A multinational phase II clinical trial of neoadjuvant imatinib for large gastrointestinal stromal tumor of the stomach

130

Background: Neoadjuvant therapy is expected to reduce the risk of primary surgery, such as rupture of the tumor, hemorrhage, and multi-visceral resection, and to improve survivals for patients with a large gastric gastrointestinal stromal tumor (GIST). This study aims to evaluate the efficacy and safety of neoadjuvant imatinib therapy for a large gastric GIST. Methods: Patients with gastric GIST, which is 10cm or larger and without metastasis, received neoadjuvant imatinib (400mg/day) for 6 months, and up to 9 months if maximal response is expected. Postoperative adjuvant imatinib was prescribed for at least 1 year and up to 3 years according to adjuvant treatment guideline. The primary endpoint was complete (R0) resection rate. A primary analysis were performed by the time all the operations were finished, to examine the efficacy and safety of the neoadjuvant treatment. Results: Between Feb 2010 and Sep 2014, 55 patients were enrolled in Japan and Korea. One patient with a jejunal GIST and one patient with PDGFRA-18 D842V mutation were excluded from analysis. Mean tumor diameter was 12cm (10-23). 86.8% of patients (46/53) completed neoadjuvant treatment. Dose reduction of imatinib was performed in 26.4% (14/53). The most frequent Grade 3 or 4 adverse events were G3 rash (5/53, 9.4%) and G3/4 neutropenia (4/53, 7.5%). Disease control rate (PR+SD) and response rate (PR) of neoadjuvant imatinib was 100% and 62.3% by RECIST, and 100% and 98.1% by Choi criteria, respectively. There was no case of CR or PD. 50 patients underwent operation, and R0 resection rate was 90.6% (n = 48, 95% CI 79.3% - 96.9%), which was significantly higher than the threshold value of 70% (p < 0.001). Combined resection of other organs (except gall bladder) was performed in 24.5% (n = 13), and 83.0% of patients (n = 44) could preserve ≥ 50% of the stomach. Postoperative complication occurred in 18.0% (9/50). Conclusions: Neoadjuvant imatinib treatment is effective and safe treatment option for a large primary GIST allowing high R0 resection rate with acceptable incidence of adverse events and postoperative complications. Clinical trial information: UMIN000003114.

Solitary Fibrous Tumor of the Stomach Treated with Laparoscopic and Endoscopic Cooperative Surgery

A 43-year-old Japanese woman with melena underwent an upper gastrointestinal endoscopy and was preoperatively diagnosed with sarcoma of the stomach. Physical examination revealed no abnormalities. Findings on the upper gastrointestinal endoscopy showed a pedunculated submucosal tumor measuring 17 mm in the antrum. An enhanced computed tomography showed wall thickening in the gastric antrum. The patient underwent a laparoscopic and endoscopic cooperative surgery (LECS) for wedge resection of the stomach. The excised tumor measured 27 × 20 × 15 mm in size. Histopathology showed spindle-shaped cells in the submucosal layer. Immunohistochemistry showed that the tumor was positive for CD34, bcl-2, and MIC-2. The final diagnosis was solitary fibrous tumor (SFT) of the stomach. The postoperative course was uneventful, and no evidence of recurrence was observed at the 8-month follow-up. We report a case of SFT arising from the stomach that was treated with wedge resection by LECS.

Combination of Laparoscopic and Endoscopic Approaches to Neoplasia with Non-exposure Technique (CLEAN-NET) for GIST with Ulceration

Combination of laparoscopic and endoscopic approaches to neoplasia with non-exposure technique (CLEAN-NET) was developed to avoid intraoperative tumor dissemination. We report two cases of gastric gastrointestinal stromal tumor (GIST) with ulceration surgically treated with CLEAN-NET at our institution. The first case was a 55-year-old male with hematemesis. Gastric endoscopy revealed a gastric GIST with ulceration of the fornix. CLEAN-NET was performed with the insertion of five trocars and a liver retractor. The operative time was 202 min (including cholecystectomy), with a perioperative blood loss volume of 29 ml; the postoperative hospital stay duration was 8 days. The second case was a 66-year-old male with a gastric submucosal tumor (SMT) with ulceration. CLEAN-NET was performed in a similar fashion to the first case. The operative time was 128 min, with a preoperative blood loss volume of 16 ml; the postoperative hospital stay duration was 9 days. In conclusion, CLEAN-NET was found to be safe and useful in the treatment of gastric GIST with ulceration.

