Screening for Hepatocellular Carcinoma and Overall Survival in a Cohort of Veterans with Cirrhosis: A Snapshot of the Post-Daa Era

ABSTRACT

Background: Screening for hepatocellular carcinoma (HCC) is associated with earlier stage at diagnosis and longer HCC-specific survival after diagnosis, but recent studies in Veterans Affairs (VA) cohorts have reached contrasting conclusions regarding this association. Few studies have evaluated screening and overall survival. Direct-acting antivirals (DAAs) for HCV have altered the natural history of HCV cirrhosis. We evaluated screening and overall survival in the post-DAA and pre-COVID era in a national cohort of veterans with cirrhosis largely due to alcohol and HCV. Method(s) All adults in VA care with CTP A or B cirrhosis for at least 1 year prior to January 1, 2015 were followed for incident HCC and all-cause mortality through December 31, 2019. Patients were censored at development of CTP C cirrhosis or maximum follow-up. Percent of time up to date with screening (PTUDS) for eligible follow up was calculated using relevant cross-sectional imaging. Time-updating age, MELD, and comorbidity score were identified for 180-day windows. We used Cox proportional hazards regression to compare survival time after HCC diagnosis by PTUDS and logistic regression to assess 3-year all-cause mortality after HCC diagnosis. Result(s) A total of 21,441 veterans were included, of whom 4.2% developed CTP C cirrhosis, 34.3% died, and 61.6% were censored at maximum follow-up. In all, 30.3% had cirrhosis from alcohol, 24.6% from HCV, 23.4% from alcohol/HCV, and 17.7% from NAFLD. There were 2,021 incident HCCs. Adjusting for time-updating age, MELD, and comorbidities, as well as etiology of cirrhosis, race, tobacco, BMI, CTP class, and GI and PCP visit density per year of follow-up, PTUDS was associated with decreased mortality in veterans diagnosed with HCC (HR for 10% increase in PTUDS 0.90, 95% CI 0.88-0.92). Restricting to those with HCV cirrhosis, PTUDS was associated with decreased mortality in those who cleared HCV (HR=0.88, 95% CI 0.85-0.91) but not in persistently HCV positive patients (HR=0.98, 95% CI 0.95-1.02). Among veterans diagnosed with HCC by December 31, 2016, PTUDS was associated with decreased 3-year overall mortality adjusting for age, MELD, and comorbidities at time of HCC diagnosis (OR for 10% increase in PTUDS 0.76, 95% CI 0.68-0.86). Restricting to those affected by HCV, the association was stronger in those who cleared HCV (OR=0.72, 95% CI 0.62-0.84) than those persistently HCV positive (OR=0.84, 95% CI 0.71-0.99). Conclusion(s) Screening is associated with both longer survival and greater 3-year overall survival in veterans with CTP A and B cirrhosis diagnosed with HCC, in particular among those who have cleared HCV.