Upper gastrointestinal tumors

• Mass screening and early detection of gastric cancer
• Gastric cancer and infection-associated cancers
• New trends in the treatment of gastric cancer: laparoscopic resection assisted by sentinel node navigation
• Mucosectomy in early gastric cancer: the risk of lymph node metastases
• Narrow-band imaging in the exploration of Barrett’s esophagus
• References

Mass screening and early detection of gastric cancer

An increasing burden of cancer is shown in the health statistics of the aging population of developed countries. More attention is now given to primary and secondary prevention of cancer worldwide. Secondary prevention is based on early detection of cancer and pre-malignant precursors at a completely curable stage in individuals who are asymptomatic or without suggestive symptoms. “Mass screening” is offered in various countries by the health authorities to all individuals within the age classes at risk of cancer, and is generally provided free of charge. Opportunistic or individual screening occurs when early detection is proposed to either symptomatic or asymptomatic patients who consult their doctor or attend a private health center.

Organized screening protocols can be evaluated through cost/benefit analyses, which do not include the individual opportunistic screenings. Questions arise about the respective impact (concurrent or complementary) of cancer prevention on population health during organized and individual screening interventions. Stomach cancer is still the first cause of cancer mortality in Japan, and since 1962 a national screening policy based on photofluorography has been in operation. This filter test is offered annually to individuals from the age of 40 years, and is performed in special radio-equipped trucks located in the municipalities. Individuals who test positive are offered gastroscopy. Each year more than 6 millions people go through this screening protocol.

Suzuki et al. [1] conducted a retrospective study on 1226 cases of early gastric cancer (EGC, confirmed by the pathologist after treatment), which were addressed to the National Cancer Center in Tokyo in 2001 - 2003. The first question was on the presence or absence of digestive symptoms: epigastric pain, nausea, vomiting. Symptoms were present in 512 (41.8 %), and 714 (58.2 %) were asymptomatic. The second question concerned diagnosis. In the symptomatic group, gastroscopy was the detection test for 91.6 %. In most cases this occurred in outpatient clinics (91.4 %), seldom in health check-up centers (7.6 %), and rarely (1 %) through mass screening. In the group of asymptomatic individuals, gastroscopy was more often used as the detection test than photofluorography (67.8 % and 32.2 %, respectively). Detection occurred in outpatient clinics in 44.8 %, in health check-up centers in 42.9 %, and through mass screening in only 12.3 %.

The results of this series are in line with national data collected in Japan. Statistics on cancer in Japan estimated that only 20 % of the eligible population accepted the governmental screening. Only 6 % of all cases of gastric cancer are detected by the government screening program, whereas 4.6 % are detected in health check-up centers, and all other cases are detected in private clinics for outpatients [2] [3]. In 2004, the Insurance Services of the Public Health system [4] confirmed the general use of gastroscopy in Japan. Each year, 7.8 million gastroscopies are carried out. Gastroscopy is performed annually in 9 % of Japanese individuals aged over 20 years. It is concluded that organized screening plays a marginal role in the early detection of gastric cancer in Japan. However, since the beginning of the national screening policy in this country, the global proportion of cases of stomach cancer diagnosed at the early superficial stage increased from 15 % to 50 %. As a consequence, the survival rate from stomach cancer is now about 45 % in Japan (at least twice the figure occurring in other countries of the world).

The temporal correlation between the screening policy and the improved survival from stomach cancer shows a good demonstration of possible misunderstanding in the evaluation of the benefit of an intervention. The organized screening policy is not failing and should be maintained in Japan; actually its benefit is linked to the massive diffusion of information about early detection of stomach cancer among medical professionals and the Japanese population. A similar analysis could be conducted on the actual benefit of an organized screening policy for colorectal cancer based on the fecal occult blood test.

Gastric cancer and infection-associated cancers

In recent years a number of bacteria or viruses, have been identified as a cause of cancer. Infection-associated tumors may be more accessible to primary prevention than those caused by other environmental factors, and predominantly occur in countries of the developing world.

