Simple model stratifies liver cancer risk in non-cirrhotic chronic hepatitis C patients after viral eradication

Current guidelines recommend hepatocellular carcinoma (HCC) surveillance only in cirrhotic chronic hepatitis C (CHC) patients after sustained virologic response (SVR). With the cost-effectiveness threshold for surveillance lowered to 0.7% annually, the need to extend surveillance to non-cirrhotic remains controversial. We aimed to develop and validate a simple model to stratify post-SVR HCC risk in non-cirrhotic CHC patients.

Multicenter cohort of 5741 direct-acting antiviral-treated CHC patients with pre-therapy FIB-4 < 3.25 (derivation = 1902; validation = 3839). A risk model for HCC was derived using Cox regression.

Over a median follow-up of 3.5 years, 85 patients developed HCC. Age emerged as the key stratifier: in patients < 65 years, a composite score performed best, whereas in ≥ 65 years, AFP alone sufficed. Age ≥ 65 years was prespecified, as incidence exceeded 0.7% annual threshold. In patients < 65 years, the SFA score (male: 1, AFP ≥ 6:1, FIB-4 ≥ 2.5: 2; range: 0-4) showed strong discrimination (C-index 0.90 derivation, 0.84 validation). SFA ≥ 3 defined a high-risk group with 2.9% annual HCC incidence. In those aged ≥ 65 years, AFP ≥ 6 ng/mL alone predicted high risk (annual incidence 3.5%; C-index 0.826 derivation; 0.792 validation). AFP ≥ 6 ng/mL and FIB-4 ≥ 2.5 thresholds were identified by Youden's index in derivation and validated thereafter.

The SFA score (for < 65 years) and AFP cut-off (for ≥ 65 years) identify a small subset of non-cirrhotic CHC patients with sufficiently high HCC risk to justify continued surveillance after SVR. This simple model supports precision HCC surveillance in an overlooked population.