A team of UNC-Chapel Hill researchers has been awarded up to $10 million in Advanced Research Projects Agency for Health (ARPA-H) funding to develop the Cancer Identification and Precision Oncology Center (CIPOC). The project is designed to improve cancer diagnosis and support personalized treatments by quickly aggregating and analyzing a wide range of health data, including electronic health records, histopathological and radiological images, insurance claims and geographic information.
Specifically, CIPOC will facilitate the development of an oncology health learning system that utilizes AI-ready data to generate real-time identification of new cancer cases, support patient recruitment for research, recommend precision cancer care, and help improve cancer care equity and quality. It also will create an accessible, adaptable system for health providers across diverse locations and resources.
The project is led by four principal investigators across Carolina:
- Ashok Krishnamurthy, PhD, director of the Renaissance Computing Institute (RENCI) and data science core lead.
- Jennifer Elston Lafata, PhD, professor in the Division of Pharmaceutical Outcomes and Policy at the UNC Eshelman School of Pharmacy and innovation and optimization partners lead.
- Caroline Thompson, PhD, MPH, associate professor of epidemiology at UNC Gillings School of Global Public Health and rapid identification core lead.
- Melissa Troester, PhD, MPH, professor of epidemiology at UNC Gillings and precision oncology core lead.
“CIPOC is a multi-disciplinary project that will significantly advance not just rapid cancer identification and precision oncology but also health data science and informatics,” said Krishnamurthy, a research professor of computer science at UNC-Chapel Hill. “The approaches we are developing can be used in other areas of health care, which is possible because CIPOC brings together diverse expertise across a number of fields to work together on a common goal.”
The project will organize and facilitate collaborative research conducted by faculty, staff and trainees from more than 12 schools, centers, departments and programs at UNC-Chapel Hill with a shared vision to create cutting-edge data tools researchers and practitioners can use at UNC – and in time across North Carolina and the United States – to improve the diagnosis and treatment of cancer.
“While precision oncology has made major advances in recent years, translation of these innovations to practice has lagged behind as has our ability to monitor, track, and therefore understand and plan for needed cancer-related services,” said Thompson, a UNC Lineberger Comprehensive Cancer Center member. “By accelerating the identification of cancer cases and developing innovative informatics tools to make improved, precision recommendations for care, this project can advance the provision of equitable care services and delivery.”
The three-year project will focus on building an oncology learning health system at UNC Health, with the potential to expand across North Carolina and nationally. A learning health system integrates scientific evidence, data and culture into daily care with a commitment to continuous improvement and innovation. The goal is to produce high-quality and high-value care that is equitable across diverse populations.
“As part of our efforts, we are forming a panel of nationally recognized experts and advisors. This panel will provide our team with ongoing feedback and serve as an independent sounding board. Their input is crucial to ensuring the usability and acceptability of our processes and products,” said Lafata, co-lead of the UNC Lineberger’s Cancer Care Quality Initiative. “This step is essential given our focus on accelerating academic discovery, optimizing cancer care delivery and supporting public health reporting. Additionally, these advisers will help us minimize any inherent biases in our work.”
CIPOC will utilize AI tools, including large language modeling, to quickly standardize, harmonize and link structured and unstructured data from multiple sources, enabling more precise tracking and treatment for different cancer types.
It also will develop an AI-driven virtual multidisciplinary tumor board to support the provision of precision oncology care. Studies have shown multidisciplinary tumor boards, in which a group of experts in different specialties review and discuss patients’ medical conditions and treatment options, can improve cancer outcomes. The board will use AI-ready datasets, including electronic health record-derived clinical data and histopathological and radiological images, to help inform prediction of risk and tumor progression as well as treatment decision making.
“We want to make precision oncology more widely available to North Carolinians. This project aims to develop tools that will use common medical record data to define care that responds to each patient’s unique tumor biology, reducing the need for additional, costly testing,” said Troester, co-leader of the UNC Lineberger Cancer Epidemiology Program.
CIPOC will make its data tools open source, allowing them to be scaled and adapted by health systems of any size, thus improving the use of clinical data for research and cancer care across a broad spectrum of communities. This innovation aligns with ARPA-H’s national goals to strengthen health care system resilience and equity.
The development and submission of the ARPA-H proposal was supported by the UNC Office of Research Development, with oversight by Nathan Blouin, MBC, CRA, assistant vice chancellor for research development, and Nate Warren, PhD, research development manager.
Breast Cancer Care in a Fragmented Health Care System
This figure shows the hypothetical journey or “pathway” of one patient that takes place in three different health care systems (thus three EHRs).
In the far left of the figure, the patient, “Mary” presents with breast symptoms to her community health care facility. Mary has a diagnostic workup by her primary care physician and radiology, and she is referred to a medical oncologist, who works at both the community center “A” and at the nearby tertiary academic center “B.”
The medical oncologist refers Mary to a surgical oncologist and a plastic surgeon, both of whom are part of an academic center. Mary has her surgery, then returns to the community center to be managed by her medical oncologist for her chemotherapy and posttreatment surveillance.
Some time passes and Mary moves to a new state (dotted line). She has a recurrence of her breast cancer and is treated by a new medical oncologist at “C”— an out-of-area center. Ultimately Mary is cancer free at the end of her journey.
Researchers wanting to understand Mary’s complete cancer pathway would need clinical details from three EHRs as well as confirmed diagnosis details from a cancer registry. The EHR data from organization A, in particular, might include important prediagnosis information about Mary, such as her frequency of screening, prior concerns, and patterns of noncancer preventive care.
Any less detail would limit our understanding of Mary’s pathway to shorter time frames, e.g., cancer diagnosis or treatment care periods. If Mary were an Oncoshare patient, we would be able to follow her through her care pathway at “A” and “B,” but we would have no information about her recurrence, which was treated at “C.”