Int J Radiat Oncol Biol Phys 2018 07 12;101(4):900-909. Epub 2018 Apr 12.
Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada; Radiation Medicine Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada. Electronic address:
Purpose: To assess the baseline albumin-bilirubin (ALBI) score as a predictor of toxicity and survival in a prospective cohort of Western patients with hepatocellular carcinoma (HCC) treated with stereotactic body radiation therapy (SBRT) in 2 prospective trials.
Methods And Materials: The study included 102 patients with Child-Pugh class A liver disease who received 6-fraction SBRT for HCC. Univariate and multivariable logistic regression investigated factors associated with toxicity, defined as an increase in Child-Pugh score ≥ 2 within 3 months of SBRT. Univariate and multivariable Cox regression analyses investigated factors predictive of overall survival (OS). The ALBI score was analyzed as a continuous and binary variable in separate analyses.
Results: On multivariable analysis of toxicity, including the ALBI score as a continuous variable, the ALBI score (odds ratio [OR] per 0.1-unit increase, 1.51; 95% confidence interval [CI] 1.23-1.85; P = .00074), mean liver dose (OR, 1.31; 95% CI 1.02-1.68; P = .036), and dose received by 800 cm of normal liver (OR, 1.10; 95% CI 1.01-1.20; P = .028) were significant. When the ALBI score was included as a dichotomous variable, the ALBI grade remained a significant predictor of toxicity (OR, 7.44; 95% CI 2.34-23.70; P = .00069). On multivariable analysis of OS, including the ALBI score as a continuous variable, the ALBI score (hazard ratio [HR] per 0.1-unit increase, 1.09; 95% CI 1.03-1.17; P = .004), tumor thrombus (HR, 1.94; 95% CI 1.23-3.07; P = .004), and treatment in trial 1 versus trial 2 (HR, 1.92; 95% CI 1.23-3.03; P = .004) were significant. Similarly, when the ALBI score was included as a binary variable, the ALBI grade, tumor thrombus, and trial were significant predictors of OS. When the ALBI score was considered, the Child-Pugh score (A6 vs A5) was not significant in multivariable models analyzing toxicity or survival. Concordance statistics indicated models containing the ALBI score were superior to those containing the Child-Pugh score.
Conclusions: The baseline ALBI score was more discriminating than the Child-Pugh score in predicting OS and toxicity in patients with Child-Pugh class A liver disease. The ALBI score should be used as a factor for stratification in future HCC SBRT trials.