© 2004 Foundation for Promotion of Cancer Research
Meeting Report |
Report of the Seventeenth International Symposium of the Foundation for Promotion of Cancer Research: Recent Advances in Gastric Cancer
1 Department of Surgery, Queen Elizabeth Hospital, Birmingham, UK, 2 Gastric Surgery Division, National Cancer Center Hospital, Tokyo, Japan and 3 Department of Gastrointestinal Medical Oncology, University of Texas, M.D. Anderson Cancer Center, Texas, USA
For reprints and all correspondence: Takeshi Sano, National Cancer Center Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan. E-mail: tksano{at}ncc.go.jp
Received June 23, 2004; accepted June 25, 2004
INTRODUCTION
The Seventeenth International Symposium of the Foundation for the Promotion of Cancer Research, "Recent Advances in Gastric Cancer", was held in Tokyo on March 15 to 17, 2004. The symposium was organized by Drs M. Sasako, J. Ajani, S. Hirohashi, A. Ohtsu, D. Saito, T. Sano and T. Ushijima with Dr T. Kakizoe as advisor.
WELCOME AND OPENING ADDRESS
Professor T. Sugimura [President Emeritus, National Cancer Center (NCC)] opened the Seventeenth International Symposium. Since 1987 over 500 speakers from around the world have been invited to discuss various cancers and this was the second symposium at which gastric cancer was discussed, the first being in 1998. Professor Sugimura stated his personal interest in this year's topic and recounted the history of his own gastric cancer, detected at screening and successfully treated at the NCC by total gastrectomy. Professor A. Ajani gave the opening address. Although there have been many successes in the diagnosis and treatment of gastric cancer there remain many unanswered questions. In epidemiology, environmental and genetic factors need to be identified and perhaps modified to reduce cancer rates. The differing incidences between the sexes and the geographical variation of sites of cancer occurence require explanation. Would it be possible to improve responses to chemotherapy and to target cancers with individualized therapy? These are major issues and this symposium offered a unique opportunity to discuss these dilemmas and challenges in detail.
SESSION 1: EPIDEMIOLOGY, GENETICS AND PREVENTION OF GASTRIC CANCER
EPIDEMIOLOGY: CHAIRMAN PROFESSOR ADRIAN LEE
Professor P. Boyle [International Agency for Research on Cancer (IARC), Lyon, France] delivered the first presentation. Gastric cancer remains the second most common cause of cancer death in men and the fifth in women. There are large variations in incidence worldwide, rates being particularly high in the Far East and Eastern Europe. Etiological factors include smoking, salt intake, and incidence of Helicobacter pylori. Early studies of H.pylori underestimated its prevalence and more recent evidence suggests an odds ratio of at least 6.0 for the development of gastric cancer, an association exceeded only by that of cigarettes and lung cancer. Since the 1950s most countries have seen the incidence of gastric cancer halved. In the West this reduction has made the largest single contribution to the overall reduction in cancer mortality. Even now the exact reasons for this decline are not clear.
Trends in the incidence of gastric cancer in Japan were presented by Dr M. Inoue (NCC, Tokyo). Although lung cancer has recently become the leading cause of cancer related death, gastric cancer continues to be more common. The age-standardized incidence in Japan has fallen dramatically while the rapid ageing of the Japanese population has resulted in an increase in the absolute number of cases, increasing until 1995 before leveling off. The proportion of early disease increased from 25% in the 1970s to over 50% now. It is predicted that both age-standardized incidence and death rates will decline but that total cases will remain static and that the proportion of patients over the age of 80 will increase from 16% to nearly 40%. This will present many challenges for the future.
The incidence of tumors of the gastro-esophageal junction (GEJ) is thought to be increasing. Dr P. Hainaut (IARC, Lyon, France) addressed this issue. Carcinomas of the lower esophagus (Siewert Type I) and those of the GEJ (Siewert Type II) are thought by many physicians to be synonymous. That they are different is witnessed by specific patterns of cytokeratin staining, differing expression of p53 mutation and the lack of Barrett's esophagitis in the type II tumors. Non-acid-secreting cardiac type mucosa is frequently seen at the GEJ and to many it represents a metaplastic change. In embryos however this cardiac mucosa is universal, marking it as a constitutive entity. After re-examining data on type I and II tumors and correcting misclassifications, the incidence of GEJ tumors has remained constant although there is little doubt that the incidence of adenocarcinomas of the lower esophagus is increasing rapidly.
