[Artificial intelligence and neural networks in urology]. / Künstliche Intelligenz und neuronale Netze in der Urologie.

Knowledge of artificial intelligence will become essential in urology and medicine in the future. Today, computer-aided solutions are used for higher-dimensioned data or complex questions. In urological research, but also in clinical application, artificial neural networks (ANN) have been increasingly used for years and are continuously being developed. ANNs are already successfully used in urology in the field of image recognition in radiology and pathology. ANNs can potentially play a role in decision-making of modern cancer treatment. Today, access to ANN has become comparatively easy.

Calgranulin A (S100A8) Immunostaining: A Future Candidate for Risk Assessment in Patients with Non-Muscle-Invasive Bladder Cancer (NMIBC).

INTRODUCTION: There is an urgent need to identify patients with bladder cancer (BC) who are at high risk of recurrence or progression. Calgranulin A is a strong marker for muscle-invasive or advanced BC and recent studies have shown its potential for identifying patients at risk even in non-muscle-invasive bladder cancer (NMIBC). The present study examines risks of recurrence and progression dependent on immunostaining with calgranulin A in NMIBC. METHODS: Calgranulin A protein expression was evaluated through the immunohistochemistry of 158 randomly selected, transurethrally resected BC specimens of separate patients (pTa 89, pT1 69) using tissue microarrays. Kaplan-Meier survival analysis and Cox regression were performed to determine whether calgranulin A expression is associated with recurrence-free survival (RFS), progression-free survival (PFS), or cancer-specific survival (CSS). RESULTS: Calgranulin A expression is significantly different between pTa and pT1 tumors (p = 0.000, Mann-Whitney U test) and between tumor grades (p = 0.015, Kruskal-Wallis test). Kaplan-Meier estimates produced significant results for low and high calgranulin A expression concerning RFS [5y-RFS 70.4 ± 4.0% vs. 35.9 ± 12.5%, median RFS not reached (NR) vs. 12.0 ± 4.4 month, p = 0.029, log-rank test], PFS (5y-PFS 90.3 ± 2.7% vs. 51.5 ± 14.0%, median PFS NR in both groups, p = 0.000, log-rank test), and CSS (5y-CSS 92.9 ± 2.6% vs. 70.7 ± 12.4%, median CSS NR in both groups, p = 0.005, log-rank test). Calgranulin A remained an independent factor for RFS (p = 0.024, HR 2.43) and PFS (p = 0.002, HR 5.92) according to the multivariate Cox regression model. CONCLUSIONS: Calgranulin A expression in NMIBC, detected through immunohistochemistry, is a promising marker for the identification of NMIBC patients at high risk of recurrence and progression.

68Ga-PSMA PET/CT Imaging Predicting Intraprostatic Tumor Extent, Extracapsular Extension and Seminal Vesicle Invasion Prior to Radical Prostatectomy in Patients with Prostate Cancer.

PURPOSE: 68Ga-labeled prostate-specific membrane antigen (PSMA) ligand positron emission tomography/computed tomography (PET/CT) has shown promising results in patients with biochemical recurrence after primary therapy for prostate cancer. In this study, we evaluated the usefulness of PSMA I&T (imaging and therapy) PET/CT prior to radical prostatectomy. METHODS: The study population consisted of 21 patients with prostate cancer who underwent 68Ga-PSMA I&T PET/CT before either open or laparoscopic radical prostatectomy. Intraprostatic tumor extent, extracapsular extension (ECE) and seminal vesicle invasion (SVI) were assessed on the PET/CT scans. Tracer uptake was quantified in terms of standardized uptake values (SUVs). Imaging findings were correlated with final whole-gland histopathology. RESULTS: Of the 21 patients, two had T stage 2b disease, nine stage 2c, six stage 3a and four stage 3b. The median Gleason score was 7. The SUVmean of the primary tumors was 9.5 ± 8.8. SUVmean was higher in tumors with ECE than in organ-confined tumors (13.8 ± 11.0 vs. 5.6 ± 3.2, p = 0.029). Peak tracer uptake was significantly positively correlated with Gleason score (rs = 0.49, p = 0.025). Sensitivity, specificity, positive predictive value and negative predictive value were, respectively, 94.7%, 75.0%, 97.3% and 60.0% for tumor infiltration of an individual prostate lobe, 75.0%, 100.0%, 100.0% and 97.4% for SVI, and 90.0%, 90.9%, 90.0% and 90.9% for ECE, using an angulated contour of the prostate as the criterion. Tumor volume derived from 68Ga-PSMA I&T PET/CT was significantly correlated with preoperative prostate-specific antigen value (rp = 0.75, p < 0.001) and tumor volume on histopathology (rp = 0.45, p = 0.039). CONCLUSIONS: 68Ga-PSMA I&T PET/CT prior to radical prostatectomy can contribute to presurgical local staging of prostate cancer. In this pilot study, 68Ga-PSMA I&T PET/CT showed promising results for prediction of lobe infiltration, ECE and SVI.