Frank, Igor

Abstract The outcome of patients with Muscle Invasive Bladder Cancer (MIBC) is highly variable with many patients experiencing progressive disease despite multimodal therapy. A key variable that contributes to poor patient outcomes lies in our inability to predict therapeutic response due to the molecular heterogeneity of MIBC tumors. There is, therefore, a significant unmet clinical need to accurately predict effective drug combinations and therapeutic doses in individual patients with MIBC. Patient-derived tumor organoids (PDOs) have the potential to serve as an ex vivo model of urothelial carcinoma (UC) capable of functionally predicting patient-specific treatment responses. However, current culture procedures typically rely on the expansion of UC organoids from resected tumors or biopsy specimens, which is inherently invasive and hence limits clinical application. We hypothesized that culture of MIBC PDOs can be achieved from urine samples, as is the case in dogs with MIBC, since histopathological features, tumor heterogeneity and subtypes, metastatic activity, and gene expression profiles of canine MIBC mimic those of human MIBC. Herein, we successfully cultured MIBC organoids from both human and canine patients using first-catch urine. Briefly, after centrifuging urine samples to pellet cells, the cells were washed with basal culture media containing FBS and Pen/Strep before being cultured in canine or human culture media containing essential growth factors based on their species of origin. Paraffin-embedded samples were used for molecular characterization by RNA in situ hybridization or immunohistochemistry. Organoids from canine MIBC patients were analyzed for expression of the urothelial differentiation marker (FOXA1) associated with high grade bladder cancer, CD44, a non-kinase transmembrane glycoprotein expressed in UC cancer stem cells, and CK7, an epithelial cell marker that is upregulated in neoplastic UC cells. FOXA1 expression was undetectable in canine MIBC, although CD44 expression was enhanced compared to healthy dog bladder organoids. All canine MIBC organoids had significantly enhanced CK7 expression. Similarly, in human-derived UC organoids, GATA binding protein 3 (GATA3), transcription factor paired box 8 (Pax8), and tumor protein p63 were utilized as markers, with a Pax8-/p63+/GATA3+ profile indicating UC origin of the organoids. Functional response to chemotherapy (cisplatin and gemcitabine) was further assessed using cell viability assay in canine and human UC organoids. Our data suggest that organoids cultured from the urine of patients with MIBC can recapitulate organ structures and maintain marker expression profiles of the original tumor tissues. 3D organoids may provide novel insights to predict therapeutic response to chemotherapy and serve as early diagnostic biomarkers of bladder cancer in both canine and human patients. Citation Format: Dipak Kumar Sahoo, Allison P. Mosichuk, Fabrice Lucien, Igor Frank, John C. Cheville, Karin Allenspach, Jonathan P. Mochel. Urine-derived urinary carcinoma organoids: A novel tool for providing new insights into human and canine bladder cancer treatment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3092.

Context.— In women, radical cystectomy includes removal of the bladder, uterus, fallopian tubes, ovaries, and anterior vaginal wall, yet contiguous extension of urothelial carcinoma to all pelvic organs is rare and routine removal may be unnecessary. Objective.— To study pelvic organ involvement in women at radical cystectomy and investigate oncologic outcomes. Design.— Women with bladder cancer who underwent radical cystectomy at the Mayo Clinic and University of Toronto (1980–2018) were evaluated. Cancer-specific survival (CSS) was estimated with the Kaplan-Meier method; comparisons were made with the log-rank test. Associations with CSS were evaluated with Cox proportional hazard modeling. Results.— A total of 70 women with pT4a and 83 with pT3b cancer were studied. Organs involved were vagina (n = 41 of 70; 58.6%), uterus (n = 26 of 54; 48.1%), cervix (n = 15 of 54; 27.8%), fallopian tubes (n = 10 of 58; 17.2%), and ovaries (n = 7 of 58; 12.1%); 22 of 58 patients (37.9%) had >1 organ involved. Of 70 with pT4a cancer, 64 were available for survival analysis by 3 pelvic organ groups: vaginal only, vaginal and/or cervical/uterine, and vaginal and/or cervical/uterine and/or fallopian tubes/ovarian involvement. Three-year CSS for vaginal involvement only was 39%; it was 14% if cervical/uterine involvement, and <1% if fallopian tube/ovarian involvement was included (P = .02). Among 20 women with pT4aN0/Nx and vaginal involvement only, 3-year CSS for vaginal involvement was 50%, whereas among 48 women with pT3bN0/Nx cancer, 3-year CSS was 58%, P = .70. Conclusions.— Isolated vaginal involvement should be separated from uterine and/or adnexal extension of urothelial carcinoma at pathologic staging. Direct ovarian extension is rare and routine removal may be unnecessary.