|
GENETICS: CHAIRMAN PROFESSOR CHRISTIAN WITTEKIND
The Lauren intestinal and diffuse types of gastric cancer differ and Dr T. Ushijima (NCC Research Institute, Tokyo) summarized many of the genetic differences. Oncogene activation (ß-catenin, K-ras and c-erbB2) and p53 mutations are frequent in intestinal type tumors (40%) compared with diffuse types (15%), but the greatest difference is in e-cadherin mutations, present in up to 50% of diffuse tumors but rare in the intestinal type. As well as direct genetic mutations, epigenetic inactivation of genes occurs and methylation of DNA, especially of promoter areas, is frequently a mechanism that prevents gene transcription. Methylated DNA is preserved after DNA replication preventing gene expression in daughter cells. Nine genes were identified as frequently methylated in gastric cancer cell lines and four were confirmed as methylated in tumors. One of these genes Lysyl Oxidase (LOX) has since been identified as a new tumor suppressor gene, suppressing the growth of tumors in LOX transfected nude mice.
Hereditary diffuse gastric cancer (HDGC) caused by germline e-cadherin mutations was first described in 1998 by Dr P. Guilford (University of Otago, Dunedin, New Zealand) who updated the symposium on its current status. Forty-four differing mutations have been identified worldwide with a penetrence of 80% in males and 70% in females. Within resected stomachs there are multiple signet ring micro-cancers (range 8–>300) with a predisposition for the transition zone between the body and antrum. There are no reports of prophylactic gastrectomy where tumor has not been found, indicating that these micro-cancers are potentially very slow growing and require a "second hit" to initiate invasion. Micro array gene mapping has identified c-src gene activation as one potential second hit. From a practical viewpoint endoscopy has previously missed many cancers but the use of Congo red dye spray has increased sensitivity such that endoscopy can be considered reliable.
Genetic alterations such as p53, ß-catenin and e-cadherin are frequent in gastric carcinogenesis and the ability of gene micro array experiments to analyze thousands of genes has allowed classification of tumors at the molecular level as well as having identified novel molecular targets for therapeutics and diagnostics. Professor H. Aburatani (University of Tokyo) compared cancerous gastric and non-cancerous gastric tissue by micro array and identified genes associated with lymph node metastasis (Oct 2) or with type of tumor (LI cadherin). To assist with the interpretation of the vast amount of data produced (up to 51 000 genes) a new visualization tool, an "expression imbalance map" has been developed to allow differing gene expression to be visualized easily.
PREVENTION: CHAIRMAN DR JAE-MOON BAE
High salt intake is a risk factor for gastric cancer. Dr S. Tsugane (NCC, Tokyo) presented two studies examining this relationship. The Eco cancer study selected 633 people aged 40 to 49 years from five geographical areas with large differences in gastric cancer mortality. There was a good correlation between salt intake and gastric cancer. A cohort study examined 100 000 people aged 40 to 69 years and again there was a good correlation between amount of salt consumed and subsequent development of cancer. A similar cohort study demonstrated a trend for reduced gastric cancer rates in people with higher consumption of fruit and vegetables. An RCT of vitamin C supplementation involving 439 randomized subjects demonstrated increased vitamin C plasma levels and a smaller change in pepsinogen I/II ratio (a surrogate maker for progression to atrophic gastritis) in the high dose group. Based on these results, reduction in salt intake and encouragement of consumption of fresh fruits and vegetables have the potential to reduce the risk of gastric cancer in Japan.
SESSION 2: HELICOBACTER PYLORI AND PATHOLOGY OF GASTRIC CANCER
HELICOBACTER PYLORI: CHAIRMAN DR PIERRE HAINAUT
Professor A. Lee (University of New South Wales, Sydney, Australia) presented evidence for H.pylori as a carcinogen. The Correa hypothesis of gastric carcinogenesis through gastritis, atrophy, intestinal metaplasia and dysplasia is widely accepted. H.pylori has been shown to be a potent initiator of inflammation and is classed as a carcinogen by the World Health Organization (WHO) although with an odds ration of only 1.92 many remain skeptical. Atrophy and intestinal metaplasia represent a hostile environment for H.pylori and early studies underestimated the extent of past infection. More recent studies suggest much higher ORs of up to 23. The location of gastritis is thought to determine the outcome of the disease with antral disease predisposing to duodenal ulcer and pan-gastritis predisposing to gastric ulcers or cancer. H.pylori cannot survive in the most acid producing areas of the stomach thriving only within a restricted pH range, most frequently found at the transition zone between body and antrum. In animal models as well as in humans reduction in pH with PPI, vagotomy or atrophy allows H.pylori to migrate proximally. There is mounting evidence that H.pylori strains that produce cagA, vacA or IL1B, as well as promoting carcinogenesis by inflammation, also have direct mitogenic effects on host cells.