Comparison between early enteral feeding with a transnasal tube and parenteral nutrition after total gastrectomy for gastric cancer

BACKGROUND/AIMS

This retrospective study evaluated 21 patients with early enteral feeding (EEF group) and 22 patients without early enteral feeding (non-EEF group) who underwent open total gastrectomy followed by Roux en Y reconstruction and were RO resectable cases. METHDOLOGY: Postoperative complications and course, postoperative/preoperative body weight, whole meal intake, and nutritional, inflammatory, and immunological parameters were recorded and evaluated in both groups.

RESULTS

Postoperative meal intake was significantly higher and the first day of defecation was significantly earlier in the EEF group than in the non-EEF group. There were no significant differences between the 2 groups in the blood laboratory data and the rate of complications. In patients with complications, lymphocyte counts and postoperative body weights were compared as indicators of immunostimulation. The lymphocyte counts 7 days after operation and postoperative/preoperative body weight were significantly higher in the EEF group than in the non-EEF group.

CONCLUSIONS

Although immunostimulation-like findings were observed in the patients with complications after surgery in the present study, the significance of EEF was not clarified because of the lack of cases whose conditions were severe. EEF should be used especially for patients in whom severe disease is possible and avoidance of TPN is desirable.

Clinicopathologic features and clinical outcomes of gastric cancer patients with bone metastasis

OBJECTIVE

We conducted a retrospective analysis to evaluate the clinical manifestations and outcomes of a population of gastric cancer patients with bone metastasis.

METHODS

The subjects were 31 gastric cancer patients who were diagnosed with bone metastasis between January 2000 and December 2010.

RESULTS

The overall median survival time (MST) was 100 days. The results of a multivariate analysis in relation to overall survival showed that the absence of extraosseous metastasis and having received chemotherapy were favorable prognostic factors. MST was 269 days in the bone metastasis alone group (n = 6) and 65 days in the extraosseous metastasis group (n = 25). We divided the extraosseous metastasis group into two subgroups according to whether the patient had received chemotherapy. Evaluation of the response in the chemotherapy group showed that the subgroup of patients with progressive disease had a significantly longer MST than the no-chemotherapy group (63 days vs. 21 days, p = 0.012).

CONCLUSIONS

We concluded that it is useful to divide gastric cancer patients with bone metastasis into two groups according to whether they have extraosseous metastasis. Aggressive chemotherapy should be considered as a means of improving the prognosis of gastric cancer patients with extraosseous metastasis.

Functional outcomes by reconstruction technique following laparoscopic proximal gastrectomy for gastric cancer: double tract versus jejunal interposition

Background

For early gastric cancer located in the upper third of the stomach, we have adopted laparoscopic 1/2-proximal gastrectomy (PG) with two types of reconstruction: double tract reconstruction (L-DT) and jejunal interposition reconstruction with crimping of the jejunum on the anal side of the jejunogastrostomy with a knifeless linear stapler (L-JIP).

Methods

Functional outcomes were prospectively compared between these two types of reconstruction following laparoscopic PG. Resection and reconstruction were performed using L-DT (n = 10) and L-JIP (n = 10) alternately. Quality of life was evaluated through a questionnaire and endoscopic examination of the ten patients in each group, and functional evaluations were carried out in five patients of each group.

Results

The postoperative/preoperative body weight ratio was significantly higher in the L-JIP group than in the L-DT group. While the incidence of reflux esophagitis was 10% in both groups, the endoscope could reach the remnant stomach in all patients. In the L-DT group, the plasma acetaminophen concentration at 15 minutes and the insulin level at 30 minutes were markedly increased after oral administration, while the increases in the blood sugar level at 30 and 60 minutes were more gradual than in the L-JIP group.

Conclusions

While L-JIP may be thought of as the ideal method for function-preserving gastrectomy, L-DT may be suitable for gastric cancer patients with impaired glucose tolerance. These results raise the possibility of individualized selection of reconstruction for gastric cancer patients with various kinds of preoperative complications.

Sentinel node mapping for gastric cancer: a prospective multicenter trial in Japan

PURPOSE

Complicated gastric lymphatic drainage potentially undermines the utility of sentinel node (SN) biopsy in patients with gastric cancer. Encouraged by several favorable single-institution reports, we conducted a multicenter, single-arm, phase II study of SN mapping that used a standardized dual tracer endoscopic injection technique.

