Abstract
Simple SummaryBiochemical recurrence after radical prostatectomy is vitally important for long-term oncological control and subsequent treatment of these patients. We applied radiomic technique to extract features from MR images of prostate cancer patients, and used deep learning algorithm to establish a predictive model for biochemical recurrence with high accuracy. The model was validated in 2 indepented cohorts with superior predictive value than traditional stratification systems. With the aid of this model, we are able to distinghuish patients with higher risk of developing biochemical recurrence at early stage, thus providing a window to initiate neoadjuvant or adjuvant therapies for prostate cancer patients.Biochemical recurrence (BCR) occurs in up to 27% of patients after radical prostatectomy (RP) and often compromises oncologic survival. To determine whether imaging signatures on clinical prostate magnetic resonance imaging (MRI) could noninvasively characterize biochemical recurrence and optimize treatment. We retrospectively enrolled 485 patients underwent RP from 2010 to 2017 in three institutions. Quantitative and interpretable features were extracted from T2 delineated tumors. Deep learning-based survival analysis was then applied to develop the deep-radiomic signature (DRS-BCR). The model’s performance was further evaluated, in comparison with conventional clinical models. The model achieved C-index of 0.802 in both primary and validating cohorts, outweighed the CAPRA-S score (0.677), NCCN model (0.586) and Gleason grade group systems (0.583). With application analysis, DRS-BCR model can significantly reduce false-positive predictions, so that nearly one-third of patients could benefit from the model by avoiding overtreatments. The deep learning-based survival analysis assisted quantitative image features from MRI performed well in prediction for BCR and has significant potential in optimizing systemic neoadjuvant or adjuvant therapies for prostate cancer patients.
Highlights
Prostate cancer (PCa) is the most common cancer among men and the second leading cause of death for men worldwide [1]
A total of 485 patients were enrolled in the study, with a median age of 69.86 [95% confidence interval (CI), 62.79–76.93] years old
The parameters were normalized by CAPRA or CAPRA-S (PSA, GG-radical prostatectomy (RP), surgical margin, extracapsular extension, seminal vesicle) criteria
Summary
Prostate cancer (PCa) is the most common cancer among men and the second leading cause of death for men worldwide [1]. The Cancer of the Prostate Risk Assessment (CAPRA/CAPRA-S) score, National Comprehensive Cancer Network (NCCN) model and Gleason grade group (GG-RP) system were three kinds of most widely adopted models [5]. According to these systems, patients were assigned into low-, intermediate- and high-risk groups, while patients in the middle group would face controversial prognosis and unclear therapeutic guidance. In this study, based on pathological ground truth registration annotations, we developed a novel model for BCR prediction, combing quantitative features of magnetic resonance imaging (MRI) and deep learning-based survival analysis. We validated the model and calculated its incremental predictive value in comparison with several acknowledgeable nomograms in a multicenter dataset
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