In Mongolian gerbils (MGs) H.pylori infection results in chronic gastritis, peptic ulcers and intestinal metaplasia. Dr M. Tatematsu (Aichi Cancer Center Research Institute) updated the symposium on the status of these animal models. H.pylori infection and treatment of the stomach with N-nitroso compounds results in gastric cancers with an intestinal appearance. High salt diets acted synergistically. The duration of infection with H.pylori correlated with the risk of cancer, infection early in life enhancing it while eradication reduces the risk. There is some disagreement with the literature regarding H.pylori as a promoter or initiator of carcinogenesis, Dr Tatematsu's group finding H.pylori infection alone, insufficient for carcinogenesis. Experiments using chimeric mice demonstrate that each gastric gland has a single progenitor cell and that cancers are likewise clonal. The submucosal tumor-like lesions reported by others as induced by H.pylori are not clonal suggesting these are regenerative phenomena and that H.pylori acts only as a promoter of carcinogenesis.
The role of H.pylori eradication in reducing cancer risk remains controversial. Dr N. Uemura (International Medical Center of Japan, Tokyo) presented a follow-up study of 1526 endoscopy patients. No gastric cancers developed in the H.pylori-negative group (n = 280) but 2.9% of the H.pylori-positive group (36/1246) developed cancers. To investigate the role of H.pylori eradication, 132 patients with early gastric cancers (EGC) treated by endoscopic mucosal resection (EMR) had either H.pylori eradication (n = 65) or no eradication (n = 67). 16% (n = 11) or the non-eradicated group developed further cancers compared with 3% (n = 2) in the eradicated group (P = 0.04). Intragastric pH in the eradicated group reduced from a mean of 6 to 3. A cohort study from China supports these findings. 1630 H.pylori-positive patients, half with dysplasia (n = 824), were randomized to H.pylori eradication (n = 820) or not (n = 810). In the patients without initial dyplasia 1.5% (6/391) patients in the control group developed cancer versus 0% (0/413) in the eradicated group. For patients with initial dysplasia the figures were 1.5% (6/391) and 1.7%(7/407), respectively.
Having seen some evidence for the benefits of H.pylori eradication, Dr D.Saito (NCC, Tokyo) discussed some outstanding questions. Worldwide it is estimated that at least 60 million people are infected with H.pylori yet the incidence rate for gastric cancer among them is only 0.4%. Many countries with high infection rates have low cancer rates. Host and environmental factors are important but which is most important is unclear. There are also potential adverse effects of H.pylori eradication related to the recovery of gastric acid secretion with acute gastro esophageal erosions and gastro esophageal reflux disease (GERD) more prominent. GERD can predispose to Barrett's metaplasia a potentially pre-malignant lesion. There are published studies of H.pylori eradication and of these a Chinese and a South American study report benefit whilst an American Study did not. An intervention study in Japan has randomized 342 patients to eradication and 340 as a control group. End points are development and progression to mucosal atrophy and the trial is due to report in March 2004. Until the reports of this and other larger scale trials are known eradication cannot be recommended to all H.pylori-positive patients.
PATHOLOGY: CHAIRMAN PROFESSOR CORNELIS JH VAN DE VELDE
There are differences in the histological classifications used in the West and in Japan and Dr T. Shimoda (NCC, Tokyo) discussed the difficulties that arise in the diagnosis and differentiation of intramucosal carcinoma and dysplasia. In Japan the diagnosis of intramucosal carcinoma is based on structural features and cellular and nuclear atypia and invasion is not required. This has led to Western pathologists diagnosing cases as low grade dysplasia that under Japanese classification were definite carcinomas. The Vienna classification allows five categories from 1 (negative for carcinoma) to 5 (invasive carcinoma). Category 4 (non-invasive high-grade dysplasia) includes as subgroups: high-grade adenoma, non-invasive carcinoma (carcinoma in situ) and suspicion of invasive carcinoma. Using this classification disagreements between Western and Japanese pathologists were limited to categories 4 and 5, both of which require treatment.
