Predicting phenotypes in benign urology

PROJECT SUMMARY When patients present with urinary urgency, frequency and nocturia, with or without urgency incontinence, we consider this the clinical phenotype of overactive bladder (OAB). Clinical observations and data suggest that clinical OAB may be comprised of several underlying subtypes. Discerning between subtypes has been challenging but is imperative to precisely guide targeted treatments. Within the last decade, several large observational cohort studies have been developed with the goal of improving our understanding of the pathophysiology underlying benign urologic symptom disorders. The NIDDK-sponsored Multidisciplinary Approach to the Study of Pelvic Pain (MAPP) network and Lower Urinary Tract Dysfunction Research Network (LURN) have collected data on urologic pain syndromes and non-painful lower urinary tract symptoms (LUTS), respectively. Both cohorts included patients with OAB. To identify clinically meaningful subtypes, MAPP and LURN used clustering methods to mathematically classify patients into probable subtypes using clinical characteristics. Though biospecimens were collected in both cohorts, biologic data from these specimens were not included in clustering. If OAB subtypes have different pathophysiology, biologic data are likely to add important distinguishing information, and possibly be predictive of differential treatment responses. Based on clinical observations and prior clustering work, we hypothesize that the syndrome of idiopathic OAB is comprised of 5 phenotypic subtypes, including dysbiotic and metabolic subtypes that will have distinct microbiome profiles from the others. To test our hypotheses, we propose an innovative approach combining MAPP and LURN clinical datasets and already collected biospecimens. Since sex-specific factors may affect subtypes, analyses will be performed on females and males separately. Our current “one-size fits all” algorithm of OAB treatments results in repeated medical visits, high health-care costs, and marginal long-term effectiveness. There is a major gap in knowledge of OAB subtypes, which in turn hampers our ability to target treatments to subgroups where they would be most effective. The proposed analyses by experienced data scientists present a unique opportunity to leverage existing resources from well-defined prospective clinical cohorts to substantially enhance our understanding of OAB.