Using mutant-allele-specific amplification procedure, patients with colorectal tumors were analyzed for K-ras point mutations in the stool and the tumor tissue. K-ras mutation of DNA purified from the stool was detected in 10 of 40 (25.0%) cancer patients, and in 3 of 10 (30.0%) adenoma patients. Otherwise, in the cases whose tumors contained the mutations, it was detected with the frequency of 71.4% in cancers, and 100% in adenomas. This frequency tended to decrease in the cancers of distal colon or small size, but there was no significant. This study suggested that stool analysis of genetic alterations would develop diagnostic method for colorectal cancer.

Oncogene is not categorized as a tumor marker in a strict sense, however, cancer related oncogens play an important role as a biomarker in hereditary malignant tumors in a wide sense. Various suppressor oncogenes have been identified in the autosomal dominant hereditary diseases such as APC, in familial adenomatous polyposis, p53 in Li-Fraumeni syndrome and BRACA 1 and 2 in breast cancer. By identifying the mutation site or deletions of germ line, it is possible to make a presymptomatic diagnosis of those hereditary malignant tumors. There is splendid progress in understanding of DNA repair mechanism. Recently, the mismatch repair genes were cloned as a causing gene of HNPCC. There are another group of genes called nucleotide excision repair genes which are causative genes of various autosomal recessive hereditary diseases such as xeroderama pigmentation. Pro and cons of presymptomatic diagnosis of familial adenomatous polyposis were discussed in a series of 72 patients among 42 family trees.

We investigated the cell clonality of 12 cases of female solitary hepatocellular carcinoma (HCC) that were associated with hepatitis virus infection. The clonal origin of HCC could be assessed by the method based on restriction fragment length polymorphism (RFLP) of X-chromosome-linked androgen receptor gene (AR) and phosphoglycerate kinase (PGK) gene, taking advantage of random inactivation of one of two X-chromosomes by methylation in females. We extracted DNA samples from both fresh and paraffin-embedded specimens of the same lesion as a source of DNA sample for polymerase chain reaction (PCR). Consequently, it was possible to use methylation-sensitive restriction enzymes and PCR to study differential methylation patterns among alleles of these genes for both DNA samples. The RFLPs of AR gene and PGK gene were found in eight of 12 cases and five of 12 cases, respectively. There were two cases which had no RFLPs in either AR gene or PGK gene. All cases of HCC which had RFLP in either AR gene or PGK gene demonstrated monoclonal origin of the tumor regardless of their histologic patterns.

To investigate the change of numerical aberrations of chromosome 17 or 18 during progression of Colorectal carcinoma, we applied FISH, using chromosome specific probes, to 3 colorectal primary carcinomas and a total of 4 metachronous metastatic lesions in the liver. We also investigated the relationship between the DNA content of whole nucleus and the number of chromosome 17 by multiparametric analysis using autostage cytofluorometry. When compared with the results for primary lesions, in 4 metastatic lesions, the population of tetrasomy 17 cells increased in 2 lesions, and that of monosomy 18 cells increased in 1 lesion. The nuclear DNA contents histograms for disomy 17 and aneusomy 17 cells analyzed individually for 3 primary and 4 metastatic lesions, showed almost no appreciable difference of peak DNA values. These results indicate that the numerical aberrations of chromosome 17 and 18 accumulate during the progression of colorectal carcinomas, and that the subpopulation of aneusomy 17 consist mainly of those cells that involve the numerical aberrations in a few chromosomes.

To investigate whether genetic changes of the p53 gene and genetic defects in DNA-mismatch repair systems are involved in progression of colorectal carcinomas (CRCs), we examined loss of heterozygosity (LOH) on 17p and mutations in exon 5 through 8 of the p53 gene as well as replication errors (RER) at four microsatellite loci in DNAs from 108 CRCs at all clinical stages (20 Dukes A, 40 Dukes b, 48 Dukes C). We observed that LOH on 17p and/or p53 mutation were detected in 93% of Dukes A carcinomas and in 84% of Dukes C carcinomas, suggesting that the p53 gene is mutated and/or deleted before carcinoma has been produced. RER-positive phenotype was observed in approximately 30% of CRCs, irrespective of clinical stage. These results suggest that genetic instability is likely to play an important role in development of a subpopulation of sporadic CRCs, but not in progression of CRCs. In addition, we found no significant association between genetic alterations and progression of CRCs. We consider that genetic defects in DNA-mismatch repair pathway do not necessarily promote genomic instability at the p53 sequences in CRCs.

