Seminars in Diagnostic Pathology (2008) 25, 274-287 Molecular targets and biological modifiers in gastric cancer Fátima Carneiro, MD, PhD,a,b,c Carla Oliveira, PhD,a,b Marina Leite, PhD,a Raquel Seruca, MD, PhDa,b From the aInstitute of Molecular Pathology and Immunology of the University of Porto (IPATIMUP), Porto, Portugal; b Medical Faculty of the University of Porto, Porto, Portugal; and the c Department of Pathology, Hospital S. João, Porto, Portugal. KEYWORDS Gastric cancer; Molecular targets; Mutation; E-cadherin (CDH1); Microsatellite instability The overall survival of gastric cancer patients remains poor despite efforts and advances in its prevention, diagnosis, and treatment. The development of new therapies is crucial for the effective control of this disease. An increasing number of genetic and epigenetic alterations have been associated with distinct histological types of gastric cancer. In this review, we will discuss the involvement of E-cadherin, EGFR, ERBB2, MMR genes, KRAS, and PIK3CA in the development and progression of gastric cancer and their role as biomarkers or as novel putative targets for therapy. © 2008 Elsevier Inc. All rights reserved. Gastric cancer (GC) is one of the most common forms of cancer in Europe.1 In 2000, there were 192,000 new cases with 158,000 deaths. In Southern Europe, Portugal shows the highest incidence of GC (as estimated by GLOBOCAN 2002).2 Despite advances in prevention and diagnosis, the overall outcome of the patients has remained poor. The overall 5-year survival is around 23%. This is largely due to the advanced stage of disease at presentation, so that most patients only receive palliative treatment. Moreover, in the various multimodal therapy regimens that are used to improve the patients’ prognosis, the majority of patients do not respond to treatment. In summary, the prediction of therapy response in GC is very limited, and the high prevalence of incurable disease produces a large burden on patients, which has a huge effect on health care resources. In this review, we will focus on putative alternative therapeutic strategies based on the identification of molecular targets and biological modifiers in GC. Address reprint requests and correspondence: Fátima Carneiro, MD, PhD, Rua Dr Roberto Frias s/n, 4200-465 Porto, Portugal. E-mail address: fcarneiro@ipatimup.pt. 0740-2570/$ -see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1053/j.semdp.2008.07.004 Overview of GC Carcinomas of the stomach are very heterogeneous from the morphologic standpoint. This heterogeneity is amply reflected in the diversity of histopathological classifications on record, which are based on different approaches: histologic profile, degree of differentiation, pattern of growth, and histogenesis.3 Laurén’s classification4 individualizes two main types of gastric carcinoma—intestinal and diffuse—which display different clinicopathologic profiles and often occur in distinct epidemiologic settings. Intestinal carcinoma is more prevalent in elderly persons of the masculine gender, whereas diffuse carcinoma tends to occur in younger individuals, mainly females, and frequently depicts hereditary conditioning. The incidence of intestinal carcinomas is steadily decreasing in most countries, in contrast to diffuse carcinomas, whose incidence is quite stable or even increasing.5 The classification proposed by Carneiro and coworkers6 highlights the heterogeneity of gastric carcinoma, individualizing mixed and solid carcinomas. By univariate analysis, the survival of the patients with mixed carcinomas Carneiro et al Gastric Cancer was shown to be significantly worse than that of patients with pure histologic type of tumors. In a multivariate analysis using the Cox’s model, Carneiro’s classification kept its independent prognostic significance, emerging as the second most important factor after TNM staging and before venous invasion.6 About 90% of GC cases appear in a sporadic setting, whereas familial clustering is observed in the remaining 10%.7 Of these, only 1% to 3% is hereditary.8 Among the cases with familial aggregation of GC, several situations can be identified: cases, in which the histopathology of the tumors is unknown, simply designated as familial GC (FGC) and cases in which it is possible to have information on the histopathological type of one or more GCs. The latter group encompasses specific syndromes/diseases as follows: hereditary diffuse GC (HDGC), familial diffuse GC (FDGC), and familial intestinal GC (FIGC).9 Histopathological features of HDGC As fully described below, truncating germline E-cadherin gene (CDH1) mutations were identified in 1998 as a causal genetic defect for HDGC by Guilford and coworkers.10 Shortly afterward, the International Gastric Cancer Linkage Consortium (IGCLC) defined the following criteria for identification of HDGC families: (1) two or more documented cases of diffuse GC in first/second-degree relatives, with at least one diagnosed before the age of 50; or (2) three or more cases of documented diffuse GC in first/seconddegree relatives, independent of age.11 Furthermore, the IGCLC recommended that carriers of CDH1 truncating germline mutations should be offered the possibility of 275 being submitted to intensive screening and/or prophylactic gastrectomy. Current knowledge on the morphologic steps underlying the development of HDGC stems from detailed studies performed in stomachs that were totally mapped, encompassing prophylactic gastrectomy specimens, and total gastrectomies performed in patients referred from chromoendoscopic surveillance programs.12-18 In most specimens, at least one focus of early invasive diffuse GC was identified. In North American families, early invasive carcinoma was not restricted to any topographic region in the stomach: foci were identified from cardia to prepyloric region, without evidence of antral clustering.12,13 In New Zealand Maori families, a predilection was observed for the occurrence of early invasive carcinomas for the distal stomach and the body-antral transitional zone.15,16 Reasons for the different anatomical localization of the cancer foci in the aforementioned studies remain to be clarified. As precursors of the invasive cancers, two distinct types of lesions were identified in prophylactic gastrectomies: (1) in situ signet ring cell carcinoma, corresponding to the presence of signet ring cells within basal membrane, generally with hyperchromatic and depolarized nuclei (Figure 1); and (2) pagetoid spread of signet ring cells below the preserved epithelium of glands/foveolae (Figure 1).14 Model of development of HDGC On the basis of the findings in prophylactic gastrectomies, a model for the development of diffuse GC in E-cadherin mutation carriers was proposed,14 encompassing the following lesions: mild nonatrophic gastritis, in situ signet ring cell carcinoma, pagetoid spread of Figure 1 Morphological steps in the development of HDGC encompassing in situ carcinoma, pagetoid spread, and early intramucosal carcinoma (upper panel). Sequencing analysis of a germline mutation in exon 14 of CDH1 gene (lower panel, left). CDH1 promoter methylation in a tumor from the same family, determined in bisulphite treated DNA by PCR and sequencing (lower panel, right). 276 Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 signet ring cells, and invasive carcinoma (Figure 1). E-cadherin immunoexpression was shown to be reduced or absent in early invasive gastric carcinomas, contrasting with the normal membranous E-cadherin expression in adjacent nonneoplastic mucosa, in keeping with a clonal origin of the cancer foci.13,14 It is accepted that the gastric mucosa in CDH1 germline mutation carriers is quite normal until the second CDH1 allele is inactivated. It is postulated that this downregulation occurs in multiple cells in the gastric mucosa, accounting for the multifocal tumor lesions. The role played by environmental factors (diet, Helicobacter pylori infection, carcinogens), ulceration, and gastritis as triggers of this downregulation event remains to be elucidated. E-cadherin and diffuse GC E-cadherin is a transmembrane glycoprotein critical for establishing and maintaining polarized and differentiated epithelia during development and in adult tissues. Adherens junctions (AJs) cluster, via homophilic interactions, through the extracellular domains of calcium-dependent E-cadherin molecules on the surface of homotypic neighbor cells.19 Experimental evidence supports a role for the E-cadherin complex, both in suppressing invasion of epithelial cells and metastasis formation.20 Somatic mutations in CDH1 gene have been identified in 40% to 83% of sporadic diffuse-type gastric cancers but not in sporadic intestinal-type gastric cancers.21 Germline mutations in CDH1 gene are the underlying genetic defect responsible for HDGC (OMIM no. 137215), an autosomal dominant syndrome characterized by high susceptibility to early-onset diffuse gastric carcinomas.10 Since then, many studies reported different types of CDH1 mutations in HDGC families (Figure 1), and it was calculated that approximately 30% to 40% of the HDGC families harbor CDH1 germline mutations.9,22 Heterozygous carriers