Neoadjuvant chemotherapy with S-1 and cisplatin followed by D2 gastrectomy with para-aortic lymph node dissection for gastric cancer with extensive lymph node metastasis.

BACKGROUND: Locally advanced gastric cancer with extensive regional and/or para-aortic lymph node (PAN) metastases is typically unresectable and associated with poor outcomes. This study investigated the safety and efficacy of S-1 plus cisplatin followed by extended surgery with PAN dissection for gastric cancer with extensive lymph node metastasis. METHODS: Patients with gastric cancer with bulky lymph node metastasis along the coeliac artery and its branches and/or PAN metastasis received two or three 28-day cycles of S-1 plus cisplatin, followed by gastrectomy with D2 plus PAN dissection. The primary endpoint was the percentage of complete resections with clear margins in the primary tumour (R0 resection). A target sample size of 50 with one-sided α of 0.105 and ß of approximately 0.2 corresponded to an expected R0 rate of 65 per cent and a threshold of 50 per cent. RESULTS: Between February 2005 and June 2007, 53 patients were enrolled, of whom 51 were eligible. The R0 resection rate was 82 per cent. Clinical and pathological response rates were 65 and 51 per cent respectively. The 3- and 5-year overall survival rates were 59 and 53 per cent respectively. During chemotherapy, grade 3/4 neutropenia occurred in 19 per cent and grade 3/4 non-haematological adverse events in 15.4 per cent. The incidence of grade 3/4 adverse events related to surgery was 12 per cent. There were no reoperations or treatment-related deaths. CONCLUSION: For locally advanced gastric cancer with extensive lymph node metastasis, 4-weekly S-1 plus cisplatin followed by surgery including PAN dissection was safe and effective for some patients. Further investigation of this treatment strategy is warranted.

KRAS and BRAF mutations are rare and related to DNA mismatch repair deficiency in gastric cancer from the East and the West: results from a large international multicentre study.

BACKGROUND: Inhibitors of the epidermal growth factor (EGFR) signaling pathway have a major role in the treatment of KRAS wild-type colorectal cancer patients. The EGFR pathway has been shown to be activated in gastric cancer (GC). However, published data on KRAS and BRAF mutation status is limited in GC and has not been compared between GC from different geographic regions. METHODS: The prevalence of KRAS and BRAF mutations was established in 712 GC: 278 GC from the United Kingdom, 230 GC from Japan and 204 GC from Singapore. The relationship between KRAS/BRAF mutation status, DNA mismatch repair (MMR) status, clinicopathological variables and overall survival was analysed. RESULTS: Overall, 30 (4.2%) GC carried a KRAS mutation. In total, 5.8% of the UK GC, 4% of Japan GC and 1.5% of Singapore GC were KRAS mutant. KRAS mutant GC had fewer lymph node metastases in the UK cohort (P=0.005) and were more frequent in elderly patients in the Japan cohort (P=0.034). KRAS mutations were more frequent in MMR-deficient GC in the UK and the Japanese cohort (P<0.05). A BRAF mutation was only detected in a single Japanese GC. CONCLUSIONS: This large multicentre study demonstrated that KRAS mutations and DNA MMR deficiency have a role in a small subgroup of GC irrespective of country of origin, suggesting that this subgroup of GC may have developed along a common pathway. Further studies need to establish whether concomitant mutations or amplifications of other EGFR signalling pathway genes may contribute to the activation of this pathway in GC.

Phase II study of oxaliplatin plus S-1 as first-line treatment for advanced gastric cancer (G-SOX study).

BACKGROUND: The efficacy and safety of oxaliplatin combined with S-1 (SOX regimen) for unresectable advanced or recurrent gastric cancer were investigated. PATIENTS AND METHODS: Oxaliplatin was administered i.v. (100 mg/m(2)) on day 1, while S-1 was administered orally (80 mg/m(2)/day, b.i.d.) for 14 days followed by a 7-day rest. This schedule was repeated every 3 weeks. RESULTS: Among 55 patients enrolled, one patient received oxaliplatin for the other study, and three patients were considered unsuitable against the inclusion criteria. Accordingly, 51 patients were assessable for efficacy. The response rate was 59%, and the disease control rate was 84%. The median progression-free survival time was 6.5 months, the 1-year survival rate was 71%, and the median survival time was 16.5 months. In 54 patients assessed for safety, the major grade 3/4 toxic effects were neutropenia (22%), thrombocytopenia (13%), anemia (9%), anorexia (6%), fatigue (6%), and sensory neuropathy (4%). CONCLUSION: These findings indicate that SOX regimen with oxaliplatin at a dose of 100 mg/m(2) is feasible and shows promising efficacy against advanced gastric cancer.

