Resected Tumor Outcome and Recurrence (RESTORE) Index for Hepatocellular Carcinoma Recurrence after Resection

Importance: Although many variables have been associated with increased risk of hepatocellular carcinoma (HCC) recurrence after resection, no simple-to-implement risk score has been developed to determine this post-resection risk. Objective: We aimed to identify risk factors for HCC recurrence and develop a risk score for predicting recurrence of HCC in patients who undergo resection with curative intent. Design: Single-center retrospective analysis Setting: Single-center tertiary care referral hospital (University of San Francisco, California). Participants: Patients who underwent resection with curative intent for HCC between January 2005 and May 2019 with complete pathologic findings and recorded follow up. Main Outcomes and Measures: Univariate and multivariate Cox regression analysis were used to identify independent risk factors for HCC recurrence. A multivariable Cox proportional-hazard regression model with listwise deletion was used to create a risk score. Results: A total of 179 patients were included in the study; 129 (72.9%) were men, and the median (IQR) age was 63 (57-67) years. Median alpha-fetoprotein (AFP) was 12.3 ng/mL at time of resection. Most patients (82%) had a single tumor nodule, and the mean aggregate nodule size was 6.75 cm; 28.4% had evidence of vascular invasion. On multivariable Cox proportional-hazards regression, AFP >= 1000 ng/mL, multinodularity, and vascular invasion were independently associated with HCC recurrence. The RESTORE index was created using stratified pre-operative AFP, vascular invasion, and the presence of a single lesion within or beyond Milan Criteria versus multiple lesions. The RESTORE index ranged from 0-9 (highest patient score was 8) and was highly predictive of HCC recurrence (C statistic 0.70). RESTORE could stratify 5-year post-resection HCC recurrence risk, ranging from less than 25% with a score of 0 to more than 80% with a score of 5-8. Conclusions and Relevance: The RESTORE index that we developed and validated is a simple-to-implement and novel risk score for patients undergoing resection for HCC and may help identify those who would benefit most from intensive surveillance strategies or adjuvant therapies.