P-190 Exploring possible predictors of survival in hepatocellular carcinoma

Abstract

Hepatocellular carcinoma (HCC) is an aggressive tumor occurring in chronic liver disease (CLD) and the 4th leading cause of cancer death worldwide. Frequently diagnosed in advanced stages, survival is poor (6-20 months). Treatment options are determined by disease extent, performance status, and severity of underlying liver disease. We aimed to correlate clinical and laboratorial findings with survival. We performed a retrospective single-center study of patients diagnosed with HCC between 2015-2019. The outcome of interest was overall survival (OS), using Kaplan-Meier Method. ROC curve was performed establishing an alpha-fetoprotein (AFP) cut-off value (dividing into two groups: low vs. high, cut-off: 86ng/mL). Spearman test and Cox regression analysis were performed. A total of 123 patients were diagnosed, 5 excluded due to follow-up (FUP) in other hospitals. Median age at diagnosis was 64 years (IQR 57.9-74.7), with 89% (N=105) male patients. Alcohol consumption was the most prevalent cause of CLD (71%; N=84), followed by hepatitis C (HCV) (36%; N=42), and hepatitis B (HBV) (11%; N=13). In 4 patients (3%), CLD originated from nonalcoholic steatohepatitis. More than one cause for CLD was found in 30% (N=35). Stage-wise, BCLC was 0 in 3% of patients (N=4), A in 7% (N=8), B in 20% (N=23), C in 40% (N=47), and D in 31% (N=36). Child-Pugh was A in 36% (N=42), B in 40% (N=47), and C in 24% (N=28). Seventy-eight percent patients (N=92) were treated with palliative intent: 77% (N=71) with best supportive care; 15% (N=14) with sorafenib and 8% (N=7) with chemoembolization. Twenty-two percent (N=26) were treated with curative intent: 65% (N=17) with chemoembolization; 12% (N=3) with surgical resection; 8% (N=2) with ablation therapy; and 1 patient underwent liver transplant. Three patients await curative treatment at time of analysis. One patient was treated with regorafenib after progressing with sorafenib. Median OS was 5.4 months [95% Confidence Interval (CI) 3.9-7] with a median FUP of 46.6 months (95%CI 23.1-70.2). Patients with HCV infection had a median OS superior compared to HBV infection (4.5 vs. 2.1 months; 95%CI 1.7-6.4; p < 0.0001). Longer median OS was observed in patients with alcoholic CLD alone compared with HCV or HBV infection alone or viral infection plus alcoholic CLD, although it was not statistically significant: 6.3 vs. 4.8 vs. 2.5 months, respectively (95%CI 3.8-7.2; p=0.629). AFP sensitivity for predicting survival outcome was 65.3% and specificity was 77% (p=0.011), with negative correlation between AFP and survival (rs=-0.43; p < 0.0001). In a multivariate Cox regression, HR for AFP was 2.1 (95%CI 1.3-3.3; p=0.001), for BCLC it was 2.4 (95%CI 1.7-3.3; p < 0.0001), for Child-Pugh it was 1.6 (95%CI 1.2-2.3; p=0.004), and for presence of extrahepatic disease it was 1.6 (95%CI 1-2.6; p=0.036), while macroscopic vascular invasion did not reach statistical significance. Regarding the etiology of CLD, alcohol was the only significant factor (HR 0.62; 95%CI 0.4-1; p=0.045). HCC frequently presents at advanced stages and prognosis remains poor. Patients with HCV-induced CLD had longer survival compared with HBV infection. Besides classical predictors of bad prognosis, such as BCLC, Child-Pugh scores and extrahepatic disease, high AFP level was an independent predictor of reduced survival.