Comparison of the New Neo-Glasgow Prognostic Score Based on the Albumin-Bilirubin Grade with Currently Used Nutritional Indices for Prognostic Prediction following Surgical Resection of Hepatocellular Carcinoma: A Multicenter Retrospective Study in Japan

Abstract

Nutritional assessment is important for predicting a prognosis in hepatocellular carcinoma (HCC). The authors examined the utility of the recently developed neo-Glasgow prognostic score (GPS) as a nutritional prognostic assessment in HCC in a multicenter retrospective study of 271 patients with HCC and Child-Pugh class A liver function who underwent R0 resection between 2011 and 2013. The median age was 72 years, 229 and 42 patients had Child-Pugh scores of 5 and 6, respectively, 223 patients had single tumors, the median tumor size was 3.6 cm, and open and laparoscopic resection were performed in 138 and 133 patients, respectively. We compared the prognostic predictive utility of the prognostic nutritional index, neutrophil/lymphocyte and platelet/lymphocyte ratios, controlling nutritional status score, GPS, and neo-GPS, which uses albumin-bilirubin grade (ALBI) instead of albumin. The c-indexes for the predictive prognostic value for overall survival (OS) and progression-free survival (PFS) were best for neo-GPS (OS: 0.571 vs. ≤0.555; PFS: 0.555 vs. ≤0.546). In multivariate analysis with the Cox proportional hazards model, elevated alpha-fetoprotein (AFP; ≥100 ng/mL; hazard ratio [HR] 2.190, 95% confidence interval [CI] 1.493–3.211, p < 0.001), multiple tumors (HR 1.784, 95%CI 1.178–2.703, p = 0.006), tumor size of ≥5 cm (HR 1.508, 95%CI 1.037–2.193, p = 0.032), and neo-GPS of ≥1 (HR 1.554, 95%CI 1.074–2.247, p = 0.019) were significant prognostic factors for OS, whereas elevated AFP (≥100 ng/mL) (HR 1.743, 95%CI 1.325–2.292, p < 0.001), multiple tumors (HR 1.537, 95%CI 1.148–2.057, p = 0.004), and neo-GPS of ≥1 (HR 1.522, 95%CI 1.186–1.954, p = 0.001) were significant prognostic factors for PFS. A neo-GPS of ≥1 was associated with a higher rate of high-grade (≥3) Clavien-Dindo complications than a neo-GPS of <1 (31.1% vs. 17.0%, p = 0.007). Neo-GPS was a good prognostic nutritional assessment tool for the prediction of postoperative complications and prognosis in patients undergoing surgical HCC resection.