A worldwide estimate of the cases of cancer attributable to infection in the year 2002 has been calculated by Parkin [5]. His analysis was based on the International Agency for Research on Cancer (IARC) database Globocan 2002 [6], which estimates the number of incident cancer in 2002 by country, age group, and sex. Data from cancer registries listed in volume VIII of Cancer Incidence in Five Continents [7] were also used for Burkitt’s lymphoma, oropharyngeal and anogenital cancer, and Kaposi’s sarcoma. For each cancer site two categories of figures were obtained: the relative risk of exposure to the infectious agent, and the prevalence of the tumor in this region of the world. The fraction of cases attributable to infection in the year 2002 was based on the formula of Cole and MacMahon [8] and adapted to the calculation of risk factors. The results were presented separately for developed countries and developing countries of the world using the United Nations population division of the world into 20 regions. Viruses were the most frequent infectious agents acknowledged as the cause of human malignancies, accounting for 12.1 % of all cancer cases. Viruses include hepatitis B virus (HBV) and hepatitis C virus (HCV; hepadnavirus and flavivirus) for liver cancer [9] (hepatocarcinoma), human papillomavirus (HPV) for cancer of the cervix and other sites (anogenital, oropharynx), Epstein-Barr virus (herpesvirus) for nasopharynx carcinoma, Burkitt’s lymphoma, Hodgkin lymphoma, adult T cell lymphoma (HTLV-1 retrovirus), and Kaposi’s sarcoma (human herpesvirus 8 [HHV8]). A single bacterium - Helicobacter pylori - was acknowledged as a causal factor for noncardia stomach cancer and for gastric lymphoma, and accounted for 5.6 % of all cancer cases; cancer at the cardia was not related to H. pylori infection. Parasitic agents were less frequent causes and accounted for only 0.1 % of all cancer cases; Schistosomia haematobium (blood fluke) was a causal factor for bladder carcinoma, and Opisthorchis viverrini (liver fluke) for bile duct carcinoma. Worldwide, an estimated 1.9 million cases were attributed to infectious agents in 2002; this represents 17.8 % of all cancers. The proportion was higher in developing (26.3 %) than in developed (7.7 %) countries.

Noncardia stomach cancer was the most frequent infection-associated cancer. In 2002 the number of cases attributable to H. pylori infection was 592 000 (5.6 % of all cancer), of which 400 000 were in developing countries. The calculation was based on the average prevalence of infection in middle-aged individuals (74 % in developing countries and 58 % in developed countries), and on the global average proportion of noncardia cases of stomach cancer (80 % in men and 87 % in women). In conclusion H. Pylori was a causal factor for 61.4 % of all stomach cancer in developed countries and for 64.4 % of all stomach cancer in developing countries.

The second infection-associated cancer in frequency was liver cancer; in 2002 the number of cases attributable to HBV and HBC was 523 000 (5.2 % of all cancer), of which 475 000 were in developing countries. The third infection-associated cancer was cervical cancer; in 2002 the number of cases attributable to HPV was 492 800 (4.9 % of all cancer), of which 409 400 were in developing countries.

Prevention of the large number of cases of noncardia gastric cancer and hepatocarcinoma attributable to infectious factors relies on the control of the responsible agents - hepatitis viruses and H. pylori. For hepatocarcinoma, primary prevention by controlling the infectious factor plays the major role. Vaccination is the most effective procedure in the prevention of hepatitis B; prevention of hepatitis C is based on an integrated strategy including screening of blood donations, safe injection practices, and avoidance of unnecessary injections. For noncardia stomach cancer, unplanned prevention occurs through changes in lifestyle, with the result that the incidence of stomach cancer declines worldwide. The Westernized lifestyle adopted by young generations includes multifactorial changes in diet and in hygiene, which lead to a reduction of H. pylori infection. In the Western world, screening for stomach cancer does not deserve a public health policy. With respect to primary prevention there is debate about whether the best approach is through dietary change, or eradication of H. pylori.

New trends in the treatment of gastric cancer: laparoscopic resection assisted by sentinel node navigation

In Japan the overall 5-year relative survival rate from gastric cancer is about 45 %, a value much higher than in Western countries. The country adopted a generalized policy of early detection, and near to 50 % of cases of gastric cancer are now detected at the early stage. EGC is staged clinically as a superficial tumor T1 (cT1). Staging of EGC is based on the superficial endoscopic appearance (and echoendoscopy if performed). There is a distinction between intramucosal cancer, which is almost always N0 and submucosal cancer, which is N+ in 15 % to 20 % of cases. Another factor explaining the high survival after surgical treatment is the systematic hunting for metastatic lymph nodes. The extensive D2 dissection of lymph nodes usually accompanies gastrectomy. Recently an innovative and less invasive strategy for the surgical treatment of gastric cancer has been developed in Japan.

