Skip to Main Content

An official website of the United States government

Principal Investigator
Name
Jennifer Prescott
Institution
Brigham and Women's Hospital / Harvard Medical School
Email
About this CDAS Project
Study
PLCO (Learn more about this study)
Project ID
2011-0122
Initial CDAS Request Approval
Sep 9, 2011
Title
Development and validation of an endometrial cancer risk prediction model
Summary
Type I endometrial cancer, the most common gynecological cancer among U.S. women, is a hormonally-regulated malignancy that is strongly influenced by modifiable lifestyle factors such as obesity and postmenopausal hormone use. Given the current obesity epidemic, we are likely to see a rise in endometrial cancer incidence. However, to date, a risk prediction model for endometrial cancer has not been developed for use in the general population that could be used to identify individuals at high risk that may benefit from targeted screening, lifestyle intervention, or chemopreventive strategies. Additionally, such a model could help project the economic burden of the disease, as well as assess the potential impact of a particular intervention. We propose to develop a risk prediction model for Type I endometrial cancer using the Nurses' Health Study followed by validation of the model among female participants in the PLCO screening trial.
Aims

AIM 1: We will develop a risk prediction model for Type I endometrial cancer using Nurses' Health Study participants. The model will allow for nonproportional hazards and account for temporal relationships between established endometrial cancer risk factors and development of disease. The model will be used to estimate the cumulative risk of developing Type I endometrial cancer to age 70. AIM 2: We will validate the Type I endometrial cancer risk prediction model among female participants of the PLCO Screening Trial. Validation will be conducted in two ways. First we will assess calibration by calculating an overall chi-square goodness of fit test. Second, we will test discriminatory ability of the model by estimating the area under the receiver operating characteristic curve.

Collaborators

Immaculata De Vivo (Brigham and Women's Hospital/Harvard Medical School)
Bernard Rosner (Brigham and Women's Hospital/Harvard Medical School)
Nicolas Wentzensen (DCEG)