| Background and Objective:Gastric cancer is one of the main diseases with high incidence and mortality in the world.Many imaging techniques are widely used in the diagnosis and preoperative staging of gastric cancer.However,each has its own advantages and disadvantages,and the overall effect is not satisfactory.As a diagnostic tool with high-tech content emerging in recent years,the application of PET/MR in the stomach is still in the groping stage.The purpose of this study was to explore the feasibility of PET/MR in the diagnosis and staging of gastric tumors,and to optimize the scanning parameters and sequences;the diagnostic value of PET/MR in preoperative T and N stages of gastric cancer and compared with PET/CT and MDCT;The correlation between postoperative pathological TNM stage,molecular biological indexes and 18F-PET/MR parameters,serological parameters and clinical features of gastric cancer was studied.The indexes of predicting tumor characteristics and pathological TNM staging were optimized.Methods:In the first time,30 patients with gastric cancer were diagnosed by gastroscope examination.The PET/MR,PET/CT examination were performed with different MR sequences,and the scanning parameters were optimized.The image artifacts,the fusion accuracy and the score of the lesion definition were evaluated for the MR different sequences and the PET/MR-T1WI、PET/MR-T2WI、PET/CT.To compare the image quality of them,the feasibility of the diagnosis of tumor by PET/MR in the stomach was discussed.Then,19 of the 30 gastric cancer patients were examined by MDCT.According to the results of PET/MR,PET/CT,MDCT,the pathological results of 27 patients undergoing radical gastrectomy were compared with the preoperative staging of gastric cancer by prospective T and N staging,respectively.The diagnostic efficacy of PET/MR in T and N stages of gastric cancer was analyzed and compared with that of PET/CT,MDCT.Finally,the imaging and clinical data of patients undergoing gastric cancer surgery were collected,and the factors that might be related to the pathological stage and molecular biological index(HER2,Ki67)after primary tumor operation,including clinical characteristics,serological parameters and imaging parameters were analyzed by univariate analysis and multi-factor analysis.Results:① The score of PET/MR-T2WI-HASTE sequence image in artifact and definition of primary focus was obviously better than that of PET/MR-T2WI-BLADE sequence image.The difference of PET/MR-T2WI、PET/MR-T1WI and PET/CT images in artifact was statistically significant(P<0.05),there are more artifacts in PET/MR-T2WI images;There is no statistical difference in the accuracy of image fusion(P>0.05).There was a significant difference in the clarity of the positive primary foci(P<0.05),PET/MR-T2WI image optimization.②Compared with postoperative pathological T staging,PET/MR and MDCT had better consistency of T staging(kappa=0.575,0.422)and poor consistency of PET/CT staging(kappa=0.283).The accuracy of T staging in PET/MR and PET/CT,MDCT was 70.4%,48.1%and 57.9%,respectively.The difference was statistically significant(Mc Nemar-test,P<0.05).The sensitivity,specificity and accuracy of each T stage are PET/MR,MDCT,PET/CT according to order of pros and cons.In addition,the accuracy of T3 and T4 phases was lower than that of other T stages in all three modal examinations(P<0.01).The sensitivity of T3 stage is low,the missed diagnosis rate is high,the T4 stage specificity is low,and the misdiagnosis rate is high.③The sensitivity,specificity and accuracy of PET/MR and PET/CT,MDCT were 88.9%,77.8%,85.2%and 72.2%,77.8%,74.1%and 66.7%,57.1%,63.2%,respectively.The sensitivity,specificity and accuracy of each N stage were N1:20%,95.5%,81.5%and 20%,81.8%,70.4%and 50%,66.7%,63.2%,N2:28.6%,80%,66.7%and 42.9%,65%,59.3%and 0%,80%,63.2%,N3:66.7%,71.4%,70.4%and 0%,100%,77.8%and 25%,100%,84.2%.The pick-out rates of PET/MR and PET/CT,MDCT in metastatic lymph nodes were 30.8%,12.3%and 15.1%,respectively.The difference was statistically significant(P<0.01).The sensitivity of PET/MR to N3 phase was higher than that of other modes,and the difference was statistically significant(P<0.05).④The greater the tumor thickness and the lower ADC value,the worse the TNM stage of gastric cancer patients.The preoperative PET/MR-TNM staging was in good agreement with the postoperative pathological TNM staging,and the preoperative PET/CT-TNM staging was less consistent with the postoperative pathological staging.There was no significant difference in sex,age,serological indexes,tumor location,differentiation degree and SUVmax、TLG in different TNM stages(P>0.05).Preoperative PET/MR staging is an independent factor associated with TNM stage III in patients with gastric cancer.There was a positive correlation between the thickness of primary tumor and Ki67,and a negative correlation between ADC and Ki67.There was no correlation between various pre-operative parameters and HER2.Conclusion:① The application of PET/MR in the diagnosis and staging of gastric cancer is feasible,and T2WI-HASTE is the key sequence of T stage.②PET/MR is superior to MDCT and PET/CT in the diagnosis of pre-operative T-staging of gastric cancer,especially in T1 stage.In addition,the accuracy of T3 and T4 phases was lower than that of other T stages in all three modal examinations.In clinical work,we should pay attention to the high rate of missed diagnosis in stage T3 and the rate of misdiagnosis in stage T4.③The diagnostic value of PET/MR in preoperative N-stage of gastric cancer is superior to that of MDCT and PET/CT.It can detect more metastatic lymph nodes,which is of great significance to the treatment plan.④The greater the thickness of the primary tumor and the lower ADC value,then the worse the TNM stage.The preoperative PET/MR-TNM staging is in good agreement with the postoperative TNM staging,and it has independent predictive value for the pathological stage III of the tumor.The value of ADC and the thickness of primary tumor can predict the proliferative activity of tumor cells by the correlation degree of ki67,which can be used as a predictor of prognosis in patients with gastric cancer. |