A self-supervised vision transformer to predict survival from histopathology in renal cell carcinoma.

PURPOSE: To develop and validate an interpretable deep learning model to predict overall and disease-specific survival (OS/DSS) in clear cell renal cell carcinoma (ccRCC). METHODS: Digitised haematoxylin and eosin-stained slides from The Cancer Genome Atlas were used as a training set for a vision transformer (ViT) to extract image features with a self-supervised model called DINO (self-distillation with no labels). Extracted features were used in Cox regression models to prognosticate OS and DSS. Kaplan-Meier for univariable evaluation and Cox regression analyses for multivariable evaluation of the DINO-ViT risk groups were performed for prediction of OS and DSS. For validation, a cohort from a tertiary care centre was used. RESULTS: A significant risk stratification was achieved in univariable analysis for OS and DSS in the training (n = 443, log rank test, p < 0.01) and validation set (n = 266, p < 0.01). In multivariable analysis, including age, metastatic status, tumour size and grading, the DINO-ViT risk stratification was a significant predictor for OS (hazard ratio [HR] 3.03; 95%-confidence interval [95%-CI] 2.11-4.35; p < 0.01) and DSS (HR 4.90; 95%-CI 2.78-8.64; p < 0.01) in the training set but only for DSS in the validation set (HR 2.31; 95%-CI 1.15-4.65; p = 0.02). DINO-ViT visualisation showed that features were mainly extracted from nuclei, cytoplasm, and peritumoural stroma, demonstrating good interpretability. CONCLUSION: The DINO-ViT can identify high-risk patients using histological images of ccRCC. This model might improve individual risk-adapted renal cancer therapy in the future.

A Matched-pair Analysis Comparing Systematic Prostate Biopsy by Conventional Transrectal Ultrasound-guidance Versus Software-based Predefined 3D-Guidance.

OBJECTIVE: To compare software-based three-dimensional-guided systematic prostate biopsy (3D-GSB) with conventional transrectal ultrasound-guided systematic biopsy (TGSB) regarding prostate cancer (PCa) detection rates (CDR). METHODS: In total, 956 patients (200 TGSB patients and 756 3D-GSB patients) without prior positive biopsies and with a prostate-specific antigen value ≤20 ng/ml were eligible for analysis. TGSB and 3D-GSB cases were matched in a 1:1 ratio using propensity score matching with age, prostate-specific antigen, prostate volume, previous biopsy status, and suspicious palpatory finding as confounders. 3D-GSB was conducted with the semi-robotic prostate fusion-biopsy system Artemis. For each patient in both groups, SB was conducted in a similar pattern with 12 cores. All cores in 3D-GSB were automatically planned and mapped on a 3D-model as well as on the real-time transrectal ultrasound imaging. Primary end points were the clinically significant (cs) and overall CDR. Secondary end point was the cancer-positive core rate. RESULTS: After matching, the csCDR was not significantly different between the 3D-GSB and the TGSB groups (33.3% vs 28.8%, P = .385). Overall CDR was significantly higher for 3D-GSB compared to TGSB (55.6% vs 39.9%, P = .002). 3D-GSB detected significantly more non-significant PCa than TGSB (22.2% vs 11.1%, P = .004). In patients with PCa, the number of cancer-positive SB cores was significantly higher by TGSB (42% vs 25%, P < .001). CONCLUSION: 3D-GSB was associated with a higher CDR than TGSB. However, no significant difference was shown in detection of csPCa between both techniques. Therefore, currently, 3D-GSB does not appear to add value to conventional TGSB.

An m6A-Driven Prognostic Marker Panel for Renal Cell Carcinoma Based on the First Transcriptome-Wide m6A-seq.