PATIENTS AND METHODS

Patients with previously untreated cT1 or cT2 gastric adenocarcinomas < 4 cm in gross diameter were eligible for inclusion in this study. SN mapping was performed by using a standardized dual tracer endoscopic injection technique. Following biopsy of the identified SNs, mandatory comprehensive D2 or modified D2 gastrectomy was performed according to current Japanese Gastric Cancer Association guidelines.

RESULTS

Among 433 patients who gave preoperative consent, 397 were deemed eligible on the basis of surgical findings. SN biopsy was performed in all patients, and the SN detection rate was 97.5% (387 of 397). Of 57 patients with lymph node metastasis by conventional hematoxylin and eosin staining, 93% (53 of 57) had positive SNs, and the accuracy of nodal evaluation for metastasis was 99% (383 of 387). Only four false-negative SN biopsies were observed, and pathologic analysis revealed that three of those biopsies were pT2 or tumors > 4 cm. We observed no serious adverse effects related to endoscopic tracer injection or the SN mapping procedure.

CONCLUSION

The endoscopic dual tracer method for SN biopsy was confirmed as safe and effective when applied to the superficial, relatively small gastric adenocarcinomas included in this study.

Modulation of fibronectin synthesis by cancer cell-fibroblast interaction

Conditioned medium (CM) of human rectal adenocarcinoma cell line RCM-1 stimulated both cellular (c-) and plasma (p-) fibronectin (FN) production by human fibroblasts and modulated the alternative splicing of its primary transcript at the EDA region to express more EDA-containing (+) mRNA. This EDA(+) mRNA-stimulating effect of CM was inhibited by treatment with an anti-human transforming growth factor (TGF)-beta antibody. TGF-beta production by RCM-1 cells was demonstrated by immunoblotting and RT-PCR. Thus, FN synthesis and splicing-in at the EDA region in fibroblasts were stimulated by cancer cells predominantly via TGF-beta. Since RCM-1 cells adhered to cFN, which contains EDA, more efficiently than pFN and adhesion to extracellular matrix proteins such as FN is the first step to migration, the cancer stroma modulated by cancer cell-fibroblast interaction may facilitate cancer invasion.

Fundic rotation technique: a useful procedure for laparoscopic exogastric resection of gastric submucosal tumors located on the posterior wall near the esophagogastric junction

OBJECTIVE

To develope a new procedure for laparoscopic exogastric resection using a so-called "fundic rotation technique (FRT)" for gastric submucosal tumors (SMTs) on the posterior wall near the esophagogastric junction (EGJ).

METHODS

Between April 2006 and February 2010, we performed laparoscopic resection for SMTs located near the EGJ (within 3.0 cm from the EGJ) in ten consecutive patients. Out of seven exogastric resections, an FRT was used in five patients with posterior tumors near the EGJ.

RESULTS

The patients comprised three men and two women, with an average age of 65 years. The maximum tumor diameter averaged 3.8 cm (range, 2.0-8.0 cm), and the average distance from the EGJ was 1.5 cm (range, 0-2.5 cm). The pathological diagnosis was GIST in all cases. One case was converted to an open surgery due to its large size (8.0 cm) and the difficult access. All the patients quickly returned to their normal activities. No patient complained any symptoms of regurgitation, and endoscopic examination revealed no remarkable reflux esophagitis. No tumor recurrences occurred during a median follow-up period of 30 months.

CONCLUSION

The indications for laparoscopic resection of SMTs located near the EGJ may be extended using an FRT.

Relationship between Helicobacter pylori and histological changes in the gastric remnant after subtotal gastrectomy

OBJECTIVE

The aim of this study was to clarify the influence of histological changes in the gastric remnant on Helicobacter pylori (H. pylori) infection after distal gastrectomy (DG) and proximal gastrectomy (PG).

METHODS

In total, 101 patients who underwent DG (n = 76) or PG (n = 25) for gastric cancer were included in the study. Three biopsy specimens from the remnant stomach were obtained during upper gastrointestinal endoscopy. Each specimen was scored according to the updated Sydney system for classifying gastritis and was examined for H. pylori infection.

RESULTS

The H. pylori infection rate after DG was 60.5% while that after PG was 20.0% (P < 0.001). The histological score for neutrophils after DG was 60.5% while that after PG was 12.9% (P < 0.001). Intestinal metaplasia after PG was 76.0% while that after DG was 22.4% (P < 0.001). No differences in mononuclear cells or atrophy were observed between the two gastrectomy groups.