Molecular Biomarker Study in a Randomised Phase III Trial of Irinotecan Plus S-1 versus S-1 for Advanced Gastric Cancer (GC0301/TOP-002).

AIMS: Gastric cancer is a common and heterogeneous disease; however, global standard and biomarkers for selecting chemotherapy regimens have not been established. This study was designed retrospectively to identify molecular biomarkers for irinotecan plus S-1 (IRI-S) and S-1 therapy from subset analyses in GC0301/TOP-002, a randomised phase III trial for advanced gastric cancer. MATERIALS AND METHODS: Paraffin-embedded primary tumour specimens were collected from 126 of 326 randomised patients in GC0301/TOP-002. The mRNA was measured for thymidylate synthase, dihydropyrimidine dehydrogenase, topoisomerase I, excision repair cross-complementing gene 1 (ERCC1) and thymidine phosphorylase; categorised into low and high to analyse their association with efficacy end points. RESULTS: There was no significant difference in each mRNA between S-1 and IRI-S groups, whereas there were differences among some clinical characteristics. Multivariate analyses for overall survival showed that mRNA levels were not correlated with prognosis. By comparison, between IRI-S and S-1 arms, low thymidylate synthase, low ERCC1 and high thymidine phosphorylase were associated with better prognosis for IRI-S versus S-1 (hazard ratio = 0.653, 0.702 and 0.709, respectively; P < 0.15 for each interaction). CONCLUSION: Low thymidylate synthase, low ERCC1 and high thymidine phosphorylase are candidates for predictive biomarkers for first-line treatment in advanced gastric cancer by IRI-S. Further study is warranted to confirm these results in other clinical trials and cohort studies.

Marked increase of trypsin(ogen) in serum of linitis plastica (gastric cancer, borrmann 4) patients.

Linitis plastica, or Borrmann 4 gastric cancer, shows very poor prognosis, and the reason has not been understood. In the present study, we examined serum levels of trypsin(ogen) in 44 gastric cancer patients, including 17 early gastric cancer, 18 non-Borrmann 4 advanced gastric cancer, and 9 Borrmann 4 gastric cancer, by using the RIA gnost Trypsin kit (Hoechst Japan, Tokyo, Japan), which was expected to detect trypsin-1, trypsin-2, trypsinogen-1, and trypsinogen-2 in sera. The trypsin(ogen) concentration was much higher in the patients with linitis plastica than in the other gross types of gastric cancer. Hypertrypsinemia was identified in approximately 60% of advanced gastric cancer cases. Lymph node involvement, liver metastasis, or poorly differentiated adenocarcinoma is an important factor of hypertrypsinemia. The serum trypsin(ogen) level in linitis plastica patients was 3484.4 +/- 2319.7 ng/ml (mean +/- SD), which was significantly higher not only than that of the early gastric cancer (384.1 +/- 92.1) but also the stage IV gastric cancer patients (578 +/- 440.4), excluding those with linitis plastica. The elevated serum trypsinogen level in linitis plastica patients may be related to the malignant behavior of this type of cancer cell. Serum trypsin(ogen) of linitis plastica shows significantly higher concentrations than do the other types of advanced gastric cancer. Therefore, serum concentration of trypsin(ogen) might be a good marker of gastric cancer of linitis plastica.

Plasma concentrations of VEGF and bFGF in patients with gastric carcinoma.

We examined plasma levels of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) in 54 patients with gastric carcinoma. Postoperative survival was significantly poorer in patients with plasma VEGF levels more than 10.0 pg/ml at the time of surgery. By an univariate analysis of the factors affecting survival, serosal invasion, lymph node metastasis, peritoneal dissemination, lymphatic vessel invasion, curability, and VEGF proteins were significant. By a multivariate analysis only VEGF levels and curability remained significant. Patients with recurrent disease, including liver metastasis, had significantly higher plasma VEGF concentrations than those with resectable primary tumors. VEGF, not bFGF, may serve as an independent prognosticator and a sensitive indicator for liver recurrence in patients with gastric carcinoma.

