Abstract
INTRODUCTION: Many single and combination blood tests that reflect local or systemic inflammation have been shown to be useful prognosticators in patients with a variety of tumor types. To try to clarify, this issue in patients with nonsurgically treatable hepatocellular carcinoma, multiple serum parameters were evaluated for their relationship to survival. METHODS: A prospectively collected database was interrogated of 487 patients with known hepatocellular carcinoma and documented survival and having all the inflammation parameters of interest in this study, together with baseline tumor characteristics from CT scans. Serum parameters included NLR, PLR, CRP, ESR, albumin, and GGT. RESULTS: All the parameters had significant hazard ratios on Cox regression model. Combination double parameters with hazard ratios >2.0 were: ESR plus GGT, albumin plus GGT, albumin plus ESR. The triplet combination of albumin plus GGT plus ESR had a hazard ratio of 6.33. Using Harrell's concordance index (C-index), the highest inflammation-based 2-parameter prognostic score was for albumin plus GGT. When clinical characteristics of patients with high values for albumin plus low values for GGT were compared to low values for albumin plus high values for GGT (worse prognosis), statistically significant differences were found for tumor size, tumor focality, macroscopic portal vein invasion, and serum alpha-fetoprotein levels. Addition of ESR did not provide additional tumor information. CONCLUSION: The combination of serum albumin plus GGT levels was the most prognostically useful among the inflammation parameters that were tested, and reflected significant differences in tumor aggressiveness characteristics.