CONCLUSIONS

H. pylori infection occurred more frequently after DG than after PG. Histological inflammation of the gastric remnant after DG was higher than that after PG. Intestinal metaplasia of the gastric remnant after PG was higher than that after DG. The intestinal metaplasia that induced inflammation indicated that H. pylori infection after PG was at a low level.

Prognostic significance of morphological distribution of metastatic foci in lymph nodes with gastric cancer

BACKGROUND/AIMS

The morphological distribution of tumor cells in metastatic lymph nodes has been investigated in positive sentinel lymph nodes in several solid cancers. The aim of this study was to clarify the effect of the distribution of metastatic foci in lymph nodes on the prognosis in gastric cancer.

METHODS

The distribution of metastatic foci in the 100 node-positive patients who had undergone curative gastrectomy were classified into two groups: (1) massive type, in which the tumor occupied the entire lymph node, and (2) non-massive type, in which the tumor did not occupy the entire lymph node.

RESULTS

There were 38 patients in the massive type group and 62 patients in the non-massive type group. The 10-year survival rate was significantly poorer in the massive type group (p = 0.001). Multivariate analysis showed that distributional type and nodal status were independent prognostic factors. UICC N stage was subcategorized by distributional type, and survival was shown to be significantly worse in the massive type in the N1 group (p = 0.035).

CONCLUSION

It seems necessary to take the morphological distribution of metastatic foci into consideration when dealing with node-positive patients who had received curative resection for gastric cancer.

Endoscopy-assisted anastomosis: a modified technique for laparoscopic side-to-side esophagojejunostomy following a total gastrectomy

INTRODUCTION

Esophagojejunostomy with a circular stapling device is sometimes difficult to perform in a laparoscopic setting. On the other hand, a side-to-side anastomosis with a linear stapling device is technically challenging.

METHODS

Between June 2002 and March 2008, 10 consecutive patients underwent a laparoscopy-assisted total gastrectomy using a side-to-side anastomosis technique. Of these patients, four underwent a laparoscopy-assisted total gastrectomy with a modified anastomosis technique. A small wound was created on the antimesenteric side of the jejunum 5 cm distal to the resected portion and then in the lower esophagus. A peroral endoscope was advanced to the hole, and the cartridge fork was introduced into the lower esophagus under endoscopic guidance. The device (45 mm, blue) was fired to create an antiperistaltic side-to-side anastomosis. The common entry hole was closed by transecting the jejunum and the esophagus with another linear stapler and by using an endoscope as a stent.

RESULTS

Four patients underwent the modified procedure and did not require an open procedure. One patient developed a pancreatic fistula, which was treated conservatively. The average operative time, reconstruction time and blood loss were 483 ± 133 minutes, 139 ± 31 minutes, and 199 ± 121 mL, respectively. An introduction of the stapler into the lower esophagus and a closure of the common entry hole were performed safely without any stress.

CONCLUSION

Although several techniques must be compared to determine the ideal procedure for laparoscopic esophagojejunostomy, the modified side-to-side anastomosis technique may be useful in clinical settings.

A phase II study of doxifluridine and docetaxel combination chemotherapy for advanced or recurrent gastric cancer

BACKGROUND

The aim of this study was to establish the efficacy and safety of doxifluridine and docetaxel for patients with advanced or recurrent gastric cancer.

METHODS

The regimen consisted of oral administration of doxifluridine 533 mg/m(2) per day on days 1-14 and an intravenous infusion of docetaxel 50 mg/m(2) on day 8. The primary endpoint was the overall response rate. The secondary endpoints were overall survival, progression-free survival, and toxicities.

RESULTS

Between June 2004 and December 2006, a total of 40 eligible patients were enrolled in this study. Seven of them showed a partial response, with an overall response rate of 17.5%. The response rate was 18.8% in 32 patients with refractory tumors. The median progression-free survival time and the median overall survival time were 2.6 months and 12.7 months, respectively, in all 40 patients; and 2.6 months and 14.0 months, respectively, in the 32 patients with refractory tumors. Grade 3/4 hematological toxicity included neutropenia in 52.5%, leukocytopenia in 17.5%, and febrile neutropenia in 7.5%. Grade 3 or more nonhematological toxicities were infrequent.

CONCLUSION

The combination chemotherapy of doxifluridine and docetaxel was well tolerated and relatively effective when used as a second-line chemotherapy for advanced or recurrent gastric cancer.

New method of laparoscopic incisional hernia repair with double circumferential transfascial sutures

BACKGROUND

Although, laparoscopic incisional hernia repair (LIHR) provides an alternative method for managing incisional hernias, the ideal procedure for reducing the incidence of postoperative complications remains unclear.