Prognostic value of tissue inhibitor of matrix metalloproteinase-1 in plasma of patients with gastric cancer.

Tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) in plasma has been reported to be related to disease progression in patients with gastric cancer. However, the prognostic significance of plasma TIMP-1 concentrations has not been clarified. Concentrations of TIMP-1 protein were measured by enzyme-linked immuno-sorbent assay in plasma samples of 147 preoperative patients who subsequently underwent gastric resection, and prognosis was compared. The cut-off value of plasma TIMP-1 concentrations was defined as 112.5 ng/ml, referring to the TIMP-1 levels in patients with intramucosal gastric cancer. Twenty-nine out of 147 patients had higher plasma TIMP-1 levels than the cut off value. When the patients were divided into those with elevated values and those with normal TIMP-1, such parameters as age, serosal invasion, metastases to lymph nodes, peritoneum, and liver, lymphatic invasion, curability, and stage were significantly different between the two. By univariate analysis of the factors affecting survival, macroscopic type, histology, serosal invasion, metastasis to lymph node, peritoneum, and liver, vessel invasions, curability, and plasma TIMP-1 were significant. However, multivariate analysis revealed that TIMP-1 was the only significant factor. In patients with gastric cancer, plasma TIMP-1 seem to be an independent and most powerful prognosticator for the survival.

Expression of nitric oxide synthase in gastric cancer.

The expression of endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) in samples of normal gastric mucosa and gastric cancer were examined by reverse transcriptase-polymerase chain reaction (RT-PCR) and semi-quantitative Western blot. In normal gastric mucosa, eNOS protein was found in all samples examined (mean, 70.2 +/- 60.1), relative to a standard protein. In gastric cancer specimens, eNOS protein was also detected in all samples, but the quantity (86.5 +/- 76.6) was not different from that found in samples of normal mucosa. The quantity of eNOS in gastric cancer tissues was negatively correlated with serosal invasion. iNbS mRNA, detected in nine of 18 cases, was slightly related to massive lymph node metastasis (n1-3 vs. n4). Neither tumor necrosis factor alpha (TNF-alpha) mRNA nor interleukin-6 (IL-6) mRNA was related to the expression of iNOS mRNA. These results suggest that iNOS not eNOS plays a role in gastric cancer tumor extension, but iNOS mRNA appears not to be induced by either TNF-alpha or IL-6.

Expression of facilitative glucose transporter 1 mRNA in colon cancer was not regulated by k-ras.

The expression of facilitative glucose transporter isoforms in colon adenocarcinoma and the possible role of k-ras in inducing GLUT (glucose transporter) mRNA were studied. RT-PCR demonstrated GLUT2 and GLUT3 expression in 100% of the ten normal colon mucosa samples but detected no GLUT1 mRNA. By contrast, GLUT1 mRNA was detected in all 20 (100%) colon cancer samples examined. GLUT4 mRNA was not detected in either normal mucosa or colon cancer tissues. Semiquantitative PCR demonstrated equal amounts of GLUT2 and GLUT3 mRNA in both normal mucosa and colon cancer samples. A point mutation in codon 12 of k-ras was detected in only six of the 20 (30%) colon cancer samples. Thus, a major difference between normal colon epithelia and colon cancer was the acquisition of GLUT1 expression, which was unlikely to have been induced by a point mutation in codon 12 of k-ras.

Suppression of facilitative glucose transporter 1 mRNA can suppress tumor growth.

We attempted to suppress glucose transporter 1 (GLUT1) expression by transfecting MKN45 cells with cDNA for antisense GLUT1. Glucose transport was significantly decreased in cells with antisense GLUT1 compared with wild-type cells or cells with vector alone. Suppression of GLUT1 mRNA resulted in a decreased number of cells in the S phase. This was accompanied by overexpression of p21 protein. Tumorigenicity in the nude mice injected with antisense GLUT1 expressing cells was significantly slower than in those with wild-type MKN45 cells. These results suggest that antisense GLUT1 mRNA inhibits tumor growth through a G(1) arrest and that expression of antisense GLUT1 mRNA via gene therapy can be used as a tool in the treatment of cancer.

Glucose uptake in the human gastric cancer cell line, MKN28, is increased by insulin stimulation.