of CDH1 germline mutations frequently remain asymptomatic at least until the second decade of life, when inactivation of the remaining wild type allele of the CDH1 gene occurs.23,24 This two-hit inactivation mechanism is believed to determine the initiation of diffuse GC, consistent with the two-hit inactivation model proposed by Knudson.25 CDH1 germline mutations found in HDGC families are spread along the whole sequence of the gene, without preferential hot-spots. About 80% of these mutations are nonsense, splice-site, and frameshift and are predicted to produce premature termination codons. The remaining 20% are missense mutations.9,22,26 Although little doubts exist about the inactivating function of truncating mutations, the pathogenic role of missense mutations remains under debate. CDH1 germline missense mutations In contrast to CDH1 germline truncating mutations, for which 80% disease-penetrance is estimated,27 missense mutations display a low-penetrance phenotype, with only few mutation carriers affected within pedigrees. This, together with the fact that these HDGC families are usually very small so that very few individuals are available for testing, has not allowed segregation analysis within E-cadherin germline missense mutations families. Lack of this clinical information is a limiting step to infer the pathogenic significance of missense mutations. In this regard, it was suggested by Fitzgerald and Caldas28 that the significance of CDH1 missense mutations should be assessed in at least four affected members in combination with functional and transcript analysis (to look for activation of cryptic splice sites). However, in some cases, this type of analysis is impossible to accomplish. To circumvent this limitation, our group developed an in vitro functional assay to infer the deleterious nature of CDH1 germline missense variants.29 In this experimental model, cells negative for E-cadherin expression are selected as model and engineered to stably express either the wild-type E-cadherin protein or its mutants. Slow aggregation on soft agar and cell invasion into matrigel or collagen matrix are used to assess the functional effect of CDH1 germline missense variants. Most missense mutations studied to date have a clear functional impact on E-cadherin function, supporting their pathogenic nature.30 Aside from the critical importance of these studies for the genetic counseling of HDGC families, the distinct stable cell lines harboring the set of missense mutations now available allow us to identify associated signaling pathways and putative molecular targets for therapy. E-cadherin second-hit inactivation mechanisms in GC In sporadic diffuse gastric carcinomas, somatic CDH1 mutations are identified in a high proportion of cases, but tumors do not commonly show loss of heterozigosity (LOH) at the CDH1 locus.31 Later, it was shown that CDH1 promoter methylation is the second-hit inactivation mechanism in more than half of the sporadic diffuse gastric carcinoma cases harboring CDH1 mutations.32 We analyzed a series of 23 sporadic gastric carcinomas for the presence of CDH1 mutations, CDH1 promoter methylation, and LOH. CDH1 mutations were detected in 56.3% of diffuse gastric carcinomas and in none of the seven intestinal gastric carcinomas. CDH1 promoter methylation was observed in 66.7% of the cases harboring CDH1 mutations. In diffuse gastric carcinomas harboring CDH1 mutations, LOH was observed in a single case. These results show that CDH1 promoter methylation is the second-hit inactivation mechanism in more than half of the sporadic diffuse gastric carcinoma cases harboring CDH1 mutations. Carneiro et al Gastric Cancer In the setting of HDGC, few tumors were analyzed for second-hit inactivation mechanisms. Somatic structural alterations of CDH1 are uncommon mechanisms of CDH1 inactivation, and promoter hypermethylation (Figure 1) was shown to be the most common mechanism.23,24,33 Recently, we extended the characterization of the second hit of CDH1 inactivation to 14 HDGC families carrying different CDH1 germline mutations (unpublished data). The combined analysis of the results of this study and those from previous studies23,24,33 showed that CDH1 promoter hypermethylation was found as a sole event in approximately 40% of the HDGC tumors analyzed, and LOH was observed to be a rather more frequent event than was initially postulated. Moreover, LOH was found concomitantly with promoter hypermethylation in one-fifth of the cases and, importantly, metastatic lesions displayed preferentially LOH. The thorough characterization of the second-hit mechanisms of CDH1 inactivation in HDGC tumors is critical to distinguish those that are irreversible (LOH, somatic mutations) from those that are reversible, such as promoter hypermethylation, the latter putatively targeted for therapy with “epigenetic” drugs (demethylating agents). This information will be pivotal for the therapeutic management of HDGC patients. E-cadherin and non-sense-mediated decay (NMD) NMD is a conserved mRNA surveillance mechanism that typically degrades mRNAs harboring premature termination codons (PTCs), using the exon junction complex (EJC), a set of proteins recruited to mRNAs near exon– exon junctions after RNA splicing.34 Evidence suggests that NMD modulates the phenotype of numerous diseases.35,36 Because in HDGC the vast majority CDH1 germline mutations (80%) are truncating and generate PTCs,22,26 we decided to examine the impact of NMD in the phenotype of HDGC patients caused by such mutations.37 To test the role of NMD in the downregulation of truncating CDH1 transcripts, we performed allele-specific expression (ASE) analysis on normal gastric mucosa from asymptomatic individuals carrying five different truncating CDH1 mutations. Four of these mutations generate a PTC that has at least one exon– exon junction downstream (in the “NMD-competent” region), and the fifth mutation generates a PTC in the last exon (in the “NMD-incompetent” region). We found that the mRNAs harboring PTCs located at the predicted NMD competent region are, in fact, expressed at a lower abundance compared with the wild type, although presenting different degradation rates. In contrast, the mRNA carrying the mutant PTC at the predicted NMD-incompetent region was not degraded but, on the contrary, was four- to fivefold more abundant than the wild type. These results suggest that NMD may promote disease in carriers of CDH1 mutations that generate PTCs in the “NMD-competent” region. To 277 assess this hypothesis, we analyzed medical data from a large set of published HDGC families (a total of 264 patients) carrying truncating germline mutations22,26 in both NMD-competent and -incompetent regions of the CDH1 gene. This analysis revealed that there is a significant association (P ⫽ 0.03) between the ability to induce NMD and the age of onset of GC. Patients carrying PTCs in the NMD-competent region developed cancer earlier than patients carrying PTCs in the NMD-incompetent region. This notion is supported by a previous study of 4 families carrying the 2398delC mutation (located in the NMD-incompetent region) with a cumulative risk of acquiring GC of only 40% for males and 63% for females, which was considerably lower than for HDGC patients in general: 67% for males and 83% for females.26 The association of predicted NMD activity with earlier age of onset of GC, as well as the observation of lower cumulative risk in patients harboring a CDH1 mutation that escapes NMD, suggests that NMD may have a detrimental effect on the clinical progression of HDGC. The simplest explanation for this is that NMD decreases the level of transcripts encoding mutated but still functional E-cadherin proteins that retain tumor suppressor activity. In agreement with this, it has previously been shown that truncated Ecadherin molecules lacking the COOH-terminal region retain partial function.38 In the future, it will be important to determine whether truncated E-cadherin proteins generated by PTCs found in CDH1 mutation carriers retain partial tumor suppressor function. This information will be critical to determine the value of inhibiting NMD in HDGC patients as a means to treat or reduce the risk of diffuse gastric carcinoma. EGFR signaling activation by loss of E-cadherin function For a long time, the functional consequences of E-cadherin loss of function have been seen as a structural cell– cell adhesion disruption rather than as a loss of E-cadherindependent regulatory events. E-cadherin acts indeed as a cell membrane receptor,20,39 and many signaling molecules have been reported to interact with E-cadherin, namely the receptor tyrosine kinases (RTKs), which have been found to localize at the basolateral membrane of epithelial cells.40,41 RTK activity in resting, normal cells is tightly controlled, but RTKs can become potent oncoproteins. Of notice, the epidermal growth factor receptor (EGFR) has been reported to be involved in a bidirectional cross-talk with E-cadherin.42 Using a set of deletion constructs along the Ecadherin gene [E-cad Db-catenin, Ecad (AAA(764) and E-cad DNT], Qian and coworkers42 suggested that the interaction of E-cadherin with EGFR would