To date, only a single transcriptome-wide m6A sequencing study of clear cell renal cell carcinoma (ccRCC) has been reported, with no validation so far. Herein, by TCGA analysis of the KIRC cohort (n = 530 ccRCC; n = 72 normal), an external expression validation of 35 preidentified m6A targets was performed. Further in-depth expression stratification enabled assessment of m6A-driven key targets. Overall survival (OS) analysis and gene set enrichment analyses (GSEA) were conducted to assess their clinical and functional impact on ccRCC. In the hyper-up cluster significant upregulation was confirmed for NDUFA4L2, NXPH4, SAA1, and PLOD2 (40%) and in the hypo-up cluster for FCHSD1 (10%). Significant downregulation was observed for UMOD, ANK3, and CNTFR (27.3%) in the hypo-down cluster and for CHDH (25%) in the hyper-down cluster. In-depth expression stratification showed consistent dysregulation in ccRCC only for 11.67%: NDUFA4L2, NXPH4, and UMOD (NNU-panel). Patients with strong NNU panel dysregulation had significantly poorer OS (p = 0.0075). GSEA identified 13 associated and significantly upregulated gene sets (all p-values < 0.5; FDR < 0.25). External validation of the only available m6A sequencing in ccRCC consistently reduced dysregulated m6A-driven targets on the NNU panel with highly significant effects on OS. Epitranscriptomics are a promising target for developing novel therapies and for identifying prognostic markers for daily clinical practice.

Predicting Complexity in Transurethral Resection of Bladder Tumours: External Validation and Modification of the Bladder Complexity Score.

INTRODUCTION: First external validation of the Bladder Complexity Score (BCS) for predicting complex transurethral resection of bladder tumours (TURBT). METHODS: For BCS calculation, TURBTs performed at our institution between January 2018 and December 2019 were reviewed for the presence of preoperative characteristics listed in the Bladder Complexity Checklist (BCC). Receiver operating characteristics (ROC) analysis was used for BCS validation. To establish a modified BCS (mBCS) with maximum area under the curve (AUC), multivariable logistic regression (MLR) analysis was performed with all BCC-characteristics for different definitions of complex TURBT. RESULTS: 723 TURBTs were included in statistical analyses. Cohort's mean BCS was 11.2 ± 2.4 points (range: 5.5-22 points). In ROC analysis, BCS could not predict complex TURBT (AUC 0.573 [95% CI: 0.517-0.628]). MLR identified tumour size (OR 2.662, p < 0.001), and tumour number > 10 (OR 6.390, p = 0.032) as sole predictors for the modified endpoint of complex TURBT defined as a procedure meeting > 1 criterion: incomplete resection, surgery > 1 h, intraoperative complication, postoperative complications Clavien-Dindo ≥ III. mBCS increased the prediction to an AUC of 0.770 (95% CI: 0.667-0.874). CONCLUSION: In this first external validation, BCS remained an insufficient predictor of complex TURBT. mBCS requires reduced parameters, is more predictive and easier to apply in clinical practice.

Patient Compliance in Assessing Electronic Patient-Reported Outcome Measures after Urologic Surgery.

INTRODUCTION: This study aimed to assess patient compliance with a newly established electronic patient-reported outcome measure (ePROM) system after urologic surgery and to identify influencing factors. METHODS: Digital surveys were provided to patients undergoing cystectomy, radical or partial nephrectomy, or transurethral resection of bladder tumor via a newly established ePROM system. Participants received a baseline survey preoperatively and several follow-up surveys postoperatively. Multivariable regression analysis was performed to identify factors predicting compliance. RESULTS: Of N = 435 eligible patients, n = 338 completed the baseline survey (78.0%). Patients who did not participate were significantly more likely male (p = 0.004) and older than 70 years (p = 0.005). Overall, 206/337 patients (61.3%) completed the survey at 1-month, 167/312 (53.5%) at 3-month, and 142/276 (51.4%) at 6-month follow-up. Lower baseline quality of life (odds ratio: 2.27; p = 0.004) was a significant predictor for dropout at 1-month follow-up. Low educational level was significantly associated with low compliance at 3- (OR: 1.92; p = 0.01) and 6-month follow-up (OR: 2.88; p < 0.001). CONCLUSION: Acceptable compliance rates can be achieved with ePROMs following urologic surgery. Several factors influence compliance and should be considered when setting-up ePROM surveys.

Prospective, Randomized Comparative Evaluation of a Novel Hands-On Endourology Training Curriculum.