The most commonly used classifications for gastric cancer are the WHO, UICC and JGCA systems. Professor C. Wittekind (University Klinikum Leipzig, Germany) presented the differences and also the latest UICC classification. The WHO classification is based on the predominant histological subtype. Although reproducible, it suffers as most tumors are heterogeneous and it has limited prognostic value. The UICC and JGCA both stage tumors according to a TNM system and have very similar T and M stages. In the N classification, the UICC adopts a numerical system based on the number of involved nodes and the JGCA use an anatomical classification based on the site of involved lymph nodes relative to the tumor. The UICC state that: to allow accurate classification the minimum number of lymph nodes retrieved should be 16 but in cases where fewer nodes are found, so long as they are all negative, the tumor should be staged as N0 not Nx. Most parties agree that the UICC system is more accurate in determining prognosis but there are concerns that it offers no guidance to surgeons as to the extent of the required lymphadenectomy.
Dr H. Katai (NCC, Tokyo) presented the results of examining 4524 patients treated at the NCC between 1969 and 1990 for potential prognostic factors. Overall survival was 70.1% ranging from 13.7% for stage IV to 92.3% for 1A. Factors that were found to be significant include (Hazard ratio and 95% CI interval): D1 dissection (1.95; 1.69–2.25); Depth (1.31; 1.26–1.36); male sex (1.21; 1.08–1.35); older age (1.03; 1.02 –1.037); type IV macroscopic type (2.13; 1.76–2.58). Hazard ratios for N stage by JGCA were: for N1 1.39 (1.18–1.62); N2 2.58 (2.22–3.00) and N3 5.18 (4.28–6.27). The Hazard ratios for N stage by UICC were similar although the UICC had better discriminating power between pN1 and pN2 and poorer discrimination between pN2 and pN3. The importance of complete examination of lymph nodes was also demonstrated as there was downward stage migration in 14.9% of cases if nodes of 5 mm or less were ignored and 1.5% if nodes of 2 mm or less were ignored.
SESSION 3: SCREENING and DIAGNOSIS OF GASTRIC CANCER
SCREENING: CHAIRMAN DR PARRY GUILFORD
Professor I. Tsuji (Tohoku University Graduate School of Medicine, Sendai) started the session with a report on screening in Japan. In 2001, 5.3 million people underwent mass screening representing 20% of the target population of all people aged 40 years or over. Each of the seven standard barium views of the stomach were reviewed by two radiologists. Sensitivity and specificity has been estimated at 70–90% and 80–90%, respectively, with a positive predictive value of 0.8–2.3%. Of those screened 10.3% were referred for endoscopy and 5275 cases of gastric cancer were diagnosed. Although the efficacy of mass screening has never been confirmed in randomized controlled trials, several observational studies demonstrate benefit. Three case–control studies show a pooled odds ratio of 0.39 (95%CI 0.29–0.52) for mortality in men from gastric cancer in the screened group and a ratio of 0.50 (0.34–0.72) for women. The reduction in mortality is attributable to the higher proportion of early gastric cancers in the screened group. Recent developments in screening revolve around changes in pepsinogen I/II ratios, measurable in a simple blood test, which are predictive of atrophic gastritis, a precursor of gastric cancer. Accuracy of this test is similar to that of contrast screening, but further proving trials are required before it can be considered for implementation.
Dr Jae-Moon Bae (NCC, Seoul, Korea) presented the rapid development of the national screening program in Korea. The program is targeted at lower income groups; those patients on Medicaid and beneficiaries of the National Health Insurance Corporation (NHIC). Screening is offered biannually to persons over the age of 40 years. As an incentive, the screening program qualified people for a 20% reduction in HNIC premiums, despite this the group's screening rate was 57.4% compared with 107.4% in the Medicaid group. In 2002, 806 699 people were screened using either double contrast barium meal or endoscopy, dependent on patient preference. The detection rate for gastric cancer was 0.11% in the Medicaid group and 0.13% in the NHIC group. The screening program was limited by a budget that, at $30.5 million, equated to $21 900 to detect each case of gastric cancer and a remarkable $38 to screen each person. The population in Korea aged over 40 years is 17 million and a randomized controlled trial of the effectiveness of the screening program was suggested before extending the screening program further.
