Published online Apr 21, 2018. doi: 10.3748/wjg.v24.i15.1641
Peer-review started: January 4, 2018
First decision: January 16, 2018
Revised: March 10, 2018
Accepted: March 25, 2018
Article in press: March 25, 2018
Published online: April 21, 2018
Processing time: 105 Days and 13.1 Hours
To investigate novel predictors of survival in hepatocellular carcinoma (HCC) patients following transarterial chemoembolization (TACE).
One hundred sixty seven patients with un-resectable HCC were retrospectively analyzed to identify factors that might contribute to their HCC biology and aggressiveness. We correlated routine laboratory results (total bilirubin, AST, ALKP, GGTP, albumin etc.) to maximum tumor diameter, number of tumor nodules, portal vein thrombosis and blood alpha-fetoprotein levels. These 4 parameters were previously combined to form an aggressiveness index (AgI). We used The Wilcoxon rank-sum (Mann-Whitney), to test the correlation between the AgI categories and liver function parameters. The Cox proportional hazards model was applied to evaluate the categories of AgI associated with overall survival.
The AgI was strongly correlated with survival in this novel patient population. Three year survival probability for AgI > or < 4 was 42.4% vs 61.8%; P < 0.0863 respectively. Several factors independently correlated with AgI using univariate multiple logistic regression of AgI with 8 laboratory parameters. Lower albumin levels had an OR of 2.56 (95%CI: 1.120-5.863 P < 0.026), elevated Alkaline phosphatase and gamma glutamyl transpeptidase (GGTP) had ORs of 1.01 (95%CI: 1.003-1.026, P < 0.017) and 0.99 (95%CI: 0.99-1.00, P < 0.053) respectively. In a Cox proportional hazard model combining mortality for AgI score and liver function parameters, only GGTP levels and the AgI were independently associated with survival. An AgI > 4 had HR for mortality of 2.18 (95%CI: 1.108-4.310, P < 0.024). GGTP’s single unit change had a HR for mortality of 1.003 (95%CI: 1.001-1.006, P < 0.016). These were considered in the final multivariate model with the total cohort. An AgI > 4 had a HR for mortality of 2.26 (95%CI: 1.184-4.327, P < 0.016). GGTP had a HR of 1.003 (95%CI: 1.001-1.004, P < 0.001).
Our study validates the AgI in a new population with un-resectable HCC patients undergoing TACE. The analysis establishes a correlation between GGTP and the AgI.
Core tip: Our cohort’s population included patients with multiple underlying liver diseases and can be widely generalized. The aggressiveness index (AgI) was correlated with survival. AgI > 4 was associated with decreased survival. Combining the AgI with elevated GGTP and ALKP levels improved its prognostic yield in our patient population. We validated the AgI as a prognostic tool to predict overall survival in a novel population of hepatocellular carcinoma patients undergoing transarterial chemoembolization.