Gastric GISTs: Analysis of c-Kit, PDGFRA and BRAF mutations in relation to prognosis and clinical pathological characteristics of patients - A GIRCG study.

BACKGROUND: Gastric gastrointestinal stromal tumors (GISTs) represent a subgroup of GISTs with a better prognosis than those located in other areas. In this retrospective study we performed a molecular characterization of a large series of patients with gastric GISTs in relation to clinical-pathological characteristics and prognosis. METHODS: DNA was extracted from paraffin-embedded sections from 221 gastric GIST patients submitted to surgery. Exons 9, 11, 13 and 17 of KIT, exons 12 and 18 of PDGFRA and exons 11 and 15 of BRAF were analyzed by direct sequencing. Cox regression analysis adjusted for clinical-pathological factors was performed to evaluate KIT and PDGFRA mutations in relation to the composite endpoint of relapse or death. RESULTS: KIT and PDGFRA mutations were observed in 119 (53.8%) and 56 (25.3%) patients, respectively, whereas 46 (20.8%) patients had wild type (wt) disease. Univariable analyses showed that a high Miettinen risk category and the presence of ulceration and KIT deletions were associated with increased risk of relapse or death (p < 0.001; p = 0.0389 and p = 0.002, respectively). After adjusting for Miettinen risk score, KIT deletions remained an independent prognostic factor (HRadj = 2.65, 95% CI [1.15-6.13], p = 0.023). Moreover, KIT deletions in exon 11 codons 557, 558 or 559 were associated with a higher risk of relapse or death than wt tumors (HRadj = 3.29 95% CI [1.64-6.64], p = 0.001). CONCLUSIONS: KIT deletions in exon 11, especially those involving codons 557, 558 or 559, were correlated with a more aggressive gastric GIST phenotype and increased risk of relapse or death.

ALK translocation detection in non-small cell lung cancer cytological samples obtained by TBNA or EBUS-TBNA.

OBJECTIVE: As the diagnosis of non-small cell lung cancer (NSCLC) is based on cytology in around 70% of cases, it is important to use the same material for molecular analyses. Fluorescence in situ hybridization (FISH) is the only approved test for the detection of the translocation and inversion of anaplastic lymphoma kinase (ALK), but the optimal procedures for the fixation or staining of the sample before FISH evaluation have not been established. We investigated whether ALK gene status determined by FISH in a prospectively enrolled case series of patients was affected by fixation and staining. METHODS: One hundred and fifteen cytological samples were obtained by transbronchial needle aspiration (TBNA) or endobronchial ultrasound (EBUS)-TBNA from 109 patients with NSCLC. All samples were evaluated for epidermal growth factor receptor (EGFR) mutation by pyrosequencing and for ALK rearrangement by FISH. Specimens for ALK determination had been fixed with Cytofix(®) and/or Carnoy's solution or 10% formalin (cell blocks) and variously stained. RESULTS: Sixteen (14%) of the 115 samples were mutated for EGFR and 99 (86%) showed wild-type EGFR status. Of these 115 samples, 79 (69%) were negative for echinoderm microtubule-associated protein like 4 (EML4)-ALK translocation, nine (8%) were positive and 27 (23%) were unevaluable. In particular, 19 (26%) of the 72 Papanicolaou-stained smears fixed with Cytofix were unevaluable because of inadequate samples or cell overlapping; neither of the two May-Grünwald-Giemsa-stained samples were evaluable. Ten of 17 smears used for rapid on-site evaluation (ROSE) and immediately post-fixed in Carnoy's solution or 80% alcohol were evaluable. CONCLUSIONS: In this series, smears were unevaluable as a result of inadequate samples, cell overlapping or lack of fixation performed immediately after FNA.

Impact of Candidate Genetic Polymorphisms in Prostate Cancer: An Overview.