Tangoku [10] has reviewed the current status of sentinel node navigation and its application in the gastrointestinal tract. The objective of sentinel node navigation is to minimize lymph node dissection. A regional lymphography is performed before the surgical treatment. Tracers are injected in the vicinity of the tumor, to be trapped in lymph nodes. Colored dyes (Evans Blue or indocyanine green) will cross the small openings in the single layer of endothelial cells of a lymph capillary and stain the nymph nodes. After injection of a radiolabeled tracer, the hot lymph node is detected with a probe. The technetium 99 is used either as small (80 nm) colloid particles with albumin or as larger particles with sulfur. The lymphography will detect the lymph node nearest to the tumor (N1 sector), which is called a sentinel lymph node. The node is excised and submitted to extemporaneous analysis. If there is no metastatic invasion lymphadenectomy is not performed; if the lymph node is metastatic lymphadenectomy is performed. False negatives occur either if the metastatic cells skip the N1 nodes and are trapped in more distal nodes (N2 territory) or if the sentinel node with a massive neoplastic invasion is not able to trap the colloid.

Sentinel node navigation was first applied to the treatment of melanoma and is now routinely used for breast cancer, where the method proved reliable and specific even for sentinel lymph nodes with micrometastases. For the surgery of gastric cancer the systematic practice of D2 lymphadenectomy is recommended in Japan as having a higher curative objective. On the other hand, the procedure is longer with a higher morbidity, and is not justified if the tumor is N0 or N1. This is why the clinical utility of sentinel node navigation is now proposed. Previous to the intervention an endoscopic submucosal injection of the dye or the radiotracer is performed near the stomach cancer. During the intervention (open abdomen or laparoscopy), the sentinel lymph node is localized and removed for study. The procedure is still on trial because the lymphatic drainage of the stomach is more complex than for ectodermal tumors, and there is a higher percentage (up to 20 %) of false negatives with skipped metastases than for breast cancer [11].

The trend for less invasive surgery for stomach cancer also includes the development of laparoscopy assisted resection [12] [13], which can be combined with the sentinel node navigation. Two interventions are possible: a wedge resection limited to the neoplastic area and a distal gastrectomy, which can be completed by a D2 lymphadenectomy. The focus is on the treatment of EGC as detected by clinical staging. Tumors classified as intramucosal cancer can be treated by laparoscopy with wedge resection of the corresponding segment of the gastric wall without lymphadenectomy. Tumors classified as submucosal cancer are treated by laparoscopy completed by sentinel node navigation. If there is invasion of the sentinel node, distal gastrectomy is performed, completed by D2 lymphadenectomy. If the sentinel node is negative, the treatment is a simple wedge resection. The legitimacy of this strategy depends on the reliability of the navigation technique applied to gastric cancer. The final results of two prospective multicenter trials are not yet available. One is conducted by the Japan Clinical Oncology Group (JCOG) using indocyanine green dye as a tracer. The other is conducted by the Japanese Society of Sentinel Node Navigation Surgery (SNNS), using a dual tracer with blue dye and a radioactive colloid.

Mucosectomy in early gastric cancer: the risk of lymph node metastases

Treatment of EGC by endoscopic mucosal resection (EMR), with or without mucosectomy (a less invasive option than surgical resection), ensures the preservation of gastric function. The legitimacy of a procedure restricted to local destruction of the tumor requires the absence of metastatic lymph nodes. The small risk of positive lymph nodes associated to EGC depends on the endoscopic appearance and on the depth of invasion in the gastric wall (mucosa or submucosa).

These factors have been analyzed in the Department of Gastroenterological Surgery in Kumamoto, Japan [14]. A retrospective review of lymph node metastases was conducted in 278 patients with EGC who were treated by gastric resection completed by D2 lymphadenectomy. In 156 patients, the cancer was classified as intramucosal, and 41 patients also had an ulcer, whereas 115 had no ulcer. The lymph nodes were negative in all 115 patients without ulcer, whereas six of the 41 with an ulcer had N1-positive lymph nodes. In 122 patients, the cancer was classified as submucosal, with 28 patients (23 %) having positive lymph nodes. Among the 122 submucosal cancers, 20 were staged as sm1 (invasion of submucosa in the upper third), and lymph nodes were positive in five of them (25 %). In conclusion the legitimacy of EMR is confirmed for nonulcerated intramucosal EGC. There is a significant risk of positive lymph nodes for mucosal cancer with an ulcer larger than 2 cm and for submucosal EGC. Then, even when the invasion of the submucosa is superficial, it is recommended to complete the treatment by gastric resection and lymphadenectomy.