The expression of the insulin-responsive glucose transporter (GLUT) 4 was studied in three histologically different human gastric cancer cell lines, MKN28, MKN45, and STSA. RT-PCR demonstrated GLUT1 and GLUT4 mRNA in all three cell lines. MKN28 cells expressed GLUT4 protein more than MKN45 and STSA cells by immunohistochemistry. Insulin stimulation of MKN28 cells resulted in a 22% increase in glucose uptake over that found under basal conditions (0.60 +/- 0.05 fmol/cell per min after insulin stimulation versus 0.53 +/- 0.07 fmol/cell per 3 min at basal). No increase in glucose uptake occurred with insulin stimulation in MKN45 or STSA cells. We conclude that the insulin responsive GLUT4 is expressed in MKN28, MKN45, and STKM1 human gastric cancer cell lines, albeit in different amounts. The greater expression of this transporter in MKN28 cells is likely responsible for the cell's ability to increase glucose uptake with insulin stimulation. However, the role played by GLUT4 in regulating the amount of glucose uptake would not be large in those human gastric cancer cell lines.

High incidence of synchronous cancer of the oral cavity and the upper gastrointestinal tract.

A high incidence of synchronous esophageal or gastric carcinoma in preoperative patients with carcinoma of the oral cavity was reported. Esophageal carcinoma was found in seven out of 56 patients (12.5%) and gastric cancer in five patients (8.9%) by videoendoscopy aided with lugol staining in the esophagus and indigocarmine solution in the stomach, although all patients were completely asymptomatic for these lesions. All patients were male, regular drinkers and heavy smokers. The depth of invasion of such tumors was limited to either mucosa or submucosa. Those esophageal and gastric lesions beside the primary oral cancers were positive for p53 protein by immunohistochemistry. Careful preoperative evaluation of not only the esophagus but also the stomach should be a routine procedure in patients with carcinoma of the oral cavity.

Insulin resistance and the alterations of glucose transporter-4 in adipose cells from cachectic tumor-bearing rats.

BACKGROUND: Insulin resistance may play an important role in cancer cachexia; however, its mechanisms remain to be clarified. METHODS: Cellular mechanisms of insulin resistance in tumor-bearing rats (TBR) were investigated in isolated adipose cells by measuring 3-O-[14C]methyl glucose transport activity and glucose transporter-4 (GLUT4) protein in low-density microsomes at a basal state and in the plasma membrane at an insulin-stimulated state. RESULTS: The insulin-stimulated glucose transport activity in adipose cells from TBR was significantly lower than that of control rats (CTR) (0.51 +/- 0.25 and 2.27 +/- 0.11 fmol/cell/min, respectively). The amount of GLUT4 in low-density microsomes at a basal state and in plasma membrane at an insulin-stimulated state was less in TBR than in CTR. CONCLUSIONS: These data suggest that the insulin resistance seen in the adipose cells of these tumor-bearing rats was caused in part by both a decreased amount of GLUT4 protein in a basal state and a decreased translocation of GLUT4 in response to insulin stimulation.

Angiogenesis inhibitor, TNP-470, suppresses growth of peritoneal disseminating foci.

BACKGROUND/AIMS: Angiogenesis is critical not only for growth of primary tumors but also for cells established at distant organs. We investigated the effects of angiogenesis inhibitor, TNP-470, on the establishment and growth of intraperitoneally inoculated human gastric cancer cell line, MKN-45, and survival of nude mice with this tumor. METHODOLOGY: Human gastric cancer cell line, MKN-45, were injected into the peritoneal cavity of an ICR nude mouse and a model of peritoneal dissemination was developed. TNP-470 was injected subcutaneously every other day from day 1 until sacrifice or death. The effects of TNP-470 on MKN-45 cells were also examined in vitro. RESULTS: Although the number of disseminated foci was not significantly different, the maximum size was significantly smaller in a TNP-treated group than those of a control. Survival time was significantly longer in a TNP-treated group. TNP-470 demonstrated no growth inhibition of MKN45 cells in vitro. CONCLUSIONS: Those results suggested that anti-angiogenic agent, TNP-470, might be effective in treating peritoneal dissemination of gastric cancer by inhibiting growth of the seeded tumor cells on the peritoneum.

Neoadjuvant chemotherapy with a combination of irinotecan and cisplatin in advanced gastric cancer--a case report.

