Development, multi-institutional external validation, and algorithmic audit of an artificial intelligence-based Side-specific Extra-Prostatic Extension Risk Assessment tool (SEPERA) for patients undergoing radical prostatectomy: a retrospective cohort study.

BACKGROUND: Accurate prediction of side-specific extraprostatic extension (ssEPE) is essential for performing nerve-sparing surgery to mitigate treatment-related side-effects such as impotence and incontinence in patients with localised prostate cancer. Artificial intelligence (AI) might provide robust and personalised ssEPE predictions to better inform nerve-sparing strategy during radical prostatectomy. We aimed to develop, externally validate, and perform an algorithmic audit of an AI-based Side-specific Extra-Prostatic Extension Risk Assessment tool (SEPERA). METHODS: Each prostatic lobe was treated as an individual case such that each patient contributed two cases to the overall cohort. SEPERA was trained on 1022 cases from a community hospital network (Trillium Health Partners; Mississauga, ON, Canada) between 2010 and 2020. Subsequently, SEPERA was externally validated on 3914 cases across three academic centres: Princess Margaret Cancer Centre (Toronto, ON, Canada) from 2008 to 2020; L'Institut Mutualiste Montsouris (Paris, France) from 2010 to 2020; and Jules Bordet Institute (Brussels, Belgium) from 2015 to 2020. Model performance was characterised by area under the receiver operating characteristic curve (AUROC), area under the precision recall curve (AUPRC), calibration, and net benefit. SEPERA was compared against contemporary nomograms (ie, Sayyid nomogram, Soeterik nomogram [non-MRI and MRI]), as well as a separate logistic regression model using the same variables included in SEPERA. An algorithmic audit was performed to assess model bias and identify common patient characteristics among predictive errors. FINDINGS: Overall, 2468 patients comprising 4936 cases (ie, prostatic lobes) were included in this study. SEPERA was well calibrated and had the best performance across all validation cohorts (pooled AUROC of 0·77 [95% CI 0·75-0·78] and pooled AUPRC of 0·61 [0·58-0·63]). In patients with pathological ssEPE despite benign ipsilateral biopsies, SEPERA correctly predicted ssEPE in 72 (68%) of 106 cases compared with the other models (47 [44%] in the logistic regression model, none in the Sayyid model, 13 [12%] in the Soeterik non-MRI model, and five [5%] in the Soeterik MRI model). SEPERA had higher net benefit than the other models to predict ssEPE, enabling more patients to safely undergo nerve-sparing. In the algorithmic audit, no evidence of model bias was observed, with no significant difference in AUROC when stratified by race, biopsy year, age, biopsy type (systematic only vs systematic and MRI-targeted biopsy), biopsy location (academic vs community), and D'Amico risk group. According to the audit, the most common errors were false positives, particularly for older patients with high-risk disease. No aggressive tumours (ie, grade >2 or high-risk disease) were found among false negatives. INTERPRETATION: We demonstrated the accuracy, safety, and generalisability of using SEPERA to personalise nerve-sparing approaches during radical prostatectomy. FUNDING: None.

Explainable artificial intelligence to predict the risk of side-specific extraprostatic extension in pre-prostatectomy patients.

INTRODUCTION: We aimed to develop an explainable machine learning (ML) model to predict side-specific extraprostatic extension (ssEPE) to identify patients who can safely undergo nerve-sparing radical prostatectomy using preoperative clinicopathological variables. METHODS: A retrospective sample of clinicopathological data from 900 prostatic lobes at our institution was used as the training cohort. Primary outcome was the presence of ssEPE. The baseline model for comparison had the highest performance out of current biopsy-derived predictive models for ssEPE. A separate logistic regression (LR) model was built using the same variables as the ML model. All models were externally validated using a testing cohort of 122 lobes from another institution. Models were assessed by area under receiver-operating-characteristic curve (AUROC), precision-recall curve (AUPRC), calibration, and decision curve analysis. Model predictions were explained using SHapley Additive exPlanations. This tool was deployed as a publicly available web application. RESULTS: Incidence of ssEPE in the training and testing cohorts were 30.7 and 41.8%, respectively. The ML model achieved AUROC 0.81 (LR 0.78, baseline 0.74) and AUPRC 0.69 (LR 0.64, baseline 0.59) on the training cohort. On the testing cohort, the ML model achieved AUROC 0.81 (LR 0.76, baseline 0.75) and AUPRC 0.78 (LR 0.75, baseline 0.70). The ML model was explainable, well-calibrated, and achieved the highest net benefit for clinically relevant cutoffs of 10-30%. CONCLUSIONS: We developed a user-friendly application that enables physicians without prior ML experience to assess ssEPE risk and understand factors driving these predictions to aid surgical planning and patient counselling (https://share.streamlit.io/jcckwong/ssepe/main/ssEPE_V2.py).