In the last few years, the presence of single nucleotide polymorphisms (SNPs) has been investigated in many tumors as predictor of disease aggressiveness and clinical outcome. We searched for relevant articles from 1998 to 2015 about the impact of SNPs in prostate cancer. Particularly, in this article, we review the pathogenetic, prognostic and predictive significance of gene polymorphisms in prostate tumor, providing a brief overview of studies in which the possible role of genetic variants was investigated in clinical settings. Because conflicting results often emerge about the impact of gene polymorphisms in prostate cancer, further larger studies are warranted in order to introduce gene polymorphism into clinical practice as biomarkers.

Circulating cell-free AR and CYP17A1 copy number variations may associate with outcome of metastatic castration-resistant prostate cancer patients treated with abiraterone.

BACKGROUND: This study aimed to investigate copy number variations (CNVs) of CYP17A1 and androgen receptor (AR) genes in serum cell-free DNA collected before starting abiraterone in 53 consecutive patients with castration-resistant prostate cancer (CRPC). METHODS: Serum DNA was isolated and CNVs were analysed for AR and CYP17A1 genes using Taqman copy number assays. The association between CNVs and progression-free/overall survival (PFS/OS) was evaluated by the Kaplan-Meier method and log-rank test. RESULTS: Median PFS of patients with AR gene gain was 2.8 vs 9.5 months of non-gained cases (P < 0.0001). Patients with CYP17A1 gene gain had a median PFS of 2.8 months vs 9.2 months in the non-gained patients (P = 0.0014). A lower OS was reported in both cases (AR: P < 0.0001; CYP17A1: P = 0.0085). Multivariate analysis revealed that PSA decline ⩾ 50%, AR and CYP17A1 CNVs were associated with shorter PFS (P < 0.0001, P = 0.0004 and P = 0.0450, respectively), while performance status, PSA decline ⩾ 50%, AR CNV and DNA concentration were associated with OS (P = 0.0021, P = 0.0014, P = 0.0026 and P = 0.0129, respectively). CONCLUSIONS: CNVs of AR and CYP17A1 genes would appear to be associated with outcome of CRPC patients treated with abiraterone.

Challenge in differential diagnosis of a liver mass histologically defined as a metastatic lesion from an occult primary intestinal tumour. The importance of clinical findings and the limitations of histology and molecular profiles. A case report.

Differential diagnosis of liver lesion in the absence of proven primary tumor is still a challenge. We experienced a case of an asymptomatic 14 cm lesion of right hemiliver in a 67 year-old man submitted to right hepatectomy in December 2010. One year before the patient underwent to endoscopic removal of a tubular adenoma of the right colon. Preoperative diagnosis was supported by ultrasound, CT scan, PET and liver biopsy. The patient received 6 cycles of preoperative chemotherapy (FOLFOX) with down-staging of the lesion diameter. Immunohistochemistry on the surgical specimen showed positivity for cytokeratins 19 and 20, CEA, MUC-2, negativity for cytokeratin 7 and a-fetoprotein. Moreover, the neoplastic cells showed a focal positivity with lower intensity for MUC-1 and MUC-5AC. The immunohistochemical profile suggested the possibility of a metastatic tumour from the large bowel, without excluding a primitive mucinous cholangiocarcinoma with intestinal phenotype. At 6 months after intervention, the patient was submitted to chemotherapy (FOLFOX). At present he is in good condition, without radiological signs of recurrence. Oncologists must evaluate the possible benefits of further adjuvant treatments based on the differential diagnosis between a primitive or metastatic liver tumour. In conclusion, correct diagnosis of liver masses is mandatory and remains a challenge that can differentiate either follow-up or surgical and adjuvant treatment. Histology and immunohistochemistry must be related to clinical findings as they may not always be sufficient to reach a correct final diagnosis, and can even be confusing. At present, molecular biology cannot be considered a helpful for diagnosis in these cases.

Combining cytology, TRAP assay, and FISH analysis for the detection of bladder cancer in symptomatic patients.

