MRI-Derived Radiomics to Guide Post-operative Management for High-Risk Prostate Cancer

Abstract

Purpose: Prostatectomy is one of the main therapeutic options for prostate cancer (PCa). Studies proved the benefit of adjuvant radiotherapy (aRT) on clinical outcomes, with more toxicities when compared to salvage radiotherapy. A better assessment of the likelihood of biochemical recurrence (BCR) would rationalize performing aRT. Our goal was to assess the prognostic value of MRI-derived radiomics on BCR for PCa with high recurrence risk. Methods: We retrospectively selected patients with a high recurrence risk (T3a/b or T4 and/or R1 and/or Gleason score>7) and excluded patients with a post-operative PSA > 0.04 ng/mL or a lymph-node involvement. We extracted IBSI-compliant radiomic features (shape and first order intensity metrics, as well as second and third order textural features) from tumors delineated in T2 and ADC sequences. After random division (training and testing sets) and machine learning based feature reduction, a univariate and multivariate Cox regression analysis was performed to identify independent factors. The correlation with BCR was assessed using AUC and prediction of biochemical relapse free survival (bRFS) with a Kaplan-Meier analysis. Results: One hundred seven patients were included. With a median follow-up of 52.0 months, 17 experienced BCR. In the training set, no clinical feature was correlated with BCR. One feature from ADC (SZE GLSZM ) outperformed with an AUC of 0.79 and a HR 17.9 ( p = 0.0001). Lower values of SZE GLSZM are associated with more heterogeneous tumors. In the testing set, this feature remained predictive of BCR and bRFS (AUC 0.76, p = 0.0236). Conclusion: One radiomic feature was predictive of BCR and bRFS after prostatectomy helping to guide post-operative management.

External Validation of an MRI-Derived Radiomics Model to Predict Biochemical Recurrence after Surgery for High-Risk Prostate Cancer

“…First, we used a feature selection method based on stability, robustness, and intercorrelation checks [19] in order to only evaluate a reduced subset of features in the model training and validation. After training using two-thirds of the cohort and validating using the rest, and additionally combining the radiomics (kept after feature set reduction) and clinical variables, three predictive models were built, i.e., a radiomics model based on a single textural feature (ADC SZE GLSZM ), a clinical model based on preoperative PSA and age at surgery, and a combined radiomics and clinical model [14].…”

“…The patients from Institution 1 were already known to us. As mentioned above, our previous article dealt with the development of the predictive models [14]. For the current analysis, we carried out retraining of the predictive models using updated follow-up information and validation using the Institution 2 cohort.…”

“…Radiomics features were extracted using homemade radiomics code implemented in MatLab ® , following the most up-to-date guidelines and benchmark values of the Image Biomarker Standardisation Initiative (IBSI) [18]. Only the previously identified feature (ADC SZE GLSZM ) [14] was considered in the present study, as explained in the statistical analysis section below.…”

“…ADC SZE GLSZM (AUC 0.82) with a cut-off value of 0.53 (the cut-off value determined in the initial development study was 0.528) [14] was the best model, with a BAcc of 79% regarding BCR predictions (Se 72%, Sp 85%, p < 0.0001) and a corresponding high stratification power according to bRFS with an HR of 8.7 (p < 0.0001, Figure 4a).…”

“…We previously reported on the development of a predictive model based on a single radiomics feature, size-zone emphasis (SZE GLSZM ), extracted from ADC maps obtained from 107 pre-therapeutic prostatic diffusion weighted imaging (DWI) MRIs [14]. Grey-Level Size Zone Matrix (GLSZM) evaluates the size of zones of the same grey-level voxels.…”