A 18-year-old woman underwent total colectomy for familial adenomatous polyposis. In order to clarify the significance of K-ras mutations in early colorectal carcinogenesis, K-ras mutations were analyzed in multiple adenomas by PCR-SSCP method. A total of 256 adenomas were found throughout the entire colon and rectum, and the distribution was a sparse type. The correlation between K-ras gene and clinicopathological factors was examined in 90 adenomas. There was no correlation among K-ras mutations and anatomical distribution, or morphological classification, but K-ras mutation was more frequent in severe compared with slight atypia. We investigated the correlation between the size of adenoma in the horizontal and vertical directions and K-ras mutation. K-ras mutation was more frequent in the horizontal size greater than 6 mm in diameter, and also more frequent in vertical size greater than 20 mm in height. It was concluded that the adenomas detecting K-ras mutations might have proliferating potential, and would be applied to determine polypectomy.

During the process of invasion, tumor cells must detach from the primary neoplasm and degrade host stroma. E-cadherin is responsible for the cell-cell adhesion and collagenase IV is the one of the matrix metalloproteinases. We determined whether the levels of mRNA for E-cadherin and collagenase IV were differently expressed within 12 cases of early and 13 cases of advanced gastric cancers using a rapid calorimetric in situ hybridization assay for mRNA. In 6 of 12 early cancers, we found a decreased expression of E-cadherin mRNA in the invasion edge compared to the main tumor. In advanced gastric cancers, 3 out of 13 cancers also exhibited this finding. Higher expression of the collagenase IV at the invasion edge of the tumor compared to the main tumor was observed in half of the early and advanced gastric cancer cases. Inverse expression levels of E-cadherin and collagenase IV mRNA were observed in 6 of 12 early cancers. However, only one of 13 advanced cancer cases expressed the same finding.

Nuclear DNA ploidy analysis was studied in patients with hepatocellular carcinoma (HCC) who underwent hepatic resections. These patients were classified three groups according to the following prescriptions. Group A (n = 100) was a group of patients of which excluded ones treated by absolute non-curative resection, Group B (n = 43) was patients who underwent absolute curative resection or relative curative resection, and Group C (n = 81) was patients whose tumor sizes were more than 2 cm in Group A. Aneuploid pattern was found in 59 cases (59.0%) in Group A, 22 cases (51.2%) in Group B and 54 cases (66.7%) in Group C. The rate of aneuploid pattern was significantly higher in patients with carcinomas more than 2 cm in diameter, fc-inf positive growth, Stage III + IV and PCNA LI > or = 40% in Group A, those with carcinomas more than 2 cm in diameter, fc-inf positive growth, im-positive and Stage III + IV in Group B, and those with PCNA LI > or = 40% in Group C. The postoperative prognoses of patients with aneuploid pattern in Group A and Group C were significantly poorer than those of the diploid one in cumulative survival rates and survival rates after recurrence. Patients with aneuploid pattern in Group B had a poorer prognosis than those with diploid one in cumulative survival rates and disease-free survival rates. These results suggest that nuclear DNA ploidy analysis was a useful marker of biological malignant potential in resected human HCCs.

We measured the cellular DNA content of paraffin-embedded tumor specimens by flow cytometry from 340 cases of resected non-small cell lung cancer, and investigated the correlation of DNA content and prognosis of these cases with long-term follow-up. These 340 cases were divided into some populations according to pathological stage, histologic type, surgical curativity and N factor, and we compared the prognosis of DNA diploidy cases and DNA aneuploidy cases in each population. DNA aneuploidy cases had a significantly less favorable prognosis than DNA diploidy cases in population of stage I adenocarcinoma, stage IIIA non-small cell lung cancer and N2 cases among stage IIIA non-small cell lung cancer, all after curative operation. But in other populations, there was no significant difference in prognosis between DNA diploidy cases and DNA aneuploidy cases. In conclusion, DNA ploidy pattern is a prognostic factor for survival in patients with stage I adenocarcinoma and N2 cases of stage IIIA non-small cell lung cancer.

Surgically resected specimens of 139 cases of large-bowel carcinoma were analysed in this study. Each carcinoma was cut out in stepwise section through the whole tumor, and flow cytometric DNA measurement was performed for each section. The intratumor DNA ploidy distribution pattern decided in this way was classified into 5 types (Type A-E). The 139 cases of carcinoma comprised 19 cases of Type A, 27 Type B, 11 Type C, 37 Type D and 45 Type E. The intratumor DNA ploidy distribution pattern showed a statistically significant correlation to tumor size, gross type, depth of invasion, growth pattern at the tumor margin, venous permeation of visceral wall, DNA Index and Dukes stage. Among these 5 types of carcinoma, carcinoma showing Type E was seen most frequently, even in the earlier stage, and found most frequently among the cases showing invasive growth pattern at the tumor margin, positive venous permeation in the visceral wall and DNA Index of more than 1.7. Therefore, the intratumor DNA ploidy distribution pattern seemed to reflect the degree of tumor malignancy as well as that of tumor advancement. Moreover, Type E pattern of carcinoma appeared to reveal the highest grade of malignancy, and early detection of this type seemed to be necessary in order to improve survival after surgery.

