An epidemiological association between the trends of nucleoside analogue prescription and liver cancer incidence from 1999 to 2012 in Hong Kong, China

Abstract

BACKGROUND AND AIMS: More than 85% of hepatocellular carcinoma (HCC) in Hong Kong, China is associated with chronic hepatitis B (CHB) infection. Nucleoside analogue therapy for CHB was introduced to Hong Kong since 1999.We determined whether there was a positive effect on the population incidence of liver cancer by the introduction of nucleoside analogue therapy up to 2012. METHODS: We obtained nucleoside analogue (lamivudine, adefovir, telbivudine, entecavir and tenofovir) prescription data (1999–2012) from the electronic health record system of the Hospital Authority, Hong Kong, which provides >90% of secondary and tertiary care to the Hong Kong population. We excluded prescriptions with concomitant anti-HIV medications. Liver cancer data from 1990 to 2012 was obtained from the population-based Hong Kong Cancer Registry. Using Poisson piecewise regression analysis, we compared rate trends of liver cancer incidence from the period of 1990 to 1998 with the period of 1999 to 2012. RESULTS: Nucleoside analogue prescription patient headcount increased from 329 prescriptions in 1999 to 5,942 prescriptions in 2006, then to 26,411 prescriptions in 2012. The increase rate of 1999–2006 and 2006–2012 were 802 and 3,411 prescriptions per year respectively (p < 0.001). Age-standardized incidence of liver cancer, using World Health Organization standard population 2000 as reference, dropped from 22.4 per 100,000 persons in 1998 to 15.4 per 100,000 persons in 2012. After adjusting for the change in local CHB prevalence over the study period, there was a decline in the age-adjusted liver cancer incidence for all patients (2.7%, p < 0.001, 95% CI 1.4–4.0%), male patients (3.5%, p = 0.009, 95% CI 1.1–5.9%) and female patients (2.0%, p < 0.001, 95% CI 1.1–2.9%). When analyzing specific age groups, the decline in liver cancer incidence was most significant among patients aged 50–59 years (male: 15.8%, p = 0.006, 95% CI 5.7–25.9%, female: 7.1%, p = 0.001, 95% 3.4–10.9%), and patients aged 30–39 years (male: 3.8%, p = 0.008, 95% CI 1.3–6.4%; female: 1.3%, p = 0.004, 95% CI 1.1–2.1%). CONCLUSIONS: From this interim analysis, we demonstrated an ecological association of nucleoside analogue prescription with the incidence of liver cancer in a CHB-prevalent region, providing epidemiological evidence that nucleoside analogue therapy for CHB could reduce the risk of HCC. Further analysis will be performed for liver cancer mortality and for more a detailed comparison of different time periods.