require an intact extracellular E-cadherin domain. Recently, this result was confirmed for naturally disease occurring CDH1 mutations, which have been identified in patients with HDGC.43 278 Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 EGFR/E-cadherin heterodimer stability was disturbed by the presence of single point mutations in the extracellular domain (T340A and A634V) but not of intracellular Ecadherin alterations (P799R and V832M). Interestingly, this reduced stability of the heterodimers correlated with increased EGFR activation on ligand (EGF) stimulation, which indicates that E-cadherin exerts an inhibitory function on EGFR activity.43 EGFR activation results in the phosphorylation of specific tyrosine residues within its cytoplasmic tail, which serve as docking sites for various adaptor proteins and signaling molecules involved in the regulation of intracellular signaling pathways that promote cell proliferation and migration, among others.44 In particular, EGFR activation has been reported to be involved in cell migration by triggering the cytoskeleton reorganization.45 On the other hand, the Rho family of small GTPases, and in particular RhoA, Rac1, and Cdc42, play a pivotal role in the reorganization of the actin cytoskeleton, thereby regulating cell migration.46,47 Our group showed very recently a correlation between increased EGFR activation, increased RhoA activation, and increased cell migration in cells expressing extracellular CDH1 missense mutations identified in HDGC families.43 On pharmacological inhibition of EGFR, E-cadherin mutant-expressing cells showed a reversion of the motile phenotype, accompanied by a decrease in RhoA activation, supporting the idea that Rho-like proteins are downstream effectors of EGFR activation. From these experiments, we suggested that, although Ecadherin normally binds to EGFR blocking its availability to EGF and consequently its activation, in cells with extracellular mutations, this interaction is disrupted and EGFR becomes abnormally activated, stimulating a series of pathways that lead to aberrant chaotic movements. In patients carrying such CDH1 extracellular mutations, this mechanism is likely to contribute to tumor progression. Of notice, most somatic CDH1 mutations identified in sporadic diffuse GC are found to cluster between exons 7 and 10 encoding part of the extracellular E-cadherin domain,31 supporting the idea that loss of function of Ecadherin may also then contribute to the frequent liganddependent activation of EGFR in tumors and to EGFR signaling activation. E-cadherin and tumor cell survival Cell survival depends on signals from the environment, such as those provided by adhesion molecules that mediate contacts between cells or between cells and the extracellular matrix.48 If these interactions are disrupted, the cells are programmed to die.49,50 E-cadherin, aside from its role in cell-to-cell adhesion, was also shown to be involved in programmed cell death.51,52 Recently, Ferreira and coworkers53 demonstrated that, in vitro, loss of functional E-cadherin interferes with cell-tocell adhesion and renders cells more resistant to apoptotic stimuli. Cell lines stably expressing mutant E-cadherins are not only capable of invading but are able also to survive and grow in the absence of contact with other cells. The apoptotic stimulus used in this in vitro study was Paclitaxel (Taxol®; Brystol–Myers Squibb Company, New York, NY), a drug widely used in the treatment of advanced cancers, including epithelial tumors harboring E-cadherin dysfunction.53 Importantly, these results question the effectiveness of such treatment in these types of tumors and highlight the need for further research on the subject. The existence of an interplay between E-cadherin and the antiapoptotic factor bcl-2 was also demonstrated,53 supporting the hypothesis that E-cadherin is more than a simple mediator of cell contact and adhesion, but is also involved in the control of programmed cell death and thus has a dual role in cancer development. Receptor tyrosine kinases (EGFR and ERBB2) activation in GC RTKs, which are found to be aberrantly activated or overexpressed in a variety of tumors, represent promising targets for therapeutical intervention. EGFR and ERBB are members of the RTK superfamily of ERBB receptors, which are glycoproteins that consist of an extracellular domain where the binding of ligands takes place, a short lipophilic transmembrane domain, and an intracellular domain carrying the tyrosine kinase activity.54,55 EGFR gene is located on chromosome 7p12 and codes for a 170-kDa receptor. Aside from the above-mentioned mechanism dependent of E-cadherin loss, EGFR activation includes amplified copy number, structural rearrangements of the receptor, and activating mutations.56 EGFR somatic mutations cluster in the kinase domain of EGFR (exons 18-21) and cause ligand-independent activation of the receptor. Mutations and amplification occur in about 20% of the patients with nonsmall cell lung cancer (NSCLC) and correlate with the clinical response to tyrosine kinase inhibitors.57,58 In GC, EGFR mutations were recently reported by our group and occur in about 3% of the cases.59 In our study, mutation and increased copy number of EGFR showed a significant association with tumor size, and all cases with alterations in EGFR were spreading into the gastric wall (T2-T4), suggesting that alterations of this gene may confer an invasive behavior to neoplastic cells. The ERBB2 gene maps to 17q12-q21 and encodes a 185-kDa transmembranar tyrosine kinase receptor (p185).60,61 In breast carcinomas, ERBB2 functions as an oncogene62 and is a marker of poor prognosis.63 Apart from breast cancer, ERBB2 amplification has been identified in several types of cancer64-66 as well as in gastric carcinomas.67,68 The evaluation of ERBB2 status in gastric carcinoma, using immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) analyses showed a frequency of overexpression ranging from 8.2% to 22.6%, whereas am- Carneiro et al Gastric Cancer plification ranged from 3.8% to 14.8%.69-72 ERBB2 amplification/overexpression has been shown to correlate with features of biological aggressiveness, such as venous invasion.67 However, the clinical significance of ERBB2 amplification/overexpression has so far not been fully clarified, because most studies did not have follow-up data or lacked statistical power for that purpose.67-69,72,73 Two major classes of anti-RTK therapies are available: ectodomain-binding antibodies and small-molecule TK inhibitors that compete with ATP in the TK domain. Many of these therapies are either in clinical use or in advanced clinical trials. As an example, Erlotinib (Roche Pharmaceuticals, Basel, Switzerland) selectively and reversibly inhibits the TK activity of EGFR, competing with adenosine triphosphate for binding in the receptor’s TK domain. Erlotinib has been approved for the treatment of chemotherapyresistant advanced NSCLC patients, and in combination with Gemcitabine (Gemzar®; Eli Lilly and Company, Indianapolis, IN) for the treatment of advanced pancreatic cancer in patients who have not received previous chemotherapy.74 However, the clinical response to tyrosine kinase inhibitors in GC treatment is yet unknown. In breast cancer, ERBB2 amplification is a predictive marker for targeted therapy with the monoclonal antibody Trastuzumab (Herceptin®, Genentech, Inc., San Francisco, California, USA),75 which is currently used as adjuvant treatment in breast carcinoma patients with ERBB2 gene amplification/overexpression.76,77 Recently it was shown that the GC cell line N87 presenting ERBB2 amplification is sensitive to Trastuzumab.78 Furthermore, Rebischung and coworkers79 reported a case of a GC patient with ERBB2 overexpression who responded to a combination of chemotherapy and Trastuzumab. An international large-scale clinical study is currently being performed to compare the response to chemotherapy combined with Trastuzumab versus chemotherapy alone in gastric carcinoma patients with ERBB2 amplification. Mismatch repair genes and microsatellite instability (MSI) in GC Germline mutations in genes responsible for the maintenance of mismatch repair (MMR) are linked to cancer predisposition in patients from the hereditary non-polyposis colorectal cancer (HNPCC) syndrome.80 MSI phenotype or mutator phenotype is caused by defects in the MMR system responsible for the correction of mismatches that occur during DNA replication.81 MSI has been reported in many human sporadic cancers, and in sporadic gastric carcinoma this phenotype is found in about 20% of the cases (Figure 2).82-93 The recognition of GC as a component of HNPCC syndrome94 raised the hypothesis of MMR genes being involved in GC development. To date, no germline mutations have been described in MMR genes in familial GC. On the other hand, somatic mutations in MMR genes 279 are rare in sporadic gastric carcinomas. More recently, our group investigated the role of MMR defects in MSI sporadic GC.95 Twenty-nine sporadic GCs with high level of MSI were screened for somatic mutations in MLH1, MSH2, MSH6, MLH3, and MBD4, and only five truncating mutations (3 in MSH6, 1 