INTRODUCTION: The aim of this randomised prospective trial was to evaluate a novel hands-on endourological training programme (HTP) and compare it to the standard endourological colloquium (SC). METHODS: A new HTP was created based on a sequence of theoretical, video-based, and practical elements emphasising contemporary teaching methods. An existing SC in which live endourological operations were attended served as a comparison. Medical students were enrolled in a ratio of 1:2 (SC:HTP). Objective knowledge questionnaires (5 questions, open answers) and subjective Likert-type questionnaires (rating 1-3 vs. 4-5) were used for evaluation. Primary endpoint was urological knowledge transfer; secondary endpoints were learning effects, progression, and urological interest. RESULTS: 167 students (SC n = 52, HTP n = 115) were included. The knowledge assessment showed a significant increase in knowledge transfer benefitting the HTP on all 5 surveyed items (mean: n = 4/5/4/3/2 vs. n = 2/3/1/1, p < 0.0001). Interest and duration of the course were rated significantly more positively by HTP students (100.0/95.0% vs. 85.0/70.0%, p < 0.0001). The HTP students were significantly more confident in performing a cystoscopy independently (HTP 43.5% vs. SC 11.5%, p < 0.0001) and significantly claimed more often to have gained interdisciplinary and urological skills during the course (HTP 90.0/96.5% vs. SC 23.1/82.7%, p < 0.0001/p = 0.003). HTP students were also more likely to take the course again (HTP 98.2% vs. SC 59.6%, p < 0.0001). CONCLUSION: Modifying endourological teaching towards hands-on teaching resulted in stronger course interest, greater confidence regarding endourologic procedures, and significantly increased urologic knowledge transfer.

Integrin Expression in Localized Prostate Cancer: A TCGA and MSKCC Cohort-based Exploratory In Silico Analysis.

BACKGROUND/AIM: Diagnostic and prognostic biomarkers in localized prostate cancer (PC) are insufficient. Treatment stratification relies on prostate-specific antigen, clinical tumor staging and International Society of Urological Pathology (ISUP) grading, whereas molecular profiling remains unused. Integrins (ITG) have an important function in bidirectional signaling and are associated with progression, proliferation, perineural invasion, angiogenesis, metastasis, neuroendocrine differentiation, and a more aggressive disease phenotype in PC. However, ITG subunit expression in localized PC and their utility as prognostic biomarkers has not yet been analyzed. This study aimed to fill this gap and provide a comprehensive overview of ITG expression as well as ITG utility as biomarkers. PATIENTS AND METHODS: The Cancer Genome Atlas (TCGA) and the Memorial Sloan Kettering Cancer Center (MSKCC) prostate adenocarcinoma cohorts were analyzed regarding ITG expression in correlation to ISUP, N- and American Joint Committee on Cancer (AJCC) stage and were correlated with disease-free survival (DFS). Statistical tests used included the Mann-Whitney U-test, logrank test and uni- and multivariable cox regression analyses. RESULTS: After grouping for ISUP (1 and 2 vs. 3-5), N0 vs. N1 and AJCC stage (≤2 vs. ≥3), multiple ITGs showed significant expression differences. The most consistent results were observed for ITGα4, ITGαX, ITGα11, ITGß2 and ITGα2. In multivariable cox regression, ITGα2, ITGα10, ITGαD, ITGαB2 (TCGA), ITGα11 and ITGß4 (MSKCC) were independent predictors of DFS. CONCLUSION: The utility of ITGs as PC biomarkers was herein shown.

Systemic inflammatory biomarkers as predictive and prognostic factors in men with metastatic castration-refractory prostate cancer treated with docetaxel therapy: a comprehensive analysis in a German real-world cohort.

PURPOSE: Advances in therapy of metastatic castration-refractory prostate cancer (mCRPC) resulted in more therapeutic options and led to a higher need of predictive/prognostic biomarkers. Systemic inflammatory biomarkers could provide the basis for personalized treatment selection. This study aimed to assess the modified Glasgow Prognostic Score (mGPS), the neutrophile-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR) and the systemic immune-inflammation index (SII) in men with mCRPC under docetaxel. METHODS: Patients with mCRPC and taxane chemotherapy at a tertiary care centre between 2010 and 2019 were screened retrospectively. The biomarkers mGPS, NLR, PLR and SII were assessed and analyzed for biochemical/radiologic response and survival. RESULTS: We included 118 patients. Of these, 73 (61.9%) had received docetaxel as first-line, 31 (26.2%) as second-line and 14 (11.9%) as third-line treatment. For biochemical response, mGPS (odds ratio (OR) 0.54, p = 0.04) and PLR (OR 0.63, p = 0.04) were independent predictors in multivariable analysis. SII was significant in first-line cohort only (OR 0.29, p = 0.02). No inflammatory marker was predictive for radiologic response. In multivariable analysis, mGPS and NLR (hazard ratio (HR) 1.71 and 1.12, both p < 0.01) showed significant association with OS in total cohort and mGPS in the first-line cohort (HR 2.23, p < 0.01). Haemoglobin (Hb) and alkaline phosphatase (AP) showed several significant associations regarding 1 year, 3 year, OS and biochemical/radiologic response. CONCLUSIONS: Pre-treatment mGPS seems a promising prognostic biomarker. A combination of mGPS, NLR and further routine markers (e.g., Hb and AP) could yield optimized stratification for treatment selection. Further prospective and multicentric assessment is needed.

