Blood-Based Screening Model Effectively Detects Ovarian Cancer

A blood-based machine learning assay may be able to differentiate patients with ovarian cancer from healthy patients with benign ovarian masses, according to a study presented at the American Association for Cancer Research Annual Meeting 2024.

Federal statistics list ovarian cancer as the fifth most common cause of cancer deaths among women in the United States, with a 5-year survival rate of approximately 50%. Part of what makes ovarian cancer so deadly is that it does not typically cause symptoms in the early stages of disease, explained Jamie Medina, PhD, a postdoctoral fellow at the Johns Hopkins Kimmel Cancer Center and co-author of the study.

Liquid biopsy, which can be used to analyze patient blood for evidence of tumor-derived DNA, has been explored as a way to noninvasively detect a variety of cancers; however, it has not always been effective in detecting ovarian cancer, the researchers noted. DELFI (DNA Evaluation of Fragments for Early Interception) uses a newer method of liquid biopsy analysis, known as fragmentomics, which has shown promise in enhancing test accuracy. DELFI works by detecting circulation changes in the size and distribution of cell-free DNA (cfDNA) fragments across the genome.

“The lack of efficient screening tools, combined with the asymptomatic development of ovarian cancer, contributes to late diagnoses when effective treatment options are limited,” said Dr. Medina. “A cost-effective, accessible detection approach could change clinical paradigms of ovarian cancer screening and potentially save lives.”

Dr. Medina and colleagues assessed fragmentomes from individuals with and without ovarian cancer using DELFI by training a machine learning algorithm to integrate the fragmentome data with plasma levels of 2 known biomarkers of ovarian cancer, proteins CA125 and HE4. They analyzed 134 women with ovarian cancer, 204 women without cancer, and 203 women with benign adnexal masses.

According to the results, the screening model showed a specificity of 99%. It identified 69%, 76%, 85%, and 100% of ovarian cancer cases staged I-IV, respectively. The investigators will test this model in a larger cohort, but they noted the current data are encouraging.

“This study contributes to a large body of work from our group demonstrating the power of genome-wide cfDNA fragmentation and machine learning to detect cancers with high performance,” said Victor Velculescu, MD, PhD, FAACR, a professor of oncology, co-director of the Cancer Genetics and Epigenetics Program at the Johns Hopkins Kimmel Cancer Center, and senior author of the study. “Our findings indicate that this combined approach resulted in improved performance for screening compared [with] existing biomarkers.”