in MLH3, and 1 in MBD4) and one missense mutation (MLH1) were identified.95 All truncating mutations were found in the coding poly-A tracts, thus suggesting that they result from the MSI phenotype rather than causing it.95 In sporadic gastric carcinomas with MSI, hypermethylation of the promoter region of MLH1 was demonstrated to be very frequent (75% in the experience of our group) and associated with decreased/absent protein expression of the MLH1 protein.88,92,93,96,97 In summary, somatic mutations in MMR genes are rare in sporadic GCs, and MLH1 hypermethylation (Figure 2) explains the development of most of cases belonging to this subtype of gastric carcinomas. Hypermethylation and GC It is well known that methylation of regulatory regions of genes acts as an important alternative to structural alteration for gene inactivation. Methylation of cytosines within CpG islands is also observed in physiological conditions as a common epigenetic event, such as chromosome X inactivation and ageing.98,99 In neoplasia, genome-wide epigenetic disturbances result in altered DNA methylation patterns. Genome-wide hypomethylation and selective hypermethylation of DNA sequences is a common event in several cancers and became recognized as a hallmark of human cancers.100 CpG island hypermethylation of normally unmethylated promoter regions correlates with loss of transcription.101 Hypermethylation of several gene promoters has been described in sporadic gastric carcinomas, namely in genes of the repair pathway, such as hMLH1 and MGMT, cell cycle regulators, such as the cyclin-dependent kinase inhibitorOp16, the mediator of epithelial cell growth COX2, and cell adhesion molecules, such as CDH1.32,100,102 Moreover, it was shown that hypermethylation of gene promoters increases along the pathway that evolves from chronic gastritis, intestinal metaplasia, and adenomas to carcinomas of the stomach.103,104 In our group, we searched for the promoter methylation status of the cancer-related genes hMLH1, CDH1, MGMT, and COX2 in sporadic GC.105 Hypermethylation was detected in all genes: about 30% for COX2 and hMLH1, and 50% to 60% for CDH1 and MGMT gene promoters. The high frequency of hypermethylation in CDH1 and MGMT genes suggests that the inactivation of these tumor-related genes may play a pivotal role in gastric tumorigenesis. The lower frequency of promoter hypermethylation detected in COX2 and hMLH1 in comparison to CDH1 and MGMT is likely to be related to the fact that, in primary GC, COX2 and hMLH1 hypermethyl- 280 Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 Figure 2 MSI status, hMLH1 promoter methylation and oncogenic mutations in sporadic gastric carcinoma. Examples of MSS and MSI gastric carcinomas; the evaluation of tumor microsatellite instability was performed using five quasimonomorphic mononucleotide repeats and pentaplex PCR (upper panel). Examples of hMLH1 promoter methylation by direct sequencing using flanking markers in bisulphite treated DNA (middle panel). Examples of oncogenic KRAS and PIK3CA mutations in MSI gastric carcinoma (lower panel). ation is associated only to a specific subset of gastric tumors. In primary gastric carcinomas, aberrant methylation of COX2 was described in 12% of the cases and linked with the loss of expression of COX2 mRNA.102,106 Regarding hMLH1, it is known that hypermethylation of its promoter region leads to MSI, as previously described in this review.93,97,104,107 The percentage of CDH1 promoter hypermethylation (51%) detected in this study is in accordance with previous results in sporadic gastric carcinomas.108 CDH1 promoter was found to occur frequently in diffuse gastric carcinoma (sporadic and hereditary) as second-hit inactivation mechanism but also in a percentage of cases in all histological types of GC.32,108,109 MGMT is a repair protein responsible for the removal of O(6)-alkyl adducts produced by several carcinogens, including N-nitrosomethylurea. In primary gastric carcinomas, as well as in many other types of malignancies, epigenetic silencing of the MGMT gene by promoter hypermethylation has been demonstrated.110-111 When MGMT is silenced, O(6)-alkyl adducts are not removed from the guanine nucleotides, leading the DNA polymerase to misread guanines as adenines. This will promote G-to-A transitions,112 increasing the mutation rate and leading to chromosomal instability.112 Through the analysis of different tumor tissues, Esteller and collaborators100 showed that, for each human cancer, there exists a unique profile of promoter hypermethylation, in which some gene changes are shared, whereas others are cancer type-specific. However, in our study, we found no preferential combination of hypermethylated loci, and no particular profile was noticed. Altogether, the results of our group and those from literature show that hypermethylation of several promoter regions occurs frequently in GC, reinforcing the idea that hypermethylation plays an important role in gastric carcinogenesis and inactivation of different gene promoters by hypermethylation is likely to lead to specific characteristics of the tumors. Carneiro et al Gastric Cancer MSI and clinicopathologic features of gastric carcinomas The clinicopathologic features of gastric carcinoma with MSI are distinct from those observed in microsatellite stable (MSS) gastric carcinomas.83,84,86,113 MSI gastric carcinomas tend to occur as large and expanding tumors of the distal stomach in relatively old patients, frequently displaying an intestinal (glandular) or atypical (solid) histotype; the frequency of lymph node metastases is low, regardless of the degree of wall invasion (usually occur with low pTNM stages); the survival curve of patients with MSI carcinomas is significantly better than that of patients with MSS carcinomas. The significant association between MSI phenotype of gastric carcinomas and the good outcome of patients may be ascribed, partly at least, to the close relationship between MSI status and staging, namely low pTNM staging of the tumors. The clinical significance of MSI was recently accessed in a large series of gastric carcinomas (n ⫽ 510).113 MSI was detected in 16% of the cancers and significantly correlated with long survival (P ⬍ 0.001). In multivariate analysis, MSI phenotype was identified as an independent factor (P ⫽ 0.005), adding prognostic information to TNM stage. The relative risk of death for MSI cancer patients was 0.6 (95% CI 0.4-0.8). Moreover, when grouped according to stage, stage II cancers showed a significant effect of MSI status on survival (P ⫽ 0.011; hazard ratio ⫽ 0.3; 95% CI 0.1-0.8). Summing up, several studies clearly demonstrate that microsatellite analysis in GC has clinical utility for prognostic evaluation of the patients. It is tempting to speculate that MSI tumors show a specific set of genes that, whenever mutated, somehow prevent or delay tumor dissemination and metastatization. Target gene mutations in GC MSI carcinomas accumulate hundreds of thousands of mutations within repetitive sequences in noncoding as well as in coding repeat sequences due to the high background of genetic instability that characterizes these tumors. It has been difficult to distinguish whether these alterations occur in genes whose inactivation contributes to tumor development (real target genes) or in genes that are inactivated by chance due to the hypermutability pathway (bystander genes) in MSI tumors. The real contribution of a large number of target genes has to rely on mutation frequencies and the functional role played by these genes, due to the lack of in vitro or in vivo functional suppressor studies. We contributed to define the molecular signatures of target genes in MSI gastric tumors. Initially, we performed a mutation screening in genes with coding mononucleotide repeats: TGF␤RII, IGFII R, BAX, and TCF-4.87,114 Mutations were almost exclusively detected in MSI tumors, thus confirming that the involvement of such genes is most likely 281 due to a mismatch repair deficiency. The higher frequency of mutations was verified in TGF␤RII gene (70%), suggesting that the alterations of the TGF␤RII gene occur as an earlier event than those of IGFIIR, BAX, or TCF-4 genes in gastric carcinogenesis. IGFIIR and BAX mutations were found in 25% and 30% of the MSI sporadic gastric carcinomas, respectively.87 The frequency of MSI tumors harboring TCF-4 gene mutations was very low (15%) when compared with the frequency of mutations in the other target genes and, moreover, never occurred as a single mutation. These findings suggest that TCF-4 is most likely to be a “bystander” gene in gastric carcinogenesis rather than a real target gene, as described in colorectal cancer.114 HPRT and CDH1 repetitive coding regions did not harbor mutations in MSI tumors (unpublished data). Mutation frequencies at 25 genes containing coding repeats were determined in MSI gastric carcinoma and compared with colorectal and endometrial MSI tumors.114 From these, TGF␤RII, BAX, IGFIIR, hMSH3, hMSH6, GRB-14, and RAD-50 genes all appeared to be real targets for instability in both gastric and colon MSI carcinomas. Recently, frameshift EPHB2 mutations were identified in 39% of MSI