MRI T2w Radiomics-Based Machine Learning Models in Imaging Simulated Biopsy Add Diagnostic Value to PI-RADS in Predicting Prostate Cancer: A Retrospective Diagnostic Study.

BACKGROUND: Currently, prostate cancer (PCa) prebiopsy medical image diagnosis mainly relies on mpMRI and PI-RADS scores. However, PI-RADS has its limitations, such as inter- and intra-radiologist variability and the potential for imperceptible features. The primary objective of this study is to evaluate the effectiveness of a machine learning model based on radiomics analysis of MRI T2-weighted (T2w) images for predicting PCa in prebiopsy cases. METHOD: A retrospective analysis was conducted using 820 lesions (363 cases, 457 controls) from The Cancer Imaging Archive (TCIA) Database for model development and validation. An additional 83 lesions (30 cases, 53 controls) from Hong Kong Queen Mary Hospital were used for independent external validation. The MRI T2w images were preprocessed, and radiomic features were extracted. Feature selection was performed using Cross Validation Least Angle Regression (CV-LARS). Using three different machine learning algorithms, a total of 18 prediction models and 3 shape control models were developed. The performance of the models, including the area under the curve (AUC) and diagnostic values such as sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), were compared to the PI-RADS scoring system for both internal and external validation. RESULTS: All the models showed significant differences compared to the shape control model (all p < 0.001, except SVM model PI-RADS+2 Features p = 0.004, SVM model PI-RADS+3 Features p = 0.002). In internal validation, the best model, based on the LR algorithm, incorporated 3 radiomic features (AUC = 0.838, sensitivity = 76.85%, specificity = 77.36%). In external validation, the LR (3 features) model outperformed PI-RADS in predictive value with AUC 0.870 vs. 0.658, sensitivity 56.67% vs. 46.67%, specificity 92.45% vs. 84.91%, PPV 80.95% vs. 63.64%, and NPV 79.03% vs. 73.77%. CONCLUSIONS: The machine learning model based on radiomics analysis of MRI T2w images, along with simulated biopsy, provides additional diagnostic value to the PI-RADS scoring system in predicting PCa.

Comparison of systematic and combined biopsy for the detection of prostate cancer.

ABSTRACT: Systematic prostate biopsy has limitations, such as overdiagnosis of clinically insignificant prostate cancer and underdiagnosis of clinically significant prostate cancer. Magnetic resonance imaging (MRI)-guided biopsy, a promising alternative, might improve diagnostic accuracy. To compare the cancer detection rates of systematic biopsy and combined biopsy (systematic biopsy plus MRI-targeted biopsy) in Asian men, we conducted a retrospective cohort study of men who underwent either systematic biopsy or combined biopsy at two medical centers (Queen Mary Hospital and Tung Wah Hospital, Hong Kong, China) from July 2015 to December 2022. Descriptive statistics were calculated, and univariate and multivariate logistic regression analyses were performed. The primary and secondary outcomes were prostate cancer and clinically significant prostate cancer. A total of 1391 participants were enrolled. The overall prostate cancer detection rates did not significantly differ between the two groups (36.3% vs 36.6%, odds ratio [OR] = 1.01, 95% confidence interval [CI]: 0.81-1.26, P = 0.92). However, combined biopsy showed a significant advantage in detecting clinically significant prostate cancer (Gleason score ≥ 3+4) in patients with a total serum prostate-specific antigen (tPSA) concentration of 2-10 ng ml -1 (systematic vs combined: 11.9% vs 17.5%, OR = 1.58, 95% CI: 1.08-2.31, P = 0.02). Specifically, in the transperineal biopsy subgroup, combined biopsy significantly outperformed systematic biopsy in the detection of clinically significant prostate cancer (systematic vs combined: 12.6% vs 24.0%, OR = 2.19, 95% CI: 1.21-3.97, P = 0.01). These findings suggest that in patients with a tPSA concentration of 2-10 ng ml -1 , MRI-targeted biopsy may be of greater predictive value than systematic biopsy in the detection of clinically significant prostate cancer.

Outcomes of radical prostatectomy in a 20-year localized prostate cancer single institution series in China.

The long-term survival outcomes of radical prostatectomy (RP) in Chinese prostate cancer (PCa) patients are poorly understood. We conducted a single-center, retrospective analysis of patients undergoing RP to study the prognostic value of pathological and surgical information. From April 1998 to February 2022, 782 patients undergoing RP at Queen Mary Hospital of The University of Hong Kong (Hong Kong, China) were included in our study. Multivariable Cox regression analysis and Kaplan-Meier analysis with stratification were performed. The 5-year, 10-year, and 15-year overall survival (OS) rates were 96.6%, 86.8%, and 70.6%, respectively, while the 5-year, 10-year, and 15-year PCa-specific survival (PSS) rates were 99.7%, 98.6%, and 97.8%, respectively. Surgical International Society of Urological Pathology PCa grades (ISUP Grade Group) ≥4 was significantly associated with poorer PSS (hazard ratio [HR] = 8.52, 95% confidence interval [CI]: 1.42-51.25, P = 0.02). Pathological T3 stage was not significantly associated with PSS or OS in our cohort. Lymph node invasion and extracapsular extension might be associated with worse PSS (HR = 20.30, 95% CI: 1.22-336.38, P = 0.04; and HR = 7.29, 95% CI: 1.22-43.64, P = 0.03, respectively). Different surgical approaches (open, laparoscopic, or robotic-assisted) had similar outcomes in terms of PSS and OS. In conclusion, we report the longest timespan follow-up of Chinese PCa patients after RP with different approaches.

Prostate Health Index (phi) and its derivatives predict Gleason score upgrading after radical prostatectomy among patients with low-risk prostate cancer.

To analyze the performance of the Prostate Health Index (phi) and its derivatives for predicting Gleason score (GS) upgrading between prostate biopsy and radical prostatectomy (RP) in the Chinese population, an observational, prospective RP cohort consisting of 351 patients from two medical centers was established from January 2017 to September 2020. Pathological reclassification was determined by the Gleason Grade Group (GG). The area under the receiver operating characteristic curve (AUC) and logistic regression (LR) models were used to evaluate the predictive performance of predictors. In clinically low-risk patients with biopsy GG ≤2, phi (odds ratio [OR] = 1.80, 95% confidence interval [95% CI]: 1.14-2.82, P = 0.01) and its derivative phi density (PHID; OR = 2.34, 95% CI: 1.30-4.20, P = 0.005) were significantly associated with upgrading to GG ≥3 after RP, and the results were confirmed by multivariable analysis. Similar results were observed in patients with biopsy GG of 1 for the prediction of upgrading to RP GG≥2. Compared to the base model (AUC = 0.59), addition of the phi or PHID could provide additional predictive value for GS upgrading in low-risk patients (AUC = 0.69 and 0.71, respectively, both P < 0.05). In conclusion, phi and PHID could predict GS upgrading after RP in clinically low-risk patients.