The effect of inflammatory-nutri̇tional prognosti̇c scori̇ng (INPS) system on treatment response and prognosi̇s i̇n pati̇ents wi̇th locally advanced gastroesophageal juncti̇on and gastri̇c cancer wi̇th neoadjuvant systemi̇c treatment.

Abstract

e16092 Background: We aimed to reveal the prognostic value of inflammatory-nutritional prognostic score (INPS), which is a new scoring system created by using frequently used immune-nutritional markers, as a biomarker, on nutrition and immune status in patients with locally advanced GEJ and gastric cancer. Methods: Patients who were treated in Dokuz Eylul University Medical Oncology Department between 2010-2022 years and diagnosed with locally advanced GEJ and gastric cancer and receiving NACT were included in this study retrospectively. We selected the most valuable biomarkers (Albumin, ABR, CRP, CAR, De-Ritis, HALP, MLR, NLR, PLR, PNI, SII) to develop INPS by the least absolute shrinkage and selection operator (LASSO) Cox regression model. A prognostic nomogram incorporating INPS and other independent clinicopathological factors was developed based on the stepwise multivariate Cox regression method. The retained features with nonzero coefficients were used to establish a novel INPS. The selected four biomarkers were used to construct the novel INPS for our patients. Results: Median follow-up time of 101 patients was 16.7 months (95% CI= 3.7-73.7), DFS was 20.5 months (95%CI= 16.2-24.8 ), OS was 36.6 months (95%CI=25.4-47.8). The 101 patients in the study were randomly assigned to train (n = 71) and test (n = 30) sets in a 7:3 ratio. Using the LASSO Cox regression model, four inflammatory-nutritional biomarkers (INB) with nonzero coefficients, namely, PLR, DeRitis, PNI and LAR, of the 13 parameters were selected. The LASSO coefficient profiles of the 13 INB and 10-fold cross-validation for tuning parameter selection in the LASSO model were presented. The optimal cut-off values for these four biomarkers based on disease-free survival were determined using the Maximally Selected Rank Statistics. The respective cut-off values for PLR, DeRitis, PNI, and LAR were obtained to be 107.93, 1.75, 33.92, and 58.17. Both the patients in the Train set and the in the Test set were classified as low-risk (INPS score ≤ 1) and high-risk (INPS score >1) groups. There was a statistically significant difference in survival probabilities between low and high risk groups in both groups (respectively; p=0.009, p=0.029). A prognostic nomogram including INPS and other independent clinicopathological factors was developed. Conclusions: The search for biomarkers with prognostic value in GEJ and gastric adenocarcinoma receiving NACT continues. With this study, it has been shown for the first time in the literature that the new screening system, INPS, created with nutritional and inflammatory markers, can be used as a prognostic tool to predict disease-free survival in this patient group.