Enhanced MRI Detects Liver Cancer Far Earlier Than Ultrasound in High-Risk Patients
A randomized trial found abbreviated MRI caught early liver cancer at nearly 3x the rate of standard ultrasound in cirrhosis patients.
Summary
A randomized clinical trial comparing screening tools for liver cancer found that an enhanced, abbreviated MRI using a contrast agent called gadoxetic acid dramatically outperformed standard ultrasound in detecting early-stage hepatocellular carcinoma (HCC) in high-risk patients with cirrhosis. Among 759 patients, the MRI approach detected early-stage cancer in 7.7% of patients versus just 2.9% with ultrasound. Even very early-stage cancers were caught at a rate of 6.1% with MRI compared to only 0.8% with ultrasound. As a bonus, MRI also produced fewer false referrals — unnecessary follow-up procedures triggered by suspicious but non-cancerous findings. These results, presented at a major liver disease conference, suggest that upgrading surveillance protocols for high-risk individuals could meaningfully improve survival outcomes.
Detailed Summary
Liver cancer caught early is far more treatable, and a new randomized trial suggests the tool used for surveillance makes a significant difference. The AMRIUS study, presented at the European Association for the Study of the Liver annual meeting, tested whether an enhanced abbreviated MRI could outperform the standard ultrasound-plus-AFP blood test approach currently recommended for cirrhosis patients at high risk of hepatocellular carcinoma (HCC).
The results were striking. Among 759 adults with cirrhosis and no prior HCC history, the MRI approach detected early-to-advanced stage cancer in 8.5% of patients versus 3.1% with ultrasound. For early-stage HCC specifically — the window where curative treatment is most viable — detection rates were 7.7% with MRI compared to 2.9% with ultrasound. At the very earliest detectable stage, MRI found cancer in 6.1% of patients versus just 0.8% with ultrasound.
Beyond detection, the MRI approach also reduced false referrals — cases where imaging flags a suspicious lesion that turns out to be benign, triggering unnecessary and costly follow-up. False referral rates were 2.1% with MRI versus 4.4% with ultrasound overall, and even lower among the healthiest patient subgroup.
The contrast agent used, gadoxetic acid, creates high liver-to-lesion contrast, which the researchers believe makes the protocol more reproducible across different clinical settings. The lead investigator acknowledged that resource-limited primary care environments may struggle to replicate these results, but noted that tertiary care and hepatology centers should achieve similar outcomes.
Despite growing evidence favoring MRI surveillance, current U.S. guidelines from the AASLD still recommend ultrasound plus AFP testing as standard care. This trial adds significant weight to calls for updating those guidelines, particularly for the highest-risk patients where early detection most directly translates to survival benefit.
Key Findings
- Abbreviated MRI detected early liver cancer at 7.7% vs 2.9% with ultrasound in cirrhosis patients
- Very early-stage liver cancer detection was 6x higher with MRI than ultrasound (6.1% vs 0.8%)
- MRI produced roughly half the false referral rate compared to ultrasound (2.1% vs 4.4%)
- Gadoxetic acid contrast agent improves liver-lesion visibility, potentially making MRI protocols more widely reproducible
- Current AASLD guidelines still recommend ultrasound; this trial supports a push to update surveillance standards
Methodology
This is a meeting coverage news report summarizing findings from the AMRIUS randomized clinical trial presented at EASL 2026. The study is a single-center RCT conducted in South Korea with 759 participants, providing moderate-to-strong evidence. The article is from MedPage Today, a credible medical journalism outlet, though the primary peer-reviewed publication has not yet been cited.
Study Limitations
This is a single-center trial from South Korea, which may limit generalizability to diverse populations and healthcare settings. The full peer-reviewed publication has not been referenced, so methodology details remain incomplete. Results may not be reproducible in lower-resource settings where MRI expertise and gadoxetic acid availability are limited.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.