|Grant Number:||5R01CA136783-03 Interpret this number|
|Primary Investigator:||Moskowitz, Chaya|
|Organization:||Sloan-Kettering Inst Can Res|
|Project Title:||Prediction Model: Breast Cancer in Women Irradiated for a Pediatric Malignancy|
DESCRIPTION (provided by applicant): Survivors of a pediatric malignancy are at risk for serious long-term morbidity and premature mortality related to their treatment. Women who were treated with chest radiation for their childhood cancer have an increased risk of developing breast cancer at a young age. We use the term chest radiation to refer to radiation that includes the following fields: mantle, mediastinal, lung, total body irradiation, or spinal. Models for predicting the absolute risk of breast cancer, such as the well-known Gail model, have been used extensively to advise patients on their individual risk of developing breast cancer and to design prevention trials. The majority of these risk calculators are not immediately applicable to survivors of a previous malignancy who have a risk that is modified by previous treatments. Moreover, there is conflicting evidence as to whether various non-treatment related factors, including the traditional breast cancer risk factors, are associated with the risk of breast cancer in these women. The traditional risk factors include age, age at menarche, age at birth of first live child, number of first-degree relatives with breast cancer, number of previous breast biopsies, and history of atypical hyperplasia. We propose to utilize the unique resources of the North American Childhood Cancer Survivor Study (CCSS) and the Dutch LATe Effect Registry (LATER) cohorts. Our first aim is to develop a risk prediction model that integrates chest radiation dose and volume, treatment-related exposures, the traditional risk factors, and other possible risk factors such as age at menopause, body mass index, the number of years with intact ovarian function after radiotherapy, and use of hormone replacement or oral contraceptive therapy, in order to estimate the individualized absolute risk of breast cancer for women who were treated with chest radiation for a childhood cancer. The model will be developed using the original CCSS cohort in which there are 1,677 female participants who were treated with chest radiation and would be included in this analysis. Of these 1,677 women, 187 women had developed breast cancer by the time of this application. Our second aim is to validate the prediction model on independent data. The validation cohort will consist of about 1225 female participants in the expanded CCSS cohort and 1087 female participants in the Dutch LATER cohort. We estimate that approximately 100 of these women will have breast cancer. The third aim is to create and disseminate a risk calculator that provides computer-assisted risk prediction. Our objective is to present a tool in an easily accessible format and disseminate it for clinical use by physicians and patients. The long-term goals of this project are to provide a means for facilitating conversations between physicians and their patients about appropriate screening and preventive strategies and to help refine screening recommendations for this population. PUBLIC HEALTH RELEVANCE: Women who were treated with chest radiation for a childhood cancer are at risk of developing breast cancer at a young age. We aim to quantify a woman's individualized risk of breast cancer as a function of the dose of radiation to the chest, other treatment-related factors, the traditional breast cancer risk factors such as family and reproductive histories, and other non-treatment related risk factors. This risk prediction model has the potential to facilitate conversations between physicians and patients and to help refine current screening recommendations for female survivors of a pediatric cancer.
Testing the incremental predictive accuracy of new markers.
Authors: Begg CB, Gonen M, Seshan VE
Source: Clin Trials, 2013 Oct;10(5), p. 690-2.
EPub date: 2013 Jul 23.
Comparing ROC curves derived from regression models.
Authors: Seshan VE, Gönen M, Begg CB
Source: Stat Med, 2013 Apr 30;32(9), p. 1483-93.
EPub date: 2012 Oct 3.
One statistical test is sufficient for assessing new predictive markers.
Authors: Vickers AJ, Cronin AM, Begg CB
Source: BMC Med Res Methodol, 2011 Jan 28;11, p. 13.
EPub date: 2011 Jan 28.