BACKGROUND: There is a need to improve the performance of urine cytology in bladder cancer diagnosis. We assessed the diagnostic performance of (i) telomerase activity detected by telomeric repeat amplification protocol (TRAP) assay, (ii) cytology and TRAP assay in parallel, (iii) cytology in parallel with the in-series combination of TRAP assay and FISH analysis, and (iv) the in-series combination of TRAP assay and FISH analysis. PATIENTS AND METHODS: Cross-sectional study of 289 consecutive patients who presented with urinary symptoms at a north Italian hospital between 2007 and 2008. All underwent cystoscopy and cytology evaluation, and conclusive results were available for TRAP assay and FISH analysis. RESULTS: Sensitivity and specificity were 0.39 and 0.83, respectively, for cytology; 0.66 and 0.72 for TRAP; 0.78 and 0.60 for the combination of cytology and TRAP; 0.78 and 0.78 for the combination of cytology, TRAP, and FISH; and 0.65 and 0.93 for the combination of TRAP and FISH. All differences versus cytology alone were significant (P ≤ 0.011). CONCLUSION: Compared with cytology alone, the combination of cytology, TRAP, and FISH provided the best trade-off between increase in sensitivity and loss in specificity, especially among non-bleeding patients, low-grade cancers, and early-stage cancers.

RT-PCR determination of maspin and mammaglobin B in peripheral blood of healthy donors and breast cancer patients.

BACKGROUND: The aim of the present study was to evaluate the accuracy of two markers, maspin and mammaglobin B, singly or in combination, to detect breast cancer. To define better the potential and limits of the two markers for diagnostic purposes, blood positivity was analyzed in relation to clinical, pathological and biological tumor characteristics. PATIENTS AND METHODS: The markers were determined in peripheral blood (PB) samples from 27 healthy donors and 140 previously untreated patients using nested reverse transcriptase polymerase chain reaction (RT-PCR). RESULTS: Positivity for maspin in blood samples was observed in 24% of patients with an 89% specificity. For mammaglobin B, positivity was observed in 7% of patients and never in healthy donors. The presence of maspin was correlated with cell proliferation of the primary tumor (P = 0.015), whereas mammaglobin B positivity correlated with pathological stage (P = 0.013). The presence of either marker was significantly related to nodal status. CONCLUSIONS: Our results indicate that the two markers in association could represent a potentially useful non-invasive tool to detect breast cancer. The validation of these markers as indicators of high risk of relapse is ongoing in a series of patients with an adequate follow-up.

Potential of telomerase expression and activity in cervical specimens as a diagnostic tool.

AIMS: To evaluate the potential use of the immunohistochemical expression of telomerase and the measurement of its activity as diagnostic tools in the uterine cervix. METHODS: The fluorescent telomeric repeat amplification protocol (TRAP) assay was used to evaluate telomerase activity in a series of 43 cervical scrapings. Twenty five cases were cytologically classified as inflammatory, and/or metaplastic, and/or acanthotic, and 18 cases presented cell alterations compatible with mild, moderate, or severe cervical intraepithelial neoplasia (CIN). Immunohistochemistry was performed on a retrospective series of 86 archival, paraffin wax embedded blocks using a recently developed anti-hTERT (human telomerase reverse transcriptase) monoclonal antibody. RESULTS: Telomerase activity was expressed as arbitrary enzymatic units (AEU). Median values were 38.0 AEU for inflammatory non-dysplastic cell specimens, 33.5 AEU for CIN I, 41.0 AEU for CIN II, and 28.0 AEU for CIN III. The median percentage of immunoreactive dysplastic cells, as detected by immunohistochemistry, was significantly (p = 0.024) lower in CIN I (45%) than in more severe dysplastic (CIN II 70%, CIN III 80%) lesions. In contrast, no differences were seen in the enzymatic activity detected by the TRAP assay among the different dysplastic lesions. CONCLUSIONS: These data indicate that, using a molecular extra situ method, the telomerase activity of inflammatory and non-dysplastic elements masks the expected differences between mild and severe dysplasia. Conversely, an in situ approach permits the accurate identification of telomerase positive dysplastic cells.