Changes in neutrophile-to-lymphocyte ratio as predictive and prognostic biomarker in metastatic prostate cancer treated with taxane-based chemotherapy.

OBJECTIVES: To assess the predictive and prognostic value of changes in longitudinal neutrophile-to-lymphocyte (NLR) ratios in men receiving taxane-based chemotherapy for metastatic prostate cancer (PC). METHODS: Retrospective, unicentric cohort study of patients treated with either docetaxel for metastatic hormone-sensitive PC (mHSPC) or docetaxel or cabazitaxel for metastatic castration-refractory PC (mCRPC) at a tertiary referral hospital between 2010 and 2019. NLR ratios were calculated for each cycle. Next, slopes over the first three (NLR3) and over six cycles (NLR6) were calculated and analysed for biochemical/radiologic response and survival. RESULTS: A total of 36 mHSPC (docetaxel), 118 mCRPC (docetaxel) and 38 mCRPC (cabazitaxel) patients were included. NLR3 was significantly associated with 1-year-survival, radiographic and biochemical response in mCRPC (docetaxel) in uni- and multivariable analyses. In mCRPC (docetaxel), positive NLR3s were associated with favourable 1-year-survival. CONCLUSION: This study demonstrated NLR3 as a prognostic marker in men receiving docetaxel for mCRPC. NLR3 might be a clinical tool to reflect the individual's response to taxane-based chemotherapy. Thereby, NLR3 could complement existing biomarkers and help to early identify treatment failure before complications arise. Further prospective and multicentric studies are needed to extend and confirm the presented results.

Baseline Modified Glasgow Prognostic Score (mGPS) Predicts Radiologic Response and Overall Survival in Metastatic Hormone-sensitive Prostate Cancer Treated With Docetaxel Chemotherapy.

BACKGROUND/AIM: To assess the baseline inflammatory markers modified Glasgow Prognostic Score (mGPS), systemic immune-inflammation index (SII), and neutrophile-to-lymphocyte ratio (NLR) as pragmatic tools for predicting response to chemohormonal therapy (docetaxel plus ADT) and prognosis in men with metastatic hormone-sensitive prostate cancer (mHSPC). PATIENTS AND METHODS: Male patients who received docetaxel at a tertiary university care center between 2014 and 2019 were screened for completion of 6 cycles. NLR, SII, mGPS, overall survival (OS), three-year survival, and radiologic response were assessed. Complete response (CR), partial response (PR), and stable disease (SD) were analyzed alone and in combination. RESULTS: Thirty-six mHSPC-patients were included. In thirty patients, baseline mGPS was assessed and was either 0 (n=22) or 2 (n=8). In Cochran-Armitage Trend Test, mGPS showed significant association with the combined radiologic endpoint of "CR, PR, or SD" (p=0.01), three-year survival (p=0.02), and OS (p<0.01). Next to prostate-specific antigen (PSA) (HR per 100 units 1.16, 95%CI=1.04-1.30, p<0.01), NLR (HR=1.31, 95%CI=1.03-1.66, p=0.03), and mGPS (2 vs. 0, HR=6.53, 95%CI=1.6-27.0, p<0.01) at baseline showed significant association with OS in univariable cox regression. However, mGPS remained the only independent predictor for OS in multivariable cox regression (p<0.01) and for the combined radiologic endpoint of "CR, PR or SD" (p=0.01) in multivariable logistic regression. SII showed no statistical relevance. CONCLUSION: Baseline mGPS seems to be a pragmatic tool for clinical decision-making in patients with mHSPC in daily routine.

