Masanori Terashima

Inactivation of the E-Cadherin Gene in Primary Gastric Carcinomas and Gastric Carcinoma Cell Lines

We investigated the E (epithelial)‐cadherin gene for mutations and loss of heterozygosity (LOH) in 24 primary gastric carcinomas (12 differentiated and 12 undifferentiated types, including 3 signet‐ring cell carcinomas), as well as 4 gastric carcinoma cell lines of the undifferentiated type (MKN‐45, GCIY, HGC‐27 and GT3TKB). We utilized PCR‐SSCP and RT‐PCR followed hy direct sequencing to detect gene mutations and skipped exons, and RT‐PCR‐SSCP to examine LOH. In primary carcinomas, gene mutations or skipped exons, were detected in 4 of 9 (44%) undifferentiated carcinomas of the scattered type, including 2 signet‐ring cell carcinomas, and in none of the 3 undifferentiated carcinomas of the adherent type and 12 differentiated carcinomas. Demonstrated mutations of the E‐cadherin gene included an 18 bp deletion (codon 418‐423) and a 3 bp deletion (codon 400, calcium‐binding domain), both located in exon 9. Skipping of exon 9 with a 1 bp insertion at codon 337, and skipping of exon 8 with a 1 bp deletion at codon 336, also were detected. LOH was confirmed in all of the carcinomas in which gene mutations or skipped exons (3/3 informative cases) were demonstrated. The MKN‐45 cell line exhibited an 18 bp deletion at the exon 6‐intron 6 boundary with loss of the wild‐type allele, and 2 of the remaining 3 cell lines (HGC‐27 and GT3TKB) had lost expression without detectable structural alteration of the E‐cadherin gene. These data provide support for classic two‐hit inactivation of the E‐cadherin gene in a high percentage of undifferentiated carcinomas of the scattered type.

ALLELOTYPE OF ADENOMA AND DIFFERENTIATED ADENOCARCINOMA OF THE STOMACH

The molecular mechanism of gastric tumourigenesis has not yet been clarified, although investigators have postulated that differentiated adenocarcinoma may arise from pre‐existing adenoma, similarly to the colorectal adenoma–carcinoma sequence. An allelotype analysis has been performed to identify chromosomal regions which are frequently deleted in gastric tumours and to examine the significance of the adenoma–carcinoma sequence in gastric tumourigenesis. Forty‐five gastric tumours, 20 adenomas, and 25 differentiated adenocarcinomas were examined for loss of heterozygosity (LOH) using 39 microsatellite markers covering each non‐acrocentric chromosome arm. Frequent LOH in the adenocarcinomas was observed on chromosomes 2q (33 per cent), 4p (33 per cent), 5q (50 per cent), 6p (33 per cent), 7q (43 per cent), 11q (36 per cent), 14q (38 per cent), 17p (45 per cent), 18q (36 per cent), and 21q (40 per cent). In contrast, the incidence of LOH in adenomas did not exceed 10 per cent at any of the loci examined. In addition to the p53 gene on 17p and the DCC gene on 18q, which are known to be frequently deleted in differentiated adenocarcinomas of the stomach, other unknown tumour suppressor genes on the above‐mentioned chromosomes may also be inactivated. These observations suggest that the adenoma–carcinoma sequence is not a major pathway in gastric tumourigenesis.

A Pilot Phase II Study of Capecitabine in Advanced or Recurrent Gastric Cancer

Objectives: To evaluate the efficacy and safety of capecitabine in patients with advanced or recurrent gastric cancer, we conducted a pilot phase II study in Japan. Methods: Patients with advanced or recurrent gastric cancer were given oral capecitabine 828 mg/m2 twice daily for 3 weeks, followed by 1 week of no treatment. Two or more cycles were administered. From July 1996 to December 1997, 32 patients were enrolled in the study. The response to capecitabine was evaluated in 31 patients, excluding 1 found to be ineligible. Results: The overall response rate was 19.4% (6/31, 95% confidence interval: 7.5–37.5%). The median duration of response was 124.5 days, the median time to disease progression 85.0 days, and the median survival time 247.5 days. Drug-related adverse events of grade 3 or higher were infrequent: in 2 patients (6.3%) total bilirubin concentration increased, and 1 patient (3.1%) each had elevation of GOT, anemia, lymphopenia, increased creatinine, and hand-foot syndrome. No patient had gastrointestinal toxicity of grade 3 or higher. Conclusion: Capecitabine was suggested to be safe and effective in the treatment of advanced or recurrent gastric cancer. Further phase II studies of capecitabine on a large scale are warranted.

