If you have cirrhosis, chronic hepatitis B, or another condition that increases your risk for hepatocellular carcinoma (HCC), you may wonder whether a blood test or an ultrasound is better for finding liver cancer early. They answer different questions: ultrasound examines the liver’s structure, while blood tests measure biological markers that can change with cancer or liver disease.
For people who qualify for routine surveillance, current U.S. professional guidance recommends liver ultrasound together with alpha-fetoprotein (AFP) blood testing approximately every six months. New molecular blood tests are promising research approaches but are not established replacements for this standard surveillance strategy. Your clinician should tailor the plan to your risk and the quality of prior imaging. [1, 2]
HCC blood test vs. ultrasound: short comparison
There is no single performance percentage that fairly ranks all blood tests against all ultrasounds. Results depend on the disease stage, the test used, the surveillance population, and the study design. The table below describes the distinct roles of each approach.
| Question | Liver ultrasound | Blood-based HCC testing |
|---|---|---|
| What does it measure? | Visible anatomy, nodules, and structural changes in the liver | AFP protein, or in investigational approaches, circulating DNA/RNA or other molecular biomarkers |
| Main clinical role | Routine surveillance imaging for eligible at-risk patients | AFP is recommended alongside ultrasound by AASLD; newer molecular panels require indication-specific validation |
| Strengths | No ionizing radiation; widely used; localizes concerning findings | Simple blood collection; not dependent on obtaining an adequate acoustic window |
| Limitations | Can miss early lesions; image quality may be limited by body habitus, liver texture and operator factors | AFP can be normal with HCC or elevated for noncancer causes; newer markers may have false positives/negatives and limited prospective evidence |
| Can it confirm HCC? | No; suspicious findings usually require diagnostic contrast-enhanced imaging | No; biomarkers alone do not establish or exclude HCC |
How does ultrasound screen for liver cancer?
An abdominal ultrasound uses high-frequency sound waves to create an image of the liver. During surveillance, the sonographer and interpreting clinician look for suspicious focal liver observations and assess how well the entire liver can be seen. Ultrasound is relatively accessible, does not expose patients to ionizing radiation, and has been a foundation of HCC surveillance for decades. [1]
Its limitations matter most when trying to detect early-stage HCC. Small tumors may not be distinguishable from nodular, scarred liver tissue. Obesity, steatotic liver disease, a markedly cirrhotic liver, and examination technique can reduce the quality of visualization. Poorly visualized examinations should not be treated as equally reassuring as high-quality negative studies. [1, 5]
What does research say about early-stage HCC detection?
A 2018 systematic review and meta-analysis analyzed 32 studies involving 13,367 participants with cirrhosis. Ultrasound sensitivity was approximately 84% for HCC at any stage but only 47% for early-stage HCC across studies reporting that outcome. In the studies that directly compared surveillance strategies, pooled sensitivity for early-stage HCC was 45% with ultrasound alone and 63% with ultrasound plus AFP. Combining ultrasound and AFP increased sensitivity but reduced specificity modestly. [3]
*Pooled estimates from directly comparing studies in the 2018 meta-analysis. These are sensitivity estimates, not overall accuracy, not a comparison with MoleculeDx, and not predictions of a particular person's result. Different study groups and methodologies produce different estimates.
AASLD's 2023 guidance cites an early-stage sensitivity estimate of about 53% for ultrasound alone from another evidence synthesis and similarly supports the combined ultrasound-plus-AFP approach. Differences between published estimates reflect study design, populations, and inclusion criteria; they should not be mixed as if they came from a single trial. [1]
What can an AFP blood test tell you?
AFP is a protein that may increase in the blood of a person with HCC, but it also can rise in some noncancer liver conditions. Normal AFP does not rule out liver cancer, and an elevated AFP value by itself does not confirm cancer. Interpreting changes over time, liver history, and ultrasound results together is more useful than treating AFP as a stand-alone yes/no answer. [1, 3]
AFP is a different kind of blood marker from RNA or DNA molecular assays. Results from one type of test cannot simply be substituted for another without direct comparative studies and appropriate clinical validation.
What about newer molecular blood tests?
Researchers are investigating circulating tumor DNA, methylation patterns, proteins, and cancer-associated RNA signals as ways to improve HCC risk assessment and surveillance. Some approaches may provide information that ultrasound and AFP do not measure. But whether adding a particular molecular test actually improves early detection, clinical decision-making, or patient outcomes must be demonstrated in appropriate studies. [1, 2]
The MoleculeDx Liver Cancer Fusion Predictor builds on research into cell-free cancer-associated fusion-transcript signals in the blood. In a 2024 American Journal of Pathology study, investigators measured nine serum fusion transcripts in cohorts of patients with HCC and individuals without HCC and developed machine-learning models. Certain study models showed encouraging performance, but these estimates were obtained in selected research cohorts and must not be described as the guaranteed sensitivity or accuracy of a marketed test, or as proof that the test detects tumors before ultrasound does. The study did not establish a randomized, prospective head-to-head comparison against routine ultrasound-based surveillance. [4]
Some assays can be collected through mobile phlebotomy, but availability, cost, eligibility, and turnaround time vary by service. Avoid assuming that any molecular blood test provides results within a fixed time or that it replaces an imaging appointment.
