# Radiomic biomarkers for clinical decision support that predict patient outcomes in serous ovarian carcinoma > **NIH NIH R56** · H. LEE MOFFITT CANCER CTR & RES INST · 2024 · $299,995 ## Abstract ABSTRACT Less than a third of women diagnosed with advanced epithelial ovarian cancer (EOC) survive five years after diagnosis. Despite therapeutic improvements over time, patient survival has remained relatively unchanged for decades. At present, we are unable to predict which EOC patients are at-risk for adverse outcomes following first-line treatment. Thus, there remains a critical unmet need to identify innovative biomarkers for clinical decision support that will achieve optimal outcomes in this deadly disease. As part of standard of care (SOC) for EOC, computed tomography (CT) is used for diagnosis, prior to initiation of treatment, and to monitor treatment response and patient outcomes. SOC CT images can be converted into quantitative data that can be used as rapid, reproducible, and accurate non-invasive biomarkers for clinical decision support in the cancer care continuum. The present proposal will utilize an extensively validated radiomics imaging pipeline developed by our group and standard medical imaging software to measure body composition to identify image-based biomarkers that predict survival among patients with high-grade serous ovarian carcinoma (HGSOC). We will focus on HGSOC as it is the most common histotype of EOC and typically responds favorably to first-line chemotherapy but >80% of patients experience recurrence. In Aim 1, we will build upon ongoing efforts to establish a multi-institutional cohort of racially and ethnically diverse women with HGSOC with pre-treatment SOC CT images and well-annotated clinical and outcomes data. Using validated radiomic pipelines developed by our team, regions of interest (i.e., the primary ovarian tumor) from pre-treatment CT scans will be identified and segmented, and image features will be calculated including radiomics and body composition depots (skeletal muscle mass and subcutaneous, visceral, intra/intermuscular, and total adipose tissue). This multi-institutional cohort will be utilized to identify and validate CT image-based features that predict survival among HGSOC patients (Aim 2). We will further test the predictive model from Aim 2 and develop de novo models among clinically relevant sub-groups of women with HGSOC, including racial and ethnic groups and women treated with neoadjuvant chemotherapy vs. upfront surgical debulking as first-line therapy (Aim 3). This work will provide readily calculable, non-invasive biomarkers from SOC CT images for clinically translational information to better predict which patients may fail first-line therapy and to support patient therapy stratification. Ultimately, with sufficient testing and validation, our long-term goal is to validate clinical-imaging models that could be incorporated into clinical care to improve outcomes among HGSOC patients. ## Key facts - **NIH application ID:** 11172693 - **Project number:** 1R56CA279060-01A1 - **Recipient organization:** H. LEE MOFFITT CANCER CTR & RES INST - **Principal Investigator:** Rikki Cannioto - **Activity code:** R56 (R01, R21, SBIR, etc.) - **Funding institute:** NIH - **Fiscal year:** 2024 - **Award amount:** $299,995 - **Award type:** 1 - **Project period:** 2024-09-13 → 2025-08-31 ## Primary source NIH RePORTER: https://reporter.nih.gov/project-details/11172693 ## Citation > US National Institutes of Health, RePORTER application 11172693, Radiomic biomarkers for clinical decision support that predict patient outcomes in serous ovarian carcinoma (1R56CA279060-01A1). Retrieved via AI Analytics 2026-05-27 from https://api.ai-analytics.org/grant/nih/11172693. Licensed CC0. --- *[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*