Development of a scoring system to predict hepatocellular carcinoma in Asians on antivirals for chronic hepatitis B

•A risk score that predicts HCC during antiviral therapy in patients with chronic hepatitis B is developed and validated. •The CAMD (cirrhosis, age, male sex, and diabetes mellitus) score requires only simple easily available information. •Patients with a CAMD score <8 and >13 points were exposed to distinctly lower and higher risks, respectively. Background & Aims The risk of hepatocellular carcinoma (HCC) during antiviral therapy in patients with chronic hepatitis B (CHB) is inadequately predicted by the scores built from untreated patients. We aimed at developing and validating a risk score to predict HCC in patients with CHB on entecavir or tenofovir treatment. Methods This study analysed population-wide data from the healthcare databases in Taiwan and Hong Kong to identify patients with CHB continuously receiving entecavir or tenofovir. The development cohort included 23,851 patients from Taiwan; 596 (2.50%) of them developed HCC with a three-year cumulative incidence of 3.56% (95% CI 3.26–3.86%). The multivariable Cox proportional hazards model found that cirrhosis, age (cirrhosis and age interacted with each other), male sex, and diabetes mellitus were the risk determinants. These variables were weighted to develop the cirrhosis, age, male sex, and diabetes mellitus (CAMD) score ranging from 0 to 19 points. The score was externally validated in 19,321 patients from Hong Kong. Results The c indices for HCC in the development cohort were 0.83 (95% CI 0.81–0.84), 0.82 (95% CI 0.81–0.84), and 0.82 (95% CI 0.80–0.83) at the first, second, and third years of therapy, respectively. In the validation cohort, the c indices were 0.74 (95% CI 0.71–0.77), 0.75 (95% CI 0.73–0.78), and 0.75 (95% CI 0.72–0.77) during the first three years, and 0.76 (95% CI 0.74–0.78) and 0.76 (95% CI 0.74–0.77) in the extrapolated fourth and fifth years, respectively. The predicted and observed probabilities of HCC were calibrated in both cohorts. A score <8 and >13 points identified patients at distinctly low and high risks. Conclusions The easily calculable CAMD score can predict HCC and may inform surveillance policy in patients with CHB during oral antiviral therapy. Lay summary This study analyses population-wide data from the healthcare systems in Taiwan and Hong Kong to develop and validate a risk score that predicts hepatocellular carcinoma during oral antiviral therapy in patients with chronic hepatitis B. The easily calculable CAMD score requires only simple information (i.e. cirrhosis, age, male sex, and diabetes mellitus) at the baseline of treatment initiation. With a scoring range from 0 to 19 points, the CAMD score discriminates the risk of hepatocellular carcinoma with a concordance rate of around 75–80% during the first three years on therapy. The risk prediction can be extrapolated to five years on treatment with similar accuracy. Patients with a score <8 and >13 points were exposed to distinctly lower and higher risks, respectively. The risk of hepatocellular carcinoma (HCC) during antiviral therapy in patients with chronic hepatitis B (CHB) is inadequately predicted by the scores built from untreated patients. We aimed at developing and validating a risk score to predict HCC in patients with CHB on entecavir or tenofovir treatment. This study analysed population-wide data from the healthcare databases in Taiwan and Hong Kong to identify patients with CHB continuously receiving entecavir or tenofovir. The development cohort included 23,851 patients from Taiwan; 596 (2.50%) of them developed HCC with a three-year cumulative incidence of 3.56% (95% CI 3.26–3.86%). The multivariable Cox proportional hazards model found that cirrhosis, age (cirrhosis and age interacted with each other), male sex, and diabetes mellitus were the risk determinants. These variables were weighted to develop the cirrhosis, age, male sex, and diabetes mellitus (CAMD) score ranging from 0 to 19 points. The score was externally validated in 19,321 patients from Hong Kong. The c indices for HCC in the development cohort were 0.83 (95% CI 0.81–0.84), 0.82 (95% CI 0.81–0.84), and 0.82 (95% CI 0.80–0.83) at the first, second, and third years of therapy, respectively. In the validation cohort, the c indices were 0.74 (95% CI 0.71–0.77), 0.75 (95% CI 0.73–0.78), and 0.75 (95% CI 0.72–0.77) during the first three years, and 0.76 (95% CI 0.74–0.78) and 0.76 (95% CI 0.74–0.77) in the extrapolated fourth and fifth years, respectively. The predicted and observed probabilities of HCC were calibrated in both cohorts. A score <8 and >13 points identified patients at distinctly low and high risks. The easily calculable CAMD score can predict HCC and may inform surveillance policy in patients with CHB during oral antiviral therapy.