Comparative Diagnostic Performance of AFP, PIVKA-II, and the ASAP Score for Hepatocellular Carcinoma: A Large Tertiary Care Referral Center Study from India

Background: Alpha-fetoprotein (AFP) and protein induced by vitamin K absence-II (PIVKA-Il) are established biomarkers for hepatocellular carcinoma (HCC). The ASAP score, incorporating age, sex, AFP, and PIVKA-II, may improve diagnosis, but data from Indian populations are limited. We compared the diagnostic performance of AFP, PIVKA-II, and ASAP in a large tertiary referral cohort from India. Methods: In this single-center cross-sectional study, consecutive adults with chronic liver disease and healthy controls evaluated at AIG Hospitals, Hyderabad, from January to December 2025 were included. HCC was diagnosed using standard imaging criteria. AFP and PIVKA-II were measured by a chemiluminescent microparticle immunoassay, and ASAP was calculated for each participant. Diagnostic performance was assessed using receiver operating characteristic (ROC) and precision–recall analyses at standard and optimized cut-offs derived from the Youden index and F1-score. Results: Among 1305 participants, 686 were cirrhotic controls, 278 had cirrhotic HCC, 130 were non-cirrhotic controls, and 37 had non-cirrhotic HCC. At standard cut-offs, AFP (20 ng/mL) showed high specificity but low sensitivity, especially for early-stage HCC (93.3% and 29.6%). PIVKA-II (40 mAU/mL) showed high sensitivity but poor specificity (82.7% and 26.4%). ASAP (0.526) provided the best balance, with 69.4% sensitivity and 52.2% specificity in early-stage HCC, 82.0% and 52.2% in cirrhotic HCC, and 81.9% and 55.5% overall. ROC analysis demonstrated superior discrimination for ASAP across cirrhotic, non-cirrhotic, and etiologic subgroups, including metabolic-dysfunction-associated steatotic liver disease (MASLD) HCC. Optimized cut-offs, particularly higher thresholds for PIVKA-II (71.3–98.1 mAU/mL) and ASAP (0.6–0.9), improved the diagnostic balance. Conclusions: The ASAP score modestly outperforms AFP and PIVKA-II, offering better overall diagnostic accuracy for HCC in a MASLD-dominant cohort. Lower cut-offs may improve early detection, but performance in very-early-stage disease remains suboptimal and hence there is limited benefit in surveillance-relevant settings. Higher population-specific cut-offs are required for Indian populations. These findings support the use of these biomarkers as complementary tools alongside imaging for risk stratification rather than standalone screening tests.