BRCA1 mutations and polymorphisms in a hospital-based consecutive series of breast cancer patients from Apulia, Italy.

INTRODUCTION: Hereditary breast cancer has been partly attributed to germline mutations in the BRCA1 gene that are deleterious for BRCA1 protein activity. This paper analyzes the incidence and characteristics of detectable BRCA1 mutations and polymorphisms in a hospital-based consecutive series of breast cancer patients from southern Italy to investigate the incidence and the association of these molecular alterations with breast cancer biology and family history. METHODS: One hundred cases with familial characteristics were selected from a consecutive series of 511 patients with a first diagnosis of breast cancer. DNA from peripheral blood was screened for whole BRCA1 gene mutations utilizing dHPLC as a pre-screening analysis and automatic DNA sequencing for the identification of specific alterations. RESULTS: In the overall series of 511 patients, 100 had a family history of breast cancer and were investigated for BRCA1 mutations. Two types of BRCA1 mutations were identified, 5382insC in six cases and 4566delA in one case. The 5382insC mutation was present in two out of six cases with ovarian cancer while 4566delA in one case of male cancer. The most frequent missense polymorphisms were E1038G, P871L, K1183R in exon 11, S1613G, M1652I in exon 16 and D1778G in exon 22. Confirming what found in previous studies, patients in whom pathological BRCA1 mutations were detected had early-onset breast cancer (p=0.05), positive nodal status (p=0.05), lower ER (p=0.02) and PgR (p=0.01) content. Interestingly, the K1183R polymorphism and, less strongly, S1613G polymorphism were associated to mutational risk (K1183R: OR 0.1 p=0.03; S1613G: OR 2.7 p=0.08). CONCLUSION: Mutations in the BRCA1 gene are frequent also in our consecutive series of patients from southern Italy. An association between two detected single nucleotide polymorphisms (SNPs) and BRCA1 mutational risk was ascertained. Finally, we confirm the fact that peculiar clinical-pathological features seem to characterize patients with a family history of breast cancer and BRCA1 alterations.

Telomerase activity detected by quantitative assay in bladder carcinoma and exfoliated cells in urine.

Early diagnosis is one of the most determining factors for patient survival. The detection of telomerase activity is a potentially promising tool in the diagnosis of bladder and other types of cancer due to the high expression of this enzyme in tumor cells. We carried out a quantitative evaluation of telomerase activity in urine samples in an attempt to determine a cut-off capable of identifying cancer patients. Telomerase activity was quantified by fluorescence TRAP assay in urine from 50 healthy volunteers and in urine and bioptic tumor samples from 56 previously untreated bladder cancer patients and expressed in arbitrary enzymatic units (AEU). Telomerase activity in urine ranged from 0 to 106 AEU (median 0) in healthy donors and from 0 to 282 AEU (median 87) in patients with cancer. A telomerase expression higher than the cut off value determined by receiver operating characteristic (ROC) analysis was observed in 78% of cases, regardless of tumor grade and in 71% (15/21) of cases of nonassessable or negative cytology. The quantitative analysis of telomerase activity in urine enabled us to define cut-off values characterized by different sensitivity and specificity. Cytologic and telomerase determination, used sequentially, enabled us to detect about 90% of tumors.

Detection of germline BRCA1 mutations by Multiple-Dye Cleavase Fragment Length Polymorphism (MD-CFLP) method.

We describe the Multiple-Dye Cleavase Fragment Length Polymorphism (MD-CFLP) method set up for a sensitive and preliminary rapid screening of BRCA1 mutations. We analysed exons 11 and 16, which are known to cover slightly more than 70% of the whole coding region of the gene, subdivided into 4 amplicons and labelled with different fluorescent dUTPs. MD-CFLP was first utilised on a panel of 30 DNA samples in which the presence of single-base substitutions or small deletions/insertions had been previously identified by direct sequencing as gold standard, in order to define the optimal conditions in terms of PCR amplification and temperature of digestion. In a second step, we blindly analysed 21 DNA samples by MD-CFLP to verify its reliability. The sensitivity and specificity of MD-CFLP were both 100% in the first study, and 80% and 94%, respectively, in the blind sample assay. Our results demonstrate the capability of the MD-CFLP method to detect DNA sequence alterations in fragments of more than 1 kb. We conclude that CFLP is a powerful tool in mutational analysis, offering reliable results in a shorter time and at a lower cost than conventional methods, and its potential can be enhanced when internal fluorescent labelling and laser detection are used.

