RSNA Grant Recipient Pioneers Research in Breast Tomosynthesis

Despina Kontos, Ph.D. University of Pennsylvania

Clinicians and radiologists may one day have an important weapon—tomosynthesis imaging biomarkers—in accurately identifying women at high risk for breast cancer thanks to pioneering research by Despina Kontos, Ph.D., integrating computer and radiologic science.

Although her most recent grant was awarded in April through the U.S. Department of Defense (DOD) Breast Cancer Research Program, Dr. Kontos' initial work pioneering the investigation of tomosynthesis imaging biomarkers for breast cancer risk was funded through Research Fellow grants awarded by the RSNA Research & Education Foundation.

"I'm investigating the use of tomosynthesis imaging as a predictive biomarker for the risk of developing breast cancer and ultimately hope to incorporate these quantitative measures into the current gold standard risk-assessment models to improve their predictive accuracy," said Dr. Kontos, a postdoctoral fellow and research associate in the Department of Radiology at the University of Pennsylvania. "Breast tomosynthesis is believed to have the potential to replace mammography in regular breast cancer screening."

Although current National Cancer Institute risk-assessment models can tell clinicians that 11 of 100 women in the population will get breast cancer, "The models are not able to identify which individual women will actually develop breast cancer," Dr. Kontos said. "That is not very useful for clinicians in their screening practices.

"The methods I'm proposing are fully automated in the sense that they could be used as adjunct software tools on the already existing mammography workstation platform," Dr. Kontos continued. "The work is translational in its approach, trying to incorporate these measures ultimately in the clinical practices and the routine mammographic examinations."

Dr. Kontos used initial grants—an Agfa HealthCare/RSNA Research Fellow Grant in 2006 and a Siemens Medical Solutions/RSNA Research Fellow Grant in 2007—to develop image analysis methods and to test preliminary pilot risk assessment models specific to breast tomosynthesis.

She has continued her research through other grants including a two-year postdoctoral fellowship from the Susan G. Komen for the Cure Foundation that builds on her RSNA-funded research to further evaluate and refine the risk-prediction models. Her latest U.S. Department of Defense Breast Cancer Research Program grant will fund a larger study incorporating quantitative imaging biomarkers for breast cancer risk assessment.

"We started this work from scratch using RSNA money," Dr. Kontos said. "RSNA R&E Foundation grants paved the way for everything that came afterward. They helped me develop preliminary data and start a project from scratch. They helped accelerate this research very much and also supported my training.

"I was coming from a computer science department and all my training and research up to that point was in science and engineering," she continued. "I really didn't have a lot of experience in clinical research. RSNA really helped me to achieve my vision to become a translational scientist and integrate my imaging science background with medical research," she said.

The combination of computer scientist and medical researcher isn't uncommon in other areas of radiology, such as brain imaging. However, breast imaging is not as integrated, leaving a gap between computer technology and clinical practices, Dr. Kontos said.

"The methods that are really used in clinical practice are more behind in terms of how computer science and imaging science in general have advanced," she said. "I think clinical practice could benefit very much from the advancements so far in the engineering and computer sciences and in the imaging sciences in general to help leverage the amount of data available and to develop quantitative imaging methods."

Dr. Kontos groundbreaking research is a step in that direction, said R. Nick Bryan, M.D., Ph.D., professor and chair of the Department of Radiology, in the University of Pennsylvania Health
System.

"Dr. Kontos has taken a basic observation, that increased breast density is a risk factor for breast cancer, and developed a sophisticated but practical way to quantify mammographic breast density and build it into a personalized algorithm for breast cancer risk," said Dr. Bryan. "This research required that she apply her skills in computer science to a daily problem in the field of medicine, of which she had no experience."

The Move Toward Quantitative Imaging

In that regard, Dr. Kontos pointed to the RSNA Quantitative Imaging Biomarkers Alliance (QIBA), which unites researchers, healthcare professionals and industry members to advance quantitative imaging and the use of biomarkers in clinical trials and practice and develops protocols to standardize the biomarkers across different platforms.

"QIBA stresses the importance of moving more toward quantitative imaging, which is basically what we—the computer scientist and engineer and the imaging science people—bring to the table," said Dr. Kontos, a QIBA member.

"From being part of this alliance, I am trying to learn and understand what is needed in the field so I can better contribute in the future and grow my research program," she said.

In her latest study, Dr. Kontos is investigating how women who are at high risk of estrogen-receptor positive (ER+) breast cancer could benefit from selective estrogen receptor modulator chemoprevention drugs. Because no method exists to accurately identify these women in advance, ER status can only be assessed with a biopsy after breast cancer is already diagnosed.

Dr. Kontos said she hopes that landing the extremely competitive DOD grant demonstrates the potential of her ideas.

"The grant is supposed to fund only high-risk, high-gain ideas that have no previous evidence," Dr. Kontos said. "It has to be an out-of-the-box idea for them. And it's blinded in the review. The reviewers don't know who I am or what institution I'm coming from, so they are only evaluating the idea.

"It definitely gives me confidence in the ideas and shows they are innovative with potential for significant clinical impact in the future," she concluded. "It's also helping me as a junior investigator to start my research program and to develop a line of research that is going to grow and expand in the future."