We report a case of advanced gastric carcinoma successfully treated with a combination of irinotecan and cisplatin as neoadjuvant chemotherapy. The patient, a 78-year-old man, had type 2 gastric cancer, which had metastasized to the paraaortic lymph nodes. He was treated with irinotecan, 70 mg on day 1 and day 15, and cisplatin, 80 mg on day 1. The course was repeated every 4 weeks. Two courses of treatment resulted in a marked reduction of both the primary tumor and lymph nodes. Subsequently, the patient underwent curative surgery consisting of total gastrectomy, splenectomy, and D3 lymph node dissection. No surgical complications were observed. On microscopic examination, only a few tumor cells were detected in the granulation tissues of the resected stomach and in the lymph nodes. This would be the first case to demonstrate the effectiveness and the safety of irinotecan and cisplatin used in the neoadjuvant setting for treatment of advanced gastric carcinoma.

Should scirrhous gastric carcinoma be treated surgically? Clinical experiences with 233 cases and a retrospective analysis of prognosticators.

BACKGROUND/AIMS: The prognosis of patients with scirrhous gastric carcinoma has been poorest. METHODOLOGY: To clarify the role of surgical treatment, 233 patients with a primary scirrhous gastric carcinoma were retrospectively analyzed. RESULTS: Of the 233 patients, 182 underwent surgical resection, while the other 51 did not. The median survival time of those with unresectable tumors was 88.0 +/- 15.3 days and that of those who underwent resection was 380.0 +/- 41.8 days. In the 182 patients who underwent resection, multivariate analysis revealed four significant factors; lymphatic invasion, serosal invasion, curability, and lymph node dissection. Of these, curability was the most significant. The median survival time of patients whose tumor were curatively resected was 727.0 +/- 116.3 days, significantly longer than 272 +/- 34.9 days for those whose resection ended noncuratively. In 65 patients whose tumor was curatively resected, subset analyses of factors by multivariate analyses revealed an absence of serosal invasion as the single significant prognosticator. The 5-year survival rate was 55.6% in patients with scirrhous cancer without serosal invasion. CONCLUSIONS: For patients with scirrhous gastric carcinoma, palliative resection should not be attempted for poor outcome. However, if curative resection seems feasible, radical surgery would be justified, especially for tumors without serosal exposure.

Insulin resistance and tissue glycogen content in the tumor-bearing state.

BACKGROUND/AIMS: Insulin resistance in cancer patients may play an important role in post-operative outcomes. METHODOLOGY: In an effort to understand the mechanism responsible, we measured glucose uptake by euglycemic hyperinsulinemic glucose clamp, glycogen content in skeletal muscle, liver, and adipose tissue by an enzymatic method, and facilitative glucose transporter mRNA by reverse transcription polymerase chain reaction (RT-PCR), in 10 patients with gastrointestinal cancers. Glucose uptake was also examined in 5 volunteers. RESULTS: The M value was significantly smaller in cancer patients (6.11 +/- 2.49 mg/kg/min) than in controls (8.55 +/- 0.91 mg/kg/min). A significant decrease in glucose storage was also observed in cancer patients compared with that of controls, 3.18 +/- 2.46 and 4.95 +/- 0.55 mg/kg/min, respectively. The glycogen content of skeletal muscle from cancer patients averaged 12.1 +/- 2.7 micrograms/mg tissue and was correlated with the M value (r = 0.75, p < 0.05). Muscle glycogen had an even stronger correlation to glucose storage (r = 0.964, p < 0.05). The glycogen content of liver averaged 105.4 +/- 79.0 micrograms/mg tissue but was not related to the M value. Glycogen in adipose tissue was not measurable by this method. Muscle and liver samples were positive for periodic and Schiff (PAS) stain, while the stain in adipose tissue was very weak. Glucose transporter 4 mRNA was the dominant facilitative glucose transporter in skeletal muscle. CONCLUSIONS: These results suggest that patients with gastrointestinal (GI) cancer and insulin resistance have decreased glycogen storage in skeletal muscle. These may be the patients who are in need of pre-operative nutritional support.

Tumor-associated metabolic alterations in patients with gastric and esophageal cancer.