PATIENTS AND METHODS

We have developed a new method of LIHR that involves a double transfascial suture and does not require the use of spiral tackers. We performed this procedure consecutively in five patients (four males and one female with a mean age of 65.6 years). We describe our new method of LIHR, and present preliminary clinical results.

RESULTS

The mean defect size was 26.2 ± 15.8 cm(2), and the mesh size that was used was 121.7 cm(2) in all cases. An occult hernia was found in one patient during laparoscopic observation. The mean operative time was 198.4 ± 49.3 minutes with a blood loss of 12.2 ± 24.6 mL. Postoperative courses were uneventful with a median postoperative hospitalization period of 8 days. No patient required mesh removal and none developed a recurrent hernia during the median follow-up period of 13 months.

CONCLUSION

Although, larger number of patients and longer follow-up will be required to prove the operative adequacy of our new procedure, it appears to represent a feasible option for LIHR.

A case of colon lymphangioma treated with laparoscopy-assisted ileocecal resection

We encountered a patient with colon lymphangioma and performed laparoscopy-assisted ileocecal resection. The patient was a 68-year-old male who visited our hospital for weight loss in May 1999. Since fecal occult blood was positive on close examination, colonoscopy was performed. A light-permeable, transparent, and pedunculated polyp was found in the ascending colon and diagnosed as submucosal tumor. The patient was admitted for close examination and treatment. Abdominal CT detected a tumor in the ascending colon accompanied by disturbed surrounding adipose tissue, suggesting extramural invasion of the tumor. Based on the diagnosis of extramural submucosal tumor of the ascending colon, ileocecal resection was performed with laparoscopic assistance. The laparoscopic findings were as follows: 1) Lymph vascular dilatation filled with lymph expanding on the serosal surface, 2) light-permeability/translucency, 3) laparoscopic cushion sign, and 4) lymph vascular dilatation in the surrounding adipose tissue. Laparoscopy-assisted surgery was useful for the diagnosis and treatment of colon lymphangioma, and characteristic laparoscopic findings were noted. Laparoscopic surgery is a useful for a diagnosis and the treatment of large lymphangioma of colon.

A case of primary jejunal cancer diagnosed by preoperative small intestinal endoscopy

The patient was a 37-year-old female. She was brought to our hospital by ambulance with nausea and vomiting. Abdominal ultra sound and abdominal enhanced CT scan showed a tumor in left side of the abdominal aorta 6 cm in size, and it showed an expanded stomach and duodenum. Upper gastrointestinal series revealed an apple core sign in upper jejunum near the Treitz' ligament. Small intestinal endoscopy (XSIF-240 endoscope, Olympus Inc.) revealed stenosis related to an epithelially protruding lesion with an irregular surface in the jejunum on the anal side of the horizontal duodenal peduncle. Biopsy suggested a well-differentiated adenocarcinoma. Scintigraphy showed hot spot in left middle abdomen. Under a diagnosis of primary jejunum cancer, Partial resection of the jejunum and partial resection of the transverse colon was performed. Histopathologically, the tumor was well differentiated adenocarcinoma exposed serosal surface. Postoperatively, the stage was evaluated as III (T3, N1, M0). Preoperative diagnosis to use small intestinal endoscopy was effectiveness. We report a patient with primary jejunum cancer in whom a definitive diagnosis was made before surgery.

Clinical outcome of proximal gastrectomy in patients with early gastric cancer in the upper third of the stomach

OBJECTIVE

To determine whether clinical outcomes after proximal gastrectomy are better than those after total gastrectomy with Roux-en Y reconstruction.

METHODS

We studied 10 consecutive patients with early gastric cancer who underwent esophagogastrostomy after proximal gastrectomy (PG group). Nutritional variables in these patients were compared with those in 10 consecutive patients who underwent Roux-en Y reconstruction after total gastrectomy (TG group). Patients were followed up for 5 years after operation.

RESULTS

There was no anastomotic leakage. The total cholesterol level 1 year after operation was higher in the PG group than in the TG group (p< 0.05). Body mass index was significantly lower than the preoperative value between 1 month and 2 years postoperation in the PG group, whereas the TG group showed decreases between 3 months to 5 years postoperation. The percent decreases in body weight at 3 and 4 years in the PG group were lower than those in the TG group (both p< 0.05). Postoperative weight loss was thus milder in the PG group than in the TG group.

