Update, validation and comparison of three different clinicopathological scores for patients with resected gastric cancer: A western experience

Introduction: Gastric cancer (GC) is a multifactorial disease. Several prognostic scores have been proposed for refining the prognostic information provided by the TNM classification. Our aim is to validate and compare the prognostic performance of different clinicopathological scores in a western cohort of patients (Marubini, Haraguchi and Kologlu scores). Material and methods: Retrospective study of all cases of GC resected in a western tertiary center (N = 377). Clinicopathological features were collected, scores were applied and statistical analyses were performed. Results: 315 cases were finally included. According to Marubini, Haraguchi and Kologlu scores, patients were stage I (18.5%, 13.3% and 49%), II (29.3%, 47.2% and 29.5%) and III (52.2%, 39.5% and 21.5%, respectively). All classifications were significantly associated with lymphovascular invasion, perineural infiltration, lymph node involvement, patient progression and death due to GC. All scores showed good patient stratification by Kaplan-Meier analyses, but OS and DFS curves depending on Haraguchi score were less evenly spaced. Kologlu classification showed prognostic superiority over Haraguchi and Marubini classifications by ROC analysis. AUC values for OS and DFS were 0.654 and 0.647 (Marubini), 0.626 and 0.618 (Haraguchi) and 0.724 and 0.709 (Kologlu). Kologlu and Marubini classifications were independent factors for both OS and DFS, but Haraguchi classification was independently associated only with DFS. Conclusions: Clinicopathological scores can be easily validated and are cost-effective. Kologlu score is the most thorough classification, and it showed the best prognostic performance for both DFS and OS in our study. More studies are needed to validate its value in other populations.