The goals of this project are to improve our understanding of carcinogenesis in gynecologic cancers and to enhance diagnostic and prognostic accuracies using quantitative analyses. During the past year, a close relationship between human papillomavirus infection (genital warts, condylomas) and cervical intraepithelial neoplasias (squamous dysplasias and carcinomas in situ) has been confirmed by nuclear DNA quantitation, immunoperoxidase stain for human papillomavirus capsid antigens, and morphologic analyses. Using these techniques, it is possible to demonstrate a transition from cervical condyloma to neoplasia. This occurrence is supported by the coexistence of papillomavirus capsid antigens and aneuploid DNA patterns in 14% of cervical condylomatous lesions. The risk of developing cervical intraepithelial neoplasia is increased in flat rather than exophytic condylomas, and in condylomas with moderate as compared to mild degree of nuclear atypia. These findings support the recent molecular hybridization studies indicating that several types of human papillomaviruses with different oncogenicities exist in the female genital tract. In 51 patients with clinical stage I or II cervical adenocarcinoma, the survival is most reliably predicted by the standard deviation of nuclear size (p=0.002), the DNA ploidy level (p=0.003), the clinical stage (p=0.0072), the percentage of glandular lumen (p=0.0074), the mean nuclear size (p=0.04), and the histologic grade (p=0.09). These findings demonstrate that the prognostic accuracies can be improved by quantitative analyses over the routine histologic examinations. (3)
Showing the most recent 10 out of 20 publications