Does prolonging the time to testicular cancer surgery impact long-term cancer control: a systematic review of the literature.

BACKGROUND: The wait times for urological cancer surgeries in Canada are beyond those recommended by the Canadian Association of Surgical Oncology. Prolonged wait times have a negative impact on patient quality of life but the effect on long-term cancer control is controversial. We conducted a systematic review of the testicular cancer literature to examine the best available evidence addressing the following key questions: What is the reported time interval for testicular cancer patients from the decision to operate until the day of testicular cancer surgery? Are there recommendations/guidelines in the urological cancer literature and, if so, how do the Canadian times compare? Is there a known association between duration of wait time beyond the recommended standard and clinical outcome (i.e. recurrence free survival, overall survival)? METHODS: A structured literature search of Medline, Pubmed, CINAHL, EMBASE, the Cochrane Database of Systematic Reviews, the Cochrane Database of Abstracts of Reviews of Effects, Healthstar and Google Scholar from January 1980 to September 2005 was conducted for published epidemiological studies and international guidelines/consensus documents that evaluated surgical wait times for testicular cancer. Data extracted from eligible studies included median time to diagnosis and to testicular cancer surgery. RESULTS: Five studies evaluating different components of wait times (e.g. delay in diagnosis, delay in orchiectomy) in testicular cancer patients were identified, four of which measured the impact of prolonged delays on relapse free and overall survival. Differences in study data availability, method of analysis and wait time definitions precluded statistical pooling of the findings. In one study from the United Kingdom, median wait time was 30 days from general practitioner referral to surgery and 4 days from diagnosis to surgery. No Canadian studies specific to testicular cancer were identified. The association between surgical delay and clinical outcomes remained controversial where only one of five epidemiological studies reported an association between treatment delay and relapse free and overall survival CONCLUSIONS: Even though the association between surgical delay and disease related clinical outcomes remains controversial, there is an ongoing concern that the psychological impact of prolonged waiting for urological cancer surgery could negatively impact patient outcomes. Additional research is needed to identify the current wait times for testicular cancer in Canada and to develop guidelines and recommendations on what appropriate wait times should be. To address these important issues, the surgical wait times (SWAT) initiative is mandated to provide the necessary guidance and recommendations to the federal and provincial governments. Through a partnership between the key stakeholders, it is the vision of SWAT to ultimately improve the care and quality of life of cancer patients.

Symptom prevalence, distress, and change over time in adults receiving treatment for lung cancer.

Adequate management of symptoms in adults with lung cancer is an important focus for clinical interventions. Knowledge of symptom prevalence and distress can be used to develop empirically based interventions that can potentially reduce distressing symptoms and improve quality of life. The purposes of this study were to describe which symptoms are most distressing, describe the prevalence of symptoms in adults receiving treatment for lung cancer, identify how symptoms change over time, and identify patient-related and clinical characteristics related to symptom distress. Data were available from 117 patients. Fatigue and pain were the most distressing symptoms for each group and at each time. Significant differences in distressing symptoms among the treatment groups were noted for nausea, fatigue, bowel pattern, and concentration at entry into the study and difficulty with appetite at 6 months. Many of the individual symptoms demonstrated a decrease in distress from 0 to 3 months and then an increase in distress levels from 3 to 6 months. Many of the individual symptoms were associated with demographic covariates and treatment group values but no consistent pattern emerged over time except for baseline symptom distress. Symptom distress at entry to the study was a strong predictor of nine distressing symptoms at 3 months and seven distressing symptoms at 6 months. Questionnaires such as the SDS may be useful as screening instruments to target those who need more intensive interventions.

A case of transitional cell carcinoma seeding of small bowel anastomosis after radical cystoprostatectomy.

We present a case in which a man developed transitional cell carcinoma of the small bowel 6 months after having undergone a cystoprostatectomy and ileal conduit urinary diversion.

Patterns of symptom distress in adults receiving treatment for lung cancer.

Knowledge of the patterns of symptom distress in adults receiving treatment for lung cancer is an important first step in developing interventions that can potentially lessen symptom distress. The purposes of this secondary analysis were to describe the changes in patterns of symptom distress over time in adults receiving treatment for lung cancer, and to examine the relationship of selected demographic and clinical characteristics to symptom distress. Complete data were available for 117 patients. The patterns of symptom distress in adults receiving treatment for lung cancer varied between treatment groups and over time. Symptom distress scores were moderate to high on entry into the study, indicating that symptom management in newly diagnosed lung cancer patients is essential and should begin early in the course of illness. Moreover, clinical interventions should be tailored to the type of treatment. Various demographic and clinical variables were weak and inconsistent predictors of symptom distress, underscoring the importance of examining the role of psychosocial factors in mediating symptom distress.