Changing epidemiology of HCC in Melbourne, Australia: primary outcomes of the prospective homer-2 cohort.

Abstract

Background: Surveillance for hepatocellular carcinoma (HCC) with six monthly ultrasounds for at-risk patients with cirrhosis or chronic hepatitis B is recommended is recommended, however rates of uptake remain unacceptably low internationally. With the HOMER-2 cohort, we aimed to describe the uptake of surveillance and the changes in HCC epidemiologic profile. Method(s): HOMER-2 is a prospective, multi-site cohort of all incident adult HCC cases from Greater Melbourne, Australia identified through HCC multidisciplinary team meetings at the eight tertiary centres with HCC specialist units between 18 Oct 2021 and 17 Oct 2022. Surveillance uptake occurred where patients were known to health services, receiving follow up and scans in the two years preceding HCC diagnosis, to account for COVID-19 disruptions to care. Optimal surveillance was defined as completion of two scans in the preceding year, and suboptimal surveillance completed less. Variables were compared to the similar HOMER cohort of HCC cases from 1 July 2012 and 30 June 2013 using Chi square or Wilcoxon rank-sum test. Result(s): There were 202 incident HCC cases identified, 78% were male (n = 159), median age of 68 years (IQR 60-74 y), 61% born overseas (n =116). Established cirrhosis was known in 47% (n =96), 33% (n = 66) newly diagnosed with cirrhosis at time of HCC diagnosis, 20% (n = 41) non-cirrhotic. Early-stage HCC (BCLC 0-A) occurred in 47% (n = 95). Surveillance uptake occurred in 39% (n = 79). Of these, 53 % (n = 42) had optimal surveillance and 47% (n = 39) had suboptimal surveillance. In patients with known cirrhosis, 67% (n= 66) were undergoing surveillance and 45% (n= 21) of eligible patients with chronic hepatitis B. Having at least one abdominal scan in the preceding 12 months for any indication was associated with increased odds of early-stage HCC (OR 2.99, 95% CI 1.68-5.32, p <0.0001). Surveillance through imaging +/- AFP was the most common mode of presentation at 37% (n =75), then symptomatic disease (pain/ decompensation) 34% (n = 69), clinician investigation of liver test derangement 16% (n = 32) and incidental findings 13% (n =27). Compared to the original HOMER cohort (2012-13), there was an overall decrease in crude HCC incidence estimates from 6.57 to 4.11/100 000, a significant decrease chronic hepatitis C as a risk factor and increase in MAFLD; a higher proportion diagnosed with early-stage HCC, although there was no improvement in surveillance uptake over time (Table 1). Conclusion(s): The incidence of HCC in Melbourne, Australia has reduced between 2012-13 and 2021-22, temporally associated with the introduction of DAAs for HCV in 2016, and an increase in MAFLD-HCC reflecting the obesity epidemic. Earlystage HCC at diagnosis was more common in 2021-22, but rates of diagnosis in surveillance remained suboptimal, highlighting the need for programs to promote community diagnosis of liver disease and enrolment in HCC surveillance. (Table Presented).