Background: With the recent increase in metabolic liver diseases and viral hepatitis on anti-viral agents, risk stratification biomarkers are urgently needed to identify cirrhosis patients who should be prioritized for hepatocellular carcinoma (HCC) screening for early detection. We previously developed the Prognostic Liver Secretome signature with alpha-fetoprotein (PLSec-AFP), which enables cost-effective risk-stratified HCC screening in patients with cirrhosis.
Methods: PLSec-AFP was further integrated with a clinical variable-based score (aMAP score) to define PAaM, which outperformed each score alone for HCC risk stratification. PAaM was validated in two phase 3 biomarker studies using the prospective statewide Texas HCC Consortium (THCCC) and nationwide NCI EDRN HCC Early Detection Strategy (HEDS) cohorts of cirrhosis patients undergoing semi-annual HCC screening, following the prospective specimen collection, retrospective blinded evaluation (PRoBE) design. The associations between baseline PAaM and incident HCC were assessed by Fine-Gray regression, with death and liver transplantation as competing events.
Results: Of 2,156 evaluable patients with cirrhosis in THCCC, 155 patients developed HCC during a median follow-up of 2.3 (interquartile range [IQR], 1.1-4.0) years. High-risk PAaM identified 404 (19%) high-risk and 903 (42%) intermediate-risk patients in the THCCC cohort, which was associated with incident HCC with sub-distribution hazard ratios (sHR) of 7.51 (95% confidence interval [CI], 4.42-12.8) and 4.20 (95% CI, 2.52-7.01) compared to low-risk patients (n = 849, 39%), respectively. Annual HCC incidence rates of high-, intermediate-, and low-risk PAaM were 5.3%, 2.7%, and 0.6%, respectively. Of 1,328 evaluable patients with cirrhosis in the HEDS cohort, 110 patients developed HCC during a median follow-up of 2.9 (IQR, 1.1-5.2) years. PAaM identified 201 (15%) high-risk and 540 (41%) intermediate-risk patients, which were associated with incident HCC with sHRs of 6.54 (95% CI, 3.85-11.1) and 1.77 (95% CI, 1.02-3.08) compared to low-risk patients (n = 587, 44%), respectively. Annual HCC incidence rates of high-, intermediate-, and low-risk PAaM were 6.2%, 1.8%, and 0.8%, respectively. In both cohorts, PAaM achieved the benchmark to enable cost-effective risk-stratified HCC screening (i.e., sHR > 2) defined in our prior Markov model-based simulation.
Conclusions: In two large-scale phase 3 biomarker validation studies, PAaM was successfully validated for HCC risk stratification. In addition to identifying high-risk patients, PAaM can identify low-risk patients (approximately 40%) with HCC incidence rate below the cost-effectiveness threshold for HCC screening.

Yujin Hoshida

University of Texas Southwestern Medical Center

Appearances