This surgical series can be compared to series based on endoscopic treatment. The endoscopic resection is performed either through the EMR technique or through the endoscopic submucosal dissection (ESD) technique. ESD will remove neoplastic areas larger than 3 cm in a single piece (i. e. “en bloc” resection) [15]. A circumferential incision of the mucosa is performed at the periphery of the lesion with an insulation tip diathermy knife (IT knife). Then the submucosal dissection is performed under the control of direct vision with the help of a transparent cap.

ESD takes more time than conventional EMR, requires greater endoscopic skill, and has a higher risk of complications (hemorrhage or perforation). On the other hand, the “en bloc” resection of large lesions gives better results (complete response and rate of recurrence) than the “piece-meal” technique [16]. The ESD technique tends to be the reference procedure in the endoscopic treatment of EGC larger than 15 mm. The results of 411 EMR and 303 ESD procedures, carried out in 655 consecutive patients with EGC were reported in a multicenter retrospective study conducted in 11 Japanese institutions, with a median follow-up of 3.2 years [17]. ESD compared with EMR was characterized by a higher proportion of “en bloc” resection (92.7 % vs. 56 %) and complete response (73.6 % vs. 61.1 %), but a higher incidence of perforation (3.6 % vs. 1.2 %), which was managed by endoscopy; there was no procedure mortality. The 3-year cumulative rate without residual tumor and without recurrence was higher following ESD (97.6 %) than after EMR (92.5 %). The 3-year survival was 94.4 % for ESD and 99.2 % for EMR.

In summary this multicenter study of endoscopic resection confirms the excellent results and high 3-year survival rate after EMR and ESD. This means that the clinical staging (endoscopic appearance and echoendoscopy) from which the depth of invasion of EGC is determined prior to endoscopic treatment is sufficiently reliable. On the other hand, the analysis of operative specimens confirms that although EMR is an adequate treatment for intramucosal EGC, for submucosal EGC, the local destruction of the tumor should be completed by lymphadenectomy when metastatic lymph nodes are present. After endoscopic treatment of EGC a careful control of the resected specimen by the pathologist is required to assess the completeness of the resection and the level of the invasion in the gastric wall. Concerning the technique of mucosectomy, the complete dissection of the submucosa to the level of the muscular layer during ESD allows a better evaluation of submucosal invasion in the operative specimen. When invasion in the submucosa is confirmed, many experts in Japan recommend gastric resection together with lymphadenectomy.

Narrow-band imaging in the exploration of Barrett’s esophagus

The increased risk of cancer in columnar metaplasia of the esophagus (Barrett’s) has been correlated with the presence of areas with a specialized type of intestinal metaplasia. Columnar metaplasia is easily detected during endoscopy as a segment of velvety, reddened mucosa, proximal to the esophagogastric junction and marked by the upper end of the gastric folds and/or the distal end of palisade vessels. On the other hand, areas of intestinal metaplasia or flat neoplasia, will often remain undetected in the standard endoscopic image. This is why a protocol of randomly obtained biopsies (Seattle protocol) has been proposed in diagnostic or surveillance strategies for Barrett’s esophagus. Unfortunately this time-consuming and costly protocol is not completely reliable because it controls only a small fraction of the target area. Several attempts have been made to improve the detection using high-resolution endoscopy, with an acceptable reliability, through selective targeted biopsies. Different classifications of the magnified surface pattern have been proposed, and further progress is expected from the coupling of narrow-band imaging (NBI) to magnification with an optical zoom.

