This study aimed to identify clinicopathological factors associated with the outcome of elderly patients with gastric cancer (GC), and to construct a nomogram for individual risk prediction. Tumor characteristics of 143 patients aged ≥ 80 years underwent surgery for GC were collected and analyzed by uni- and multivariate analyses. A prognostic nomogram was constructed using the factors which resulted to be significantly associated with overall survival. Discrimination of nomogram was tested by Kaplan-Meier (KM) curves and boxplots. With a median follow up of 18.37 months, overall 1-year survival rate was 51% and it was 60 and 40% for older and younger than 83 years, respectively (P = 0.003). Univariate analysis indicated that age (P = 0.008), pre-operatory performance status (P < 0.001), depth of invasion (P = 0.007), lymph nodes involvement (P < 0.001), and residual tumor (P < 0.001) were significant prognostic factors. Based on these variables, a nomogram to predict 3, 6, 12, and 24 months survival probability after GC surgery was developed. KM and boxplots according to the range of nomogram total points highlighted the appropriateness of distinguish the patients' survival in all the subgroups. Moreover, this nomogram exhibited superior prognostic discrimination between intermediate stages (II-III) than AJCC-TNM classification. This study showed that after good surgical selection, the prognosis of elderly GC patients may be influenced by several clinicopathological factors. Therefore, a predictive nomogram to distinguish more accurately fit patients may allow physicians to individualize treatments and to detect those patients who may benefit from an intensive multidisciplinary approach.
Keywords: Chemotherapy; Elderly; Gastric cancer; Nomogram; Older adults; Prognosis.