H. Katai

Background: Although incidence and mortality have decreased, gastric cancer (GC) is still a public health issue globally. The international collaborative study for cancer survival using population-based cancer registry showed that the survival of GC was higher in Korea and Japan than other countries, including the United States of America (US). Aim: We examined the determinant factors of the high survival in Japan, compared with the US, with particular focus on the impact of the number of lymph nodes (LNs) examined for surgically treated patients. Methods: We obtained data on 88,447 cases from the nationwide GC registration project, the Japanese Gastric Cancer Association (JGCA), from 2004-2007. We also obtained 18,995 GC cases from US population-based cancer registry data from the Surveillance, Epidemiology, and End Results Program (SEER), diagnosed from 2004-2010. We estimated five-year relative survival and applied a multivariate excess hazard model to compare the two countries. We considered the effect of LNs examined on differences in survival. Results: Five-year relative survival in Japan was 79.8%, compared with 40.1% in the US. After controlling for confounding factors, we still observed significantly higher survival in Japan. The distribution of number of LNs examined showed notable differences between two countries. In over 50% of patients in the US, only 1 to 15 LNs were examined. A higher number of LNs examined showed better survival in both countries. The differences in excess death from cancer between countries were reduced in the category when more than 30 LNs were examined. Conclusion: Although it is difficult to remove biases to compare the two countries, stage migration, related to the more detailed retrieving strategy for LNs in Japan, is a key explanation for high survival in Japan.

Japan has a huge number of patients with gastric cancer and has developed various surgical treatments for this disease.This paper intends to introduce our strategies against gastric cancer. The Japanese Gastric Cancer Association was established in 1962. Its major purposes are promotion of basic and clinical researches and popularization of the latest knowledge and technologies. For the purposes, the association organized the annual scientific meeting and the nationwide registry by member hospitals, and published the Japanese Classification of Gastric Cancer (1) and the Treatment Guide Line (2). The nationwide registry reported that proportion of Stage-I cancer was 22.5% in 1963-66,which increased to 59.3% in 2008 (3,4,5). 11,261 patients with gastric resection were registered by 187 hospitals in 2008. 63 patients were died within 30 postoperative days and the direct death rate was 0.55%. 5 year survival rate (5YSR) was 37.5% for resected cases in 1963-66, which was improved to 70.1% in 2008. 5YSR was improved from 55.1% to 74.1% for Stage-II, and from 39.1% to 48.8% for Stage-III in the period. According to remarkable increase of early stage cancer, principle of surgical treatments was shifted from "€œextended and standardized surgery for radicality" €to "€œreasonable and individual surgery considering safety and quality of life"€. This trend produced a large variation in surgical treatments; namely 1) minimally invasive surgeries,2) function preserving surgeries, 3) optimal extent of lymph node dissection, and 4) aggressive but safe surgeries.Intention of this paper is to explain these procedures, the intentions, the indications, and the treatment results.

90 Background: GC with ELM (bulky N2 metastasis and / or para-aortic lymph node metastases [PAN]) is commonly regarded unresectable, while in JCOG combined modality treatment has been tested since 2000 (JCOG0001 and JCOG0405). Both trials met their primary endpoints (i.e., 3 year-survival of 27.3% in JCOG0001 and R0 resection of 82.4% in JCOG0405). The survival and the toxicity profile were quite different between the trials despite the similar eligibility with an outstanding 3-year survival of 58.8% in JCOG0405. This study is conducted to explore if survival is still better in JCOG0405 after adjusting baseline factors and if there is a subset of patients (pts) who benefit more from either treatment. Methods: Eligibility criteria for both included histologically proven gastric adenocarcinoma; bulky nodal involvement around major branched arteries to the stomach and/or PAN; cM0 (except PAN); negative lavage cytology; not linitis plastica type; PS of 0 or 1. Pts received two or three cycles of induction chemotherapy of IP: irinotecan (70 mg/m² on day 1 and day 15) and cisplatin (80 mg/m² on day 1) in JCOG0001, or SP: S1 (80 mg/m² from day 1 to 21) and cisplatin (60 mg/m² on day 8) in JCOG0405, followed by D3 gastrectomy. Multivariate analysis for overall survival adjusting baseline factors and treatment (IP/SP) was performed with a Cox regression model. Interaction tests were also carried out between baseline factors and treatment. Results: After adjusting baseline factors, SP was superior than IP for overall survival (HR=0.335: 0.184 – 0.612). There was only interaction effect between treatment and the status of lymph node metastases (bulkyN+/PAN- vs bulkyN-/PAN+ vs bulkyN+/PAN+; p=0.1306). Conclusions: SP was shown to be the favorable treatment for GC with ELM by multivariate analysis, while poor prognosis in pts having both bulky N+ and PAN+ may necessitate further treatment improvement. No significant financial relationships to disclose.

