Summary: An AJR study found that LI-RADS Ultrasound v2024 improves sensitivity but reduces specificity for detecting hepatocellular carcinoma, with differences linked to incorporating rising alpha-fetoprotein levels and initial VIS-C scores predicting imaging limitations.

Key Takeaways

  1. Improved Sensitivity with v2024: The ACR’s LI-RADS Ultrasound Surveillance v2024 showed higher sensitivity for detecting hepatocellular carcinoma (HCC) compared to v2017, attributed to incorporating increasing, rather than just elevated, alpha-fetoprotein (AFP) levels. However, this improvement came with reduced specificity.
  2. Predictive Value of Initial VIS-C Scores: Initial visualization scores (VIS-C) were the only significant predictor of limitations in subsequent ultrasound exams, emphasizing the importance of initial imaging quality in guiding follow-up surveillance.
  3. Need for Multidisciplinary Frameworks: The study supports a multidisciplinary approach integrating ultrasound results and AFP values to optimize management strategies for at-risk patients, particularly as medical professionals aim to refine follow-up protocols for VIS-C observations.

———————————————————————————————————————————————————————————

A study published in the American Journal of Roentgenology (AJR) found that the ACR’s LI-RADS Ultrasound Surveillance v2024 offers higher sensitivity but lower specificity compared to v2017 for detecting hepatocellular carcinoma (HCC). These differences were attributed to incorporating increasing, rather than just elevated, alpha-fetoprotein (AFP) levels.

“The findings support a multidisciplinary surveillance framework that integrates ultrasound results and AFP values to guide management for at-risk patients,” writes corresponding author Sang Hyun Choi, MD, PhD, from the Research Institute of Radiology, Ulsan College of Medicine, and Asan Medical Center in Seoul, Korea. The study revealed that initial VIS-C scores were the only significant predictor of limitations (VIS-C) in subsequent ultrasounds.

Insights from v2024 Algorithm Study

The study included 407 patients with cirrhosis (median age 56 years; 230 males, 177 females) who underwent semi-annual surveillance ultrasounds in a prospective trial (November 2011–December 2012). Two radiologists independently categorized round-1 ultrasounds and scored visualization for both round-1 and round-2, with a third radiologist resolving disagreements. AFP was positive if elevated or increasing from baseline values, per v2024 criteria. HCC diagnosis was confirmed by biopsy or LR-5 observation on MRI. Diagnostic performance was then compared between v2017 and v2024.

Results showed that v2024 demonstrated higher sensitivity for HCC detection (reader 1: 64.3% vs. 42.9%; reader 2: 64.3% vs. 39.3%) but lower specificity (reader 1: 82.0% vs. 92.6%; reader 2: 82.3% vs. 92.9%). The adjusted odds ratio for VIS-C prediction based on initial scores was 21.0.

“This is the first study to compare v2024 to the earlier algorithm,” says Kathryn McGillen, MD, a radiologist at Penn State Health’s Milton S. Hershey Medical Center. She emphasized that understanding predictors of repeat VIS-C scores could improve follow-up decisions, stating, “Such predictive ability would guide decisions on how and when follow-up after VIS-C scores should occur.”