Radiomics Of Multiparametric MRI For Pretreatment Prediction Of Progression-Free Survival In Advanced Nasopharyngeal Carcinoma (Stage:?-?b)

[Objective]To identify MRI-based radiomics that predict progression-free survival(PFS)in patients with advanced nasopharyngeal carcinoma(NPC)and investigate associations between radiomics features and clinical data using heat maps.[Materials and Methods]PatientsOur Institutional Review Board approved this retrospective study and waived the need to obtain informed consent from the patients.A total of 118 consecutive patients with advanced NPC met the criteria were identified and divided into two cohorts at a ratio of 3:1 using computer-generated random numbers.Eighty-eight patients were allocated to the training cohort,while 30 patients were allocated to the independent validation cohort.Demographic and pretreatment clinical characteristics were collected from PACS,including age,gender,histology,T-stage,N-stage,overall stage,hemoglobin,and platelet counts.We achieved axial contrast-enhanced T1-weighted Digital Imaging and Communications in Medicine(DICOM)images and T2-weighted DICOM images in the institutional Picture Archiving and Communication System(PACS,Carestream,Canada).We used ITK-SNAP software for three-dimensional manual segmentation(open source software;www.itk-snap.org).The region of interest covered the whole tumor and was delineated on both the axial T2-weighted images and contrast-enhanced T1-weighted images on each slice.All feature extraction methods were implemented using MatLab 2014a(MathWorks,Natick,MA,USA).Radiomics features include first-order statistics features,shape-and size-based features,statistics-based textural features,and wavelet features.A total of 970 features were extracted for each patient.We used least absolute shrinkage and selection operator(LASSO)method to select features that were most significant and then built a cox model including selected variates.The Rad-score was calculated for each patient as a linear combination of selected features that were weighted by their respective coefficients.The patients were divided into low risk and high risk groups according to median Radscore.Stratified Kaplan-Meier survival analyses were performed to estimate progression-free survival(PFS)in various subgroups,comparing high-risk patients and low-risk patients.The relative hazard ratio(HR)of clinical data and radiomics signature as risk factors of PFS was calculated on the basis of the G-rho rank test in a Cox proportional hazards model.We built clinical nomogram and radiomics nomogram to predict PFS.The predictive performance of the clinical nomogram and radiomics nomogram was evaluated in the training cohort and then tested in the validation cohort,C-index and calibration curve were obtained from multivariable Cox proportional hazard regression analyses.We performed a heatmap analysis to evaluate associations between radiomics features and clinical data.[Results]The radiomics signatures were significantly associated with PFS.In the training cohort,the radiomics signature derived from CET1-w images yielded a C-index of 0.690(95%confidence interval[CI]:0.593 to 0.787).The radiomics signature from T2-w images yielded a C-index of 0.648(95%CI:0.551 to 0.745).The radiomics signature from joint CET1-w and T2-w images yielded the highest C-index,which was 0.758(95%CI:0.661 to 0.856).One radiomics nomogram combined a radiomics signature from joint CET1-w and T2-w images with the TNM staging system.This nomogram showed a significant improvement over the TNM staging system in terms of evaluating PFS in the training cohort(C-index,0.761 vs 0.514;p<2.68 × 10-9).Another radiomics nomogram integrated the radiomics signature with all clinical data and thereby outperformed a nomogram based on clinical data alone(C-index,0.776 vs 0.649;p<1.60 × 10-7).The calibration curves showed good agreement between nomogram-evaluated and actual survival.Findings were then confirmed in the validation cohort.Kaplan-Meier survival analysis showed that PFS was significantly lower in low risk patients than high risk patients(p<0.05).A Cox regression analysis identified radiomics signature as an independent risk factors in the training cohort(HR:4.13,95%CI:2.45 to 6.95,p = 3.73 × 10-8)and the validation cohort(HR:2.82,95%CI:1.35 to 5.88,p = 0.004).Unsupervised clustering revealed clusters of NPC patients with similar radiomics expression patterns.Heat maps revealed associations between radiomics features and tumor stages.[Conclusions]In summary,the present study developed and validated multiparametric MRI-based radiomics as a convenient approach to predicting PFS pre-treatment in patients with advanced NPC(stage III-IVb).The radiomics signature that we presented added value to both the TNM staging system and clinical data as method of providing individualized predictions of PFS.Prognostic models based on quantitative radiomics could potentially be useful for precision medicine and affect the treatment strategies that are used for patients with NPC.However,the validation of radiomics may largely dependent on future positioning of the radiomics approach relative to other multiparameter prognostic and predictive assays,such as genomics and proteomics.