gastric carcinomas, showing that EPHB2 is selectively targeted in MSI GCs.115 Clinicopathologic features associated to target gene mutations in MSI GC An indirect way to elucidate the putative contribution of target genes to cancer development is to correlate the mutations in these genes with the clinicopathologic features of tumors and patients. Data from our group and others showed that TGF␤RII poly(A)10 tract is a real mutation target in MSI gastric carcinoma, and mutations in the TGF␤RII gene were significantly associated with the intestinal histotype. The resulting inactivating frameshift mutations are predicted to truncate the TGF␤ receptor, altering the TGF␤ pathway and providing a means for tumor cells to escape TGF␤-mediated growth restriction.116 TGF␤ conveys signals via two serine/ threonine kinase receptors, type I and type II, both of which are necessary for signal transduction.117,118 In this complex, the serine/threonine kinase domain of the type II receptor phosphorylates serine residues in a serine- and glycine-rich region adjacent to the amino-terminal boundary of the kinase domain (GS domain) of the type I receptor. This phosphorylation step activates the kinase domain of the type I receptor, which can subsequently propagate the signal to downstream effectors.119,120 Mutations of the type II receptor can interrupt the signal transduction pathway by dominant-negative mechanisms,119,121 leading to growth stimulation rather than growth restriction. The significant association found between TGF␤RII mutations and the MSI phenotype of the tumors somehow reflects the result of disruption of the TGF␤ pathway, considering that MSI 282 Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 tumors are larger and more expanding when compared with MSS tumors. Our finding of a close relationship between TGF␤RII mutations and the intestinal histotype might be supported by the high percentage of TGF␤RII mutations found in colon cancer,122 if one believes that the intestinal type of gastric carcinoma mimics the morphologic appearance of colon cancer, but further studies are needed to confirm this finding. IGFIIR poly(G)8 tract was shown to be mutated in about 30% of MSI gastric tumors, and a trend (P ⫽ 0.06) was found between IGFIIR mutations and lower prevalence of lymph node metastases in the setting of MSI cases.87 IGFIIR mutations are predicted to generate inactive receptors. The biological role of IGF II is mediated through both IGFI and II receptors, with higher affinity to IGFIR.123 IGFIIR is a multifunctional protein that plays a role in lysosomal enzyme trafficking, endocytosis, and activation of TGF␤.124,125 The IGFIIR also inhibits cell proliferation mediated by the IGFII ligand, itself a potent growth stimulant, by internalizing and degrading this protein.126 Thus, IGFIIR, by antagonizing the growth stimulatory effect of IGFII and activating the growth inhibitory effect of TGF␤, serves as a growth suppressor gene.123 We observed a trend toward an association between IGFIIR mutations and low prevalence of lymph node metastases. Recently, IGFIIR mutations were described to be inversely associated to serosa invasion.127 Altogether, these results are in accordance with the role on cell motility and metastatization advanced for IGFIIR.128 Basic and clinical studies are necessary to confirm whether or not there is a relationship between IGFIIR mutations and decreased nodal metastatic ability of gastric carcinoma. KRAS, BRAF mutations in GC Our group has been focusing on determining the frequency and spectrum of mutations of diverse oncogenes in MSI carcinomas, specifically on KRAS, BRAF, and PI3KCA genes (Figure 2). We and others verified that HNPCC, sporadic MSI gastric, and colorectal carcinoma progress through RAS/RAF/MAPK pathway, though targeting distinct kinases and with specific profiles of activating mutations. Recently, a high frequency of activating mutations in BRAF, a gene from the RAS regulated kinase-encoding RAF family, was found in colorectal tumors and associated to MSI tumors.129-131 Further, data on BRAF showed a hotspot mutation within exon 15 (V600E), which is the most frequent somatic substitution ever identified in MSI colon cancers. In addition, these mutations were inversely associated to KRAS mutations, showing that KRAS and BRAF genes are alternative genetic events in colorectal cancer and reinforcing the idea that both MSI and MSS colorectal tumors progress through the same RAS/RAF/MAPK pathway.130 However, it was not clear whether MSI colon and gastric tumors share the alterations of the RAS/RAF/MAPK genes found in MSI colon tumors, nor if KRAS and BRAF genes are also alternative genetic events in GC. We investigated whether KRAS gene was mutated at a similar low frequency in MSI and MSS gastric tumors. Mutations were found in 28% of MSI carcinomas and absent in MSS carcinomas.132 The significant association between KRAS and MSI (P ⫽ 0.0005) is exclusive of GC because, in colon carcinomas, KRAS mutations occur in MSS and MSI carcinomas. In contrast to KRAS, BRAF mutations do not occur in MSI GC and are also extremely rare in MSS GC. In fact, we only detected a BRAFV600E mutation in a MSS carcinoma (1/124, 0.8%).133 Altogether, these results suggest that KRAS but not BRAF mutations contribute to the tumorigenesis of MSI GC and that KRAS and BRAF mutations are not alternative genetic events in GC. PIK3CA and GC In GC, PIK3CA mutations were identified in 4% to 11% of the cases.134-136 A higher frequency (25%) was found by Samuels and coworkers,137 but this study was performed in a very small series of tumors. In gastric carcinoma, the MMR status of the tumors is an important parameter to consider when determining the frequency of PIK3CA mutations. The mutational distribution of PIK3CA in GC is similar to the one observed for KRAS. As for KRAS, we were unable to find PIK3CA mutations in MSS gastric carcinoma.135 Another group reported a single mutation in a series of 73 MSS gastric carcinomas.136 It is likely that PIK3CA mutations do not represent an important oncogenic event in MSS gastric carcinogenesis analogously to what is observed for KRAS and BRAF mutations in MSS gastric carcinoma.132,133 In MSI gastric carcinoma, we found PIK3CA mutations in 19.2% of the cases135 and verified that PIK3CA and KRAS mutations were mutually exclusive events.135 These findings are not surprising because PI3K may function as a downstream effector of the RAS pathway.138,139 It was suggested that genes involved in the same signaling pathway may manifest mutations in cancer cells in a mutually exclusive manner, presumably due to the lack of selective growth advantage in having a second hit in the already altered pathway.136 In GC, PIK3CA alterations were found in both early and advanced specimens,134 which could indicate an important role in the development and progression of this tumor type. Genomic amplification of PIK3CA was found to occur in 36.4% of gastric tumors, strongly associated with increased expression of PIK3CA transcript and elevated levels of phosphor-AKT.140 Moreover, PIK3CA amplification was predominantly detected in tumors with no PTEN alterations, suggesting that loss of PTEN and activation of PIK3CA genes are mutually exclusive events in gastric tumorigenesis.140 PIK3CA constitutes an attractive molecular marker for early detection or for the monitoring of tumor progression in Carneiro et al Gastric Cancer diverse types of cancer. More importantly, PIK3CA is a promising therapeutic target for the development of specific inhibitors of the p110␣ subunit. Some inhibitors of PI3K pathway have been extensively used in vitro and in vivo, such as wortmannin and LY294002, which inhibit the catalytic activity of p110 subunits of PI3K.141-143 These compounds were shown to inhibit cell proliferation and/or induce apoptosis in cancer cells via inhibition of the PI3K–AKT pathway. They also play a role in enhancing the effectiveness of radio- or chemotherapy.143-147 These inhibitors may also be applied to mutant forms of PIK3CA, considering that the transforming potential of PIK3CA mutants can still be inhibited by broadly acting PI3K inhibitors, such as LY294002.148 Despite their efficacy toward inhibition of active PI3K pathway, such inhibitors have the disadvantage of nonselectivity.145 PI3K is ubiquitously present in many mammalian cells,145 thus the use of wortmannin and LY294002 may be toxic to the cells that do not display any alteration of the pathway. The lack of stability of wortmannin and the lack of solubility of LY294002 are additional things that hampered further clinical studies of these agents.145 Another possibility of therapy is the use of inhibitors of the downstream targets of the PI3K pathway. PIK3CA mutations lead to an increase in the phosphorylation of AKT with its concomitant activation.148-152 The repression of the activation of AKT represents an attractive possibility of inhibiting PI3K pathway, although no small AKT inhibitors have been established yet.143 Further, the inhibition of PI3K–AKT target genes may represent a fighting chance. mTOR is an example of this possibility. Some inhibitors of mTOR are under preclinical and clinical