BACKGROUND: Immunoreactivity for p53 tumor suppressor gene product is commonly found in human malignancies and some premalignant lesions, but its role in cancer development and its value as a marker of tumor biologic behavior is still unclear. OBJECTIVES: This study was undertaken to assess p53 immunoexpression in esophageal squamous cell carcinoma and attempts to determine its correlation with morphological features associated with tumor behavior. METHODS: Immunohistochemical study was performed on archival paraffin-embedded tissue of 37 esophageal squamous cell carcinomas and respective adjacent mucosa. RESULTS: Twenty-one tumors (56.8%) demonstrated specific staining for p53. Sixteen areas of dysplasia were present in 14 out of the 35 cases. p53 positivity was found in one low-grade dysplasia and in six high-grade dysplasias. By univariate analysis, p53 immunoexpression correlated positively with local invasion (P = 0.01) and perineural spread (P = 0.04). Multivariate analysis with logistic regression showed that tumor invasion was the only factor that discriminated between p53 positive and p53 negative cases (OR: 15.6, P < 0.02). No relationship was found between p53 expression and tumor grade, DNA nuclear ploidy, and S-phase fraction. CONCLUSIONS: These data suggest that p53 dysfunction may be implicated in early, preinvasive, stages of esophageal cancer as well as in the tumor progression related to a more invasive phenotype.
BACKGROUND: Immunoreactivity for p53tumor suppressor gene product is commonly found in humanmalignancies and some premalignant lesions, but its role in cancer development and its value as a marker of tumor biologic behavior is still unclear. OBJECTIVES: This study was undertaken to assess p53 immunoexpression in esophageal squamous cell carcinoma and attempts to determine its correlation with morphological features associated with tumor behavior. METHODS: Immunohistochemical study was performed on archival paraffin-embedded tissue of 37 esophageal squamous cell carcinomas and respective adjacent mucosa. RESULTS: Twenty-one tumors (56.8%) demonstrated specific staining for p53. Sixteen areas of dysplasia were present in 14 out of the 35 cases. p53 positivity was found in one low-grade dysplasia and in six high-grade dysplasias. By univariate analysis, p53 immunoexpression correlated positively with local invasion (P = 0.01) and perineural spread (P = 0.04). Multivariate analysis with logistic regression showed that tumor invasion was the only factor that discriminated between p53 positive and p53 negative cases (OR: 15.6, P < 0.02). No relationship was found between p53 expression and tumor grade, DNA nuclear ploidy, and S-phase fraction. CONCLUSIONS: These data suggest that p53 dysfunction may be implicated in early, preinvasive, stages of esophageal cancer as well as in the tumor progression related to a more invasive phenotype.