Impact of etiology of hepatocellular carcinoma on treatment outcomes in a real-world database.

508 Background: Hepatocellular carcinoma (HCC) is a common and deadly form of liver cancer. Systemic therapy with combination atezolizumab and bevacizumab (A/B) is the new treatment standard; yet fewer than 30% experience treatment response. The etiology of HCC (viral, alcohol, or NASH) and its impact on treatment response is unknown. Thus, we sought to determine the impact of etiology on survival outcomes of patients treated with A/B. Methods: This study used the nationwide Flatiron Health electronic health record (EHR)-derived de-identified database. Patients with HCC receiving first-line A/B after January 1, 2020 were eligible for inclusion. We used a validated algorithm to adjudicate cohort by likely etiology (viral, alcohol and NASH). The primary outcome was median real-world overall survival (OS), defined from initial receipt of A/B until death or censored at last receipt of A/B by etiology. The secondary outcome was real-world time to treatment discontinuation (TTD -a validated surrogate that approximates progression-free survival) defined as time from initial receipt of A/B to progression, death, or treatment discontinuation for more than 120 days or censored at last clinical contact by etiology. The Kaplan-Meier method, Cox proportional hazards model and the log-rank test was used to test for differences between etiologies. Results: In total, 429 eligible HCC patients were included (n=216, 50.3% Viral; n=68, 15.8% Alcohol; n=145, 33.8%, NASH). In the total cohort, median OS was 9.4 months (95% CI 7.1-10.9) and median TTD was 5.7 months (5.0-7.0); by etiology, OS medians were 10.1 (7.2-14.0), 9.8 (6.1-13.9), 7.5 (5.3-11.0) months, and TTD medians were 6.1 (5.0-8.0), 5.8 (3.7-7.8), and 5.5 (2.6-7.0) months, for Viral-HCC, Alcohol-HCC, and NASH-HCC respectively. No significant difference in the hazard of death or TTD by etiology were observed in the unadjusted or adjusted Cox regression analyses (see table). The multivariate model included age, ECOG performance status, receipt of prior locoregional therapy, albumin-bilirubin (ALBI) grade, and baseline aspartate aminotransferase. Conclusions: In this large cohort study of US patients with HCC, we did not identify a significant association between etiology and survival or time to treatment discontinuation outcomes, although outcomes were numerically worse in patients with NASH. Although the underlying microenvironment and oncogenesis may differ by etiology, further research is needed to identify those most likely to benefit from A/B. [Table: see text]