A multi-target blood test detected early liver cancer three times as often as ultrasound, with more false positives

Design

Prospective multicentre diagnostic accuracy study, paired comparison (ALTUS)

Population

2,293 US adults with cirrhosis or chronic HBV under HCC surveillance

Primary outcome

Early-stage HCC sensitivity (vs ultrasound) and specificity, CT/MRI reference

Effect

Sensitivity 66.7% vs 22.2% (difference 44.4%, 95% CI 20.2–63.8); specificity 81.8% vs 98.6%

ALTUS was a prospective multicentre US study comparing a multi-target HCC blood test (mt-HBT) with standard surveillance ultrasound in adults with cirrhosis or chronic hepatitis B. Everyone had CT or MRI as the reference standard. Of 3,089 enrolled, 2,293 had paired results; 37 had HCC, 27 of them early-stage by Milan criteria.

Sensitivity for early-stage HCC was 66.7% (95% CI 47.8–81.4%) with the blood test and 22.2% (10.6–40.8%) with ultrasound — a paired difference of 44.4 points. Any-stage sensitivity was 73.0% versus 29.7%. But specificity was 81.8%, just missing the 82% co-primary target, against 98.6% for ultrasound.

Ultrasound surveillance performs poorly, especially in obese patients with steatotic liver, and this is the strongest evidence yet that a blood test can find more early cancers. The price is about one false positive in five tests, each needing cross-sectional imaging. With only 27 early cancers, the sensitivity estimates are imprecise. The test was studied only in the US, and its place alongside AFP and ultrasound is not yet defined.

• Continue six-monthly ultrasound, ideally with AFP, for HCC surveillance in cirrhosis and high-risk hepatitis B.
• Recognise that ultrasound misses most early cancers, particularly in obesity — consider CT or MRI when views are poor.
• Blood-based HCC tests are promising but produce about one false positive per five tests; they are not yet standard care.
• Improve surveillance adherence — a missed scan detects nothing.

Why it matters

It is the first head-to-head prospective evidence that a blood test can outperform ultrasound for early liver cancer detection.

Don't overread it

Only 27 early cancers, and specificity missed its target — sensitivity estimates are wide and the test has not been shown to reduce mortality.

The statistics, in plain English

With 27 early cancers, each missed or detected case moves sensitivity by nearly 4 points, which is why the interval runs from 48% to 81%. Specificity of 82% means 18 of every 100 people without cancer would test positive.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.