The value of PSMA-PET radiomics as a prognostic biomarker for prognosis prediction in prostate cancer has only been recently explored in oligometastatic prostate cancer patients. However, the investigation of PSMA-PET-based prognostic radiomics biomarkers for primary prostate cancer patients remains lacking for this new imaging technique. In our study, we aimed to fill this gap by accruing a cohort of 134 patients with prostate adenocarcinoma (PCa) (primary prostate tumors present in all cases, with 28 patients having pelvic nodal involvement), treated with androgen deprivation therapy and external radiotherapy. Our purpose is to predict metastasis-free survival (MFS) for the patients using clinical factors and radiomics extracted from primary tumor and node volumes in pre-treatment PSMA PET/CT scans.
Our cohort includes 134 PCa patients (nodal involvement in 28 patients). Gross tumor volumes of primary tumor (GTVp) and nodes (GTVn) on CT and PET scans were segmented. A 5mm expansion ring area outside primary tumor was defined. Z-score normalization was applied to radiomics features extracted from tumor and ring; dimension reduction was performed using Principal Components Analysis (PCA). For patients with only primary tumor, we took 3 principal components (PCs) from GTVp and one ring PC as representative radiomics components from CT and PET scans. For patients with nodes, we calculated
weighted average (by volume) of radiomics from primary tumor and nodes, computed the first 3 PCs, and combined it with the 1st PC from the ring. Radiomics PCs and clinical variables (age, Gleason score, initial prostate specific antigen value (i PSA), PSA_relapse) formed the predictors. Due to MFS data imbalance (metastasis-24, no metastasis-110), we performed a 70:30 train-test split and applied imbalance correction to the training data. Univariate Cox regression was used to select top predictors (logistic regression p < 0.05). Multivariate Cox regression was performed on imbalance-corrected training data and fit on testing data (using predictors selected from training). Model 2 was built using clinical variables and radiomic PCs from primary tumors (GTVp, ring). Model 3 was built using clinical variables only. Binary classification analysis for the prediction of five-year MFS was also performed.
Results of time-to-event analysis (MFS) were: Cox-regression c-scores: model1: train- 0.77 [0.72, 0.78]; test- 0.69 [0.64, 0.70]; model2: train- 0.72 [0.66, 0.73]; test- 0.63 [0.58, 0.64]; model3: train- 0.62 [0.57, 0.63]; test- 0.54 [0.51, 0.56]. The results of 5 year MFS classification analysis were [sensitivity, specificity, AUC]: model 1: train- [83.6%, 91.3%, 0.88]; test [76.3%, 82.5%, 0.81]; model 2: train- [77.4%, 85.1%, 0.84]; test- [71.5%, 78.2%, 0.76]; model 3: train- [69.3%, 78.2%, 0.76]; test- [64.7%, 72.6%, 0.68].
The two cohorts of patients classified by model 1 showed statistically significant differences in their actual survival curves, demonstrating the efficacy of the classification. Integration of node with primary tumor-radiomics provides the best prognostic performance in MFS prediction.
This is one of the first studies to explore the prognostic value of pre-treatment PSMA-PET, a relatively recent advancement in the care of prostate adenocarcinoma patients. It addressed an important clinical gap by investigating the prognostic utility of radiomic biomarkers derived from pre-treatment PSMA-PET/CT scans, combined with clinical parameters, to predict metastasis-free survival in patients with primary prostate adenocarcinoma. Results from the study demonstrate the potential of using imaging biomarkers from PSMA-PET/CT images for prognosis prediction before the treatment, which provides clinicians with valuable information for customizing the treatment paradigm to improve the outcomes for primary prostate cancer patients. These findings underscore the potential of PSMA-PET/CT-based radiomics as a non-invasive, image-derived biomarker to support personalized treatment strategies and improve clinical decision-making in primary prostate cancer management.
Figure 1. The workflow involves segmentation of regions of interest, radiomics feature extraction, combining radiomics and clinical information into a prognostic signature, and performing time-to-survival event and binary classification analysis of metastasis-free survival.
Figure 2. Representative figures showing the regions of interest, including (a) the primary prostate tumor volume (GTVp) and expansion ring region and (b) the node regions (GTVn1-n4) on the pre-registered PSMA-PET/CT scan slices of a patient.
Figure 3. KM survival curves (b1-b3) stratified by five-year metastasis-free survival classification results (MFS event vs no event) for models 1, 2, and 3.
Written by: Apurva Singh, PhD,1 William Silva Mendes, MD,1 Sang-Bo Oh, MD,1,4 Ozan Cem Guler, MD,2 Aysenur Elmali, MD,3 Birhan Demirhan, MD,2 Amit Sawant, PhD,1 Phuoc Tran, MD, PhD,5 Cem Onal, MD,2,3* Lei Ren, PhD1*
- Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, USA
- Baskent University Faculty of Medicine, Adana, Dr. Turgut Noyan Research and Treatment Center, Department of Radiation Oncology, Adana, Turkey
- Baskent University Faculty of Medicine, Department of Radiation Oncology, Ankara, Turkey
- Division of Medical Oncology and Hematology, Department of Internal Medicine, Pusan National University Yangsan Hospital, Pusan National University School of Medicine, Yangsan, Republic of Korea
- Professor and Chair of Genitourinary Radiation Oncology at MD Anderson Cancer Center, Houston, TX, USA