investigations.145 It is the case of rapamycin and its derivatives (Rad001, CCI-779, and AP23573),153 that were shown to work as potent radiosensitizers of endothelial cells in vitro and led to improved tumor-growth delay of glioma xenographs in vivo.145 Together, these results imply that inhibition of the downstream targets may be a valuable approach to indirectly block the oncogenic ability of active PI3K–AKT pathway. However, it should be taken into account that the downstream effectors of PI3K that are activated on deregulation of the pathway may depend on the cell type or tumor model, and in the case of GC, it will be of limited use. In the future, the development, if possible, of inhibitors that target specifically the mutant forms of oncogenes that are activated in MSI GC would be of particular interest. Concluding remarks and future challenges In this review, we focused on several genes and downstream pathways shown to play a role in gastric carcinogenesis, aiming at analyzing putative targets for therapy and biological modifiers in gastric carcinoma. From the experience of our group and the data on record, we would highlight the following key issues: 283 ● Identification and characterization of germline CDH1 mutations play a pivotal role in decision making on prophylactic gastrectomies in the setting of HDGC. ● The characterization of the second-hit mechanisms of CDH1 inactivation in HDGC tumors will be relevant for the therapeutic management of HDGC patients. ● EGFR becomes (abnormally) activated in cells with Ecadherin loss of function. Thus, it is tempting to test the use of EGFR pharmacological inhibitors in patients with tumors harboring CDH1 mutations, namely whenever such mutations are localized in the extracellular domain. ● CDH1 mutations lead to an increased resistance to apoptosis mediated by Taxol. This experimental result questions the effectiveness of Taxol in the treatment of advanced gastric patients with tumors harboring CDH1 mutations and calls for the need for further research on the subject. ● Activation of RTKs such as EGFR and ERBB2 by mutations or amplification is rare in GC. However, they are good molecular targets for anti-RTKs therapies with ectodomain-binding antibodies or small-molecule TK inhibitors. ● In GC, MSI is a molecular marker of good prognosis. ● A subset of MSI GC carries mutations in KRAS or PIK3CA genes. These molecules or their downstream effectors are promising therapeutic targets. Acknowledgments This work is supported in part by grants from The Portuguese Foundation for Science and Technology (FCT) (POCTI/SAU-OBS/58111/2004 and POCI/SAU-OBS/ 57275/2004) and Sixth Framework Programme from EUFP6 (LSHC-CT-2005-018754). References 1. Parkin DM, Bray F, Ferlay J, et al: Global cancer statistics, 2002. CA Cancer J Clin 55:74-108, 2005 2. Ferlay J, Bray F, Pisani P, et al (eds): GLOBOCAN 2002. Cancer Incidence, Mortality and Prevalence Worldwide. IARC Cancer Base No. 5, Version 2.0. Lyon, France, IARC Press, 2004 3. Carneiro F: Classification of gastric carcinomas. Curr Diagn Pathol 4:51-59, 1997 4. Laurén P: The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. An attempt at a histoclinical classification. Acta Pathol Microbiol Scand 64:31-49, 1965 5. Henson DE, Dittus C, Younes M, et al: Differential trends in the intestinal and diffuse types of gastric carcinoma in the United States, 1973-2000: increase in the signet ring cell type. Arch Pathol Lab Med 128:765-770, 2004 6. Carneiro F, Seixas M, Sobrinho–Simões M: New elements for an updated classification of the carcinomas of the stomach. Pathol Res Pract 191:571-584, 1995 7. Ekstrom AM, Serafini M, Nyren O, et al: Dietary antioxidant intake and the risk of cardia cancer and noncardia cancer of the intestinal 284 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 and diffuse types: a population-based case-control study in Sweden. Int J Cancer 87:133-140, 2000 Palli D, Galli M, Caporaso NE, et al: Family history and risk of stomach cancer in Italy. Cancer Epidemiol Biomarkers Prev 3:15-18, 1994 Carneiro F, Oliveira C, Suriano G, et al: Molecular pathology of familial gastric cancer, with an emphasis on hereditary diffuse gastric cancer (HDGC). J Clin Pathol 61:25-30, 2008 Guilford P, Hopkins J, Harraway J, et al: E-cadherin germline mutations in familial gastric cancer. Nature 392:402-405, 1998 Caldas C, Carneiro F, Lynch HT, et al: Familial gastric cancer: overview and guidelines for management. J Med Genet 36:873-880, 1999 Chun YS, Lindor NM, Smyrk TC, et al: Germline E-cadherin gene mutations: is prophylactic total gastrectomy indicated? Cancer 92: 181-187, 2001 Huntsman DG, Carneiro F, Lewis FR, et al: Early gastric cancer in young, asymptomatic carriers of germ-line E-cadherin mutations. N Engl J Med 344:1904-1909, 2001 Carneiro F, Huntsman DG, Smyrk TC, et al: Model of the early development of diffuse gastric cancer in E-cadherin mutation carriers and its implications for patient screening. J Pathol 203:681-687, 2004 Charlton A, Blair V, Shaw D, et al: Hereditary diffuse gastric cancer: predominance of multiple foci of signet ring cell carcinoma in distal stomach and transitional zone. Gut 53:814-820, 2004 Blair V, Martin I, Shaw D, et al: Hereditary diffuse gastric cancer: diagnosis and management. Clin Gastroenterol Hepatol 4:262-275, 2006 Rogers WM, Dobo E, Norton JA, et al: Risk-reducing total gastrectomy for germline mutations in E-cadherin (CDH1): pathologic findings with clinical implications. Am J Surg Pathol 32:799-809, 2008 Lynch HT, Kaurah P, Wirtzfeld D, et al: Hereditary diffuse gastric cancer: diagnosis, genetic counseling, and prophylactic total gastrectomy. Cancer 112:2655-2663, 2008 Shore EM, Nelson WJ: Biosynthesis of the cell adhesion molecule uvomorulin (E-cadherin) in Madin-Darby canine kidney epithelial cells. J Biol Chem 266:19672-19680, 1991 Mareel M, Leroy A: Clinical, cellular, and molecular aspects of cancer invasion. Physiol Rev 83:337-376, 2003 Berx G, Becker KF, Höfler H, et al: Mutations of the human Ecadherin (CDH1) gene. Hum Mutat 12:226-237, 1998 Oliveira C, Seruca R, Carneiro F: Genetics, pathology and clinics of familial gastric cancer. Int J Surg Pathol 14:21-33, 2006 Grady WM, Willis J, Guilford PJ, et al: Methylation of the CDH1 promoter as the second genetic hit in hereditary diffuse gastric cancer. Nat Genet 26:16-17, 2000 Oliveira C, de Bruin J, Nabais S, et al: Intragenic deletion of CDH1 as the inactivating mechanism of the wild-type allele in an HDGC tumour. Oncogene 23:2236-2240, 2004 Knudson AG Jr: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820-823, 1971 Kaurah P, MacMillan A, Boyd N, et al: Founder and recurrent CDH1 mutations in families with hereditary diffuse gastric cancer. J Am Med Assoc 297:2360-2372, 2007 Lynch HT, Grady W, Suriano G, et al: Gastric cancer: new genetic developments. J Surg Oncol 90:114-133, 2005 Fitzgerald RC, Caldas C: Clinical implications of E-cadherin associated hereditary diffuse gastric cancer. Gut 53:775-778, 2004 Suriano G, Oliveira C, Ferreira P, et al: Identification of CDH1 germline missense mutations associated with functional inactivation of the E-cadherin protein in young gastric cancer probands. Hum Mol Genet 12:575-582, 2003 Suriano G, Seixas S, Rocha J, et al: A model to infer the pathogenic significance of CDH1 germline missense variants. J Mol Med 84: 1023-1031, 2006 Becker KF, Höfler H: Frequent somatic allelic inactivation of the E-cadherin gene in gastric carcinomas. J Natl Cancer Inst 87:10821084, 1995 32. Machado JC, Oliveira C, Carvalho R, et al: E-cadherin gene (CDH1) promoter methylation as the second hit in sporadic diffuse gastric carcinoma. Oncogene 20:1525-1528, 2001 33. Corso G, Roviello F, Paredes J, et al: Characterization of the P373L E-cadherin germline missense mutation and implication for clinical management. Eur J Surg Oncol 33:1061-1067, 2007 34. Chang YF, Imam JS, Wilkinson MF: The nonsense-mediated decay RNA surveillance pathway. Annu Rev Biochem 76:51-74, 2007 35. Holbrook JA, Neu–Yilik G, Hentze MW, et al: Nonsense-mediated decay approaches the clinic. Nat Genet 36:801-808, 2004 36. Wilkinson MF: A new function for nonsense-mediated mRNAdecay factors. Trends Genet 21:143-148, 2005 37. Karam R, Carvalho J, Bruno I, et al: The NMD mRNA surveillance pathway downregulates aberrant E-cadherin transcripts in gastric cancer cells and in CDH1 mutation carriers. Oncogene 2008 Apr 21 [Epub ahead of print] 38. Sasaki CY, Lin H, Morin PJ, et al: Truncation of the extracellular region abrogates cell contact but retains the growth-suppressive activity of E-cadherin. Cancer Res 60:7057-7065, 2000 39. Yap AS, Kovacs EM: Direct cadherin-activated cell signaling: a view from the plasma membrane. J Cell Biol 160:11-16, 2003 40. Crepaldi T, Pollack AL, Prat M, et al: Targeting of the SF/HGF receptor to the basolateral domain of polarized epithelial cells. J Cell Biol 125:313-320, 1994 41. Hoschuetzky H, Aberle H, Kemler R: Beta-catenin mediates the interaction of the cadherin– catenin complex with epidermal growth factor receptor. J Cell Biol 127:1375-1380, 1994 42. Qian X, Karpova T, Sheppard AM, et al: E-cadherin-mediated adhesion inhibits ligand-dependent activation of