Fibrinogen-to-lymphocyte Ratio As A Novel Preoperative Plasma Indicator For Predicting Lymph Node Metastasis In Patients With Clinically Node-negative Advanced-stage Gastric Cancer

Objective: Various hematological indicators have been reported to predict lymph node metastasis(LNM)in gastric cancer(GC)patients,but the relationship between fibrinogen-to-lymphocyte ratio(FLR)and LNM has not been studied.Therefore,the aim of this study was to evaluate the role of preoperative FLR in predicting LNM in patients with clinically node-negative(c N0)advanced gastric cancer(AGC).Methods: We retrospectively reviewed 1207 patients with primary AGC adenocarcinoma who underwent radical gastrectomy in our hospital from January 2018to January 2021.Firstly,clinicopathological features and hematological parameters of patients within 1 week before surgery were collected and sorted out.Receiver operating characteristic(ROC)curve analysis was used to evaluate the predictive value of FLR for LNM.And patients were divided into high and low FLR groups according to the optimal cutoff value determined by Youden index.The independent predictors of LNM were determined by multivariate logistic regression analysis.The nonlinear relationship between FLR and LNM risk was assessed using restricted cubic spline curve.Sensitivity analyses were performed according to FLR quartiles to further assess the robustness of the results.The nomogram was built based on FLR and clinicopathological characteristics,and the model was evaluated by calibration curves,area under curve(AUC)and decision curve analysis(DCA).In addition,we also resampled 1000 times through the bootstrap method for internal verification,and calculated the corrected concordance index(C-index).Results: A total of 571 eligible patients with c N0 AGC were screened through inclusion and exclusion criteria in this study.The LNM positive rate was significantly increased in patients with tumor size ≥ 5cm,CA199 level ≥ 37,low differentiation and high FLR.The area under the curve(AUC)value for FLR to predict LNM was 0.592,and the sensitivity and specificity were 66.9% and 48.3%,respectively.Patients were divided into the high and low groups based on the optimum FLR cutoff value(1.78).Multivariate analysis showed that T stage(P < 0.001),grade(P < 0.001)and FLR(P =0.009)were independent predictors of LNM.The results of RCS curve analysis showed that there was a linear relationship between the FLR value and the risk of LNM,and the risk of LNM increased with FLR value.The results of subsequent sensitivity analysis further confirmed that high FLR level was an independent risk factor for LNM(P for trend = 0.013).The clinicopathological characteristics between the high FLR group and the low FLR group were compared.The results showed that there were significant differences in age,tumor size,T stage,CA199 level and LNM incidence between the two groups.The distribution difference of FLR value between subgroups classified according to T stage(P = 0.027)and tumor size(P < 0.001)was statistically significant.According to the results of multivariate analysis,a nomogram was constructed to predict LNM in c N0 AGC patients.The correction curve can show that the predicted probability of the model was consistent with the actual probability(Hosmer-Lemeshow test,P = 0.503).In addition,the C-index of internal verification was 0.755,indicating that the prediction accuracy of the model was good.The model predicted that the AUC of LNM was 0.741(95% CI = 0.703-0.777),which was significantly better than the single T stage(P < 0.001)and grading(P < 0.001).DCA also shows that this model has higher clinical practicability than T staging,grading or FLR alone.Conclusion: FLR has promising clinical applications for predicting LNM in patients with c N0 AGC.The nomographs based on FLR,T staging and grading performed well in the risk assessment of LNM,and can be used to guide the selection of individualized treatment plans.