CONCLUSION

Esophagogastrostomy after PG may produce better clinical outcomes than Roux-en Y reconstruction after TG in patients with early gastric cancer arising in the upper third of the stomach.

Esophagogastrostomy before proximal gastrectomy in patients with early gastric cancers in the upper third of the stomach

OBJECTIVE

This study was designed to assess the outcome of esophagogastrostomy before proximal gastrectomy in patients with early gastric cancers in the upper third of the stomach.

METHODS

From 1997 through 2004, we studied 10 consecutive patients. A stapler was introduced into the stomach, and an esophagogastrostomy was performed before proximal gastrectomy. Hill's posterior gastropexy and Dor's anterior fundic wrap were performed to prevent reflux esophagitis.

RESULTS

The operation time was 171 ± 44 minutes, and the intraoperative bleeding volume was 294 ± 228 mL. There was no anastomotic leakage. Anastomotic stenosis, occurring in 40% of the patients, required endoscopic balloon dilatation. Symptoms of reflux esophagitis, occurring in 40% of the patients, resolved within 2 years after operation. As compared with the preoperative value, body mass index was significantly decreased 1 and 2 years after operation, but was similar at 3 to 5 years. The percent decrease in body weight after operation fluctuated between 6% and 8% between 2 and 5 years. Postoperative weight loss was thus mild.

CONCLUSIONS

Esophagogastrostomy before proximal gastrectomy may be less invasive, simpler, and produce better outcomes than conventional procedures for the surgical treatment of early gastric cancer in the upper third of the stomach.

cDNA microarray analysis of cancer associated gene expression profiles in malignant peripheral nerve sheath tumours

BACKGROUND

Malignant peripheral nerve sheath tumour (MPNST) is a highly aggressive malignancy that arises within peripheral nerves, and is associated with poor prognosis. Little is known about the underlying biology of MPNST, especially the mechanisms involved in cell proliferation, invasion, or escape from apoptosis.

AIMS

To identify genes differentially expressed in MPNST compared with benign tumours, such as neurofibromas and schwannomas, by means of cDNA microarray analysis.

METHODS

Six MPNST cases and five benign cases (three schwannomas and two neurofibromas) were analysed.

RESULTS

Six genes (keratin 18, survivin, tenascin C, adenosine deaminase, collagen type VIa3, and collagen type VIIa1) were significantly upregulated in MPNST, whereas one gene, insulin-like growth factor binding protein 6, was downregulated in MPNST. Survivin and tenascin C expression was validated by reverse transcription polymerase chain reaction. Immunohistochemistry confirmed upregulation of survivin in MPNST at the protein level in six of eight cases compared with benign tumours. Tenascin C was also expressed at the invasive front and tumorous stroma in all MPNST cases. MPNST cells expressed tenascin C in four of nine cases.

CONCLUSIONS

Survivin and tenascin C may be associated with the malignant potential of MPNST and could be considered as potential therapeutic targets.

Focus on the conditions of resection and reconstruction in gastric cancer. What extent of resection and what kind of reconstruction provide the best outcomes for gastric cancer patients?

To assess the roles of the extent of gastric resection and duodenal food passage reconstruction in gastric cancer, we examined a consecutive series of 1,061 patients who underwent total or partial (proximal and distal) gastrectomies with or without duodenal food passage reconstruction between August of 1974 and January of 2002, and received gastrectomies with D2-3 lymph node dissection. Patients who underwent distal or proximal gastrectomy were found to have significantly better survival rates than those who underwent total gastrectomy in stages 1A (10-year survival: 86.6 and 78.9 vs. 61.6%), 2 (56.5 and 65.6 vs. 34.4%), 3A (45.9 and 33.3 vs. 15.2%), and 4 (5-year survival rates: 23.7 and 50.0 vs. 7.1%). Additionally, patients with duodenal food passage reconstruction or double tract reconstruction also showed significantly better survival rates than those without duodenal food reconstruction in stages 1A (10-year survival: 86.4 and 82.5 vs. 61.7%), 1B (69.9 and 90.6 vs. 54.1%), 2 (60.5 and 63.3 vs. 16.5%), and 3A (39.9 and 47.4 vs. 23.1%). In multivariate analysis, the independent prognostic factors were age at operation, depth of tumor, duodenal food passage reconstruction, and lymph node metastasis. Our results indicate that both the extent of gastric resection and duodenal food passage reconstruction were important factors in the outcome of gastric cancer patients, and that surgeons should perform minimal gastric resection with preservation of the duodenal food passage when the gastric stump is tumor-free.