DIAGNOSIS: CHAIRMAN PROFESSOR PETER BOYLE
Dr M. Kida (Kitasato University East Hospital, Sagamihara) presented recent developments in endoscopic ultrasound (EUS). EUS is 91.3% accurate for the diagnosis of intramural cancer but suffers in the diagnosis of submucosal (sm) cancer (67.7%) and invasion of the deeper muscle layers (65.7%). This is of crucial importance as new Japanese guidelines allow tumors with no likelihood of metastasis to be treated by endoscopic mucosal resection. These cancers are well-differentiated, intramural cancers and well-differentiated tumors less than 3 cm diameter with sm invasion <500 µm. A new development, three-dimensional EUS (3D-EUS) promises to improve accuracy significantly. In 3D-EUS a small scanning head rotates as the head is automatically withdrawn within a sheath. The operator can set the speed of rotation and withdrawal and the resulting spiral image is akin to a spiral CT scan. Data can be manipulated to allow reconstruction in various planes including linear images and radial series. Accuracy is estimated at 5–10% higher than conventional EUS and importantly, for lesions larger than 0.5 cm diameter, depth of sm invasion can be determined with 75–100% accuracy.
Another new technology was presented by Dr G. Iinuma (NCC, Japan). Modern multi-detector row CT (MDCT) scanners perform highly detailed scans within a single breath hold. Computer manipulation of MDCT data allows the display of virtual endoluminal images (VEIs), views familiar to any endoscopist, as well as 3D reconstructions of the stomach. In 2003, 86 gastric cancers in 84 patients were assessed using VEIs and 3D views and compared with conventional diagnostic and staging data. Of the early lesions VEIs revealed 21/44 (47.7%) and 3D reconstructions 18/44 (59.5%). For advanced lesions 25/42 (59.5%) were identified on VEIs and 32/42 (76.2%) by 3D reconstruction. Whilst accuracy is presently insufficient for clinical use, the development of CT scanners with 16 scanning arrays compared with the current four, will lead to great improvements in resolution and the potential to revolutionize diagnosis and staging of gastric cancer.
Positive cytology of peritoneal washes is a poor prognostic indicator equating to M1 (cy1). However peritoneal recurrences occur in patients with negative cytology. To explain these recurrences Dr Y. Kodera (Nagoya University Graduate School of Medicine) reported the results of reverse transcriptase polymerase chain reaction (RT-PCR) with CEA as the target using a real time, Light Cycler instrument to detect and quantify CEA mRNA. Positive results were strongly predictive of peritoneal recurrence (83%) but this technique was unreliable in primary tumors not expressing CEA. To improve this accuracy RT-PCR against cytokeratin 20 (CK20) was performed in 195 patients but results were disappointing with accuracy of just 54% in predicting peritoneal recurrence. CEA RT-PCR remains the most sensitive method for predicting free cancer cells in the peritoneal cavity.
The effect of isolated tumor cells (ITC) detectable only by immunohistochemistry (IHC) or RT-PCR in lymph nodes has caused considerable controversy. They are frequently detected even in pT1A tumors (12–18%) and several papers report poorer prognosis when present. Dr M. Sasako (NCC, Tokyo) presented definitive data dismissing their importance. 402 patients from four centers with pT2pN0 (n = 221) and pT2pN1 (n = 181) tumors underwent D2 or greater surgery. ITC was detected by IHC in 187 cases, 81 (37%) of the pT2pN0 group and 106 (58%) of the pT2pN1 group. For ITC(–) the 5-year survival rate was 84.4% and 10-year survival 70.4%, for ITC(+) the rates were 83.9% and 72.9% (NS). For the pT2pN0 subgroup rates were 90.7% and 76.6% for ITC(–) and 91.4% and 78.2% for ITC(+) (NS). There were no significant differences between any of the survival curves and in multivariate analysis ITC was not an independent prognostic factor.
SESSION 4: SURGERY AND ADJUVANT THERAPY FOR GASTRIC CANCER
SURGERY 1 AND 2: CHAIRMEN PROFESSOR JAFFER A. AJANI AND DR ALAN ANTHONEY
The extent of lymph node dissection during gastrectomy continues to cause controversy with high mortality and lack of efficacy reported by the Dutch and British MRC randomized controlled trials (RCTs). Interim analysis of two D1/D2 and the final report on the Dutch RCT trials were reported.
Dr M. Degiuli (Division of Surgery, Turin, Italy) presented convincing evidence that D2 gastrectomy can be performed safely in the West. The Italian Gastric Cancer Study Group (IGCSG) RCT of D1 versus D2 resection involves five centers and has recruited 203 patients (97 D1 and 106 D2). To avoid excess mortality the protocol forbids distal pancreatectecomy and splenectomy unless there is direct organ invasion. Recruitment is ongoing. Morbidity is low at 4.1% in the D1 group and 6.6% in the D2 group. Only 3/203 (1.45%) patients have died, one from the D1 group and two after D2. There is a high proportion of T1 tumors (33%) and as these are less likely to have lymph node metastasis, there was some suggestion that this may cause the trial to be underpowered.
