Feasibility Study On Predicting Pathological Grade Of Rectal Adenocarcinoma Based On Extracellular Volume Fraction Of CT

Objective: The purpose of this study was to explore the feasibility of predicting the pathological grade of rectal adenocarcinoma based on extracellular volume fraction(extracellular volume,ECV)of enhanced CT.Methods: From October 2021 to October 2022,enhanced CT scanning was performed in the affiliated Hospital of Inner Mongolia Medical University.25 cases of high-grade and 71 cases of low-grade rectal adenocarcinoma were confirmed by operation and pathology.According to the histological classification of colorectal adenocarcinoma(non-special type)(according to the 2010 WHO standard),patients were divided into low-grade group and high-level group.The clinicopathological features,routine CT parameters and ECV were compared between the two groups.The clinicopathological features included sex,age,longest tumor diameter,preoperative CEA and ECV.The routine CT parameters include plain CT value of ROI,arterial phase CT value,venous phase CT value,equilibrium phase CT value,arterial phase absolute enhanced CT value(ΔS1;ΔS1 = HU arterial phase-HU plain scan phase),venous phase absolute enhanced CT value(ΔS2;ΔS2 = HU venous phase-HU plain scan phase),and absolute enhanced CT value during equilibrium period(ΔS3).ΔS3 = HU equilibrium phase-HU plain scan),the ECV of the primary focus was calculated by measuring the CT values of the areas of interest in the non-contrast scan and the balance period.The data were analyzed by SPSS statistical analysis software(version 24.0),and the clinicopathological features,routine CT parameters and ECV were compared between the two groups.Those who obey normal distribution are compared by independent sample t-test,and those who do not obey normal distribution are compared by independent sample nonparametric test.Intra-group correlation coefficient(ICC)was calculated to evaluate the consistency of ECV scores between observers.Spearman correlation analysis was used to evaluate the correlation between the ECV value and the pathological grade of rectal adenocarcinoma based on multi-phase enhanced CT.The variables with statistical differences selected by univariate analysis were analyzed by multivariate analysis.The independent predictors for predicting the pathological grade of rectal adenocarcinoma were screened by binary Logistic regression,and the joint predictors were constructed.The diagnostic efficacy of independent predictors and joint predictors was evaluated by ROC curve.When the area under the ROC curve(AUC)≥0.5,the closer the AUC is to 1,the higher the diagnostic efficiency.Bland-Altman diagram and subject operating characteristics(ROC)were analyzed by Med Calc software.De Long test was used to compare the AUC of different parameters.Results: A total of 96 subjects with an average age of 64.4 years were enrolled in this study,including 25 in the high-level group(26.04%)and 71 in the low-level group(73.96%).1.There was no significant difference in gender(0.065),age(0.547)and preoperative CEA(0.343)between the two groups.There was significant difference in the longest diameter of tumor between the two groups(P < 0.05).2.Among the routine CT parameters,there were significant differences in arterial phase(P < 0.012),equilibrium phase(P < 0.007),S 1(P <0.001)and S 3(P < 0.006).3.There was significant difference in ECV between the two groups(P < 0.05).4.Among the above indexes,equilibrium phase,S3 and longest tumor diameter were positively correlated with pathological grade,ECV was moderately positively correlated with pathological grade,and arterial phase and S1 were negatively correlated with pathological grade.5.Binary Logistic regression analysis showed that ECV and S1 were independent predictors of high-grade rectal adenocarcinoma.6.ROC curve comparison results: the AUC of ECV is 0.796,with 34.54% as the cutoff value,the sensitivity is 80%,the specificity is 77.46%;the AUC of S1 is 0.714,with 21.1HU as the cutoff value,the sensitivity is 88%,the specificity is 53.52%;the AUC of joint prediction factor Y is 0.869%,with 0.365 as the cutoff value,sensitivity is 76%,specificity is 88.73%.Conclusion: The longest diameter of tumor,CT value of arterial phase,CT value of equilibrium phase,ΔS1,Δ S3 and ECV are the influencing factors to predict the pathological grade of rectal adenocarcinoma.ΔS1 and ECV are the independent factors to predict the pathological grade of rectal adenocarcinoma.ΔS1,ECV and combined predictor Y have certain accuracy in predicting the pathological grade of rectal adenocarcinoma.