Search

Background: Serum prealbumin (PALB) can predict the prognosis of patients with gastric cancer (GC). However, the prognostic value of combination of C-reactive protein and PALB (CRP/PALB) remains unclear. Methods: A total of 419 gastric cancer patients included in a clinical trial (NCT02327481) were analyzed. The present study is a substudy of the trial. Receiver operating characteristic (ROC) curves were generated, and by calculating the areas under the curve (AUC) and the C-index, the discriminative ability of each inflammatory index was compared, including CRP/PALB, C-reactive protein/albumin, Glasgow prognostic score (GPS), modified GPS, systemic immune-inflammation index, neutrophil–lymphocyte ratio, and platelet–lymphocyte ratio. Results: Ultimately, 401 patients were included in this study. The optimal cutoff value of CRP/PALB was 17.7. According to this cutoff point, the entire sample was divided into a CRP/PALB < 17.7 (LCP) group and a CRP/PALB ≥ 17.7 (HCP) group, comprising 245 and 156 patients, respectively. There were 54 and 22 patients experienced recurrence in the HCP and LCP group, respectively, p < 0.001. Compared with traditional inflammatory indices, CRP/PALB had the highest AUC (0.707) and C-index (0.716), all p < 0.05. The post-recurrence survival (PRS) of patients in the HCP group was significantly shorter than that in the LCP group (p = 0.010), especially for pathological stage III patients (p = 0.015) or patients with distant (p = 0.018) or local (p = 0.023) recurrences. Conclusions: The predictive value of preoperative CRP/PALB for the recurrence of GC is significantly better than traditional inflammatory indices. HCP significantly reduces the PRS, especially for pathological stage III patients or patients with distant or local recurrences. [ABSTRACT FROM AUTHOR]