Screening and diagnosis are different questions
Screening tests people without symptoms to reduce illness or death. Diagnostic evaluation investigates symptoms, a mass or known high risk. The USPSTF recommends against screening asymptomatic average-risk women for ovarian cancer because CA-125, transvaginal ultrasound or their combination did not reduce mortality and false-positive results led to unnecessary surgery [1]. This recommendation does not apply to people with a known high-risk hereditary syndrome.
Persistent bloating, pelvic or abdominal pain, early satiety, urinary change, unexplained weight change or a concerning examination still warrant clinical evaluation. The absence of a population screening programme is not advice to ignore symptoms. Similarly, an ovarian-mass risk score is designed to help evaluate an existing adnexal mass and is not approved as a general screening test [2].
- Clarify whether a test is for screening, diagnosis, triage or recurrence monitoring.
- Evaluate persistent or progressive symptoms rather than ordering a consumer screen.
- Assess inherited risk separately from average-risk population screening.
Why CA-125 and ultrasound have not solved early detection
CA-125 can rise in benign conditions and may not rise in early ovarian cancer. Ultrasound can find benign abnormalities that cannot be confidently classified without follow-up or surgery. The UKCTOCS randomised trial followed participants for a median 16.3 years and found that neither multimodal CA-125 screening nor ultrasound screening significantly reduced ovarian and tubal cancer deaths [3].
A screening programme must do more than shift stage or find cancer. It must show that the entire pathway improves meaningful outcomes and that harms are acceptable. Those harms include anxiety, repeated imaging, invasive procedures and surgery in people without cancer. Survival measured from the date of diagnosis can also look longer simply because a test moved diagnosis earlier, even when time of death is unchanged.
- Demand mortality or validated clinical-outcome evidence, not sensitivity alone.
- Include downstream diagnostic procedures in the harm assessment.
- Do not interpret an isolated CA-125 result outside clinical context.
Liquid biopsy is a family of technologies
Liquid biopsy can analyse circulating tumour DNA, methylation, proteins, extracellular vesicles, microRNA or circulating cells in blood and other fluids. A test used to guide treatment in someone with known cancer is a different product from a test intended to screen healthy people. Early ovarian tumours may release very little signal, making sensitivity at the stage where screening matters especially challenging.
Multi-cancer detection tests may combine molecular signals and a model that predicts the tissue of origin. NCI states that no multi-cancer detection test has FDA approval and that more research is needed [4]. Some are offered as laboratory-developed tests, which does not establish that screening reduces mortality. A commercially available result must not be represented as a clinical recommendation.
- Ask what biological signal the test measures and for which intended use.
- Request stage-specific sensitivity, specificity and positive predictive value.
- Check regulatory status in the country where care is delivered.
Low prevalence makes false positives consequential
Even a highly specific test can generate many false positives when disease prevalence is low. Positive predictive value depends on the tested population, not only test engineering. A result also needs a defined diagnostic pathway. If a molecular signal points toward ovary but imaging is normal, evidence is limited on how often to repeat testing, which procedure to perform and when investigation becomes more harmful than helpful.
NCI notes that no multi-cancer assay has yet shown mortality reduction in a randomised trial [5]. Trials such as the NCI Cancer Screening Research Network are being designed to evaluate feasibility and clinical outcomes [6]. Until evidence changes, emerging assays should be discussed as investigational or unproven for population screening, not as replacements for proven breast, cervical, colorectal or lung screening.
- Model false positives using the intended population prevalence.
- Predefine confirmatory testing and stopping rules.
- Continue guideline-recommended screening for other cancers.
Where technology can responsibly move the field
Research can improve study design by collecting serial specimens, including diverse ancestry and age groups, validating assays in independent cohorts and reporting early-stage performance. It should compare outcomes with usual care and measure diagnostic burden, overdiagnosis, quality of life, equity and mortality. Transparent reporting must separate training, validation and prospective clinical performance.
For high-risk hereditary groups, prevention and surveillance decisions should be made with genetics and gynaecologic-oncology specialists. NCI states that transvaginal ultrasound and CA-125 have not been shown to detect BRCA-associated ovarian tumours early enough to improve long-term survival [7]. Risk-reducing strategies therefore cannot be replaced by an attractive dashboard or a negative liquid-biopsy result.
- Report performance by stage, subtype and demographic subgroup.
- Publish all prespecified outcomes, including harms and indeterminate results.
- Never use an investigational negative result to override symptoms or inherited risk.
What the evidence cannot yet answer
- Current evidence does not support ovarian-cancer population screening in asymptomatic average-risk women.
- Liquid-biopsy studies often use case-control samples that overestimate performance compared with real screening populations.
- No multi-cancer detection assay has yet demonstrated a mortality benefit in a randomised screening trial.
Questions worth taking into care
- Is this test intended for screening, diagnosing a mass, selecting treatment or monitoring known cancer?
- What is its sensitivity for stage I ovarian cancer and its positive predictive value in this population?
- What confirmatory pathway follows a positive or indeterminate result?
- Has the test received regulatory approval for this exact use?
- Could a negative result delay evaluation of symptoms or hereditary risk management?
Source record
Evidence used in this review
Sources were selected for clinical authority, methodological relevance and traceability. Links open the original guidance, public-health record or research publication.
- [1]Final Recommendation Statement: Ovarian Cancer Screening
US Preventive Services Task Force · 2018
- [2]Ovarian Adnexal Mass Assessment Score Test System
US Food and Drug Administration · 2011
- [3]
- [4]What Cancer Screening Tests Check for Cancer?
National Cancer Institute · 2024
- [5]Cancer Screening Overview
National Cancer Institute · 2023
- [6]Questions and Answers About Multi-Cancer Detection Tests
National Cancer Institute · 2025
- [7]BRCA Gene Changes: Cancer Risk and Genetic Testing
National Cancer Institute · 2024
This evidence synthesis is for general information. It does not diagnose a condition or replace care from a qualified health professional. Treatment choices depend on individual history, examination, local guidance and informed preference. Emergency or rapidly worsening symptoms need urgent local medical assessment.