MRI-based Texture Analysis Of The Primary Tumor For The Prediction Of Bone Metastases,therapeutic Evaluation Of Androgen Deprivation Therapy In Prostate Cancer

Part Ⅰ MRI-based Texture Analysis of the Primary Tumor for Pre-treatment Prediction of Bone Metastases in Prostate CancerObjective:To identify MRI-basedtexture features for pre-treatment prediction of bone metastases(BM)in patients with prostate cancer(PCa).Methods:One-hundred and seventy-six patients with clinicopathologically confirmed PCa were enrolled,and the data was gathered from January 2008 to January 2018.A total of 976 radiomics features were extracted from T2-weighted(T2-w)and dynamic contrast-enhanced T1-weighted(DCE T1-w)MRI.Step regression,ridge regression and the least absolute shrinkage and selection operator(LASSO)regression methods model were applied to select features and develop the predicting model for BM.The performance of the radiomics features,PSA level and Gleason Score were explored with the respect to the receiver operating characteristics(ROC)curve.Multivariable logistic regression analysis starting with the following clinicopathological risk factors(PSA level,gleason score and age)and imaging biomarkers were applied to develop predicting model for BM in PCa.Results: The texture features,which consisted of 15 kindsofselected features,were significantly associated with BM(P<0.01).The combined MRI features derived from T2-w and DCE T1-w showed better performance(AUC=0.898)than features derived from single sequence(T2-w AUC=0.875、DCE T1-w AUC=0.870)and Gleason Score(AUC=0.731)for pre-treatment prediction of BM in PCa.MRI-based imaging biomarker combined with clinicopathological risk factors(free PSA,age and Gleason score)yielded the highest AUC(AUC=0.916).Multivariate regression analysis showed that the imaging biomarker was an independent risk factor for the detection of bone metastases,along with f-PSA level(free PSA)and Gleason score.Conclusion: MRI-based texture feature was significant predictor for BM in PCa.Clinicopathological risk factors combined with MRI-based texture features could further improve the prediction performance,which provide an illustrative example of precision medicine and may affect treatment strategies.Part Ⅱ MRI-based Texture Analysis for Therapeutic Evaluation in Prostate Cancer with Bone Metastases Treated with Androgen Deprivation TherapyObjective: To identify MRI-based texture features of the primary tumor before treatment as prognostic biomarker in patients with bone metastases(BM)from prostate cancer(PCa)treated with androgen deprivation therapy(ADT).Methods:225 patients with BM from PCa were enrolled,and the data was gathered from January 2008 to August 2019.A total of 976 MRI-based radiomics features were extracted from axial T2-w and DCE T1-w MRI.Step regression,ridge regression and LASSO regression method model were applied to select features and develop the predicting model for castration resistant prostate cancer(CRPC)and progression-free survival(PFS).We built predicting model combining clinicopathologic characteristics and MRI-based texture features to predict CRPC following ADT for PCa with BM,the performance of the predicting model was explored with the respect to the receiver operating characteristics(ROC)curve.We built predicting model combining clinicopathologic characteristics and MRI-based texture features to predict PFS following ADT for PCa with BM,the risk score was calculated,the patients were divided into low-risk and high-risk groups according to median risk score,stratified Kaplan-Meier survival analysis was performed to estimate PFS in low-risk group and high-risk group,the relative hazard ratio(HR)of MRI-based texture features and clinicopathologic characteristics as risk factors of PFS was calculated in Cox proportional hazards model.C-index was obtained from Cox proportional hazards regression analysis.Results:MRI-based imaging biomarker(including 8 texture features)combined with clinicopathologic risk factors(total prostate specific antigen,age and Gleason score)demonstrated good performance in predicting CRPC in the training cohort with AUC=0.903,which was then surprisingly confirmed in the validation cohort with AUC=0.885.The MRI-based radiomics signature,which consisted of 6 selected texture features,were significantly associated with PFS(P < 0.01).Kaplan-Meier survival analysis showed that PFS was significantly lower in high-risk patients than low-risk patients(P<0.05).The predicting model yielded a C-index of 0.713(95% CI:0.535-0.901).Cox proportional hazards regression analysis identified Gleason score(hazard ratio[HR]: 1.598),and T2-w 0-1 Inverse different moment(HR: 0.185)were significant independent predictors of PFS.Conclusion:Biparametric MRI-based texture analysis provided improved prognostic ability in patients with BM from PCa treated with ADT,and might help guide individualized treatment in such patients.