Upper Tract Urinary Cancer Recurrence after Radical Cystectomy: Risk Assessment of Intraoperative Frozen Section.

INTRODUCTION: Upper tract urinary cancer recurrence (UTUCR) after radical cystectomy (RC) is outcome-limiting. Surgical recommendations on intraoperative performance of frozen section analysis (FSA) and management of positive ureteral margin (PUM) are lacking. METHODS: 634 RC cases were identified (2010-2018). In patients with PUM, sequential ureteral resections up to a negative margin were performed. We investigated the accuracy of FSA, significance of PUM, and identified risk factors (RFs) to stratify patients for UTUCR. RESULTS: FSA was performed in 355 patients, including a total of 693 ureters. FSA sensitivity was 0.93 and specificity 0.99. PUM conversion was possible in 52 (91.2%) cases. UTUCR occurred in 17 (4.8%) patients. Identified UTUCR RFs are non-muscle invasive bladder carcinoma (NMIBC) (OR 3.8, 95% confidence intervals [CI] 1.4-10.2, p = 0.008), multifocal bladder cancer in cystectomy specimen (OR 4.7, CI 1.1-20.8, p = 0.042), and recurrent NMIBC (OR 4.1, CI 1.5-10.9, p = 0.006). Risk-group stratification showed a six-fold increase in UTUCR risk (low-to high-risk). CONCLUSION: FSA is a sensitive and specific method to identify PUM. UTUCR occurs significantly more often in patients with recurrent, multifocal NMIBC at the time of RC. Patients can be risk stratified for UTUCR. In case of NMIBC-PUM, surgeons can safely opt for a kidney preserving strategy.

Long-term Reinterventions after Thulium Laser Enucleation of the Prostate: 12-Year Experience with more than 1000 Patients.

BACKGROUND: Thulium laser enucleation of the prostate (ThuLEP) is an established treatment option for benign prostatic enlargement (BPE), but long-term outcomes have not yet been reported. OBJECTIVE: To prove the durability of ThuLEP by investigating its long-term efficacy and morbidity. DESIGN, SETTING, AND PARTICIPANTS: All patients who underwent ThuLEP at a German tertiary referral center between 2009 and 2021 were retrospectively followed up for reinterventions for persistence or regrowth of prostate adenoma (ReIP) or long-term complications (ReIC). INTERVENTION: ThuLEP. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: We calculated the cumulative incidence for ReIP and ReIC at 10 yr. Univariate and multivariate Cox regression models were constructed to identify predictors of ReIP and ReIC. RESULTS AND LIMITATIONS: Overall, 1097 patients underwent ThuLEP. The median overall follow-up was 6.0 yr (interquartile range [IQR] 2.4-9.2). For one-third of patients (n = 369), median follow-up of 10 yr (IQR 9.1-11.2) was available. A total of 42 patients (3.8%) underwent ReIP after a median of 2 yr (IQR 0.3-4.9). The rate of long-term ReIC was 2.6% (n = 29) and the median time to ReIC was 0.5 yr (IQR 0.3-1.7). The most frequent ReIC was urethrotomy (n = 16, 1.5%). The cumulative incidence of ReIP and ReIC at 10 yr was estimated at 5.6% and 3.4%, respectively. Enucleation weight ≥60 g was a significant predictor of ReIP (hazard ratio 1.2, p = 0.014). The retrospective study design and the lack of functional outcomes are the main limitations. CONCLUSIONS: ThuLEP is a durably effective and safe procedure with low reintervention rates within 12 yr. PATIENT SUMMARY: This study investigated long-term outcomes of thulium laser enucleation of the prostate for benign enlargement of the prostate (BPE). Low rates of repeat treatment for BPE recurrence or for other complications were observed. Our results show the safety and efficacy of this treatment over a period of 12 years.

A novel endoimaging system for endoscopic 3D reconstruction in bladder cancer patients.