In Taiwan an RCT of D1 versus D2 has completed recruiting. Uniquely this trial with 110 D1 and 111 D2 cases was performed in one center by three surgeons allowing excellent surgical quality control with no post-operative mortality. Morbidity was higher in the D2 group (17.1% versus 7.3%), the difference largely attributable to intra-abdominal abscess formation (8.1% versus 0%) and minor anastamotic leak (4.5% versus 0%). Professor C.W. Wu (National Yang-Ming University, Taipei, Taiwan) presented the interim survival analysis. At a median follow-up of over 5 years there appeared to be a small survival advantage for the D2 group. After correcting for confounding factors that biased the results against the D2 group, 5-year survival was stated to be 15% better in the D2 group. The trial is due to complete when all patients have completed 5-year follow-up.
Professor C.J.H. van de Velde (Leiden University Medical Center, Netherlands) discussed the long term results of the Dutch RCT of D1 versus D2 trial. Follow-up is now available to 11 years and survival remains the same in both groups (31% versus 35% NS). The excess surgical mortality and morbidity in the D2 group was largely attributable to the distal pancreatectomy (DP) and splenectomy (S) but what is frequently not realized is the contribution that these procedures make to ongoing mortality, biasing results against the D2 group. In subgroup analysis excluding patients with DP and S, the higher mortality still seen in the D2 group is not significant (3.8% versus 6.3%) and survival differences at 5 (47% versus 56%) and 10 years (33% versus 47%) were significantly better in the D2 group. Even though numbers are small the benefit of D2 gastrectomy was largely in patients staged as N2 by the new UICC staging system.
Three large RCTs have been organized by the Japanese Clinical Oncology Group (JCOG) examining the extent of surgery required in the treatment of gastric cancer and Dr T. Sano (NCC, Tokyo) appraised these trials. JCOG 9501 recruited 523 patients to a trial of D2+ para-aortic lymphadenectomy versus D2 for T2b/T3/T4 tumors. Morbidity was higher in the group undergoing more extensive surgery (29% versus 21%) but mortality was identical with only two deaths in each group. The final analysis is due in 2006. The current trial (JCOG 0110) commenced in 2002 and will recruit 500 patients with T2 or greater proximal cancers. Randomization will be between splenectomy and spleen preservation. Although splenectomy is performed without increased mortality in Japan JCOG 0110 will compare reduction in morbidity associated with spleen preservation against the risk of local recurrence. JCOG 9502 was a trial comparing abdominal (A) approach versus thoracoabdominal (TA) approach with dissection of some mediastinal nodes for gastric cancer with esophageal invasion. The TA approach was expected to prove superior. After 8 years the first interim analysis was performed on 165 patients. There was only one death in the TA group but morbidity was significantly higher (47% versus 34%). Survival was better at all times in the A group (NS). The estimate for the possibility of a significant survival advantage occurring in the TA group if the trial was completed was 3.65% and the trial was halted.
Following trials the discussion turned to surgical techniques. In 2001 in Japan 959 cases of laparoscopy assisted distal gastrectomy (LADG) were performed, up from just 59 in 1996. Professor S. Kitano (Oita University) presented his department's results with LADG. The technique involves a laparoscopic dissection including D1 + 
and mobilization of the stomach. Division of the stomach and reconstruction are then performed via a mini laparotomy. Of 136 patients treated mean operating time was 238 minutes with 7% morbidity. Two non-gastric cancer deaths occurred and there has been no recurrent disease. Professor Kitano suggested that for an individual experienced in both open gastric surgery and laparoscopic surgery the learning curve would be relatively short, about 10 cases.
Continuing the theme of less extensive surgery, Dr Y. Kitagawa (Keio University School of Medicine) talked on sentinel node biopsy and its applications in gastric cancer. In a proving study with 270 patients with T1/2 N0 disease, 99 m-Technetium Tin colloid was injected endoscopically in to the tumor basin prior to conventional D2 gastrectomy. Sentinel nodes were detected using a hand held gamma probe and confirmation was by histology using routine H&E stain to detect the colloid. Frequently more than one sentinel node was found (average 4.1) and the detection rate for sentinel nodes was 97% with 99% accuracy. In 37% of cases sentinel nodes were found in JGCA N2 nodes. Laparoscopic wedge resection with confirmation of N0 status requires absolute accuracy from sentinel node mapping and the application of real time intraoperative RT-PCR is being investigated.