Loss of heterozygosity during the development and progression of differentiated adenocarcinoma of the stomach

In a recent allelotypic analysis of differentiated adenocarcinoma of the stomach, loss of heterozygosity (LOH) was found frequently on chromosomes 2q, 4p, 5q, 6p, 7q, 11q, 14q, 17p, 18q, and 21q. To clarify the sequence of these chromosomal losses during gastric carcinogenesis, microsatellite analysis of the chromosome arms described above was performed in 25 early and 29 advanced differentiated adenocarcinomas of the stomach. LOH on these chromosome arms fell within a range of 20–50 per cent. On 4p, 7q, 14q, 17p, and 21q, LOH was detected at a similar frequency in both early and advanced carcinomas, while LOH on 2q, 5q, 6p, 11q, and 18q was observed more than twice as frequently in advanced than in early lesions. Mean fractional allelic losses (FALs) were 0·221 in early and 0·413 in advanced carcinomas, representing a significant difference (P<0·05). These results suggest that LOH on 4p, 7q, 14q, 17p, and 21q is a relatively early event, while LOH on 2q, 5q, 6p, 11q, and 18q typically accumulates during the progression of gastric carcinogenesis.

Analysis of genetic and epigenetic alterations of the PTEN gene in gastric cancer

Abstract. The PTEN tumor suppressor gene on 10q23.3, responsible for the Cowden and Bannayan-Zonana syndromes, encodes a dual-specificity phosphatase able to dephosphorylate both tyrosine phosphate and serine/threonine phosphate residues. Mutational inactivation of PTEN has been reported in various malignancies, including endometrial cancers, ovarian cancers, and glioblastomas. In this study, we investigated PTEN gene mutations in 10 gastric cancer cell lines and 58 primary gastric cancers by polymerase chain reaction single strand conformation polymorphism (PCR-SSCP). Hypermethylation of promoter region CpG islands, an alternative mechanism of gene inactivation to coding region mutations, was also evaluated by methylation specific PCR (MSP). Only one (1.7%) of the 58 primary tumors carried a somatic 5-bp deletion in intron 7 of PTEN, which did not alter the mRNA sequence, and no mutations were detected in any of the cell lines. Similar levels of PTEN mRNA expression were observed in all cell lines and primary tumors studied by RT-PCR, and PTEN promoter CpG islands remained unmethylated. Therefore, we conclude that PTEN does not participate in gastric carcinogenesis as a tumor suppressor gene.

Commonly deleted regions on the long arm of chromosome 21 in differentiated adenocarcinoma of the stomach

During an allelotype analysis of differentiated adenocarcinoma of the stomach, we observed frequent loss of heterozygosity (LOH) on several chromosomes including the long arm of chromosome 21 (21q). Therefore, we analyzed DNA isolated from 45 tumors for LOH at 10 loci on 21q by using polymorphic microsatellite markers. In 20 (44%) of 45 tumors, we detected LOH at single or multiple loci on 21q. Deletion mapping of these 20 tumors revealed two separate commonly deleted regions. Our findings suggest that 21q contains at least two potential tumor suppressor genes which play crucial roles in the development of differentiated adenocarcinoma of the stomach. Genes Chromosom. Cancer 18:318–321, 1997.