To investigate the prognostic significance of DNA ploidy pattern and DNA index (DI), DNA contents were measured by flow cytometer in 412 patients with colorectal cancer and correlation between their prognoses and DNA contents were analyzed on the same clinical stage. There were significant differences in the survival rate and the incidence of tumor recurrence between diploid and aneuploid tumors, especially the poor survival rate and frequent tumor recurrence in the aneuploid tumor with DI above 1.5. Cox's multiple regression proportional hazard model was used to investigate the prognostic value of DNA ploidy pattern, DI and clinicopathological findings. From these analyses, DI 1.5 was found to be the most significant prognostic factor. These results suggest that flow cytometrically evaluated DI values have a relevant independent power for predicting the clinical outcome of colorectal cancer patients.

Significance of flow cytometric DNA analysis for assessing malignant potential and survival of colorectal cancer was investigated using paraffin-embedded materials from 163 patients who underwent resection of curability A from 1971 to 1985, excluding intramucosal carcinoma. DNA diploid was confirmed in 46% (75 cases) of the patients and DNA aneuploid in 54% (88 cases). No significant correlation was seen between DNA ploidy and clinicopathological factors, such as tumor location, macroscopic type, histological type, depth of invasion, lymph node metastasis and stage. Cumulative survival rates after curable resection of colorectal cancer were significantly lower in patients with DNA aneuploid tumor than those with DNA diploid tumor. Furthermore, in patients in stage of II and III, survival rates were lower in DNA aneuploid patients than DNA diploid patients, respectively. A multivariate analysis of survival data using Cox's proportional hazard model showed that DNA ploidy was the significant discriminating factor on survival in stage II and III cancer. In conclusion, nuclear DNA ploidy in patients with stage II and III colorectal cancer undergoing curable resection may represent malignant potential and may be an independent prognostic factor.

Curatively resected specimens of 124 cases of colorectal carcinoma, and 6 cases of colorectal carcinoma showing hematogenous metastasis were used in this study. Each carcinoma was cut out in stepwise section through the entire tumor. DNA ploidy for each section was determined by flow cytometry, and the intratumor DNA ploidy distribution pattern was decided, and divided into 5 types (Type A-E). These 5 types were broadly divided into 2 types; predominantly diploidy type (Type A, C) and predominantly aneuploidy type (Type B, D, E). A statistically significant difference was seen in the 5-year survival between 25 cases of the predominantly diploidy type (100%) and 99 cases of the predominantly aneuploidy type (76.7%). However, there was no statistically significant difference in survival between these 2 types in any Dukes stage. The intratumor DNA ploidy distribution pattern in 6 cases of colorectal carcinoma showing hematogenous metastasis comprised 3 cases of Type E, 2 Type D and 1 Type B. And that of 8 cases of curatively resected colorectal carcinoma showing hematogenous metastasis postoperatively comprised 5 cases of Type D, 2 Type E and 1 Type B. Thus, a close correlation between the aneuploidy predominant type, especially Type D as well as Type E, and hematogenous metastasis, was suggested.

The clinical and histological factors predicting survival of patients with head and neck cancer were discussed. Recent reports with multivariate analysis focusing on clinical prognostic factors revealed that staging of nodal metastases is most significant. Number and size of metastatic nodes can be evaluated as factors to predict the outcome of advanced cancer. The number of positive nodes over 4 may be associated with poor prognosis. A scoring system of histological grading has been developed at this primary site. The mode of invasion or depth of invasion of cancer may predict the metastases to regional nodes. The research on molecular pathology is ongoing in head and neck cancer. No independent prognostic factor is yet available in clinical practice.

It is almost impossible to diagnose early cancers by elevated tumor markers in serum. However, it is possible by detecting tumor associated biomarkers in the excreted specimens such as nipple discharge, urine and feces. Elevated CEA in nipple discharge is indicative of non-palpable "To" early breast cancer, elevated fecal CEA indicated hemoccult test-negative colon cancer, basic fetoprotein and beta-hCG core fragment in the urine for urogenital malignancies, and urinary VMA or HVA for asymptomatic neuroblastomas. Detection of abnormal genes is more sensitive than cytology and suggests cancers or precancerous changes. PCR-SSCP method enabled to determine Ki-ras gene mutation in pancreatic juice and in duodenal juice in pancreatic cancer patients.

Breast carcinoma is considered to occur as non-invasive carcinoma (DCIS) and to progress to the stage of invasive cancer. In DCIS, multiple gene and chromosome alterations are shown to occur already, and the pattern of alterations are concordant between DCIS and invasive components in most of individual tumors. Therefore, unknown molecular alterations are considered to exist that are involved in tumor invasion. A lot of alterations are also shown to occur in breast cancer cells, e.g., overexpression of growth factor receptors, cyclins and the molecules regulating cell-cycle, and cell-adhesion molecular. Clarification of signal-transducing mechanism of these molecules in cells and alterations of the system in cancer cells will be important as molecular mechanism of proliferation and progression of breast cancer. It is expected that the therapeutic strategy targetted to the control of these molecular events is developed in future.