ADJUVANT THERAPY
CHAIRMAN: DR AL BENSON
The results of the South Western Oncology Group trial of adjuvant chemoradiotherapy (SWOG9008) have caused a re-examination of the role of adjuvant and neoadjuvant therapies in gastric cancer. Dr D. Ciot (Memorial Sloan-Kettering Cancer Center, New York, USA) and Professor C.J.H. van de Velde (Leiden University Medical Center, Netherlands) discussed the history of such treatments, the implications of SWOG9008 and future directions for trials.
Most prospective RCTs of adjuvant therapy have failed to demonstrate a benefit and have been criticized as being underpowered, using out dated treatments and having heterogeneous patient populations. Recent meta-analyses of these trials have suggested a small benefit in the order of 3–5%. SWOG9008 randomized 566 patients to either observation or postoperative 5-FU/leucovorin and 4500 cGy of fractionated radiotherapy. Improvements in median survival (27 versus 36 months) and 3-year survival (41% to 50%) were seen. Toxicity was a problem and despite the trial being limited to fit patients a third of the chemoradiotherapy arm were unable to complete treatment. The benefit of treatment appears to be largely related to improved locoregional control but both in this and in survival, the observation arm did particularly badly compared with patients in the Dutch D1/D2 trial, and the treatment arm had only comparable results with the Dutch group. This may be explained by the majority of cases having had inadequate clearance of even the N1 nodes (D0). Application of the Muruyama Index (MI) that rates the probability of lymph node metastasis based on the primary tumor's characteristics suggested considerable residual disease. Paradoxically in the subgroup where the probability of residual disease was low the beneficial effects of chemoradiotherapy appeared greater. Both surgeons concluded that although chemoradiation demonstrated benefit after poor surgery, the result could not be extrapolated to show benefit after surgery with a more radical lymph node clearance. Professor van de Velde outlined the proposal for a new Pan European Gastric Adjuvant Study with Uniform Surgery (PEGASUS), comparing adequate (D1 or greater with preservation of spleen and pancreas) surgery alone with adequate surgery with chemoradiotherapy.
The Japanese perspective on adjuvant therapy was given by Dr T. Kinoshita (NCC East, Tokyo). Two large RCTs of adjuvant treatment versus observation JCOG 8801 (MMC + 5-FU + oral UFT) with 579 patients and JCOG 9206 (MMC + 5-FU + CA + oral 5-FU) with 252 patients failed to demonstrate benefit. The results of JCOG 9206-2 with 282 patients comparing surgery + cisplatin + 5-FU + oral UFT with surgery alone are expected soon. S-1 is a new orally available agent with low toxicity and high single agent response rate. The ACTS-GC trial (Adjuvant Chemotherapy Trial of S-1 for Gastric Cancer) involving nearly 100 participating centers compares adjuvant S-1 with observation after surgery. The trial has recruited 800 of the target 1000 and will complete recruitment this year. Three phase II trials of neoadjuvant treatment are ongoing. Like his Western colleagues Dr Kinoshita's opinion was that SWOG9008 was not directly applicable to Japanese institutions where D2 lymph node dissections are routine.
SESSION 5: EMR AND MEDICAL ONCOLOGY
EMR: CHAIRMAN PROFESSOR CHEW-WUN WU
A recurrent problem for endoscopic mucosal resection (EMR) when dealing with lesions larger than 15 mm in diameter is the assessment of the depth of invasion where pre-treatment diagnosis is incorrect in 20% of cases. Conventional EMR techniques suffer from removing tumors piecemeal compromising pathological assessment in up to 30% and resulting in up to 17% local recurrence rate. En block resection suffers less from these problems and Dr T. Gotoda (NCC, Tokyo) demonstrated EMR with an "Insulated Tip" (IT) knife. Lesions up to 13 cm in diameter have been treated and of 2000 cases there has been only one local recurrence. The complication rate was 10% (7% bleeding and 3% perforation) and virtually all have been treated successfully endoscopically (the perforations with clips). Only 8 patients have required emergency surgery.
MEDICAL ONCOLOGY: CHAIRMEN DR MAURIZIO DEGIULI AND DR DANIEL COIT
Professor J.A. Ajani (University of Texas MD Anderson Cancer Center, USA) began the session on the North American perspective with a brief overview of the current agents used worldwide. There are significant problems in evaluating treatments including a deficit of phase III trials and problems with trial methodology. Trials are now international and methodology and reporting should be standardized to allow conclusions to be universally applicable. Differences in treatments would continue as cultural differences on the acceptability of side effects varied. As examples he cited higher acceptance of hair loss in the USA and Europe compared to Japan and India and tolerance of higher doses of S-1 by Japanese compared to European patients. He stated his belief that current evidence favored newer agents such as CPT-111, S-1 and oxaliplatin over 5-FU and cisplatin.