diverse receptor tyrosine kinases. EMBO J 23:1739-1748, 2004 43. Mateus AR, Seruca R, Machado JC, et al: EGFR regulates RhoAGTP dependent cell motility in E-cadherin mutant cells. Human Mol Genet 16:1639-1647, 2007 44. Yarden Y, Sliwkowski MX: Untangling the ErbB signaling network. Nat Rev Mol Cell Biol 2:127-137, 2001 45. den Hartigh JC, van Bergen en Henegouwen PM, Verkleij AJ, et al: The EGF receptor is an actin-binding protein. J Cell Biol 119:349355, 1992 46. Hall A: Rho GTPases and the actin cytoskeleton. Science 279:509514, 1998 47. Nobes CD, Hall A: Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol 144:1235-1244, 1999 48. Comoglio PM, Boccaccio C, Trusolino L: Interactions between growth factor receptors and adhesion molecules: breaking the rules. Curr Opin Cell Biol 15:565-571, 2003 49. Ruoslahti E, Reed JC: Anchorage dependence, integrins, and apoptosis. Cell 77:477-478, 1994 50. Frisch SM, Ruoslahti E: Integrins and anoikis. Curr Opin Cell Biol 9:701-706, 1997 51. Ihara A, Koizumi H, Hashizume R, et al: Expression of epithelial cadherin and a- and b-catenins in nontumoral livers and hepatocellular carcinomas. Hepatology 23:1441-1447, 1996 52. Hirata K, Ajiki T, Okazaki T, et al: Frequent occurrence of abnormal E-cadherin/␤-catenin protein expression in advanced gallbladder cancers and its association with decreased apoptosis. Oncology 71:102110, 2006 53. Ferreira P, Oliveira MJ, Beraldi E, et al: Loss of functional Ecadherin renders cells more resistant to the apoptotic agent taxol in vitro. Exp Cell Res 310:99-104, 2005 54. Ullrich A, Schlessinger J: Signal transduction by receptors with tyrosine kinase activity. Cell 61:203-212, 1990 55. Klapper LN, Kirchbaum MH, Sela M, et al: Biochemical and clinical implications of the ErbB/HER signaling network of growth factor receptors. Adv Cancer Res 77:25-79, 2000 56. Hynes NE, Lane HA: ERBB receptors and cancer: the complexity of targeted inhibitors. Nat Rev Cancer 5:341-354, 2005 57. Lynch TJ, Bell DW, Sordella R, et al: Activating mutations in the epidermal growth factor receptor underlying responsiveness of non- Carneiro et al 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. Gastric Cancer small-cell lung cancer to gefitinib. N Engl J Med 350:2129-2139, 2004 Paez JG, Janne PA, Lee JC, et al: EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 304: 1497-1500, 2004 Moutinho C, Mateus AR, Milanezi F, et al: Epidermal growth factor receptor structural alterations in gastric cancer. BMC Cancer 8:10, 2008 Akiyama T, Sudo C, Ogawara H, et al: The product of the human c-erbB-2 gene: a 185-kilodalton glycoprotein with tyrosine kinase activity. Science 232:1644-1646, 1986 Popescu NC, King CR, Kraus MH: Localization of the human erbB-2 gene on normal and rearranged chromosomes 17 to bands q12-21.32. Genomics 4:362-366, 1989 Slamon DJ, Godolphin W, Jones LA, et al: Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244: 707-712, 1989 Dowsett M, Bartlett J, Ellis IO, et al: Correlation between immunohistochemistry (HercepTest) and fluorescence in situ hybridization (FISH) for HER-2 in 426 breast carcinomas from 37 centres. J Pathol 199:418-423, 2003 McKenzie SJ, DeSombre KA, Bast BS, et al: Serum levels of HER-2 neu (C-erbB-2) correlate with overexpression of p185neu in human ovarian cancer. Cancer 71:3942-3946, 1993 Hirashima N, Takahashi W, Yoshii S, et al: Protein overexpression and gene amplification of c-erb B-2 in pulmonary carcinomas: a comparative immunohistochemical and fluorescence in situ hybridization study. Mod Pathol 14:556-562, 2001 Cohen JA, Weiner DB, More KF, et al: Expression pattern of the neu (NGL) gene-encoded growth factor receptor protein (p185neu) in normal and transformed epithelial tissues of the digestive tract. Oncogene 4:81-88, 1989 David L, Seruca R, Nesland JM, et al: c-erbB-2 expression in primary gastric carcinomas and their metastases. Mod Pathol 5:384-390, 1992 Nakajima M, Sawada H, Yamada Y, et al: The prognostic significance of amplification and overexpression of c-met and c-erb B-2 in human gastric carcinomas. Cancer 85:1894-1902, 1999 Takehana T, Kunitomo K, Kono K, et al: Status of c-erbB-2 in gastric adenocarcinoma: a comparative study of immunohistochemistry, fluorescence in situ hybridization and enzyme-linked immuno-sorbent assay. Int J Cancer 98:833-837, 2002 Varis A, Zaika A, Puolakkainen P, et al: Coamplified and overexpressed genes at ERBB2 locus in gastric cancer. Int J Cancer 109: 548-553, 2004 Park DI, Yun JW, Park JH, et al: HER-2/neu amplification is an independent prognostic factor in gastric cancer. Dig Dis Sci 51:13711379, 2006 Kim MA, Jung EJ, Lee HS, et al: Evaluation of HER-2 gene status in gastric carcinoma using immunohistochemistry, fluorescence in situ hybridization, and real-time quantitative polymerase chain reaction. Hum Pathol 38:1386-1393, 2007 Kimura M, Tsuda H, Morita D, et al: A proposal for diagnostically meaningful criteria to classify increased epidermal growth factor receptor and c-erbB-2 gene copy numbers in gastric carcinoma, based on correlation of fluorescence in situ hybridization and immunohistochemical measurements. Virchows Arch 445:255-262, 2004 Vulfovich M, Rocha–Lima C: Novel advances in pancreatic cancer treatment. Expert Rev Anticancer Ther 8:993-1002, 2008 Slamon DJ, Leyland–Jones B, Shak S, et al: Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344:783-792, 2001 Romond EH, Perez EA, Bryant J, et al: Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer. N Engl J Med 353:1673-1684, 2005 Piccart–Gebhart MJ, Procter M, Leyland–Jones B, et al: Trastuzumab after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med 353:1659-1672, 2005 285 78. Tanner M, Hollmen M, Junttila TT, et al: Amplification of HER-2 in gastric carcinoma: association with Topoisomerase IIalpha gene amplification, intestinal type, poor prognosis and sensitivity to trastuzumab. Ann Oncol 16:273-278, 2005 79. Rebischung C, Barnoud R, Stefani L, et al: The effectiveness of trastuzumab (Herceptin) combined with chemotherapy for gastric carcinoma with overexpression of the c-erbB-2 protein. Gastric Cancer 8:249-252, 2005 80. Peltomaki P: Deficient DNA mismatch repair: a common etiologic factor for colon cancer. Hum Mol Genet 10:735-740, 2001 81. Liu B, Nicolaides NC, Markowitz S, et al: Mismatch repair gene defects in sporadic colorectal cancers with microsatellite instability. Nat Genet 9:48-55, 1995 82. Peltomaki P, Lothe RA, Aaltonen LA, et al: Microsatellite instability is associated with tumors that characterize the hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res 53:5853-5855, 1993 83. Seruca R, Santos NR, David L, et al: Sporadic gastric carcinomas with microsatellite instability display a particular clinicopathologic profile. Int J Cancer 64:32-36, 1995 84. dos Santos NR, Seruca R, Constância M, et al: Microsatellite instability at multiple loci in gastric carcinoma: clinicopathologic implications and prognosis. Gastroenterology 110:38-44, 1996 85. Chung YJ, Park SW, Song JM, et al: Evidence of genetic progression in human gastric carcinomas with microsatellite instability. Oncogene 15:1719-1726, 1997 86. Hayden JD, Cawkwell L, Quirke P, et al: Prognostic significance of microsatellite instability in patients with gastric carcinoma. Eur J Cancer 33:2342-2346, 1997 87. Oliveira C, Seruca R, Seixas M, et al: The clinicopathological features of gastric carcinomas with microsatellite instability may be mediated by mutations of different “target genes”: a study of the TGFbeta RII, IGFII R, and BAX genes. Am J Pathol 153:1211-1219, 1998 88. Fleisher AS, Esteller M, Wang S, et al: Hypermethylation of the hMLH1 gene promoter in human gastric cancers with microsatellite instability. Cancer Res 59:1090-1095, 1999 89. Halling KC, Harper J, Moskaluk CA, et al: Origin of microsatellite instability in gastric cancer. Am J Pathol 155:205-211, 1999 90. Kang GH, Shim YH, Ro JY, et al: Correlation of methylation of the hMLH1 promoter with lack of expression of hMLH1 in sporadic gastric carcinomas with replication error. Lab Invest 79:903-909, 1999 91. Leung SY, Yuen ST, Chung LP, et al: hMLH1 promoter hypermethylation and lack of hMLH1 expression in sporadic gastric carcinomas with high-frequency microsatellite instability. Cancer Res 59:159164, 1999 92. Yamamoto H, Perez–Piteira J, Yoshida T, et al: Gastric cancers of the microsatellite mutator phenotype display characteristic genetic and clinical features. Gastroenterology 116:1348-1357, 1999 93. Pinto M, Oliveira C, Machado JC, et al: MSI-L gastric carcinomas share the hMLH1 methylation status of MSI-H carcinomas but not their clinicopathological profile. Lab Invest 80:1915-1923, 2000 94. Aarnio M, Salovaara R, Aaltonen LA, et al: Features of gastric cancer in hereditary non-polyposis colorectal cancer syndrome. Int J Cancer 74:551-555, 1997 95. Pinto M, Wu Y, Mensink RG, et al: Somatic mutations in mismatch repair genes in sporadic gastric carcinomas are not a cause but a consequence of the mutator phenotype. Cancer Genet Cytogenet 18:110-114, 2008 96. Suzuki H, Itoh F, Toyota M, et al: Distinct methylation pattern and microsatellite instability in sporadic gastric cancer. Int J Cancer 83:309-313, 1999 97. Toyota