Multiple detection of genetic alterations in tumors and stool.

Detection of genetic alterations in exfoliated intestinal cells in stool could represent an alternative, noninvasive tool for the screening of colorectal tumors. To verify this, we analyzed p53 and K-ras mutations and microsatellite instability on 46 cases of colorectal cancer and compared the presence of molecular alterations in tumor tissue and stool samples from individual patients. p53 exons 5-8 and K-ras exons 1-2 were analyzed by denaturing gradient gel electrophoresis. For the microsatellite instability, a set of 5 microsatellite markers (D2S123, D5S346, D17S250, BAT25, and BAT26) was evaluated. In the 18 healthy individuals, no genetic alterations in either tissue or stool were detected. p53 mutations were detected in 17 (37%), K-ras alterations in 15 (33%), and microsatellite instabilities in 5 (11%) of the 46 tumors analyzed. In a side study, we analyzed the correlation in genetic alteration profiles between tumors and macroscopically normal or healthy tissue from the same patient. The presence of at least one molecular alteration in tumor was observed in 31 (67%) of the cases. p53, K-ras mutations, and microsatellite instabilities were detected in stool samples in 18, 40, and 60% of patients with tumors harboring the same alterations. Due to the largely complementary presence of p53 and K-ras mutations in tumors, the use of highly sensitive procedures for stool analysis could offer a means competitive with colonoscopy and the fecal occult blood test.

Correlation between p53 gene mutations and p53 protein accumulation evaluated by different methodologies.

Mutations in the p53 gene are the most common genetic alterations in many tumour histotypes. Many of these mutations induce conformational changes resulting in p53 protein stabilisation and consequently an accumulation detectable with immunochemical methods. Available data on the correlation between p53 gene alterations and p53 overexpression widely vary. In this study we analysed the correlation between p53 gene alterations detected by DGGE, SSCP and sequencing and protein expression detected by flow cytometric and immunohistochemical approaches by using PAb 1801 antibody. The study was performed on 21 bladder tumours and 10 cell lines derived from different tumour histotypes as representative of different methodologic problems which can be met starting from different types of biological material. The best correlation (81%) was observed between p53 mutations and FCM results, using a double evaluation criterion for the latter which includes the percentage of positive cells and "delta values", evaluated as the difference between the mean values of Pab 1801 stained cells and isotypic control. The high correlation obtained between results from this FCM double criterion and p53 gene mutations is a good starting point for the analysis on large series of tumours and for a multiparameter FCM analysis including p53 protein levels.

Relation between food habits and p53 mutational spectrum in gastric cancer patients.

p53 tumour suppressor gene mutations were analysed in gastric cancer in relation to food habits and social class in 56 patients from a high risk region of Italy. Exons 5-8 were analysed with DGGE method on amplified DNA from formalin-fixed paraffin-embedded samples. All p53 mutations were observed in patients belonging to low social class and the majority of mutations were found in intestinal type cancers. A positive association was also found with low raw vegetables, fresh, dried and preserved fruits, and ascorbic acid intake. Moreover, specific types of mutations were significantly associated with particular factors, thus suggesting the presence of specific molecular etiologic process in stomach carcinogenesis.

Microsatellite instability in colorectal-cancer patients with suspected genetic predisposition.

