ABSTRACT
BACKGROUND: Systemic inflammation is a key factor in tumor growth. C-reactive protein and albumin are parameters of systemic inflammation from the Glasgow Prognostic Score (GPS). The purpose was to evaluate the prognostic role of GPS in a homogeneous population of gastric cancer patients undergoing surgical treatment only. METHODS: Patients underwent operations between 2009 and 2014. Those who had received perioperative treatment or had other malignancies or inflammatory diseases were excluded. Eighty-eight patients met all inclusion criteria (age >18 years, documented preoperative serum levels of albumin and C-reactive protein, histologically proven gastric cancer, curative operation, including lymphadenectomy). C-reactive protein and albumin levels were retrieved from our prospective database. GPS was correlated with clinicopathologic characteristics and outcome. RESULTS: Increasing GPS was linked to aggressive tumor biology in terms of tumor size (GPS 0: 51.2% T1 and T2, 48.8% T3 and T4; GPS 1: 23.8% T1 and T2, 76.2% T3 and T4; GPS 2: 23.1% T1 and T2, 76.9% T3 and T4; P = .026), synchronous distant metastases (GPS 0: 47.1% M0, 0.0% M1; GPS 1: 25.9% M0, 0.0% M1; GPS 2: 27.1% M0, 100.0% M1; P = .030), venous vessel invasion (GPS 0: 91.2% V0, 8.8% V1; GPS 1: 66.7% V0, 33.3% V1; GPS 2: 55.0% V0, 45.0% V1; P = .008), resection margin status (GPS 0: 97.4% R0, 2.6% R1; GPS 1: 90.0% R0, 10.0% R1; GPS 2: 77.3% R0, 22.7% R1; P = .044), reduced overall survival (GPS 0: median 25.2 months [range 0.4-106.0]; GPS 1: 15.3 [0.2-59.5]; GPS 2: 5.8 [0.1-55.3]; P = .016) with median overall survival in the whole cohort being 16.2 months (range 0.1-106.0) and perioperative mortality (GPS 0: 0.0% of perioperative deaths, GPS 1: 20.0%, GPS 2: 80.0%; P = .036). Furthermore, GPS was identified as an independent prognosticator of overall survival (P = .033). A gradual decrease in survival between GPS subgroups was evident. CONCLUSION: GPS represents an independent prognostic factor for long-term outcome in resected gastric cancer patients without perioperative treatment and is strongly associated with perioperative mortality.