INTRODUCTION: The methods employed to document cystoscopic findings in bladder cancer patients lack accuracy and are subject to observer variability. We propose a novel endoimaging system and an online documentation platform to provide post-procedural 3D bladder reconstructions for improved diagnosis, management and follow-up. MATERIAL AND METHODS: The RaVeNNA4pi consortium is comprised of five industrial partners, two university hospitals and two technical institutes. These are grouped into hardware, software and clinical partners according to their professional expertise. The envisaged endoimaging system consists of an innovative cystoscope that generates 3D bladder reconstructions allowing users to remotely access a cloud-based centralized database to visualize individualized 3D bladder models from previous cystoscopies archived in DICOM format. RESULTS: Preliminary investigations successfully tracked the endoscope's rotational and translational movements. The structure-from-motion pipeline was tested in a bladder phantom and satisfactorily demonstrated 3D reconstructions of the processing sequence. AI-based semantic image segmentation achieved a 0.67 dice-score-coefficient over all classes. An online-platform allows physicians and patients to digitally visualize endoscopic findings by navigating a 3D bladder model. CONCLUSIONS: Our work demonstrates the current developments of a novel endoimaging system equipped with the potential to generate 3D bladder reconstructions from cystoscopy videos and AI-assisted automated detection of bladder tumors.

Ex vivo validation of a real-time multispectral endoscopic system for the detection and biopsy of bladder tumors.

BACKGROUND: Real-time multispectral imaging (rMSI) simultaneously provides white light (WL), photodynamic diagnosis (PDD) images, and a real-time fusion of both. It may improve the detection of bladder tumors. However, rMSI has not been used for transurethral biopsy or resection so far. The aim of this ex vivo study was to test the feasibility of bladder tumor biopsies using the rMSI system and compare it to a conventional endoscopic system. METHODS: A 3D printed rigid bladder phantom was equipped with small and flat (5 mm × 1 mm) mock-bladder-tumors made of silicone and fluorescent Qdots655 (Thermo Fisher Scientific, Germany). Urologists (n=15) were asked to perform a rigid cystoscopy and biopsy of all identified lesions (n=6) using a prototype rMSI system and the Image1 S system (Karl Storz, Tuttlingen). Success rate and completion time were measured. The image quality of both systems and the usability of the rMSI system according to the system usability scale (SUS) were evaluated with a task-specific questionnaire. RESULTS: Tumor detection and biopsy rate were 100% (90/90) for the rMSI system and 98.9% (89/90) for the Image1 S system (P=0.3). The biopsy completion time did not differ significantly between the systems (P=0.48). Differentiation between healthy and suspect mucosa with the rMSI system was rated as comparable to the Image1 S system by 53% of surgeons and as better by 33% of the surgeons. The median SUS score for the rMSI system was 87.5%. CONCLUSIONS: Accurate transurethral biopsies are feasible with the rMSI system. Furthermore, the rMSI system has an excellent SUS. This study paves the way to the first in-human transurethral resections of bladder tumors (TUR-B) using rMSI technology.

Soft Urinary Bladder Phantom for Endoscopic Training.

Bladder cancer (BC) is the main disease in the urinary tract with a high recurrence rate and it is diagnosed by cystoscopy (CY). To train the CY procedures, a realistic bladder phantom with correct anatomy and physiological properties is highly required. Here, we report a soft bladder phantom (FlexBlad) that mimics many important features of a human bladder. Under filling, it shows a large volume expansion of more than 300% with a tunable compliance in the range of 12.2 ± 2.8 - 32.7 ± 5.4 mL cmH2O-1 by engineering the thickness of the bladder wall. By 3D printing and multi-step molding, detailed anatomical structures are represented on the inner bladder wall, including sub-millimeter blood vessels and reconfigurable bladder tumors. Endoscopic inspection and tumor biopsy were successfully performed. A multi-center study was carried out, where two groups of urologists with different experience levels executed consecutive CYs in the phantom and filled in questionnaires. The learning curves reveal that the FlexBlad has a positive effect in the endourological training across different skill levels. The statistical results validate the usability of the phantom as a valuable educational tool, and the dynamic feature expands its use as a versatile endoscopic training platform.

Deep learning approach to predict lymph node metastasis directly from primary tumour histology in prostate cancer.