The present study of the fine mucosal pattern of columnar metaplasia in the esophagus is published by the Department of Endoscopy in the Jukei School of Medicine in Tokyo [18]. In this study the fine architecture of the mucosal surface with NBI coupled to magnification is compared to that of the capillary network. The authors analyzed columnar metaplasia with the video-endoscope model GIF-Q 240Z (Olympus Optical Co., Ltd, Tokyo, Japan) in 58 patients: 51 had a short segment (5 - 29 mm) of columnar metaplasia and seven had a long segment (30 mm and more). A superficial neoplastic area was present in four patients. The fine structure of the mucosa in NBI coupled to magnification was assessed at 184 sites in the 51 patients with a short segment and at 33 sites in the seven patients with a long segment. The mucosal pattern was categorized into one of five categories (1 - round or oval; 2 - long and straight; 3 - villous; 4 - cerebriform; 5 - irregular). Prior to biopsy, the capillary pattern at each site was also analyzed (1 - honeycomb; 2 - vine-like; 3 - coiled; 4 - DNA spiral-like; 5 - irregular). Tissue samples were obtained at each site and examined by the pathologist for correlation with the endoscopic image. Out of the 217 corresponding tissue samples, six showed neoplasia with intramucosal cancer, and 211 were classified as either fundic epithelium, cardiac epithelium with or without foveolar hyperplasia, or specialized intestinal metaplasia.

Endoscopic diagnosis of intestinal metaplasia: correspondence of histology to the fine mucosal pattern was highest in cerebriform type (61 %) and in villous type (43 %). Correspondence of histology to the capillary pattern was highest in DNA spiral-like type (88 %). Adding the fine mucosal pattern and the capillary pattern increased significantly the diagnostic value of NBI coupled to magnification, which was expressed as a ratio of sensitivity to specificity.

Endoscopic diagnosis of cancer: correspondence of histology to both the fine mucosal pattern and the capillary pattern was highest with the irregular type. A 100 % correspondence was confirmed in all six tissue specimens with cancer.

In addition, a 3D study of the capillary network was conducted in 45 biopsy specimens from 12 patients by confocal laser scanning microscopy (LSM 510 Carl Zeiss Ltd., Tokyo, Japan), after immunostaining of the capillaries by a monoclonal antibody (CD34) and a fluorescein-labeled immunoglobulin G antibody. The capillary pattern observed in NBI was extremely similar to the 3D reconstruction achieved by the confocal laser.

The correlation established in this study between the pattern of the magnified mucosal surface and that of the capillary network further increases the reliability of the endoscopic image in the detection of intestinal metaplasia and neoplasia. The same group had previously made a similar observation for the endoscopic diagnosis of EGC [19]. Further advances may also lead to the substitution by targeted biopsies of the time-consuming protocol of random biopsies. Other techniques [20] [21] have been proposed to improve the topographic diagnosis of intestinal metaplasia or neoplasia in columnar metaplasia. However, these techniques require sophisticated material and technology, whereas endoscopic imaging with magnification coupled to NBI is increasingly accessible to routine procedures.

