PERFORMANCE OF A MULTI-TARGET BLOOD TEST VERSUS ULTRASOUND IN DETECTING EARLY HEPATOCELLULAR CARCINOMA: RESULTS FROM THE ALTUS PROSPECTIVE STUDY
Recommended Citation
Rezaee-Zavareh MS, Wang T, Tabrizian P, Yeo Y, Ji F, Adetyan H, Luu M, Zhong S, Marino R, Ward SC, Abdelrahim M, Esmail A, Li S, Fang J, Zhu Z, Zhang S, Lei Q, Chen H, Deng F, Liu X, Shi J, Wang B, Hou Y, Zhu L, Tang R, Yang M, Wu J, Sun Q, Zheng Z, Salgia R. PERFORMANCE OF A MULTI-TARGET BLOOD TEST VERSUS ULTRASOUND IN DETECTING EARLY HEPATOCELLULAR CARCINOMA: RESULTS FROM THE ALTUS PROSPECTIVE STUDY. Hepatology 2026; 83(1).
Document Type
Conference Proceeding
Publication Date
1-1-2026
Publication Title
Hepatology
Abstract
Background: Abdominal ultrasound (US) is a central component of hepatocellular carcinoma (HCC) surveillance despite suboptimal early-stage sensitivity and poor adherence. In previous studies, a multi-target HCC blood test (mt-HBT) incorporating methylated DNA markers, AFP, and patient sex showed promise in early detection of HCC. Here, we report the mt-HBT sensitivity and specificity with a prospective head-to-head comparison to US for early-stage HCC detection. Methods: ALTUS (NCT: 05064553) is a prospective multicenter study in the United States that enrolled adults with cirrhosis (92.9%) or chronic HBV. Participants underwent standard-of-care HCC surveillance (92.9% US, 7.1% CT/MRI) and concurrent mt-HBT testing. Reference triphasic CT/MRI was performed within 30 days. HCC was defined by central radiology adjudication of LI-RADS 5 or tumor-in-vein observations by two blinded, independent radiologists, or by pathology. Early-stage HCC was defined by Milan Criteria. The mt-HBT algorithm was locked before sample processing by a blinded central lab. The primary objectives were to assess early-stage HCC sensitivity in a non-inferiority comparison to US with sequential testing of superiority, and overall specificity. The secondary objective was to assess overall sensitivity for HCC. Results: Among 3089 enrolled participants, 2467 (79.9%) were evaluable (mean age, 63.1 years; 42.8% female; 20.5% Hispanic). The most common cirrhosis etiologies were MASLD (40.5%), ALD (27.2%), HCV (14.0%), and HBV (2.2%). Forty HCCs were identified, of which 28 were early-stage. Sensitivity [95% CI] for detecting early-stage HCC was higher for mt-HBT vs.US (66.7% [47.8-81.4] vs.22.2% [10.6-40.8], respectively; p=0.002, Figure 1). Overall HCC sensitivity was 70.0% [54.6-81.9] for mt-HBT, 29.7% [17.5-45.8] for US, and 52.6% [37.3-67.5] for US + AFP (>=20 ng/mL), while mt-HBT, US, and US + AFP specificities were 81.9% [80.3-83.4], 98.6% [98.0-99.0], and 97.7% [96.9-98.2], respectively. mt-HBT sensitivities by tumor size (<2 cm, 2-5 cm, >5 cm) were 58.3% [32.0-80.7], 78.3% [58.1-90.3], and 75.0% [30.1.-95.4], respectively. Conclusion: mt-HBT achieved higher sensitivity for early-stage HCC compared to US while maintaining a clinically relevant specificity above the expert-consensus recommended threshold of 80%. mt-HBT's enhanced ability to detect very-early and early-stage HCC provides an opportunity to improve the effectiveness of HCC surveillance in at-risk patients.
Volume
83
Issue
1