OBJECTIVE: To develop a new digital biomarker based on the analysis of primary tumour tissue by a convolutional neural network (CNN) to predict lymph node metastasis (LNM) in a cohort matched for already established risk factors. PATIENTS AND METHODS: Haematoxylin and eosin (H&E) stained primary tumour slides from 218 patients (102 N+; 116 N0), matched for Gleason score, tumour size, venous invasion, perineural invasion and age, who underwent radical prostatectomy were selected to train a CNN and evaluate its ability to predict LN status. RESULTS: With 10 models trained with the same data, a mean area under the receiver operating characteristic curve (AUROC) of 0.68 (95% confidence interval [CI] 0.678-0.682) and a mean balanced accuracy of 61.37% (95% CI 60.05-62.69%) was achieved. The mean sensitivity and specificity was 53.09% (95% CI 49.77-56.41%) and 69.65% (95% CI 68.21-71.1%), respectively. These results were confirmed via cross-validation. The probability score for LNM prediction was significantly higher on image sections from N+ samples (mean [SD] N+ probability score 0.58 [0.17] vs 0.47 [0.15] N0 probability score, P = 0.002). In multivariable analysis, the probability score of the CNN (odds ratio [OR] 1.04 per percentage probability, 95% CI 1.02-1.08; P = 0.04) and lymphovascular invasion (OR 11.73, 95% CI 3.96-35.7; P < 0.001) proved to be independent predictors for LNM. CONCLUSION: In our present study, CNN-based image analyses showed promising results as a potential novel low-cost method to extract relevant prognostic information directly from H&E histology to predict the LN status of patients with prostate cancer. Our ubiquitously available technique might contribute to an improved LN status prediction.

Endourological Training Using 3D-Printed Bladder Phantoms: Development and Prospective Evaluation.

Background: To create and evaluate a realistic, anatomically accurate, and user-friendly bladder phantom for reproducible endourological training purposes and endoscope mastery. Materials and Methods: The anatomy of full bladders was mapped from human computed tomography datasets. After a 3D model development process, content evidence and response process evidence (RPE) of the phantom were evaluated using the system usability scale (SUS), 5-point Likert scale questionnaires, and task execution of experienced urologists (U) and endoscopy-naive medical students (MS) in two training sessions (first vs second). Required validation cohort sizes (1:10) of the evaluating urologists (n = 12) and students (n = 115) were precalculated. Time measurements were recorded. Students were additionally evaluated by a validated global psychomotor assessment score (GPSS). Group comparisons were calculated by the Mann-Whitney U test. All tests were two sided with p < 0.05 considered statistically significant. Results: Content evidence was assessed by urologists with an "excellent" SUS score of 89.4 ± 5.9 and an average "agreement" of ≥4 pts in the Likert scale questionnaires. RPE was assessed by intra- and intergroup time comparison for the execution of endoscopic tasks (cystoscopy [CY], guidewire insertion, and tumor biopsy). For CY, U: first 17.6 ± 4.4 seconds vs second 12.4 ± 2.0 seconds, p = 0.002; MS: first 56.6 ± 28.2 seconds vs second 28.6 ± 14.7 seconds, p < 0.001; U vs MS: first U 17.6 ± 4.4 seconds vs first MS 56.6 ± 28.2 seconds, p < 0.001, second U 12.4 ± 2.0 seconds vs second MS 28.6 ± 14.7 seconds, p < 0.001. Significant time differences were documented for all tasks and sessions (p < 0.001). Additionally, significant GPSS differences were recorded between the sessions (GPSS: first 20.4 ± 5.1 pts vs second 24.7 ± 4.0 pts, p < 0.001). Conclusions: Our low-fidelity 3D-printed bladder, called BladCap, is an easy-to-assemble, inexpensive, and robust phantom. We present data, which establish construct validity to support use as a clinical training device.

The Comprehensive Complication Index (CCI) for improved reporting of complications in endourological stone treatment.

The Clavien-Dindo Classification (CDC) lacks a combined score of multiple complications in one patient. The comprehensive complication index (CCI) circumvents this problem making it a valuable instrument to optimize quality control. We aimed to introduce and validate CCI in the treatment of urolithiasis. 60 day postoperative complications of 327 consecutive patients undergoing percutaneous nephrolitholapaxy (PNL) and ureterorenoscopy (URS) between 2017 and 2019 were retrospectively assessed and graded according to CDC. CCI was calculated for each patient. Overall morbidity scores of CCI and CDC were compared. Correlation analyses between the two scores and length of hospital stay (LOS) were performed. A multivariate analysis was performed to identify predictive factors for complications. Sample size calculation for an imaginary clinical trial was compared between CCI and CDC. A significant difference in overall morbidity between CCI and CDC was revealed for PNL (p < 0.001) and URS (p = 0.001). There was no statistically significant difference in comparing correlations between cumulative CCI and LOS versus non-cumulative CDC and LOS for both cohorts. Operating time > 90 min, maximum stone size, positive preoperative urine culture and PNL type (p < 0.001) were predictive factors for postoperative complications in PNL, while urine culture (p = 0.02) was for URS. Sample size calculation based on CCI resulted in a significant reduction of required patients for PNL (- 48%) and URS (- 84%) compared to CDC. CCI could successfully be validated in endourological stone treatment with the advantage of assessing complications in their entirety compared to CDC. CCI can significantly reduce the required sample size in future clinical trials.

