Researchers at the Johns Hopkins Kimmel Cancer Center, alongside collaborators from institutions across the U.S. and Europe, have developed an innovative blood test that combines artificial intelligence (AI) with genetic and protein biomarkers to detect early signs of ovarian cancer. Their findings, published in Cancer Discovery on September 30, highlight a new screening method that could potentially improve early detection for a cancer that often goes unnoticed until advanced stages.
Integrating AI with Biomarker Analysis
The study utilized an AI-driven approach to analyze DNA fragments circulating in the bloodstream, alongside two established protein biomarkers: cancer antigen 125 (CA-125) and human epididymis protein 4 (HE4). While both CA-125 and HE4 have been identified as markers for ovarian cancer, their effectiveness when used alone has been limited, especially in distinguishing between malignant tumors and benign ovarian growths. By combining these protein biomarkers with AI analysis of DNA fragment patterns, the researchers were able to significantly improve detection accuracy.
“The combination of artificial intelligence, cell-free DNA fragmentomes, and a pair of protein biomarkers in a simple blood test improved detection of ovarian cancer even in patients with early-stage disease,” says Victor Velculescu, senior author of the study and co-director of the Cancer Genetics and Epigenetics Program at the Johns Hopkins Kimmel Cancer Center.
The AI-powered technology utilized in this study, known as DELFI (DNA Evaluation of Fragments for early Interception), analyzes the structural patterns of DNA fragments released by dying cancer cells. Unlike healthy cells, which break down into predictable, organized DNA fragments, cancer cells leave behind irregular and disordered fragments. This irregularity forms a “fragmentomic” signature, which the AI system can detect, allowing for the identification of cancers at early stages.
Early Detection Could Save Lives
Ovarian cancer ranks as the fifth leading cause of cancer-related deaths among women in the U.S., largely due to its late diagnosis in many cases. According to the Centers for Disease Control and Prevention (CDC), the five-year survival rate hovers around 50%, but this number can be much lower when the cancer is detected at an advanced stage.
“Early detection of ovarian cancer may save lives but most women are diagnosed late in the course of the disease when survival rates are much lower,” says Jamie Medina, co-first author and postdoctoral fellow at Johns Hopkins. “The lack of specific symptoms early in the course of the disease or effective biomarkers has hindered earlier detection efforts.”
The study involved blood samples from 94 women diagnosed with ovarian cancer, 203 women with benign ovarian tumors, and 182 women without any known ovarian growths. The researchers used a modified version of the DELFI test, called DELFI-Pro, to examine the samples. This test, which integrates AI analysis of cell-free DNA with CA-125 and HE4 biomarker testing, was able to detect a much higher percentage of ovarian cancer cases than tests relying on the biomarkers alone.
For example, DELFI-Pro identified 72%, 69%, 87%, and 100% of ovarian cancer cases in stages I–IV, respectively. In contrast, tests using CA-125 alone detected only 34%, 62%, 63%, and 100% of cancers at those same stages. Importantly, DELFI-Pro was able to accomplish this with minimal false positives, which is critical for reducing unnecessary surgeries and follow-up procedures.
Confirming Results and Distinguishing Tumor Types
To further validate the test’s accuracy, the researchers applied DELFI-Pro to a second set of samples from American women. This smaller cohort included 40 patients with ovarian cancer, 50 with benign ovarian growths, and 22 women without known ovarian abnormalities. The test achieved similar success rates in this population, detecting 73% of all cancers and 81% of high-grade serous ovarian carcinoma, the most aggressive form of the disease. Again, false positives were nearly absent in women without cancer.
One of the most critical capabilities of DELFI-Pro was its ability to distinguish between cancerous tumors and benign growths, a distinction that current imaging methods like ultrasounds often fail to make. “Ovarian cancers have a unique DNA fragmentation signature that is not present in benign lesions,” explains Akshaya Annapragada, co-first author and M.D./Ph.D. student at Johns Hopkins. This distinction is particularly important because exploratory surgery is typically the next step when ovarian growths are detected. The ability to avoid unnecessary surgeries for benign conditions could provide significant benefits for patients.
Next Steps
While these results are promising, Velculescu and his colleagues stress that more work is needed to validate the test in larger, randomized clinical trials. “This study provides further evidence demonstrating the benefit of genome-wide, cell-free DNA fragmentation, and artificial intelligence to detect cancers with high accuracy,” Velculescu notes. “Our results show that this combined approach has higher performance for screening than existing biomarkers.”
Publication Details
Medina JE, Annapragada AV, Lof P, et al. Early detection of ovarian cancer using cell-free DNA fragmentomes and protein biomarkers. Cancer Discov. Published online September 30, 2024. doi:10.1158/2159-8290.CD-24-0393.