Which is better for HCC screening: a blood test or ultrasound?
For eligible at-risk patients, the clinically supported approach today is regular ultrasound plus AFP, not an either/or choice between imaging and a blood test. The methods provide different information: ultrasound searches for a visible lesion, and AFP provides an additional blood-based warning signal. New molecular assays may someday improve this combination, but their role depends on further prospective validation and current regulatory and clinical context. [1, 2]
If previous ultrasounds have had limited visualization, ask your clinician whether a different surveillance modality is appropriate. In selected patients, MRI or multiphase CT may be considered; these are not routine substitutions for every person. If a suspicious lesion is seen or AFP becomes concerning, diagnostic multiphase CT or contrast-enhanced MRI is generally the next evaluation rather than relying on another screening blood test. [1, 5]
Questions to ask your doctor
- Does my liver condition and overall health make me eligible for regular HCC surveillance?
- Was my most recent ultrasound of adequate quality to assess the whole liver?
- Should AFP be measured alongside ultrasound, and how should a rising or normal value be interpreted?
- Would an MRI-based approach be appropriate if repeated ultrasounds have poor visualization?
- What follow-up will I need if a blood marker is abnormal or a new nodule is found?
- If I am considering an additional molecular blood test, what evidence supports using it in patients like me?
Frequently asked questions
Can an ultrasound miss early liver cancer?
Yes. Ultrasound sensitivity is lower for early-stage HCC than for HCC at any stage, and technically limited examinations can miss lesions. If an ultrasound report says visualization was limited, discuss appropriate follow-up with your clinician. [1, 3]
Can a normal blood test exclude HCC?
No. Normal AFP does not exclude HCC. A negative result from a newer molecular assay likewise cannot justify stopping recommended surveillance without adequate clinical evidence and a clinician's assessment.
Can an HCC blood test detect cancer that ultrasound misses?
Some blood markers may be abnormal when imaging has not revealed a lesion, but this does not establish that a specific marker reliably detects otherwise occult HCC. For emerging molecular tests, that claim requires prospective studies using appropriate reference standards and patient follow-up.
Do I need both ultrasound and AFP?
AASLD recommends both about every six months for patients eligible for surveillance. Your clinician may adjust the plan if ultrasound is inadequate or you have a new concerning finding. [1]
Are CT or MRI better than ultrasound?
Contrast-enhanced CT and MRI are essential for working up suspicious findings and may be used for surveillance in some patients with inadequate ultrasound visualization. They are not automatically the first routine surveillance test for everyone because cost, availability, contrast considerations, and other clinical factors matter. [1, 2]
Does MoleculeDx replace guideline-recommended ultrasound?
No. A MoleculeDx molecular blood test is not a substitute for a clinician-recommended surveillance or diagnostic imaging plan. A screening result by itself cannot establish or exclude HCC.
References and further reading
- Singal AG, Llovet JM, Yarchoan M, et al. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023;78:1922–1965. DOI: 10.1097/HEP.0000000000000466.
- European Association for the Study of the Liver. EASL Clinical Practice Guidelines on the management of hepatocellular carcinoma. J Hepatol. 2025;82:315–374. DOI: 10.1016/j.jhep.2024.08.028.
- Tzartzeva K, Obi J, Rich NE, et al. Surveillance imaging and alpha fetoprotein for early detection of hepatocellular carcinoma in patients with cirrhosis: a meta-analysis. Gastroenterology. 2018;154:1706–1718.e1. PMID: 29425931.
- Yu YP, Liu S, Geller D, Luo JH. Serum Fusion Transcripts to Assess the Risk of Hepatocellular Carcinoma and the Impact of Cancer Treatment through Machine Learning. Am J Pathol. 2024;194:1262–1271. DOI: 10.1016/j.ajpath.2024.02.017.
- American College of Radiology. LI-RADS US Surveillance v2024: management and visualization score guidance. ACR LI-RADS US guidance.
Learn about MoleculeDx's molecular blood test
MoleculeDx studies fusion-transcript markers that provide different information from imaging and AFP. Discuss whether additional testing is appropriate with your healthcare professional. Continue all guideline-recommended surveillance regardless of a separate molecular screening result.