questionnaire about surgical techniques and approaches. The majority (79%) of respondents performed >21 gastrectomies per year. Open gastrectomy was preferred over laparoscopic gastrectomy (distal: 35% for early and 91% for advanced cancer; total: 52% for early and 94% for advanced cancer). Resection of the greater omentum was favored by the majority (89%) of respondents. A Roux-en-Y reconstruction with a jejunal pouch was preferably performed after total gastrectomy by a minority of respondents (17%). Thus far, as a reflection of the current practice, the minimally invasive techniques have not been generally implemented in clinical guidelines. Chapter 4 In the UMC Utrecht cohort of patients with an identified germline ecadherin-1 (CDH1) mutation, prophylactic laparoscopic total gastrectomy was performed to eliminate the high risk of developing diffuse gastric cancer. The median operative time was 4:19 (3:15-6:03) hours and the median blood loss was 200 (20-400) ml. Median length of hospital stay was 10 (7-27) days. The 60-day mortality rate was 0%. Multiple foci of intramucosal diffuse gastric signet ring cell carcinoma were found in the resection specimen of 9/11 (82%) patients. All resections were microscopically radical. Enhanced postoperative recovery and reduced surgical trauma may especially be relevant to patients with a germline CDH1 mutation, since they have a higher life expectancy than patients with actual gastric cancer. Chapter 5 Laparoscopic gastrectomy was performed in our cohort of Western European patients with predominantly locally advanced gastric cancer. The median intraoperative blood loss was 305 (30-2700) milliliters. The median postoperative hospital stay was 11 (5-91) days. The 30-day mortality was 4.3%. A radical resection was achieved in the vast majority of patients (90%). The median number of dissected lymph nodes was 17 (2-62). We demonstrated that laparoscopic gastrectomy can safely be performed in Western European patients with advanced gastric cancer and meets the oncologic standard. Chapter 6 Omentectomy additional to gastrectomy is considered to be the standard surgical technique in the curative treatment of patients with resectable gastric cancer. In our prospective series of patients undergoing gastrectomy the greater omentum was marked during operation and pathologically analyzed to evaluate the presence of omental lymph nodes, tumor deposits and patterns of lymphatic spread. The omental lymph nodes contained metastases in 1 (2%) patient with stage IB gastric cancer. Omental tumor deposits were found in 4 (8%) patients stage IB-IIIA, of which 3 underwent perioperative chemotherapy. No significant differences in 1-year overall survival (p=0.106) and 1-year disease-free survival (p=0.258) for patients with and without omental lymph node metastases or tumor deposits were found. No predictive factors for omental tumor involvement could be identified. Therefore, omentectomy should be the standard in gastrectomy for all gastric cancer patients. Chapter 7 In order to evaluate the laparoscopic gastrectomy compared to open gastrectomy in a large national setting a non-blinded, multicenter, prospectively randomized controlled superiority trial was designed. Patients (n=210) with resectable gastric cancer will be enrolled. The primary outcome is postoperative hospital stay (days). Secondary outcome parameters include postoperative morbidity and mortality, oncologic outcomes, readmissions, quality of life and cost-effectiveness. It is hypothesized that the functional recovery is faster in the laparoscopic group. Also, it is expected that laparoscopic gastrectomy will be associated with a lower postoperative morbidity, less readmissions, higher cost-effectiveness, better postoperative quality of life, with similar mortality and oncologic outcomes, compared to open gastrectomy. Chapter 8 A major complication of gastrectomy is leakage of the esophagojejunostomy. Sealing the anastomosis with a fibrin patch (TachoSil) containing a human fibrinogen and thrombin, may improve mechanical strength of the anastomosis. A feasibility study of 15 patients with an esophageal anastomosis was performed. The sealant patch could be applied successfully in all patients. A median of 2 (1-6) attempts were necessary to reach successful application. The median duration was 7 (3-26) minutes before successful application was accomplished. In patients that underwent total gastrectomy, the patch was folded into a harmonica shape and wrapped around the esophagojejunostomy. PART 2 – GASTRO-ESOPHAGEAL