Correlation Study Between ¹⁸F-FDG PET/CT Metabolic Parameters And Clinicopathological Characteristics Of Lung Cancer

Objective Through a large sample analysis,explore the relationship between the PET metabolic parameters(SUVmax,SUVmean,SUVpeak,MTV,TLG)and the clinicopathological characteristics of lung cancer patients(age,gender,smoking,long diameter of tumor,location,CT value,pathological type,stage,and anatomical classification),explore the predictive value of PET metabolic parameters combined with tumor markers for lung cancer pathological types.Methods A retrospective analysis of 250 lung cancer cases underwent PET/CT examination at the Cancer Hospital of Tianjin Medical University from September2018 to January 2020.The clinicopathological characteristics and PET/CT image data of these patients were collected.The PET metabolic parameters of the lesions were measured by PET VCAR image post-processing software.Mann-Whitney U test and Kruskal-Wallis H test were used to analyze the correlation between PET metabolic parameters and clinicopathological characteristics of lung cancer patients.The area under the ROC curve was used to obtain the diagnostic efficiency of PET metabolic parameters and tumor markers.The binary logistic regression model was used to calculate the combined index calculation formula of multiple indexes,and the diagnostic efficacy of each single index and the combined index was compared.We took non-small cell lung cancer and small cell lung cancer as dependent variables,five PET metabolic parameters as independent variables,and performed binary logistic regression analysis to obtain the PET metabolic parameter joint group model.The same method takes 5 tumor markers,5 PET metabolic parameters and 5 tumor markers as independent variables,and constructs a tumor marker joint group model and a PET metabolic parameter + tumor marker joint group model,using separate indicators and joint The indicators predict NSCLC and SCLC respectively.And use these three models to predict squamous cell carcinoma and adenocarcinoma in NSCLC.By comparing the diagnostic efficacy of these indicators and the model,it is analyzed whether the combined indicator model can better predict the pathological type of lung cancer relative to a single indicator.Results The correlation between PET metabolic parameters and clinicopathological characteristics of lung cancer patients showed that the differences in 5 PET metabolic parameters were statistically significant in different genders,lesion length,anatomical classification,pathological type,and clinical stage grouping(P<0.05).In addition,the PET metabolic parameters in male,long diameter ≥ 3cm,central type,and squamous cell carcinoma groups were higher.With the gradual increase of clinical stage,PET metabolic parameters also gradually increased.In the smoking group,the differences of SUVmax,SUVmean,SUVmax,and TLG were statistically significant(P<0.05),and these metabolic parameters were higher in the smoking group.In the different CT value groups,the differences of MTV,TLG,and SUVpeak were statistically significant(P<0.05),and these metabolic parameters were higher in the CT value ≥30Hu group.There was no statistically significant difference in the five PET metabolic parameters among different ages and lesion location groups(P>0.05).Predictive studies on the pathological types of lung cancer show that the tumor markers NSE,Pro GRP and the three combined group models have good diagnostic efficacy for NSCLC and SCLC(AUC>0.75).Moreover,the combined diagnostic model of PET metabolic parameters + tumor markers has the best diagnostic efficiency,while the five metabolic parameters of PET and SCC,CEA,and Cyfra21-1in tumor markers have poor diagnostic efficacy.The five metabolic parameters of PET,SCC,Cyfra21-1 and three combined group models in tumor markers are all effective in diagnosing adenocarcinoma and squamous cell carcinoma(AUC>0.75).Moreover,the combined diagnosis model of PET metabolic parameters and tumor markers has the best diagnostic efficiency,while Pro GRP,NSE,and CEA in tumor markers have poor diagnostic efficacy for adenocarcinoma and squamous cell carcinoma.Conclusions In different genders,lesion length,anatomical classification,pathological type,and clinical stage grouping,the differences in the five PET metabolic parameters were statistically significant(P<0.05).The difference of SUVmax,SUVmean,SUVpeak and TLG in the smoking group was statistically significant(P<0.05).In the CT value ≥30Hu and <30Hu group,the difference of MTV,TLG,SUVpeak was statistically significant(P<0.05).There was no statistically significant difference in the five PET metabolic parameters between different ages and different lesion locations(P>0.05).The use of PET metabolic parameters and serum tumor markers alone has limited value in predicting the histological type of lung cancer.When the two are combined,the diagnostic efficacy can be significantly improved.