ResultsContinence was attained in 79 patients (79%) after a median of 126 days. The DL model achieved a C-index of 0.6 and an MAE of 85.9 in predicting continence. APMs were ranked higher by the model than clinicopathological features. In the historical cohort, patients in Group 1/APMs had superior rates of urinary continence at 3 and 6 months postoperatively (47.5 vs 36.7%, P = 0.034, and 68.3 vs 59.2%, P = 0.047, respectively). ConclusionUsing APMs and clinicopathological data, the DeepSurv DL model was able to predict continence after RARP. In this feasibility study, surgeons with more efficient APMs achieved higher continence rates at 3 and 6 months after RARP.Keywords robotic surgical procedures, prostatectomy, artificial intelligence, urinary incontinence, quality of life

BackgroundPresenting features of inflammatory bowel disease (IBD) are non-specific. We hypothesized that mRNA profiles could (1) identify genes and pathways involved in disease pathogenesis; (2) identify a molecular signature that differentiates IBD from other conditions; (3) provide insight into systemic and colon-specific dysregulation through study of the concordance of the gene expression.MethodsChildren (8–18 years) were prospectively recruited at the time of diagnostic colonoscopy for possible IBD. We used transcriptome-wide mRNA profiling to study gene expression in colon biopsies and paired whole blood samples. Using blood mRNA measurements, we fit a regression model for disease state prediction that was validated in an independent test set of adult subjects (GSE3365).ResultsNinety-eight children were recruited [39 Crohn’s disease, 18 ulcerative colitis, 2 IBDU, 39 non-IBD]. There were 1,118 significantly differentially (IBD vs non-IBD) expressed genes in colon tissue, and 880 in blood. The direction of relative change in expression was concordant for 106/112 genes differentially expressed in both tissue types. The regression model from the blood mRNA measurements distinguished IBD vs non-IBD disease status in the independent test set with 80% accuracy using only 6 genes. The overlap of 5 immune and metabolic pathways in the two tissue types was significant (p<0.001).ConclusionsBlood and colon tissue from patients with IBD share a common transcriptional profile dominated by immune and metabolic pathways. Our results suggest that peripheral blood expression levels of as few as 6 genes (IL7R, UBB, TXNIP, S100A8, ALAS2, and SLC2A3) may distinguish patients with IBD from non-IBD.

ObjectivesTo evaluate automated performance metrics (APMs) and clinical data of experts and super-experts for four cardinal steps of robot-assisted radical prostatectomy (RARP): bladder neck dissection; pedicle dissection; prostate apex dissection; and vesico-urethral anastomosis. Subjects and MethodsWe captured APMs (motion tracking and system events data) and synchronized surgical video during RARP. APMs were compared between two experience levels: experts (100-750 cases) and super-experts (2100-3500 cases). Clinical outcomes (peri-operative, oncological and functional) were then compared between the two groups. APMs and outcomes were analysed for 125 RARPs using multi-level mixed-effect modelling. ResultsFor the four cardinal steps selected, super-experts showed differences in select APMs compared with experts (P < 0.05). Despite similar PSA and Gleason scores, super-experts outperformed experts clinically with regard to peri-operative outcomes, with a greater lymph node yield of 22.6 vs 14.9 nodes, respectively (P < 0.01), less blood loss (125 vs 130 mL, respectively; P < 0.01), and fewer readmissions at 30 days (1% vs 13%, respectively; P = 0.02). A similar but nonsignificant trend was seen for oncological and functional outcomes, with super-experts having a lower rate of biochemical recurrence compared with experts (5% vs 15%, respectively; P = 0.13) and a higher continence rate at 3 months (36% vs 18%, respectively; P = 0.14). ConclusionWe found that experts and super-experts differed significantly in select APMs for the four cardinal steps of RARP, indicating that surgeons do continue to improve in performance even after achieving expertise. We hope ultimately to identify associations between APMs and clinical outcomes to tailor interventions to surgeons and optimize patient outcomes.

ObjectiveTo assess the utility of routine frozen section analysis of ureters at the time of radical cystectomy (RC) for urothelial cancer (UC), and the long-term outcomes of adverse ureteric pathology. Patients and MethodsPathological data on 2 047 patients undergoing RC for UC with routine frozen section analysis of ureters (January 1971 to December 2009) were analysed. Univariate and multivariable logistic and Cox proportional hazards models were used to determine the risk of upper tract UC (UTUC) recurrence, local recurrence and overall survival in those identified as having adverse pathology (severe atypia/ carcinoma in situ [CIS] or UC) at time of frozen section analysis. ResultsAdverse pathology was identified by frozen section analysis in 178 patients (8.6%). Frozen section analysis was found to have poor sensitivity in identifying adverse pathology (59.1%), which was improved in patients with preoperative CIS (68.0%). After a median (interquartile range) follow-up of 12.4 (1.9-10.1) years, 28 patients (1.4%) developed UTUC recurrence. There were no uretero-enteric anastomotic recurrences. Adverse pathology on frozen section analysis was associated with UTUC recurrence on univariate analysis (hazard ratio [HR] 6.2, 95% confidence interval [CI] 2.9-13.5), but 15/28 patients (54%) with UTUC recurrence had benign ureteric frozen section analysis on initial sectioning. Adverse pathology on frozen section analysis was not independently associated with the risk of local recurrence (HR 1.08, 95% CI 0.61-1.89) or overall survival (HR 1.12, 95% CI 0.94-1.35) in multivariate models. ConclusionsUreteric frozen section analysis has poor sensitivity and may be marginally improved in pre-existing CIS. UTUC recurrence is rare and can occur despite negative frozen section analysis. Our data question the utility of routine frozen section analysis of the distal ureteric margin at the time of RC.