Analysis Of Prognostic Factors And Establishment Of Prognostic Evaluation Model For Patients With Gastric Cancer Based On Inflammation-Nutrition Related Indicators

Background and ObjectiveGastric cancer is one of the malignant tumors with high morbidity and mortality in China.Most of the existing treatment decisions are based on the TNM stages and pathological types of patients.However,due to the high degree of heterogeneity of gastric cancer,even patients in the same stage and the same pathological type still have large differences in response to treatment and prognosis.Therefore,the establishment of simple and feasible auxiliary strategies to improve the accuracy of prognosis prediction has become an urgent need for individualized diagnosis and treatment of gastric cancer and improved prognosis.In recent years,the relationship between systemic inflammation and nutritional status and the prognosis of gastric cancer has attracted the attention of many researchers.High systemic inflammation or poor nutritional status has become an independent related factor for early warning of poor prognosis of gastric cancer.However,the current research on the relationship between inflammation and nutrition-related indicators and the prognosis of gastric cancer is not sufficient and there are inconsistent conclusions.Most studies only evaluate the relationship between single inflammation or nutritional markers and the prognosis of gastric cancer,and lack of systematic research on preoperative and postoperative inflammation and nutrition-related indicators.Therefore,we plan to use the same population cohort to systematically evaluate the relationship between preoperative and postoperative inflammatory and nutritional indicators and the prognosis of gastric cancer,to screen for inflammation and nutrition markers with higher predictive efficiency,and to establish a prognostic evaluation model for gastric cancer to assist clinical decision-making.Methods1.From October 2008 to January 2017,continuously included in the advanced gastric adenocarcinoma patients who underwent radical resection of gastric cancer at the Department of Digestive Diseases,Xijing Hospital of Digestive Diseases,Air Force Military Medical University,and received adjuvant chemotherapy in the Department of Gastroenterology within 3-5 weeks after the operation.Patients were excluded from neoadjuvant chemotherapy,positive resection margins,lack of clinicopathological information,loss to follow-up,and other gastric malignancies(neuroendocrine tumors,squamous cell carcinoma,etc.),and 668 cases were finally included.Then,collect the patient’s general information,clinicopathological characteristics,and preoperative and postoperative laboratory indicators(within 1 week before the first chemotherapy).2.Through univariate and cox multivariate analysis to evaluate the relationship between preoperative and postoperative nutrition-related indicators: Body mass index(BMI),hemoglobin,albumin,inflammation-related indicators: lymphocyte count(LC),Lymphocyte-monocyte ratio(LMR),neutrophil-lymphocyte ratio(NLR),platelet-lymphocyte ratio(PLR)and Onodera’s prognostic nutritional index(O-PNI)and the prognosis of gastric cancer.The calculation method of O-PNI is as follow:O-PNI = albumin value(g/l)+ 0.005× lymphocyte count(/μL).3.According to the results of COX multivariate regression model,the prognostic models were established based on inflammation-nutrition-related indicators preoperatively and postoperatively,and the internal calibration curves were drawn with 3 years and 5years survival as the outcome events to evaluate the prediction fit of the two models respectively.The pros and cons of the two models were compared based on the C-index of each model.4.Statistical methods: SPSS 24.0 software was used for general statistical description and statistical analysis,and rate and composition ratio were used for statistical description of data.Chi-square test or Fisher exact probability method was used for statistical analysis;survival curve drawing and comparison used Kaplan-Meier method.The Cox regression model was used to predict single-factor and multi-factor hazard ratios.X-tile software was used to calculate the best cut-off value of inflammation and nutrition related markers;R software was used to calculate the consistency index(C-index)and draw the nomogram.Result1.Among the 668 patients,523 were males and 145 were females,with an average age of 54.9 years.Histopathological differentiation: There were 504 cases(75.4%)of histopathologically poorly differentiated or mucinous carcinoma,and 164 cases(24.6%)were histopathological middle-highly differentiated.Surgical methods: There were 45cases(6.7%)with proximal gastrectomy,258 cases(38.6%)with distal gastrectomy,and365 cases(54.6%)with total gastrectomy.Among them,12 patients(1.8%)were in stage Ⅰ,171 patients(25.6%)were in stage II and 485(72.6%)were in stage III,and 277 patients died.The 1-year,3-year,and 5-year survival rates were 88.3%,65.2%,and 56.8%,respectively.And the median follow-up was 4.87 years.2.Univariate analysis showed that preoperative nutrition and inflammation related indicators: anemia(P = 0.043),albumin(P = 0.001),lymphocyte count(P < 0.001),LMR(P = 0.004),NLR(P <0.001),PLR(P = 0.022),BMI(P = 0.001)and O-PNI(P <0.001)are related to prognosis;postoperative nutritional and inflammation related indicators:anemia(P = 0.004),albumin(P <0.001),Lymphocyte count(P <0.001),LMR(P = 0.012),NLR(P = 0.003),PLR(P = 0.005),BMI(P = 0.003),and O-PNI(P <0.001)are all related to prognosis.3.Multivariate analysis showed that in addition to T stage and N stage,preoperative high BMI(HR = 0.680,95%CI: 0.521～0.887,P = 0.004)and high O-PNI(HR = 0.683,95%CI: 0.527～0.886,P = 0.004)were the independent prognostic factors for overall survival;postoperative high BMI(HR = 0.756,95%CI: 0.583～0.981,P = 0.035),high O-PNI(HR = 0.577,95%CI: 0.438～0.761,P <0.001),high lymphocyte count(HR =0.685,95%CI: 0.502～0.936,P <0.017)were independent prognostic factors for overall survival.4.The preoperative indexes BMI and O-PNI and the postoperative indexes BMI,lymphocyte count and O-PNI index combined with p T and p N stage to drawn nomograms to predict the 3-years and 5-years survival rate after radical of gastric cancer.It can be seen from the drawn calibration curve that the predicted survival period of the model and the actual survival period have a better fit,and can more accurately predict the 3-year and5-year survival rate of patients.By calculating the C-index value,it can be seen that the preoperative prognostic model’s C index is 0.6877,and the postoperative model is 0.7028.The prognostic predictive power of the postoperative model is slightly better than that of the preoperative model.ConclusionsBoth preoperative and postoperative BMI and O-PNI are independent prognostic factors for patients with gastric cancer.And patients with low BMI or low O-PNI have a poor prognosis.The nomogram model established based on postoperative BMI,O-PNI,lymphocyte count,T stage and N stage has better prognostic efficacy than the preoperative model,which can partially compensate for the lack of TNM stage and more accurately evaluate prognosis and guide individualized treatment.