A radiographic tumor burden score-based preoperative model to predict overall survival in patients with hepatocellular carcinoma undergoing resection

Abstract Purpose Tumor size, number, alpha-fetoprotein (AFP), and cirrhosis are well-known prognostic factors in patients with hepatocellular carcinoma (HCC) undergoing liver resection (LR). We aimed to develop a preoperative model to predict overall survival (OS) in these patients based on these variables. Methods Patients who underwent LR for Barcelona Clinic Liver Cancer (BCLC) stage 0, A, or B HCC were enrolled. Results Among 743 enrolled patients, 193 (26.0%) patients had a low tumor burden score (TBS) (< 2.6), 474 (63.8%) had a medium TBS (2.6–7.9), and 75 (10.1%) had a high TBS (> 7.9). Multivariate analysis showed that TBS (low: referent; medium: HR = 2.91; 95% CI: 1.62–5.24; p < 0.001; high, HR = 7.73; 95% CI: 3.90–15.3; p < 0.001), AFP (< 400 ng/mL: referent; ≧400 ng/mL: HR = 1.64, 95% CI: 1.10 − 2.45, p = 0.015), and cirrhosis (absence: referent; presence: HR = 1.88, 95% CI: 1.30–2.73, p < 0.001) were associated with OS. A simplified risk score was superior to BCLC in concordance index (0.688 vs. 0.617), Akaike information criteria (1438 vs. 1466), and homogeneity (57.5 vs. 31.6). It also further stratified patients within each BCLC group relative to five-year OS: BCLC 0: very low = 95%, and low = 78%; BCLC A: very low = 93%, low = 79%, medium = 63%, and high = 52%; and BCLC B: low = 82%, medium = 38%, and high = 21%. Conclusion We have developed a simple preoperative model that performs better in predicting OS than the BCLC system.