Hereditary non-polyposis colorectal cancer (HNPCC) is a dominantly inherited syndrome linked to DNA-mismatch-repair (MMR) gene defects, which also account for microsatellite instability (MSI) in tumour tissues. Diagnosis is based mainly on family history, according to widely accepted criteria (Amsterdam Criteria: AC). Aim of this work was to assess MSI in colorectal-cancer patients with suspected genetic predisposition, and to verify whether MSI represents a tool to manage MMR gene (hMSH2 and hMLH1) mutation analysis. We investigated 13 microsatellites (including the 5 NCI/ICG-HNPCC markers) in 45 patients with suspected hereditary predisposition (including 16 subjects from HNPCC families fulfilling the AC). We found MSI-H (high frequency of instability, i.e., in > or =30% of the markers) in 85% of the HNPCC patients and in 16% of the non-HNPCC subjects. The 5 NCI/ICG-HNPCC microsatellites proved to be the most effective in detecting MSI, being mononucleotide repeats the most unstable markers. We investigated the association between hMSH2- and hMLH1 gene mutations and MSI. Our results indicate that AC are highly predictive both of tumour instability and of MMR-gene mutations. Therefore, as the most likely mutation carriers, HNPCC subjects might be directly analyzed for gene mutations, while to test for MSI in selected non-HNPCC patients and to further investigate MMR genes in MSI-H cases, appears to be a cost-effective way to identify subjects, other than those from kindred fulfilling AC, who might benefit from genetic testing.

Gastric cancer: epidemiologic and biological aspects.

Although temporal trends and regional-racial variations in gastric cancer incidence have led to the formulation of different hypotheses, no definite association has been seen between this disease and any behavioural or genetic determinant. In fact, several aetiological factors have been associated with risk of gastric cancer, but not without controversy. Various studies have suggested that genetic factors might be of importance in the pathogenesis of gastric tumours. In fact, stomach carcinoma occurs more frequently among close relatives of affected individuals than in the general population and some of the most common pre-cancerous lesions seem to be genetically determined. In the light of this circumstantial evidence, we decided to investigate the role of various genetic alterations in gastric cancer in order to study their relationship with aetiopathogenesis and disease progression and their value as indicators of risk and prognosis. Our main areas of interest were: i. c-Ha-ras locus polymorphism; ii, truncated c-myc gene variant; iii. loss of heterozygosity, iv. p53 gene mutations; v. oncogene amplification; vi. oncogene amplification proliferative activity and their relation to prognosis in gastric cancer.

Mean age of tumor onset in hereditary nonpolyposis colorectal cancer (HNPCC) families correlates with the presence of mutations in DNA mismatch repair genes.

Fourteen Italian families affected with hereditary nonpolyposis colorectal cancer (HNPCC) were screened for germline mutations at three DNA mismatch repair (MMR) genes, MSH2, MLHI, and GTBP, by using a combination of different methods that included an in vitro synthesized protein assay, single-strand conformation polymorphism analysis, and direct sequencing. DNA alterations were observed in six instances, including a single base deletion in MSH2 exon 14, an A-to-G transition in the splice donor site of MLHI exon 6, and two missense mutations in MLHI exons 5 and 9. A previously reported common mutation affecting the splice donor site of MSH2 exon 5 was identified in two families. No mutations were detected in the GTBP gene. In total, eight of 16 Italian HNPCC families (50%), including two previously reported kindreds, were found to carry a mutation in MMR genes. We compared the mean age of colorectal cancer onset in the index cases (three patients for each family) between the two groups of kindreds, those with identified mutation vs. those without, and found that the first had a significantly lower value (43.0 vs. 53.7 years, P = 0.014). This finding suggests that HNPCC families with a more advanced age of tumor onset are less likely to be associated with known MMR genes.

Microsatellite instability in intestinal- and diffuse-type gastric carcinoma.

