Background: Pre-treatment frailty is associated with unfavorable short and long-term outcomes among patients with hepatocellular cancer (HCC) post liver resection.However, there is a knowledge gap in identifying the subset of older adults with HCC at high risk for inferior survival outcomes. Methods: We evaluated the association between patient-reported geriatric assessment (GA)-based frailty index and overall survival (OS) among older adults (≥60y) diagnosed with HCC enrolled in a single institutional prospective registry (Cancer and Aging Resilience Assessment (CARE) registry). All patients underwent a patient-reported GA at enrollment that encompassed multiple health domains related to aging. To determine frailty, we utilized a 44-item CARE Frailty Index (CARE-FI) based on the principles of deficit accumulation (Giri S et al JAGS 2021). We categorized frailty into robust, pre-frail and frail. The primary outcome was OS from the time of GA. Comparison of survival between groups was made using log-rank statistic. A multivariable Cox regression model measured the independent association between frailty and OS adjusted for age, etiology, Child-Pugh score and cancer stage. Results: A total of 116 older adults with HCC were identified with a median follow-up 2.1 year, median age was 67y (63-72) with 82% males, 27% blacks, 64% with Child-Pugh-A and 78% with stage III/IV disease. Overall, 50.1% (n=58) were frail, 33.6% (n=39) pre-frail and 16.3% (n=19) robust. No significant clinico-demographic differences across the 3 frailty groups were observed. In univariable analysis, there was a marginally significant difference between the three frailty groups (p value 0.06). However, in multivariable analysis, being frail vs robust was associated with an increased risk of mortality (HR 2.6 [95% CI 1.03 – 6.56]; p=0.04) after adjustment for aforementioned confounders. Conclusions: Frailty status assessed by patient-reported GA using CARE-FI is associated with an increased risk of mortality in older adults with HCC. Frailty assessment should be encouraged and considered in prospective clinical trials to estimate the impact of treatment on the most vulnerable adults.

*The age is continuous variable in 1-year increments.