Scientists say they have developed a simple blood test to spot ovarian cancer early that could “significantly improve” outcomes for women with the disease, reports The Guardian.
More than 300 000 women, mostly over 50, are diagnosed worldwide each year, according to the World Cancer Research Fund. Ovarian cancer is often diagnosed late, making treating the condition more difficult.
The test trialled by UK and US researchers looks for two different types of blood markers in those showing symptoms of the disease, which include pelvic pain and a bloated tummy. It then uses machine learning to recognise patterns that would be difficult for humans to detect.
Currently, the disease is usually diagnosed using a mix of scans and biopsies, like an ultrasound scan, a CT scan, a needle biopsy, a laparoscopy or surgery to remove tissue or possibly the ovaries.
Often, it is detected late because symptoms such as bloating, feeling full quickly after eating or frequent urination are not always obvious potential signs of cancer.
The blood test looks for what ovarian cancer sheds into the bloodstream, even in its early stages.
Cancer cells release fragments into the blood that carry tiny, fat-like molecules known as lipids, along with certain proteins. This combination of lipids and proteins are like a biological fingerprint for ovarian cancer, according to AOA Dx, which developed the test.
It also uses an algorithm that has been trained on thousands of patient samples to recognise subtle patterns across these lipids and proteins that signal ovarian cancer.
“The test can detect the disease at early stages and with greater accuracy than current tools,” said Alex Fisher, the chief operating officer and co-founder of AOA Dx.
Dr Abigail McElhinny, chief science officer at AOA Dx, added: “By using machine learning to combine multiple biomarker types, we’ve developed a diagnostic tool that detects ovarian cancer across the molecular complexity of the disease in sub-types and stages.
“This platform offers a great opportunity to improve the early diagnosis of the disease, potentially resulting in better patient outcomes and lower costs to the healthcare system.”
A study, led by the universities of Manchester and Colorado and published in the American Association of Cancer Research journal Cancer Research Communications, tested 832 samples using the AOA Dx platform.
In samples from the University of Colorado, the test accurately detected ovarian cancer across all stages of the disease 93% of the time, and 91% in the early stages.
In samples from the University of Manchester, the test showed 92% accuracy at all stages and 88% accuracy in early stages.
Emma Crosbie, a Professor at the University of Manchester and an honorary consultant in gynaecological oncology at Manchester University NHS Foundation Trust, said: “AOA Dx’s platform has the potential to significantly improve patient care and outcomes for women diagnosed with ovarian cancer.
“We are eager to continue advancing this important research through additional prospective trials to further validate and expand our understanding of how this could be integrated into existing healthcare systems.”
Study details
Utilizing serum-derived lipidomics with protein biomarkers and machine learning for early detection of ovarian cancer in the symptomatic population
Brendan M. Giles, Rachel Culp-Hill, Robert A. Law et al.
Published in Cancer Research Communications on 13 August 2025
Abstract
Ovarian cancer (OC) is the fifth leading cause of cancer-related deaths among women. Most patients are diagnosed at late-stage (III/IV), resulting in a five-year survival rate below 30%. This is driven by the presentation of vague abdominal symptoms (VAS) that confound diagnosis at early stages (I/II) and a shortage of robust biomarkers. We are taking a novel approach for earlier OC detection, leveraging lipids as biomarkers. We utilised untargeted UHPLC-MS to analyse sera from two large, independent cohorts (N=433, N=399) designed to reflect the symptomatic population, including: individuals with benign adnexal masses, early- and late-stage OC, gastrointestinal disorders, and otherwise healthy women seeking care for symptoms. We identified a significantly altered lipid profile in OC and early-stage OC specifically across both cohorts, compared with controls. We also profiled select protein biomarkers (CA125, HE4, FOLR1, MUC1) and, utilising machine learning-based modelling, identified a proof-of-concept multi-omic model consisting of less than 20 top-performing lipid and protein features. This model was trained on Cohort 1 and tested on Cohort 2, achieving AUCs of 92% (95% CI: 87-95%) for distinguishing OC from controls and 88% (95% CI: 83-93%) for distinguishing early-stage OC from controls. These findings demonstrate the clinical utility and robustness of lipids as proof-of-concept diagnostic biomarkers for early OC within the clinically complex symptomatic population, particularly when applied in a multi-omic approach.
See more from MedicalBrief archives:
Ovarian cancer screening test less effective for black women, review finds
Blood test has potential to detect ovarian cancer
SA team devises cost-effective genetic breast and ovarian cancer screening