Application Of Gemstone Spectral Ct And Intravoxel Incoherent Motion MRI In Pathological Grading Of Esophageal Squamous Cell Carcinoma

Background and purposeEsophageal carcinoma is a common malignant tumor of the digestive system,which is mainly squamous cell carcinoma,accounting for about 90%.In China,the incidence and mortality of esophageal carcinoma are fifth and fourth,respectively.The prognosis of esophageal carcinoma is poor,and studies have shown that pathological grades of tumors have a significant impact on patient’s prognosis.Therefore,accurate pathological grading of esophageal carcinoma before operation has important clinical significance for its prognosis evaluation.Routine CT examination,MRI routine scan and other imaging techniques are commonly used for the diagnosis,location and preoperative staging of esophageal carcinoma,but lack of functional imaging methods,which has some limitations in differentiating the pathological grading of esophageal carcinoma.Gemstone spectral CT and intravoxel incoherent motion diffusion weighted imaging can get the parameters reflecting the internal characteristics of tumor and make quantitative analysis,which provides valuable information for pathological grading.Therefore,the purpose of this study is to explore the value of gemstone spectral CT and intravoxel incoherent motion of MRI in the pathological grading of esophageal squamous cell carcinoma.Part1 Clinical value of gemstone spectral CT imaging in preoperative evaluation of pathological grading of esophageal squamous cell carcinomaObjectiveTo investigate the value of gemstone spectral computed tomography quantitative parameters in the assessment of pathological grade of esophageal squamous cell carcinoma before operation.Materials and methodsThe imaging findings of 79 patients confirmed esophageal squamous cell carcinoma by surgery and pathology were prospectively analyzed.All the patients were divided into three groups based on their pathological finding: well differentiated group(n=23),moderately differentiated group(n=24),poorly differentiated group(n=32).They all underwent chest plain scan and double phase enhanced scan ofgemstone spectral computed tomography.the enhancement attenuation(HU),the average of the slope of the spectral Hounsfield Unit curve(λHU),normalized iodine concentration(NIC),normalized effective atomic number(Zeff-a)were measured and calculated in arterial phase and venous phase.The difference of HU,λHU,NIC,Zeff-a among different grades were statistically analyzed.The receiver operating characteristic curve was used to evaluate the diagnostic efficiency of single and combined parameters in the differentiation of poorly and well-moderately differentiated esophageal squamous cell carcinoma.ResultsThe differences of HU,λHU,NIC,Zeff-a in different pathological grading of the esophageal squamous cell carcinoma in arterial phase and venous phase were statistically significant(F/H=10.675,8.713,27.331,22.388 and 6.467,14.616,12.885,7.708,respectively,all P<0.05).The best single parameter to differentiate poorly differentiated from well-moderately differentiated esophageal squamous cell carcinoma was NIC in arterial phase with AUC(Areas under the ROC curves),the cutoff value,sensitivity,specificity,accuracy of 0.840,0.130,96.9%,57.4%,73.4%;The best combination of parameters was(NIC+λHU+Zeff-a+HU)in arterial phase with AUC,the threshold of predicted probability,sensitivity,specificity,accuracy of 0.927,0.401,87.5%,80.9%,83.5%.ConclusionGemstone spectral imaging quantitative parameters can be used to evaluate esophageal squamous cell carcinoma pathological grading,the NIC and(NIC+λHU+Zeff-a+HU)in arterial phase have the highest differential diagnostic efficiency.Part2 The application value of intravoxel incoherent motion diffusion-weighted MRI in the pathological grading of esophageal squamous cell carcinomaObjectiveTo investigate the value of the parameters of mono-exponential and intravoxel incoherent motion double-exponential models in the pathological grading of esophageal squamous cell carcinoma.Materials and methods70 patients with histopathologically confirmed esophageal squamous cell carcinoma were prospectively analyzed.All the patients were divided into three groups based on their pathological finding:well differentiated group(n=22),moderately differentiated group(n=23),poorly differentiated group(n=25).They all underwent preoperative MR examination.The scanning sequences included conventional magnetic resonance imaging,intravoxel incoherent motion diffusion-weighted imaging with 12 b values.The Apparent Diffusion Coefficient(ADC),slow Apparent Diffusion Coefficient(slow ADC,ADCslow),fast Apparent Diffusion Coefficient(fast ADC,ADCfast),f(fraction of fast apparent diffusion coefficient)were generated and measured in post-processing software by two physicians,independently.Intra-class correlation coefficient(ICC),Bland-Altman analysis are used to evaluated the reliability and consistency of the parameters measured by two physicians.The difference of ADC,ADCslow,ADCfast,f among different grades were statistically analyzed by one-way analysis of variance,LSD-t test.Spearman correlation analysis was used to evaluate the correlation between the parameters and pathological grades.The receiver operating characteristic curve was used to evaluate the diagnostic efficiency of the parameters in the differentiation of poorly and well-moderately differentiated esophageal squamous cell carcinoma.ResultsThe intra-class correlation coefficient(ICC)value between the two physicians were 0.996 for ADC,0.999 for ADCslow,0.891 for ADCfast and 0.944 for f.The differences of ADC,ADCslow in different pathological grading of the esophageal squamous cell carcinoma were statistically significant(F= 19.177,67.951;all P < 0.01).The differences of ADCfast,f in different pathological grading of the esophageal squamous cell carcinoma have no statistically significant(F=2.613,0.948;all P > 0.05).A significantly correlation was obtained between ADC,ADCslow with histological grade: r=-0.603(P < 0.05),r =-0.812(P< 0.05),respectively.There is no correlation between ADCfast,f and pathological grade: r=-0.077,r= 0.040,respectively,all P > 0.05.The best parameter to differentiate poorly differentiated from well-moderately differentiated esophageal squamous cell carcinoma was ADCslow with AUC(Areas under the ROC curves),the cutoff value,sensitivity,specificity,accuracy of 0.909,0.00081,88.9%,80%,84.3%.ConclusionThe parameters of mono-exponential and intravoxel incoherent motion double-exponential models canbe used for clinical evaluation in pathological grading of esophageal squamous cell carcinoma,ADC and IVIM-derived ADCslow showed the better diagnostic performance in differentiating esophageal squamous cell carcinoma pathological grading.