CEB Projects |
![]() ![]() |
![](https://webarchive.library.unt.edu/eot2008/20090130185503im_/http://archive.nlm.nih.gov/cebimages/bullet_header42.gif)
Cervical cancer is the second most common cancer among women worldwide and the first in many developing countries.
Cervical cancer is caused by persistent infection with one of approximately 15 carcinogenic types of human papillomavirus (HPV),
which is sexually transmitted. The success of cervical cancer screening is based on the capability of detecting and treating
precancer during its prolonged natural history. The tests used for cervical cancer screening are able to detect women suspicious
of having precancerous or cancerous lesions, but also detect HPV infection and its associated cytologic (low-grade squamous
intraepithelial lesion [LSIL]) and histologic (CIN 1) lesions. Therefore, to avoid overtreatment, women with positive screening
results are evaluated visually by colposcopy to evaluate any abnormal epithelium and to obtain directed biopsy samples with
the goal of distinguishing precancerous or cancerous lesions from milder HPV-related changes. Colposcopy is a critical part
of gynecologic practice but has documented deficiencies, including lack of correlation between the colposcopic appearance and
the severity of underlying neoplasia, limited reproducibility, and difficulty in the optimal placement of colposcopically
directed biopsies.
In a collaborative effort to improve colposcopy, NLM is developing tools to aid in the analysis of digitized cervigram images from the National Cancer Institute (NCI) funded studies: ASCUS-LSIL Triage Study (ALTS) and the Guanacaste Project. ALTS, which stands for the ASCUS/LSIL Triage Study, was conducted at four major medical centers: the University of Alabama at Birmingham; the University of Oklahoma in Oklahoma City; Magee-Womens Hospital of the University of Pittsburgh Medical Center in Pittsburgh, Pa.; and the University of Washington in Seattle. In ALTS, 5,060 women with community-based cytologic results of atypical squamous cells of undetermined significance or LSIL were randomized to one of the three management arms: immediate colposcopy (all women referred to colposcopy), HPV DNA testing (colposcopy if the HPV DNA test results were positive or the enrollment liquid-based cytologic results were HSIL) and conservative follow-up (follow-up with cytologic examination and colposcopy if results are HSIL). The Guanacaste Project, a population based study developed in a rural area of Costa Rica with high incidence of cervical cancer, enrolled 10,000 women who were followed for up to 7 years. During this study, women were screened using visual (cervigrams), cytologic (Pap smear and liquid-based cytologic examination), and molecular tests to explore which factors influence HPV infection persistence and risk to develop CIN and cancer. Some of the data collected during these projects are visual, such as cervicography, digital images taken at colposcopy (Denvu), cytologic analysis, and histologic analysis. Cervicography is a 35-mm photograph of the cervix taken using a specially designed camera implemented with a 100-mm macro lens and ring flash. In addition to the cervigram images, this database contains clinical, cytologic, and molecular information at multiple examinations of 15,000 women, with password and ID labeling strategies to protect patient privacy.
During the ALTS and Guanacaste projects, close to 100,000 cervigrams were obtained from 50,000 patient-visits and included patients with invasive cancer or CIN, women without CIN at enrollment but in whom disease developed at follow-up, healthy women in whom pathologic changes in the cervix never developed, and so forth. This enormous visual database gives us a unique opportunity to study the normal and abnormal uterine cervix, changes related or not related to HPV infection, and its evolution to clearance of HPV or the development of precancer. To manage, evaluate, and collect information from the pictures, we decided to translate the conventional 35-mm pictures into digital files.