Competing interests: None

References

• 1 Suzuki H, Gotoda T, Sasako M, Saito D. Detection of early gastric cancer: misunderstanding the role of mass screening.  Gastric Cancer. 2006;  9 315-319
  CrossrefPubMedSearch in Google Scholar
• 2 Ministry of Health Labour and Welfare of Japan .Report on Health Center Activities and Health Services for the Aged. Tokyo; Administrative document 2003: 550-554
  Search in Google Scholar
• 3 Foundation for promotion of cancer research, .Cancer statistics in Japan - 2005. Tokyo; Administrative document 2005
  Search in Google Scholar
• 4 Statistics and Information Department, .Survey of the medical care activities in Public Health Insurance 2004, Vol 2. 
• 5 Parkin D M. The global health burden of infection-associated cancers in the year 2002.  Int J Cancer. 2006;  118 3030-3044
  CrossrefPubMedSearch in Google Scholar
• 6 Ferlay J, Bray F, Pisani P, Parkin D M. GLOBOCAN 2002. Cancer incidence, mortality and prevalence. IARC CancerBase No. 5, version 2.0. Lyon; IARC Press 2004
  Search in Google Scholar
• 7 Parkin D M, Whelan S L, Ferlay J. et al (eds) .Cancer incidence in five continents, Vol. VIII. IARC Scientific Publications No. 155. Lyon; IARC 2002
  Search in Google Scholar
• 8 Cole P, MacMahon B. Attributable risk percent in case-control studies.  Br J Prev Soc Med. 1971;  25 242-244
  PubMedSearch in Google Scholar
• 9 Raza S A, Clifford G M, Franceschi S. Worldwide variation in the relative importance of hepatitis B and hepatitis C viruses in hepatocellular carcinoma: a systematic review.  Br J Cancer. 2007;  96 1127-1134
  CrossrefPubMedSearch in Google Scholar
• 10 Tangoku A, Seike J, Nakano K. et al . Current status of sentinel lymph node navigation surgery in breast and gastrointestinal tract.  J Med Invest. 2007;  54 1-18
  CrossrefPubMedSearch in Google Scholar
• 11 Liu C G, Lu P, Lu Y. et al . Distribution of solitary lymph nodes in primary gastric cancer: a retrospective study and clinical implications.  World J Gastroenterol. 2007;  13 4776-4780
  PubMedSearch in Google Scholar
• 12 Nomura S, Kaminishi M. Surgical treatment of early gastric cancer.  Dig Surg. 2007;  24 96-100
  PubMedSearch in Google Scholar
• 13 Kitano S, Shiraishi N, Uyama I. et al . Japanese Laparoscopic Surgery Study Group. A multicenter study on oncologic outcome of laparoscopic gastrectomy for early cancer in Japan.  Ann Surg. 2007;  245 68-72
  CrossrefPubMedSearch in Google Scholar
• 14 Ishikawa S, Togashi A, Inoue M. et al . Indications for EMR/ESD in cases of early gastric cancer: relationship between histological type, depth of wall invasion, and lymph node metastasis.  Gastric Cancer. 2007;  10 35-38
  CrossrefPubMedSearch in Google Scholar
• 15 Takeuchi Y, Uedo N, Iishi H. et al . Endoscopic submucosal dissection with insulated-tip knife for large mucosal early gastric cancer: a feasibility study (with videos).  Gastrointest Endosc. 2007;  66 186-193
  CrossrefPubMedSearch in Google Scholar
• 16 Gotoda T. Endoscopic resection of early gastric cancer.  Gastric Cancer. 2007;  10 1-11
  CrossrefPubMedSearch in Google Scholar
• 17 Oda I, Saito D, Tada M. et al . A multicenter retrospective study of endoscopic resection for early gastric cancer.  Gastric Cancer. 2006;  9 262-270
  CrossrefPubMedSearch in Google Scholar
• 18 Goda K, Tajiri H, Ikegami M. et al . Usefulness of magnifying endoscopy with narrow band imaging for the detection of specialized intestinal metaplasia in columnar-lined esophagus and Barrett’s adenocarcinoma.  Gastrointest Endosc. 2007;  65 36-46
  CrossrefPubMedSearch in Google Scholar
• 19 Nakayoshi T, Tajiri H, Matsuda K. et al . Magnifying endoscopy combined with narrow band imaging system for early gastric cancer: correlation of vascular pattern with histopathology.  Endoscopy. 2004;  36 1080-1084
  Thieme ConnectPubMedSearch in Google Scholar
• 20 Evans J A, Bouma B E, Bressner J. et al . Identifying intestinal metaplasia at the squamocolumnar junction by using optical coherence tomography.  Gastrointest Endosc. 2007;  65 50-56
  CrossrefPubMedSearch in Google Scholar
• 21 Kara M A, Da Costa R S, Streutker C J. et al . Characterization of tissue autofluorescence in Barrett’s esophagus by confocal fluorescence microscopy.  Dis Esophagus. 2007;  20 141-150
  CrossrefPubMedSearch in Google Scholar

