| Part Ⅰ:Analysis of therapeutic effect of neoadjuvant chemotherapy and neoadjuvant radiotherapy for ⅠB3-ⅡA2 stage cervical cancerObjective:To compare the efficacy of neoadjuvant chemotherapy and neoadjuvant radiotherapy in preoperative treatment of IB3-IIA2 cervical cancer and the incidence of postoperative recurrence risk factors.Analyzing the related factors affecting the curative effect.To provide basis for the selection of neoadjuvant therapy for ⅠB3-ⅡA2 cervical cancer.Methods:The clinical and pathological data of patients with cervical cancer who met the inclusion and exclusion criteria in our hospital from January 2016 to December 2021 were retrospectively analyzed.Data related to age,FIGO stage,pathological type,maximum diameter of newly diagnosed tumor,neoadjuvant therapy,degree of pathological remission and risk factors for postoperative recurrence were analyzed.SPSS 26.0 was used for statistical analysis,t-test was used for inter-group comparison of measurement data,chi-square test was used for inter-group comparison of counting data,and P<0.05 was considered as statistical difference.Results:A total of 120 patients with cervical cancer were enrolled,including 68 patients in the neoadjuvant chemotherapy group and 52 patients in the neoadjuvant radiotherapy group.There was no statistically significant difference in the general clinical data between the two groups(P>0.05).The probability of ideal pathological response in neoadjuvant chemotherapy group and neoadjuvant radiotherapy group was 60.3%and 46.2%,with no statistical significance(P>0.05).There was statistically significant difference in the score of pathological withdrawal degree between the two groups(P=0.001).There was no significant difference in the incidence of postoperative recurrence risk factors between the two groups(P>0.05).There was no significant difference in postoperative adjuvant therapy rate between the two groups under FIGO standard(P>0.05).There was no significant difference in the efficacy of cervical squamous carcinoma,adenocarcinoma and adenosquamous carcinoma(P>0.05).Conclusions:1.There is no significant difference in response sensitivity between neoadjuvant chemotherapy and neoadjuvant radiotherapy for stage IB3-IIA2 cervical cancer.2.Compared with the neoadjuvant chemotherapy group,the neoadjuvant radiotherapy group had a better degree of pathological regression,indicating that it had a better effect in weakening the activity of tumor cells,making tumor cells degenerate and necrotic,and thus fading tumor cells.3.There was no significant difference in the incidence of postoperative recurrence risk factors and the postoperative adjuvant therapy rate under FIGO standard between the two treatments.4.There was no significant difference in the therapeutic response among different pathological types of squamous cell carcinoma,adenocarcinoma and adenosquamous cell carcinoma.Part Ⅱ:Prediction of neoadjuvant radiotherapy response for ⅠB3-ⅡA2 cervical cancer based on CT radiomic modelObjective:To evaluate the effectiveness of CT radiomics model in predicting neoadjuvant radiotherapy response for ⅠB3-ⅡA2 cervical cancer.Methods:The pre-treatment enhanced CT images and clinical data of patients with cervical cancer who met the inclusion and exclusion criteria in our hospital were retrospectively collected.Monaco system was used for manual ROI delineation,and 3D Slicer(version 4.11)software was used for feature extraction.Mutual information method and LASSO method and random forest recursive feature elimination method were used for feature screening.Sample size was increased by SMOTE method.All cases were assigned corresponding labels according to efficacy evaluation criteria and randomly divided into training sets and test sets in a ratio of 7:3.Logistic regression algorithm was used to conduct training and establish models.Clinical data such as patient age,FIGO stage,tumor size and tumor pathological type were collected to establish the clinical model.ROC curve was used to evaluate the diagnostic performance of different models,and the AUC,accuracy,sensitivity and specificity of the ROC curve were calculated.The Delong test was used to evaluate the performance differences between the models.Results:A total of 46 patients with cervical cancer were included in this study,20 patients were sensitive to neoadjuvant radiotherapy and 26 patients were insensitive.Four features were extracted respectively from the Intratumoral and extratumoral areas of 5mm,and the logistic regression algorithm was used to establish the radiomics