M, Ahuja N, Suzuki H, et al: Aberrant methylation in gastric cancer associated with CpG island methylator phenotype. Cancer Res 59:5438-5442, 1999 98. Latham KE: X chromosome imprinting and inactivation in the early mammalian embryo. Trends Genet 12:134-138, 1996 286 Seminars in Diagnostic Pathology, Vol 25, No 4, November 2008 99. Issa JP, Ottaviano YL, Celano P, et al: Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat Genet 7:536-540, 1994 100. Esteller M, Corn PG, Baylin SB, et al: A gene hypermethylation profile of human cancer. Cancer Res 61:3225-3229, 2001 101. Nguyen C, Liang G, Nguyen TT, et al: Susceptibility of nonpromoter CpG islands to de novo methylation in normal and neoplastic cells. J Natl Cancer Inst 93:1465-1472, 2001 102. Kikuchi T, Itoh F, Toyota M, et al: Aberrant methylation and histone deacetylation of cyclooxygenase 2 in gastric cancer. Int J Cancer 97:272-277, 2002 103. Kang GH, Shim YH, Jung HY, et al: CpG island methylation in premalignant stages of gastric carcinoma. Cancer Res 61:2847-2851, 2001 104. Fleisher AS, Esteller M, Tamura G, et al: Hypermethylation of the hMLH1 gene promoter is associated with microsatellite instability in early human gastric neoplasia. Oncogene 20:329-335, 2001 105. Carvalho B, Pinto M, Cirnes L, et al: Concurrent hypermethylation of gene promoters is associated with a MSI-H phenotype and diploidy in gastric carcinomas. Eur J Cancer 39:1222-1227, 2003 106. Song SH, Jong HS, Choi HH, et al: Transcriptional silencing of cyclooxygenase-2 by hyper-methylation of the 50 CpG island in human gastric carcinoma cells. Cancer Res 61:4628-4635, 2001 107. Nakagawa H, Nuovo GJ, Zervos EE, et al: Age-related hypermethylation of the 50 region of HMLH1 in normal colonic mucosa is associated with microsatellite-unstable colorectal cancer development. Cancer Res 61:6991-6995, 2001 108. Tamura G, Yin J, Wang S, et al: E-Cadherin gene promoter hypermethylation in primary human gastric carcinomas. J Natl Cancer Inst 92:569-573, 2000 109. Lee TL, Leung WK, Chan MW, et al: Detection of gene promoter hypermethylation in the tumor and serum of patients with gastric carcinoma. Clin Cancer Res 8:1761-1766, 2002 110. Esteller M, Hamilton SR, Burger PC, et al: Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res 59:793-797, 1999 111. Park TJ, Han SU, Cho YK, et al: Methylation of O(6)-methylguanine-DNA methyltransferase gene is associated significantly with K-ras mutation, lymph node invasion, tumor staging, and disease free survival in patients with gastric carcinoma. Cancer 92:2760-2768, 2001 112. Mitra G, Pauly GT, Kumar R, et al: Molecular analysis of O6substituted guanine-induced mutagenesis of ras oncogenes. Proc Natl Acad Sci USA 86:8650-8654, 1989 113. Beghelli S, de Manzoni G, Barbi S, et al: Microsatellite instability in gastric cancer is associated with better prognosis in only stage II cancers. Surgery 139:347-356, 2006 114. Duval A, Reperant M, Compoint A, et al: Target gene mutation profile differs between gastrointestinal and endometrial tumors with mismatch repair deficiency. Cancer Res 62:1609-1612, 2002 115. Davalos V, Dopeso H, Velho S, et al: High EPHB2 mutation rate in gastric but not endometrial tumors with microsatellite instability. Oncogene 26:308-311, 2007 116. Amendt C, Schirmacher P, Weber H, et al: Expression of a dominant negative type II TGF-beta receptor in mouse skin results in an increase in carcinoma incidence and an acceleration of carcinoma development. Oncogene 17:25-34, 1998 117. Attisano L, Wrana JL, Lopez–Casillas F, et al: TGF-beta receptors and actions. Biochim Biophys Acta 1222:71-80, 1994 118. Derynck R: TGF-beta-receptor-mediated signaling. Trends Biochem Sci 19:548-553, 1994 119. Wrana JL, Attisano L, Carcamo J, et al: TGF beta signals through a heteromeric protein kinase receptor complex. Cell 71:1003-1014, 1992 120. Heldin CH, Miyazono K, ten Dijke P: TGF-beta signaling from cell membrane to nucleus through SMAD proteins. Nature 390:465-471, 1997 121. Carcamo J, Zentella A, Massague J: Disruption of transforming growth factor beta signaling by a mutation that prevents transphosphorylation within the receptor complex. Mol Cell Biol 15:15731581, 1995 122. Markowitz S, Wang J, Myeroff L, et al: Inactivation of the type II TGF␤ receptor in colon cancer cells with microsatellite instability. Science 268:1336-1338, 1995 123. Souza RF, Appel R, Yin J, et al: Microsatellite instability in the insulin-like growth factor II receptor gene in gastrointestinal tumours. Nature Genet 14:255-257, 1996 124. Dahms NM, Lobel P, Kornfeld S: Mannose 6-phosphate receptors and lysosomal enzyme targeting. J Biol Chem 264:12115-12118, 1989 125. Dennis PA, Rifkin DB: Cellular activation of latent transforming growth factor beta requires binding to the cation-independent mannose 6-phosphate/insulin-like growth factor type II receptor. Proc Natl Acad Sci USA 88:580-584, 1991 126. Kornfeld S: Structure and function of the mannose 6-phosphate/ insulin-like growth factor type II receptors. Annu Rev Biochem 61:307-330, 1992 127. Falchetti M, Saieva C, Lupi R, et al: Gastric cancer with high-level microsatellite instability: target gene mutations, clinicopathologic features, and long-term survival. Hum Pathol 39:925-932, 2008 128. Minniti CP, Kohn EC, Grubb JH, et al: The insulin-like growth factor II (IGF-II)/mannose 6-phosphate receptor mediates IGF-II-induced motility in human rhabdomyosarcoma cells. J Biol Chem 267:90009004, 1992 129. Davies H, Bignell GR, Cox C, et al: Mutations of the BRAF gene in human cancer. Nature 417:949-954, 2002 130. Rajagopalan H, Bardelli A, Lengauer C, et al: Tumorigenesis: RAF/ RAS oncogenes and mismatch-repair status. Nature 418:934, 2002 131. Yuen ST, Davies H, Chan TL, et al: Similarity of the phenotypic patterns associated with BRAF and KRAS mutations in colorectal neoplasia. Cancer Res 62:6451-6455, 2002 132. Brennetot C, Duval A, Hamelin R, et al: Frequent Ki-ras mutations in gastric tumors of the MSI phenotype. Gastroenterology 125:1282, 2003 133. Oliveira C, Pinto M, Duval A, et al: BRAF mutations characterize colon but not gastric cancer with mismatch repair deficiency. Oncogene 22:9192-9196, 2003 134. Lee JW, Soung YH, Kim SY, et al: PIK3CA gene is frequently mutated in breast carcinomas and hepatocellular carcinomas. Oncogene 24:1477-1480, 2005 135. Velho S, Oliveira C, Ferreira A, et al: The prevalence of PIK3CA mutations in gastric and colon cancer. Eur J Cancer 41:1649-1654, 2005 136. Li V, Wong C, Chan T, et al: Mutations of PIK3CA in gastric adenocarcinoma. BMC Cancer 5:29, 2005 137. Samuels Y, Wang Z, Bardelli A, et al: High frequency of mutations of the PIK3CA gene in human cancers. Science 304:554, 2004 138. Katso R, Okkenhaug K, Ahmadi K, et al: Cellular function of phosphoinositide 3-kinases: Implications for development, immunity, homeostasis, and cancer. Annu Rev Cell Dev Biol 17:615-675, 2001 139. Wymann MP, Pirola L: Structure and function of phosphoinositide 3-kinases. Biochim Biophys Acta 1436:127-150, 1998 140. Byun DS, Cho K, Ryu BK, et al: Frequent monoallelic deletion of PTEN and its reciprocal association with PIK3CA amplification in gastric carcinoma. Int J Cancer 104:318-327, 2003 141. Vivanco I, Sawyers CL: The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer 2:489-501, 2002 142. Samuels Y, Velculescu VE: Oncogenic mutations of PIK3CA in human cancers. Cell Cycle 3:1221-1224, 2004 143. Osaki M, Kase S, Adachi K, et al: Inhibition of the PI3K-Akt signaling pathway enhances the sensitivity of Fas-mediated apoptosis in human gastric carcinoma cell line. MKN-45. J Cancer Res Clin Oncol 130:8-14, 2004 Carneiro et al Gastric Cancer 144. Osaki M, Oshimura M, Ito H: PI3K-Akt pathway: its functions and alterations in human cancer. Apoptosis 9:667-676, 2004 145. Kim DW, Huamani J, Fu A, et al: Molecular strategies targeting the host component of cancer to enhance tumor response to radiation therapy. Int J Radiat Oncol Biol Phys 64:38-46, 2006 146. Ng SSW, Tsao M-S, Chow S, et al: Inhibition of phosphatidylinositide 3-kinase enhances gemcitabine-induced apoptosis in human pancreatic cancer cells. Cancer Res 60:5451-5455, 2000 147. Sarkaria JN, Tibbetts RS, Busby EC, et al: Inhibition of phosphoinositide 3-kinase related kinases by the radiosensitizing agent wortmannin. Cancer Res 58:4375-82, 1998 148. Samuels Y, Diaz LAJ, Schmidt-Kittler O, et al: Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 7:561-573, 2005 287 149. Ikenoue T, Kanai F, Hikiba Y, et al: Functional analysis of PIK3CA gene mutations in human colorectal cancer. Cancer Res 65:45624567, 2005 150. Isakoff SJ, Engelman JA, Irie HY, et al: Breast cancer-associated PIK3CA mutations are oncogenic in mammary epithelial cells. Cancer Res 65:10992-11000, 2005 151. Zhao JJ, Liu Z, Wang L, et al: The oncogenic properties of mutant p110{alpha} and p110{beta} phosphatidylinositol 3-kinases in human mammary epithelial cells. Proc Natl Acad Sci USA 102:1844318448, 2005 152. Kang S, Bader AG, Zhao L, et al: Mutated PI 3-kinases: cancer targets on a silver platter. Cell Cycle 4:578-581, 2005 153. Chan S: Targeting the mammalian target of rapamycin (mTOR): a new approach to treating cancer. Br J Cancer 91:1420-1424, 2004