Performance of ASAP, LG2m, and GP73 for HCC Detection in Patients with Cirrhosis

Background: The sensitivity of ultrasound plus alpha-fetoprotein (AFP) for early-stage hepatocellular carcinoma (HCC) detection remains suboptimal, prompting interest in blood-based biomarkers as an alternative strategy. The ASAP [age, sex, AFP, PIVKA-II] score has promising performance, although validation in contemporary Western populations is needed. Methods: We conducted a case-control study in which the cases were patients with HCC and the controls had cirrhosis without HCC. AFP, PIVKA-II, LG2m, and GP73 were measured using the Abbott Alinity platform, and ASAP scores were calculated. Sensitivity and specificity for early-stage HCC (BCLC stage 0/A) were compared between ASAP and ultrasound plus AFP using McNemar's chi-square test. Results: Among 294 patients (median age of 61 years, 67.7% men), 50.7% had HCC (69.1% BCLC 0/A), and 49.3% had cirrhosis. The performance of individual biomarkers for early-stage HCC was moderate to poor (AUROC: AFP 0.71, PIVKA-II 0.73, LG2m 0.51, and GP73 0.53). ASAP, at a cutoff of ≥0.526, demonstrated a sensitivity of 88.3% for early-stage HCC, with specificity of 54.5%. Sensitivity was higher in men and viremic HCV, whereas specificity was higher in younger individuals, women, and Child-Pugh A cirrhosis. In subgroup analysis (n = 178), ASAP had comparable sensitivity to ultrasound plus AFP (85.5% vs. 90.3%; p = 0.51) but lower specificity (53.4% vs. 93.1%; p < 0.001). Using the Youden-optimized cutoff (≥0.804) improved specificity to 83.4% while maintaining a sensitivity of 68.0% for early-stage HCC. Conclusions: ASAP demonstrated high sensitivity for early-stage HCC detection but low specificity. These findings suggest a need for optimized cutoffs prior to implementation in clinical practice.