Dr A. Ohtsu (NCC East, Tokyo) presented the results of JCOG 9205, a trial of 5-FU versus 5-FU and cisplatin versus UFT + mitomycin C in advanced cancer. Although 5-FU + cisplatin gave higher response (34%) rates than 5-FU alone (11%), there was no difference in survival and 5-FU remains the standard for control arms of new trials. A brief review of previous JCOG trials demonstrated slight survival benefits for chemotherapy compared with best supportive treatment. New agents are undergoing phase II trials and S-1 appears promising with a single agent response rate of 45%. Non-controlled audit data suggests that, since the approval of S-1 and Taxanes, 3-year survival rates have increased from 2% to 9%. S-1 and CPT-111 are currently undergoing phase III trials in advanced cancer. JCOG 9912 will recruit 450 patients and compares 5-FU versus cisplatin + CPT-111 versus S-1 with survival as the primary endpoint. 322 patients have accrued since October 2000. A post marketing RCT of S-1 versus S-1 + cisplatin has recruited 210 of the target 300. The final outcomes of both trials are expected in 2006.
Dr A. Anthoney (University of Leeds, UK) presented a brief history of trials within Europe. Over the last 15 years 5-FU has provided the mainstay of treatment in combination with other drugs. In a large British trial of ECF versus FAMTX, ECF was superior with a response rate of 40% and a median survival time of 9.4 months. In the UK ECF is mainly used to palliate symptoms, a role where high response rates and rapid response times are of clinical importance. In Italy phase II trials have shown superior response rates for the PELF regimen and this has become routine care in Italy. As elsewhere, new agents such as the taxanes, oxaliplatin and capecitabine are undergoing phase II trials but 5-FU and cisplatin remain the mainstays of treatment throughout Europe.
The final two presentations focused on the use of tests for chemosensitivity and individualized treatment. Dr T. Kubota (Keio University) summarized retrospective trial data on chemosensitivity suggesting that laboratory tests of a tumor's chemosensitivity could predict outcome. Prospective data on 128 patients with stage III and IV cancers undergoing surgery and chemotherapy, demonstrated 85% 3-year survival rates in sensitive versus 52% in insensitive patients. Gene array mapping of tumors have implicated COX2, VEGF, a putative potassium channel, retinoblastoma binding protein-1 isoform, thymidylate kinase and dihydropyrimidine dehydrogenase as genes important for chemosensitivity. Transfection of COX2 and VEGF into cell lines not natively expressing these genes, demonstrated that COX2 confers chemoresistance but VEGF did not. In mice however COX2 inhibitors were unable to increase sensitivity to chemotherapy.
Tumors are not homogeneous and Professor A. Benson III (Northwestern University's Feinberg School of Medicine, Chicago, USA) discussed targeting treatment based on molecular profiles and avoiding drugs where there was a high probability of resistance. The use of gene arrays was providing large numbers of potential molecular markers but before being used as predictive markers they must be demonstrated to have significant and independent value and be validated by clinical testing. The use of novel agents should also be explored. Examples include vaccination to gastrin 17, a growth promoter for a large number of tumors, where production of antibodies predicted a longer survival, and matrix metalloproteinase inhibitors, which appeared to offer benefits to patients post chemotherapy. In summary as well as new agents; new targets, new markers, and new methods to establish response to treatment should all be investigated further.
CLOSING ADDRESS: CHAIRMAN PROFESSOR A. AJANI
Dr M. Sasako (NCC, Tokyo) gave the closing address where he thanked both speakers and attendants. The word encompassing the meeting was heterogeneity. Heterogeneity among cancers caused difficulties in staging and classification and with the epidemiology of the varying sites and types of tumor. Heterogeneity required tailoring surgery to the type and stage of cancer and to the patient. Most importantly heterogeneity between tumors and the identification of more homogeneous subgroups might allow individualized chemotherapy with the promise of better results. He called for a third meeting of this symposium in 6 years time to discuss further outcomes.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Emura, D. Saito, T. Kakizoe, and M. J. Blaser Infection, Cancer and Prevention: Report of the 19th International Symposium of the Foundation for Promotion of Cancer Research Jpn. J. Clin. Oncol., November 1, 2006; 36(11): 745 - 755. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||