JUNCTION CANCER SURGERY Chapter 9 Adenocarcinomas of the Gastro-Esophageal Junction (GEJ) are located at the transition zone between the esophagus and the stomach. The optimal surgical treatment of patients with this type of cancer has not been established yet. In a systematic review cohort studies comparing gastrectomy versus esophagectomy were evaluated. Radical resection rates varied between 72–93% for esophagectomy and 62%–93% for gastrectomy. Morbidity was 33–39% after esophagectomy versus 11–54% after gastrectomy. The 30-day mortality ranged between 1.0–2.3 after esophagectomy and 1.8–2.7% after gastrectomy. At 6 months after surgery, health-related quality of life was higher after total gastrectomy than after esophagectomy. The 5-year survival rates varied between 30–42% for esophagectomy and 18–38% for gastrectomy, but were not significantly different. In the selection of the most optimal surgical strategy, it is important to choose a surgical procedure that allows for radical resection of the GEJ tumor with dissection of adjacent lymph nodes. If both surgical strategies are deemed possible, a gastrectomy appears to offer the best long-term quality of life perspective. Chapter 10 To evaluate the international preferences in the surgical treatment of GEJ tumors a survey was performed amongst surgical members of the International Society for Diseases of the Esophagus, the World Organization for Specialized Studies on Disease of the Esophagus and the International Gastric Cancer Association. The preferred surgical approach for Siewert type 1 tumors (5-1 cm proximal of the GEJ) was esophagectomy in 93% of respondents, whereas 6% favored gastrectomy and 3% combined a distal esophagectomy with a proximal gastrectomy. For Siewert type 2 tumors (1-2 cm from the GEJ) an extended gastrectomy was favored by 66% of respondents, followed by esophagectomy in 27% and total gastrectomy in 7%. Siewert type 3 tumors (2-5 cm distal of the GEJ) were preferably treated with gastrectomy in 90% of respondents, esophagectomy in 6% and extended gastrectomy in 4%. For choosing optimal treatment for GEJ tumors, adequate determination of tumor location is pivotal because the location affects the surgical procedure. Esophagogastroscopy was deemed most important by 81% of respondents, followed by CTscan in 14%, EUS in 2%, PETscan in 1%, and diagnostic laparoscopy in 1%. Also, the classification that is used affects surgical treatment. A combination of the Siewert classification and the TNM7 classification was preferred by 45% of respondents, whereas 39% used the Siewert classification only and 16% solely favored the TNM7 classification. Chapter 11 In the prospective UMC Utrecht database of 266 consecutive patients with surgically resectable GEJ adenocarcinomas the diagnostic staging was analyzed. The accuracy of EUS and CT regarding tumor localization according to Siewert, nodal status and its consequences on treatment strategy were assessed. Overall accuracy in determining tumor localization was 73% for EUS and 61% for CT (p=0.018). For EUS, the accuracy decreased when patients were treated with neoadjuvant therapy (82% to 70%, p=0.023), with CT no difference was found (62% to 61%, p=0.884). Accuracy for determining a positive nodal station in patients without neoadjuvant therapy was 77% for EUS and 71% for CT (p=0.001). Accuracy for detecting upper mediastinal lymph nodes was 80–92%, compared to 50–80% for peritumoral and abdominal nodes in both EUS and CT. A radical resection was performed in 88% of patients. Despite the suboptimal accuracy of determining tumor localization with EUS and CT, in only a small number of patients (3%) an intraoperative change of surgical treatment was needed. EUS is superior to CT in determining nodal status and tumor localization in GEJ tumors. Chapter 12 In the prospective UMC Utrecht database of 266 consecutive patients with surgically resectable GEJ adenocarcinomas the surgical treatment was analyzed. Post-operative histopathological analysis revealed that 25% of patients had a type I tumor, 66% a type II tumor, 6% a type III tumor. I n total, 86% were treated with esophagectomy and 14% with gastrectomy. In patients with a type II GEJ adenocarcinoma, a positive circumferential resection margin was more common with gastrectomy than esophagectomy (29 vs. 11 %; p = 0.025). However, the type of operation did not significantly influence overall survival on multivariate analysis (p = 0.606). Upper mediastinal nodal involvement was present in 11% of patients with a type II tumor.