Currently, call-back rates for screening mammography in the U.S. are high at about 10%. Due to the effect of tissue superimposition on the 2D projection images, many women, especially women with dense breasts, are recalled for additional imaging of "pseudo-lesions", essentially suspicious-looking superimpositions of normal tissues which, after diagnostic workup, prove to be normal. Digital breast tomosynthesis (DBT) is a new 3D x-ray imaging modality in which tomographic breast images are reconstructed from multiple low-dose source projections. In DBT, the effects of tissue superposition are largely removed from the image set, thereby providing superior breast tissue visualization compared to mammography. Early clinical trials suggest up to a 40% reduction in false positive recalls when DBT is incorporated in the screening setting. There is however, a higher radiation dose when DBT is incorporated into the screening paradigm. Therefore this potential benefit of DBT must be carefully weighed against a potential increase of dose. The identification of a subset of women who would benefit most from DBT imaging is critical. To address this concern, our study will compare the effect of breast density and overall breast parenchymal complexity on the recall decision in breast cancer screening with digital mammography (DM) versus DBT. Our hypothesis is that complex parenchymal patterns (i.e., dense and/or texturally complex breasts) have a higher likelihood of being recalled for false positive findings with DM than when DBT is incorporated in the screening process. Towards this end, we propose to develop an imaging index for characterizing breast parenchymal tissue complexity. Currently, there is no standard lexicon to comprehensively reflect parenchymal complexity. Breast density is the only such image-based descriptor in the standard mammography BIRADS lexicon. Therefore, we propose to combine the standard density measures with advanced image texture features into a quantitative breast complexity index (BCI) that can be used to identify a subset of women who would benefit most from DBT screening. The rapidly evolving technology and the potential for superior performance will determine the role of DBT in clinical practice. If our hypothesis proves to be true, DBT could replace or complement DM for the screening of women with dense and/or texturally complex breasts, to reduce unnecessary recalls and additional diagnostic imaging procedures.

Public Health Relevance

The recent FDA approval coupled with the rapidly evolving technology and a potential for superior performance will determine the role of DBT in clinical practice. If our hypothesis proves to be true, DBT could replace or complement DM for the screening of women with dense and/or texturally complex breasts, and ultimately reduce unnecessary recalls and additional diagnostic imaging procedures. Our proposed breast complexity index (BCI) could be used as an imaging marker to identify women that can benefit most from DBT screening, understanding that the improvement in tissue visualization comes with an increase in radiation dose.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Research Project (R01)
Project #
5R01CA161749-02
Application #
8465846
Study Section
Special Emphasis Panel (ZRG1-DTCS-U (81))
Program Officer
Baker, Houston
Project Start
2012-05-04
Project End
2016-03-31
Budget Start
2013-04-01
Budget End
2014-03-31
Support Year
2
Fiscal Year
2013
Total Cost
$247,323
Indirect Cost
$91,283
Name
University of Pennsylvania
Department
Radiation-Diagnostic/Oncology
Type
Schools of Medicine
DUNS #
042250712
City
Philadelphia
State
PA
Country
United States
Zip Code
19104
Keller, Brad M; Chen, Jinbo; Conant, Emily F et al. (2014) Breast density and parenchymal texture measures as potential risk factors for Estrogen-Receptor positive breast cancer. Proc SPIE Int Soc Opt Eng 9035:90351D
Keller, Brad M; Nathan, Diane L; Gavenonis, Sara C et al. (2013) Reader variability in breast density estimation from full-field digital mammograms: the effect of image postprocessing on relative and absolute measures. Acad Radiol 20:560-8