Development of a machine learning-based predictive model for prediction of success or failure of medical management for benign prostatic hyperplasia.

OBJECTIVE: To develop a novel predictive model for identifying patients who will and will not respond to the medical management of benign prostatic hyperplasia (BPH). METHODS: Using data from the Medical Therapy of Prostatic Symptoms (MTOPS) study, several models were constructed using an initial data set of 2172 patients with BPH who were treated with doxazosin (Group 1), finasteride (Group 2), and combination therapy (Group 3). K-fold stratified cross-validation was performed on each group, Within each group, feature selection and dimensionality reduction using nonnegative matrix factorization (NMF) were performed based on the training data, before several machine learning algorithms were tested; the most accurate models, boosted support vector machines (SVMs), being selected for further refinement. The area under the receiver operating curve (AUC) was calculated and used to determine the optimal operating points. Patients were classified as treatment failures or responders, based on whether they fell below or above the AUC threshold for each group and for the whole data set. RESULTS: For the entire cohort, the AUC for the boosted SVM model was 0.698. For patients in Group 1, the AUC was 0.729, for Group 2, the AUC was 0.719, and for Group 3, the AUC was 0.698. CONCLUSION: Using MTOPS data, we were able to develop a prediction model with an acceptable rate of discrimination of medical management success for BPH.

Treatment patterns and survival of low and intermediate-risk prostate cancer in end-stage kidney disease: A retrospective population cohort study.

BACKGROUND: In accordance with guidelines, observation with or without active surveillance for low-risk prostate cancer increased in recent years in the general population. We compared treatment patterns and mortality for low- and intermediate-risk prostate cancer and mortality rates among end-stage kidney disease (ESKD) and non-ESKD patients. METHODS: This is a retrospective population-based observational cohort study of Surveillance, Epidemiology, and End Results-Medicare data of men aged 66 years and older with localized prostate cancer (2004-2015). ESKD status was determined using Medicare billing codes. Multivariable logistic regression models and Cox-proportional hazards models were used to study definitive treatment patterns and mortality, respectively. RESULTS: For low-risk prostate cancer, dialysis patients (N = 83) had lower but not statistically significant odds (OR, 0.74; 95% CI: 0.48-1.16) of receiving definitive treatment than non-ESKD patients (N = 24,935). For those with intermediate-risk prostate cancer, dialysis patients (N = 254) had lower odds to receive definitive treatment (OR, 0.54; 95% CI: 0.42-0.72) than non-ESKD patients (N = 60,883). From 2004-2010 to 2011-2015, for patients with low-risk prostate cancer, while the receipt of definitive treatment for non-ESKD patients trended down from 72% to 48%, it trended up for dialysis patients from 55% to 65%. Kidney transplant patients (N = 33 for low-risk and N = 91 for intermediate-risk) had lower rates of definitive treatment for low-risk and similar rates of treatment for intermediate-risk prostate cancer compared to non-ESKD patients. CONCLUSIONS: The disparity in definitive treatment rates for low-risk prostate cancer among dialysis patients exists despite their high mortality, compared to the general population.