To investigate the role of genetic instability in the development of intestinal- and diffuse-type gastric cancers, six microsatellite loci were analysed in 98 carcinomas of the two main histotypes, at both early and advanced stages of progression, and in five preneoplastic lesions. RER+ phenotype frequency proved to be significantly higher (P = 0.013) in intestinal (23 per cent) than in diffuse cancers (5 per cent) and slightly higher in advanced (19 per cent) than in early (12 per cent) tumours. When comparing early and advanced tumours of the same histotype, a similar frequency was found for diffuse tumours (4 per cent vs. 6 per cent), and an increase from 19 to 30 per cent for intestinal cancers. Instability at more than one locus was limited to intestinal tumours and replication errors were also detected in an intestinal dysplasia. On the whole, these data suggest that genetic instability has an important and early role in gastric carcinogenesis of the intestinal type and a less important role in gastric carcinogenesis of the diffuse type. Most tumours of this panel had previously been characterized for p53 gene mutations. p53 screening was extended to all samples, to investigate the possible association between gene mutations and microsatellite instability. Analysis showed a trend (P = 0.07, Fisher's exact test) towards a negative association between these two genetic lesions in tumours of the intestinal type.

Microsatellite instability and mutations of p53 and TGF-beta RII genes in gastric cancer.

To investigate the molecular mechanism of gastric carcinogenesis, we analyzed genetic instability and p53 gene mutations in 40 primary gastric carcinomas. Tumor samples were from untreated patients with no family history suggestive of genetic predisposition to cancer. We screened six microsatellite loci by the polymerase chain reaction (PCR) method, and exons 5-8 of the p53 gene by the PCR-based denaturing gradient gel electrophoresis and sequencing techniques. Microsatellite instability was detected in 32.5% (13/40), and gene mutations in 40% (16/40), of the tumors analyzed. No statistically significant associations were found between genetic alterations and clinico-pathological variables (with the exception of diffusion of lymph node metastases, which was inversely associated with the presence of microsatellite alterations; P < 0.01). Interestingly, a negative association was found between genetic instability and p53 gene mutations: 11 out of 13 tumors showing instability proved to carry a nonmutated p53 gene versus 2/13 carrying a mutated gene (P = 0.03). These observations suggest that genetic instability and p53 gene mutations play a crucial role in the gastric carcinogenic process, but likely act through distinct pathways during cancer development. However, genetic instability is not in and of itself neoplastic. Therefore, we investigated whether insertion/deletion mutations of the polyadenine tract within the transforming growth factor-beta type II receptor gene (TGF-beta RII) were frequently present in gastric tumors with an RER+ (replication error) phenotype. We found RII mutations in 8/40 (20%) samples: mutations were present in 7/13 (54%) RER+ tumors versus 1/27 (4%) RER- cases (P < 0.001).

Intestinal and diffuse gastric cancers arise in a different background of Helicobacter pylori gastritis through different gene involvement.

Investigation of extensively sampled nontumor gastric mucosa from 205 early gastric cancers showed Helicobacter pylori colonization in 85% of cases, including 100% of diffuse and 78% (83% in 97 cases with Swiss rolls) of glandular or mixed cancers. Intestinal metaplasia, including its type III variant, was prominent in the mucosa associated with glandular and mixed (but not diffuse) early cancers. Both glandular (usually called "intestinal") and diffuse-type cancers showed admixtures of intestinal and gastric tumor cell phenotypes. Both p53 gene mutations and p53 protein immunostaining were essentially restricted to glandular or mixed cancers and associated dysplastic lesions. Their appearance in the advanced stage of diffuse cancer was partly due to a change of the histologic pattern from glandular to diffuse during progression of some tumors. Loss of laminin, beta I integrin, or zonula adherens junctions was a common finding in both early and advanced diffuse cancer. It is concluded that two main pathways operate in gastric carcinogenesis, both starting from H. pylori gastritis and both leading to phenotypically variable, often mixed gastric/intestinal tumor growth. However, only one of the two pathways involves intestinal metaplasia, its type III variant, p53 gene alteration, and dysplasia to end in glandular cancer. In the other pathway, diffuse cancer apparently arises directly from hyperplastic, sometimes atypical necks of mostly nonmetaplastic gastric glands, through primary involvement of genes affecting cell-cell and cell-matrix junctional proteins.