Evaluation of the ratio of lymph node metastasis as a prognostic factor in patients with gastric cancer

Background. Lymph node metastasis in patients with gastric cancer is one of the important prognostic factors. However, there is no consensus concerning the best classification for lymph node metastasis as a prognostic factor. So, to evaluate the ratio of the number of metastatic lymph nodes to the total number of dissected lymph nodes (the ratio of LN meta) as a prognostic factor, we compared the ratio of LN meta with lymph node status according to the Japan Classification of Gastric Carcinoma and the total number of metastatic lymph nodes with multivariate analysis. Methods. Between 1991 and 1997, a total of 360 patients with primary gastric cancer who underwent gastrectomy with D2 or more extended lymph node dissection were included in this study. Ten kinds of prognostic factors and three types of different classifications for lymph node metastasis were analyzed by multivariate analysis using the Cox regression. Results. The average number of dissected lymph nodes and metastatic lymph nodes were 55.0 (range, 11–184) and 2.6 (range, 0–86), respectively. There were significant differences of the 5-year cumulative survival rates among each group of the ratio of LN meta (0%, 1%–9%, 10%–24%, and more than 25%). Age, tumor size, curability, and the ratio of LN meta were selected as independent prognostic factors by forward stepwise selection. The ratio of LN meta showed the highest hazard ratio by Cox regression. Conclusion. The ratio of LN meta appears to be an important prognostic factor and the best classification factor for lymph node metastasis.

The usefulness of the ATP assay with serum-free culture for chemosensitivity testing of gastrointestinal cancer

In order to predict a patients response to a drug prior to chemotherapy for gastrointestinal cancer, we developed an adenosine triphosphate (ATP) assay with serum-free culture (SF-ATPA) which enables fibroblast overgrowth to be suppressed. A total of 244 gastrointestinal cancer tissue samples were obtained from surgical resection. After enzymatic digestion, cells (2 x 10(4)/well) were cultured for 72 h with continuous exposure to drugs (single or combined use), and cell viability was evaluated by measuring the intracellular ATP level. 208 of 244 samples (85%) were considered to be evaluable in terms of drug response. There were no differences in the evaluability rates among the tumour types. A drug was judged as active using the criterion of < or = 50% reduction of the intracellular ATP level in a single-agent treated group compared with the level of the control. Similarly, in the combined drug treated group, drugs were considered as active using two different criteria (< or = 30% reduction of the intracellular ATP level for two drugs and < or = 20% for three drugs). Each tumour type had its own spectrum for chemosensitivity. Of 25 patients evaluated for assay-clinical correlations, 16 were examined by both single-agent and combined drug administration and the predictive accuracy of the assay was higher for the combined drug than for the single agent (88% versus 69%, respectively). In 25 patients, the true positive rate was 64% and the true negative rate was 100%, yielding an overall predictive accuracy of 84%. These results suggest that SF-ATPA may be useful for predicting drug response in patients with gastrointestinal cancer.

Microvessel count predicts metastasis and prognosis in patients with gastric cancer.

In order to evaluate the clinical relevance of angiogenesis in patients with gastric cancer, we investigated the microvessel count in gastric cancer tissues and compared the results with several clinicopathologic factors and prognosis.

Analysis of the fragile histidine triad gene in primary gastric carcinomas and gastric carcinoma cell lines

The FHIT (fragile histidine triad) gene has been isolated from the chromosome region 3p14.2, which includes the fragile site locus FRA3B and the breakpoint of the t(3;8) of familial renal carcinoma. FHIThas been suggested to be a candidate tumor suppressor gene for digestive tract carcinomas. To evaluate the significance of FHIT gene abnormalities in gastric carcinogenesis, we examined the allelic status and transcripts of the gene in 23 primary gastric carcinomas as well as in 7 gastric carcinoma cell lines. Four of the seven (57%) cell lines exhibited homozygous deletions of variable sizes at 3p14.2, all of which included D3S1300, which is located close to, or within, FRA3B. However, only 2 of 16 (13%) informative cases showed loss of heterozygosity at D3S1300 in the primary tumors. Direct analysis by reverse transcriptase polymerase chain reaction failed to reveal abnormal transcripts, including exon skipping and sequence changes, in the primary tumors or in the cell lines without homozygous deletions. These results suggest that FHIT gene abnormalities are infrequent in primary gastric carcinomas and that the frequent homozygous deletions seen in cell lines might simply reflect the plasticity of the genome at FRA3B under culture conditions. Genes Chromosom. Cancer 20:98–102, 1997.