R. Lambert, MD

Screening Group IARC

150 cours Albert Thomas
Lyon 69372
cedex 8
France

Fax: +33-4-72738518

Email: lambert@iarc.fr

References

• 1 Suzuki H, Gotoda T, Sasako M, Saito D. Detection of early gastric cancer: misunderstanding the role of mass screening.  Gastric Cancer. 2006;  9 315-319
  CrossrefPubMedSearch in Google Scholar
• 2 Ministry of Health Labour and Welfare of Japan .Report on Health Center Activities and Health Services for the Aged. Tokyo; Administrative document 2003: 550-554
  Search in Google Scholar
• 3 Foundation for promotion of cancer research, .Cancer statistics in Japan - 2005. Tokyo; Administrative document 2005
  Search in Google Scholar
• 4 Statistics and Information Department, .Survey of the medical care activities in Public Health Insurance 2004, Vol 2. 
• 5 Parkin D M. The global health burden of infection-associated cancers in the year 2002.  Int J Cancer. 2006;  118 3030-3044
  CrossrefPubMedSearch in Google Scholar
• 6 Ferlay J, Bray F, Pisani P, Parkin D M. GLOBOCAN 2002. Cancer incidence, mortality and prevalence. IARC CancerBase No. 5, version 2.0. Lyon; IARC Press 2004
  Search in Google Scholar
• 7 Parkin D M, Whelan S L, Ferlay J. et al (eds) .Cancer incidence in five continents, Vol. VIII. IARC Scientific Publications No. 155. Lyon; IARC 2002
  Search in Google Scholar
• 8 Cole P, MacMahon B. Attributable risk percent in case-control studies.  Br J Prev Soc Med. 1971;  25 242-244
  PubMedSearch in Google Scholar
• 9 Raza S A, Clifford G M, Franceschi S. Worldwide variation in the relative importance of hepatitis B and hepatitis C viruses in hepatocellular carcinoma: a systematic review.  Br J Cancer. 2007;  96 1127-1134
  CrossrefPubMedSearch in Google Scholar
• 10 Tangoku A, Seike J, Nakano K. et al . Current status of sentinel lymph node navigation surgery in breast and gastrointestinal tract.  J Med Invest. 2007;  54 1-18
  CrossrefPubMedSearch in Google Scholar
• 11 Liu C G, Lu P, Lu Y. et al . Distribution of solitary lymph nodes in primary gastric cancer: a retrospective study and clinical implications.  World J Gastroenterol. 2007;  13 4776-4780
  PubMedSearch in Google Scholar
• 12 Nomura S, Kaminishi M. Surgical treatment of early gastric cancer.  Dig Surg. 2007;  24 96-100
  PubMedSearch in Google Scholar
• 13 Kitano S, Shiraishi N, Uyama I. et al . Japanese Laparoscopic Surgery Study Group. A multicenter study on oncologic outcome of laparoscopic gastrectomy for early cancer in Japan.  Ann Surg. 2007;  245 68-72
  CrossrefPubMedSearch in Google Scholar
• 14 Ishikawa S, Togashi A, Inoue M. et al . Indications for EMR/ESD in cases of early gastric cancer: relationship between histological type, depth of wall invasion, and lymph node metastasis.  Gastric Cancer. 2007;  10 35-38
  CrossrefPubMedSearch in Google Scholar
• 15 Takeuchi Y, Uedo N, Iishi H. et al . Endoscopic submucosal dissection with insulated-tip knife for large mucosal early gastric cancer: a feasibility study (with videos).  Gastrointest Endosc. 2007;  66 186-193
  CrossrefPubMedSearch in Google Scholar
• 16 Gotoda T. Endoscopic resection of early gastric cancer.  Gastric Cancer. 2007;  10 1-11
  CrossrefPubMedSearch in Google Scholar
• 17 Oda I, Saito D, Tada M. et al . A multicenter retrospective study of endoscopic resection for early gastric cancer.  Gastric Cancer. 2006;  9 262-270
  CrossrefPubMedSearch in Google Scholar
• 18 Goda K, Tajiri H, Ikegami M. et al . Usefulness of magnifying endoscopy with narrow band imaging for the detection of specialized intestinal metaplasia in columnar-lined esophagus and Barrett’s adenocarcinoma.  Gastrointest Endosc. 2007;  65 36-46
  CrossrefPubMedSearch in Google Scholar
• 19 Nakayoshi T, Tajiri H, Matsuda K. et al . Magnifying endoscopy combined with narrow band imaging system for early gastric cancer: correlation of vascular pattern with histopathology.  Endoscopy. 2004;  36 1080-1084
  Thieme ConnectPubMedSearch in Google Scholar
• 20 Evans J A, Bouma B E, Bressner J. et al . Identifying intestinal metaplasia at the squamocolumnar junction by using optical coherence tomography.  Gastrointest Endosc. 2007;  65 50-56
  CrossrefPubMedSearch in Google Scholar
• 21 Kara M A, Da Costa R S, Streutker C J. et al . Characterization of tissue autofluorescence in Barrett’s esophagus by confocal fluorescence microscopy.  Dis Esophagus. 2007;  20 141-150
  CrossrefPubMedSearch in Google Scholar

R. Lambert, MD

Screening Group IARC

150 cours Albert Thomas
Lyon 69372
cedex 8
France

Fax: +33-4-72738518

Email: lambert@iarc.fr