70 Background: GC with ELM, especially in case of M1(LYM), is regarded unresectable in many countries, while challenge is ongoing in Japan to treat them by intensive chemotherapy followed by super extended surgery. In our previous phase II study (JCOG0001) for the same population, iritotecan plus P (IP) chemotherapy followed by surgery showed remarkable survival of 27.0% at 3 years but its toxicity led to rather high treatment related death (TRD). Development of safer and more effective treatment is urged. Methods: Eligibility criteria included histologically proven gastric adenocarcinoma; bulky nodal involvement around major branched arteries to the stomach and/or para-aortic nodal metastases; cM0 (except para-aortic nodes); negative lavage cytology; not linitis plastica type; PS of 0 or 1. Patients (pts) received two or three 28- day cycles of induction chemotherapy of S (80 mg/m² from day 1 to 21) and P (60 mg/m² on day 8), and then underwent D3 gastrectomy. Primary endpoint was R0 resection and key secondary endpoint was 3-year survival. Other secondary endpoints included response rate and adverse events. Results: Between 02/2005 and 06/2007, 53 pts were enrolled and 2 pts were ineligible. Only surrogate endpoints of efficacy and feasibility were reported in ASCO-GI 2008, which included clinical response of 64.7%, R0 of 82.4%, pathological response of 51.0%, and no TRD with low toxicities. The 3-year overall survival, which we first clarified in this report, was 58.8% (95% CI, 44.1-70.9%). Conclusions: Preoperative SP followed by D3 gastrectomy demonstrated excellent 3-year survival with low mortality compared with preoperative IP, which was far better result than expected against this unresectable population. This multimodal treatment is highly promising for GC with ELM. [Table: see text]

Background: Total gastrectomy with D2 dissection is the standard treatment in Japan for early upper-third gastric cancer. The purpose of this study was to confirm the safety and radicality of proximal gastrectomy as an alternative operation. Methods: Between 1993 and 1999, 45 patients with an apparent primary early gastric cancer in the upper third of the stomach were treated by proximal gastrectomy with jejunal interposition. The spleen was preserved, but the suprapancreatic nodes were cleared. Results: Histology confirmed early gastric cancer in 37 patients while eight had advanced disease. Nodal metastasis was observed in six patients. There was no hospital death and no early postoperative complications. One patient died from nodal recurrence and two from unrelated causes. Two patients had reflux symptoms without endoscopic oesophagitis. Mean weight loss was 11·5 per cent of initial bodyweight. Conclusion: Proximal gastrectomy for early upper-third gastric cancer can be performed safely with an excellent cure rate. This procedure deserves further clinical evaluation to assess patients' quality of life.

Treatment results of gastric cancer patients differ considerably between Japan and the West, even at the same stage of disease. This may be partly explained by the varying extent of local control and the "stage-migration" theory. There is an important conceptual difference concerning lymphadenopathy and lymphadenectomy: Japanese surgeons believe that lymph nodes are the governors of metastatic disease and thus that lymphadenectomy will result in cure, while in the West lymph nodes are seen merely as the indicators of systemic metastasis and thus lymphadenectomy serves only for staging. The UICC TNM system has recently abandoned the anatomical N-classification and adopted the numeric N-classification. Although this is a good prognostic indicator, it does not provide surgeons with any information for surgical decision-making. Japanese surgeons will continue to use the Japanese classification that has served as a guideline for standard D2 lymphadenectomy. To establish an international standard for gastric cancer treatment, both sides should make efforts to understand each other and discuss most practical and beneficial treatment modalities for patients in the respective medical environments.