The Comprehensive Complication Index for Advanced Monitoring of Complications Following Endoscopic Surgery of the Lower Urinary Tract.

Purpose: To evaluate the Comprehensive Complication Index (CCI) for reporting complications in lower urinary tract transurethral procedures and compare it with the Clavien-Dindo classification (CDC). Materials and Methods: A total of 450 consecutive patients were included into the analyses [150 each of transurethral resection of bladder tumors (TURBT), transurethral resection of the prostate (TURP), and transurethral enucleation of the prostate using Tm:YAG, (ThuLEP)]. Complications were assessed according to the modified CDC. The CCI was calculated using a freely accessible online tool. Descriptive statistics and correlation analyses were applied to quantify operational differences and length of stay (LOS) between CDC and CCI. Sample size calculations for hypothetical clinical trials were contrasted for CDC and CCI application. Results: Overall n = 150 patients with complications (33.3%) within the first 60 days after operation were identified. Of these, n = 125 (83.4%) were minor complications up to CDC grade IIIa. Of patients with complications, n = 57 patients (12.6%) experienced more than one complication. Here, the cumulative CCI led to an upgrade of at least one CDC grade in 33 patients. Hence, in 22.0% of cases, the highest CDC grade underestimated the degree of complications. CCI showed higher correlation with LOS compared with CDC (all r > 0.2, all p-values ≤0.0001). Using CCI instead of CDC for sample calculation resulted in a strong reduction of the required number of patients for all three interventions (percentage of patient number decrease: -93.2% for TURBT, -71.8% for TURP, and -81.1% for ThuLEP). Conclusion: The CCI gives a more precise interpretation of the postinterventional complications of TURBT, TURP, and ThuLEP. CCI application may reduce the required sample size for clinical trials and will relieve their recruitment in the future.

Chronic Kidney Disease After Partial Nephrectomy in Patients With Preoperative Inconspicuous Renal Function - Curiosity or Relevant Issue?

BACKGROUND: Chronic kidney disease (CKD) is a severe long-term complication after partial nephrectomy (PN). Clinical and scientific focus lies on patients with impaired renal function at the time of surgery. Little data is available on patients with normal preoperative renal function (NPRF). PATIENTS AND METHODS: Patients who underwent PN with a preoperative estimated glomular filtration rate > 60 mL/min/1.73m2 were retrospectively examined at 8 European urologic centers. The occurrence of new onset CKD ≥ stage III after surgery (sCKD) was defined as the primary endpoint. Group comparisons and risk correlations were determined. Based on this data, a risk stratification model for sCKD was developed. RESULTS: Of the 1315 patients with NPRF included, 249 (18.9%) developed sCKD after a median follow-up of 44 months (range, 6-255 months). Pair analysis and univariable regression revealed age, arterial hypertension, American Society of Anesthesiologists score, tumor stage, surgical approach, intraoperative blood loss, perioperative blood transfusions and preoperative CKD stage as predictors for sCKD development. Multivariate analysis confirmed perioperative blood transfusion (hazard ratio [HR], 2.96; P ≤ .0001), age (≥ 55 years; HR, 2.60; P = .0002), tumor stage (> pT1; HR, 2.15; P = .025), and preoperative CKD stage (stage II vs. I; HR, 3.85; P ≤ .0001) as independent risk factors. A model that stratified patient risk for new onset CKD was highly significant (P < .0001). CONCLUSION: Every fifth patient with NPRF developed sCKD following PN. Elderly patients with higher tumor stage and who require blood transfusion appear to be at increased risk. Based on our risk stratification, patients with ≥ 2 risk factors are candidates for an early, nephrologic follow-up.