AS178 – Mammographic density and invasive breast cancer in the estrogen-plus-progestin trial
This page provides study documentation for AS178. The AS178 datasets are also available on this site for qualified investigators (sign in and a completed Data Distribution Agreement are required; see details on the Data site).
Investigator Names and Contact Information
Etta D. Pisano, MD - Principal Investigator
Medical University of South Carolina, Charleston, SC
E-mail: pisanoe@musc.edu
Project Contact (Dr. Pisano's Laboratory Manager): Elodia Cole, coleeb@musc.edu
Introduction/Intent
The primary aims of this project were to estimate the relative risk for breast cancer associated with a change in mammographic density (MD), to determine whether the increased breast cancer risk associated with MD change is independent of baseline MD, to determine the proportion of breast cancer cases attributable to MD change among women assigned to estrogen-plus-progestin therapy (EPT), and to determine whether post-EPT MD change explains the increased breast cancer risk associated with EPT within the entire study population.
This study is an extension of WHI Ancillary Study (AS) 36, "Hormone Replacement Therapy and Mammographic Density Change."
Materials/Methods
Data: The study was a quasi nested case-control study within the WHI estrogen-plus-progestin trial, where eligible cases and controls were first identified, and later asked to participate in this study. It was necessary for participants that did not participate in AS36, to provide consent so that mammographic films could be obtained and digitized.
Eligible cases included all incident invasive breast cancers through December 2004 among participating clinics. Invasive breast cancers that occurred after the trial stopped on July 7th 2002 were also included as eligible cases. Eligible controls were randomly selected at a 3:1 ratio to cases, from the risk set of the corresponding case's event, and matched by age (+/- 2 years) and race/ethnicity; preference was given to women that participated in AS36 to leverage their already digitized films. A total of 1596 eligible participants (399 cases and 1197 controls) were identified where 288 of 1596 were leveraged from AS36. Although specified in the study protocol, eligible controls were not matched by randomization group or by clinic. Deviations were made to allow for an association to exist between EPT and invasive breast cancer, and reduced staff burden at participating clinics.
Participation was less than expected. Less than 60% of eligible non-AS36 participants provided AS178 consent (766 of 1308). Mammographic films could not be obtained for all consenters. Ultimately, data were received by the WHI Coordinating Center from 903 participants (n=173 cases and 730 controls).
Methods: Mammograms were taken at baseline (pre-treatment), and the first screening examination taken at least 12 months after the start of treatment. Four experienced readers blinded to treatment and outcome assessed MD using two different validated computer-assisted techniques (Cumulus and Madena). The four readers' results were well correlated and averaged in this study. For cases, density change was assessed on the breast contralateral to the affected breast; for controls, a randomly selected left or right breast was used. The selection probability used to select the breast to be measured for controls reflected the laterality of affected breast among cases. Cases affected in both breasts were excluded. To avoid inter-study bias, MD was recomputed by the four readers for participants whose mammographic films that were previously digitized in AS36.
Usage notes: The AS36 and AS178 data were intended to stand-alone. Researchers who wish to combine MD measurements should consider a calibration model to standardize MD measurements between studies, because different computer-assisted interactive thresholding techniques were used. Per design, there are many participants (n=266) with MD measurements from both studies; see https://www.whi.org/researchers/data/Documents/Specimen%20Results%20Read%20Me.pdf for commentary on calibration.
Two datasets are provided for AS178 where MD measures are either averaged over reader (AS178_data.csv) or given by reader (AS178_data_by_reader.csv). The later dataset allows researchers to account for systematic differences between readers.
Main results and publications
In this study, the associated change in MD was a significant predictor of breast cancer risk and statistically accounted for the increased breast cancer risk associated with EPT. Thus change in MD may be a useful intermediate marker of increased breast cancer risk with use of EPT.
Some of the publications related to this ancillary study are:
Byrne, Celia, et al. [presentation abstract] "Change in mammographic density with estrogen and progestin therapy: A measure of breast cancer risk in the Women's Health Initiative." Cancer Research 70.8 Supplement (2010): 5741-5741.
For a complete, up-to-date list of WHI papers related to this ancillary study, please use the searchable Bibliography section of this website. To search for papers by study number, access the Simple Search, and enter the study number in the “Related Studies” field.
