The following is a summary of “PAGE-B SCORE TO ESTIMATE THE HEPATOCELLULAR CARCINOMA RISK IN PEOPLE WITH HIV AND HBV”, presented by Bernard Surial, Adria Ramirez Mena, Marie Roumet, Andreas Limacher, Colette Smit, Olivier Leleux, Amanda Mocroft, Marc van Der Valk, Fabrice Bonnet, Lars Peters, Jurgen K. Rockstroh, Huldrych F. F. Gunthard, Annalisa Berzigotti, Andri Rauch, Gilles Wandeler.
Coinfection with the hepatitis B virus (HBV) is widespread in people with HIV (PWH) and is the leading cause of hepatocellular carcinoma (HCC) globally. The PAGE-B risk score, which takes into account factors like age, sex, and platelets, is suggested for the prediction of HCC in people with HBV monoinfection, although it has not been tested in PWH. Researchers conducted a cross-cultural external validation of PAGE-B in Europeans with HIV/HBV co-infection. Before initiating tenofovir, they included PWH who tested positive for HBsAg but had no evidence of HCC using data from four European cohorts (the Swiss HIV Cohort Study, EuroSIDA, ATHENA, and Aquitaine). Investigators calculated the PAGE-B score’s ability to foretell the development of HCC in the first 15 years following tenofovir treatment. Cox regression with the prognostic index as a covariate, and the calculation of Harrell’s c-index, were used to evaluate the discriminatory power of the models following multiple imputation. Cumulative incidence estimates were compared to the original PAGE-B derivation research using the Kaplan-Meier method to evaluate model calibration.
About 2,963 had both HIV and HBV. Of these people, the median age was 41 (interquartile range: 35-47), there were 466 (16%) females, 2,023 (68%) were Caucasian, and 314 (11%) showed signs of liver cirrhosis. PAGE-B was less than 10 in 26.5% of patients, between 10 and 17 in 57.7%, and more than equal to 18 in 15.7%. A total of 68 people were diagnosed with HCC during a median follow-up period of 9.6 years (2.58/1000 person-years, 95% CI 2.03-3.27). The pooled c-index was 0.77 (range 0.73-0.80), and the regression slope of the prognostic index for developing HCC within 15 years was 0.93 (95% CI 0.61-1.25), suggesting good model discrimination equivalent to the original model derivation study.
Those with a PAGE-B score more than equal to 18 in the analysis had a 5.6% cumulative incidence of HCC over 5 years, but this number was 17% in the derivation study, indicating that the 2 studies used different calibration methods. An HCC risk of less than 12 remained below the generally established screening threshold and a PAGE-B cut-off of less than 10 had a negative predictive value of 99.4% for HCC within 5 years. PAGE-B is a reliable and valid technique for estimating the risk of HCC in people who are HIV-positive and HBV-positive co-infected. Those whose PAGE-B score is 10 or lower have a very low risk of having HCC. Accurate risk prediction has the potential to boost surveillance uptake in high-risk persons and reduce healthcare costs by the elimination of screening for those at extremely low risk of developing HCC.