prediction model.The accuracy,sensitivity and specificity of intratumoral model were 0.671,0.833 and 0.500,respectively and the AUC values of training set and test set were 0.772 and 0.735.The accuracy,sensitivity and specificity of extratumoral areas of 5mm model were 0.728,0.791 and 0.666 and the AUC values of the training set and test set were 0.840 and 0.838.The prediction accuracy of both models is higher than that of clinical model.The two radiomics models had good predictive ability,and there was no significant difference in AUC(P>0.05).Conclusions:In this study,the two imaging radiomics models established in intratumoral and extratumoral 5mm had no significant difference in prediction ability,and both had good classification ability,which could be noninvasive distinguish patients with sensitive and insensitive efficacy.Providing a basis for clinicians to make clinical decisions.Part Ⅲ:Prediction of neoadjuvant chemotherapy response for ⅠB3-ⅡA2 cervical cancer based on MRI radiomics model.Objective:To evaluate the effectiveness of MRI radiomics model in predicting neoadjuvant chemotherapy response for ⅠB3-ⅡA2 cervical cancer.Methods:The pre-treatment MRI images and clinical data of patients with cervical cancer who met the inclusion and exclusion criteria in our hospital were retrospectively collected.Monaco system was used for manual ROI delineation,and 3D Slicer(version 4.11)software was used for feature extraction.Mutual information method and LASSO method and random forest recursive feature elimination method were used for feature screening.Sample size was increased by SMOTE method.All cases were assigned corresponding labels according to efficacy evaluation criteria and randomly divided into training sets and test sets in a ratio of 7:3.Logistic regression algorithm was used to conduct training and establish models.Clinical data such as patient age,FIGO stage,tumor size and tumor pathological type were collected to establish the clinical model.ROC curve was used to evaluate the diagnostic performance of different models,and the AUC,accuracy,sensitivity and specificity of the ROC curve were calculated.The Delong test was used to evaluate the performance differences between the models.Results:A total of 43 patients with cervical cancer were enrolled.Among them,26 cases were sensitive to neoadjuvant chemotherapy,17 cases were insensitive.Four features were extracted respectively from T1WI Intratumoral,T1WI tumor expansion 5mm,T2WI Intratumoral,T2WI tumor expansion 5mm,respectively.Single sequence and multi-sequence combined radiomics prediction model were established using logistic regression algorithm.There were 4 single sequence models,including T1WI Intratumoral、T1WI tumor expansion 5mm、T2WI Intratumoral、T2WI tumor expansion 5mm radiomics models.There were 4 multiple sequence combination models,including(T1WI intratumoral+T2WI intratumoral)、(T1WI tumor expansion 5mm+T2WI tumor expansion 5mm)、(T1WI intratumoral+T2WI tumor expansion 5mm)、(T1WI tumor expansion 5mm+T2WI intratumoral)radiomics models.The lowest AUC value of all the eight models was 0.761.In the training set and test set,there was no statistical difference in AUC values between T1WI and T2WI intratumoral models(P>0.05),and no statistical difference in AUC values between T1WI and T2WI tumor expansion 5mm models(P>0.05).In the test set,there were statistically significant differences in AUC values between T1WI intratumoral and T2WI intratumoral models and T1WI and T2WI models with tumor expansion of 5mm(P<0.05).The combination model(T1WI intratumoral+T2WI intratumoral)showed statistical differences with the other three combination models in the training set and test set(P<0.05),while the other three combination models showed no statistical differences in the pairwise comparison between the training set and test set(P>0.05).Conclusions:The eight single-sequence and multi-sequence combination models established in this study have good classification and prediction ability.In single-sequence models,T1WI and T2WI intratumoral models had better predictive performance,and there was no significant difference between the two models.The four combined models all had good predictive performance,among which(T1WI intratomoral+T2WI intratomoral)model had the highest performance and the best predictive performance among the four combined models,which could effectively and noninvasively predict the efficacy of neoadjuvant chemotherapy forⅠB3-ⅡA2 cervical cancer,providing a basis for clinical individualized treatment. |
PDF Full Text Request