Blood test shows 93% accuracy in detecting ovarian cancer, study finds

Researchers at the Universities of Manchester and Colorado have validated a blood test that can identify ovarian cancer in symptomatic women with 93% accuracy—a significant leap beyond the diagnostic tools clinicians have relied on for three decades. The test, developed by diagnostics company AOA Dx, analyzes two distinct categories of biological markers—proteins and lipids—from a single blood sample, then uses machine learning to determine whether ovarian cancer is present. The findings appear in Cancer Research Communications.

The study examined the test's performance across more than 950 patients drawn from both Colorado and Manchester. In the Colorado cohort, the test achieved 93% accuracy when detecting ovarian cancer across all disease stages, and 91% accuracy specifically for early-stage cases. The Manchester samples yielded similarly strong results: 92% accuracy for all stages and 88% for early-stage disease. These numbers matter because they substantially outpace the single-marker blood tests that have been standard practice since the 1990s, which have never exceeded 90% accuracy.

The clinical problem the test addresses is stark. Ovarian cancer ranks as the fifth leading cause of cancer death among women, a ranking driven largely by the fact that most cases are caught late. Nearly all women with Stage I disease experience symptoms—bloating, abdominal pain, digestive disturbance—yet only one in five ovarian cancers are diagnosed at Stage I or II. The symptoms are vague enough that they often masquerade as benign conditions, and existing diagnostic methods, which tend to be invasive or unreliable, frequently miss early disease entirely. A woman arriving at her doctor's office with these complaints faces a diagnostic maze that can stretch for months.

AOA Dx says the test's design reflects an attempt to capture the molecular complexity of ovarian cancer across its various subtypes and stages. By combining multiple biomarker categories rather than relying on a single marker, the platform can detect disease that simpler tests might miss. The company is now pursuing regulatory approval in the United States and Europe, with an eye toward eventual launch through the NHS. If the test reaches clinical practice, AOA suggests it could streamline diagnosis, reduce costs to health systems, and most importantly, accelerate the moment when a symptomatic woman receives clarity about what is happening in her body.

Professor Emma Crosbie of Manchester University, who is also an honorary consultant in gynecological oncology at Manchester University NHS Foundation Trust, described the platform as offering significant promise for early detection and called it a practical solution for women presenting with symptoms. She noted that the research team is eager to pursue additional prospective trials to validate the findings further and understand how the test might be integrated into existing clinical workflows. The next phase will determine whether the laboratory accuracy translates into real-world benefit—whether women actually reach diagnosis sooner, whether treatment outcomes improve, and whether health systems can afford to deploy the test widely.