BACKGROUND/AIMS: This study was conducted to elucidate tumor-induced alterations in pace maker enzymes and possible effects of pre-operative nutritional support on those enzymes in gastrointestinal (GI) cancer patients. METHODOLOGY: Three pacemaker enzymes; phosphoenolpyruvate carboxykinase (PEPck) for gluconeogenesis, malic enzyme for de novo fatty acid synthesis, and lipoprotein lipase (LPL) for triglyceride clearance in liver and adipose tissues were examined in 22 patients with gastrointestinal cancers, both with or without pre-operative total parenteral nutritional support. Five patients with normal nutrition and liver function but undergoing cholecystectomy were used as controls. RESULTS: Activity of hepatic malic enzyme was significantly decreased (p < 0.05) and activity of hepatic PEPck was slightly elevated in patients with advanced stage cancer. LPL activity in the abdominal subcutaneous tissue was increased by advanced tumor bearing (p < 0.05). Pre-operative total parenteral nutrition for 1 week resulted in significant stimulation of LPL activity in adipose tissue from patients with advanced cancer compared to that from controls (6.53 +/- 4.99 U/g tissue and 1.32 +/- 0.18 U/g tissue, respectively). CONCLUSIONS: Effects of tumor bearing on key regulatory enzymes were evident only at the advanced stages. Pre-operative hyperalimentation may increase fat storage at the peripheral tissue by stimulating tissue LPL activity even at the advanced stage disease.

Gastrectomized patients are in a state of chronic protein malnutrition analyses of 23 amino acids.

BACKGROUND/AIMS: Malnutrition is one of the major postoperative complications of radical subtotal or total gastrectomy for gastric cancer. This study was conducted to clarify the nutritional consequences of radical gastrectomy with respect to protein metabolism. METHODOLOGY: To evaluate the nutritional status and the abnormalities in protein metabolism in such cases, serum concentrations of 23 amino acids were measured by high performance liquid chromatography in 40 patients who had undergone either subtotal (n = 20) or total (n = 20) gastrectomy more than 6 months prior to this analysis. RESULTS: Serum concentrations of total amino acids and nonessential amino acids were the same between gastrectomized patients and healthy controls (n = 50). However, concentrations of essential amino acids, essential amino acid/nonessential amino acid and branched-chain amino acid/total amino acid ratios were significantly lower in patient groups than in normal controls. Each essential amino acid was decreased and concentrations of glutamate and citrulline were increased in both patient groups compared with controls. The major differences between patients with subtotal and total gastrectomies included an increased ornithine and a decreased arginine concentration in patients with subtotal gastrectomy. CONCLUSIONS: These changes suggest that malabsorption of protein from the intestinal tract causes persistent proteolysis in the skeletal muscle for long periods of time after surgery in these patients and that changes in ornithine and citrulline levels may reflect more severe alterations in those with total gastrectomy.

Expression of facilitative glucose transporters in gastric tumors.

BACKGROUND/AIMS: Although cancer cells are known to have an increased rate of glucose metabolism, the complete mechanism for increased glucose uptake in tumor cells is unknown. METHODOLOGY: The presence of mRNA for 5 facilitative glucose transporter (GLUT) isoforms was evaluated by reverse transcription-polymerase chain reaction (RT-PCR) in paired samples of normal gastric mucosa and gastric tumor from 20 individuals. Expression of GLUT proteins was immunohistochemically determined in 70 resected gastric cancer specimens. RESULTS: By using RT-PCR, GLUT2 mRNA was detected in 80% of normal gastric mucosal samples, while GLUT4 mRNA was seen in only 40%, GLUT1 mRNA was not detected in normal gastric mucosa. In gastric carcinoma samples, GLUT1 mRNA was detected in 19 out of 20 cases (95%) and GLUT2, GLUT3, and GLUT4 mRNAs in all samples. By immunohistochemistry, GLUT1 protein was detected in 19% of the tumors. A majority of tumors (61%) expressed 1 or more transporter protein. The presence of GLUT1 protein in a tumor was positively correlated with the tumor's invasion into the gastric wall, lymphatics or blood vessels and with lymph node metastases. The post-operative survival of patients with tumor expressing GLUT1 protein was significantly worse than those with tumor without GLUT1 protein. CONCLUSIONS: Gastric cancer cells may acquire the ability to produce GLUT1 mRNA by malignant transformation. Increased expression of the high-affinity glucose transporters, GLUT1 and/or GLUT4, in tumor cells may drain glucose preferentially to the tumor at the expense of the tumor-bearing host.