Objectives. To compare the clinicopathologic characteristics of urothelial carcinoma (UC), urothelial carcinoma with squamous differentiation (UCSD), and squamous cell carcinoma (SCC) of the bladder, which have been suggested to differ in terms of risk factors, immunophenotype, and prognosis. Methods. We evaluated the clinicopathologic features of radical cystectomy specimens between 1980 and 2015 with a diagnosis of SCC, UCSD, and UC. PD-L1 immunohistochemistry (clinically available clones 22C3, SP142, and SP263) was performed on SCC and UCSD. Multivariate Cox regression was used to identify prognostic factors. Kaplan–Meier curves were plotted to assess cancer-specific survival (CSS). Results. Of the 1478 cases, there were 1126 UC (76%), 217 UCSD (15%), and 135 SCC (9%). Bladder cancer was more common in men than women (80% vs 20%, P < .0001). However, a higher proportion of SCC and UCSD occurred in women (SCC-36%, UCSD-22%, UC-18%). Women were significantly more likely to be never smokers in all 3 cohorts (UC: 45% vs 16%, UCSD: 44% vs 12%, SCC: 40% vs 18%, P < .0001). Patients with SCC and UCSD were at a higher pathologic stage (>pT2) at the time of cystectomy (UCSD-74%, SCC 71%, UC-44%, P < .0001) and had worse CSS compared to patients with UC ( P = 0.006). SCC had higher PD-L1 scores (all clones) than UCSD ( P < .0001). PD-L1 22C3 ( P = .02, HR: 0.36) and SP142 scores ( P = .046, HR: 0.27) predicted CSS on Kaplan–Meier analysis for SCC cases. Conclusions. UC, UCSD, and SCC are associated with different risk factors, gender distributions, and clinical outcomes. PD-L1 is expressed in SCC and UCSD, suggesting some patients may benefit from targeted therapy.

421 Background: Reported rates of adherence to post-radical cystectomy (RC) surveillance guidelines in real-world practice have been as low as 9%, in part reflecting a nihilistic view held by many of the value of routine follow-up. Indeed, conflicting data exist regarding the outcomes of patients with cancer recurrence detected by scheduled surveillance versus symptom-directed evaluation. Herein, we assessed comparative outcomes of patients with symptomatic recurrence (SR) versus asymptomatic recurrence (AR) after RC. Methods: We reviewed our Institutional Registry of RC patients to identify patients with cancer recurrence following RC. Presenting symptoms in the SR cohort included pain, constitutional symptoms, gastrointestinal symptoms, and hematuria/voiding symptoms, whereas AR was defined as recurrence detected on routine surveillance in the absence of symptoms. Baseline demographic and clinical characteristics were compared between study groups using chi-square and t-test. Kaplan-Meier and Cox survival analyses were performed to compare cancer-specific survival (CSS) and overall survival (OS) between AR and SR groups. Results: Of 3822 patients who underwent RC from 1980-2018 (with a median follow-up after RC of 2.4 years (IQR 1.1,5.5), a total of 1100 were subsequently diagnosed with recurrence, including 311 (28.3%) with AR and 789 (71.7%) with SR. Median time from RC to recurrence was longer in the AR group (13.2 months) than in the SR group (10.8 months; p = 0.01). Presenting symptoms included pain (70.2%), constitutional symptoms (50.7%), gastrointestinal symptoms (23.3%), and urinary symptoms (23.3%). Median follow-up after recurrence was 2.4 years (IQR 1.1-5.5), during which time 997 patients died, including 840 who died of bladder cancer. Compared to patients with SR, patients with AR had a longer median CSS (54.5 months vs 27.3 months, p < 0.001) and OS (43.0 months vs 25.8 months, p < 0.001). On multivariable Cox proportional hazards models adjusting for demographic and clinical factors, SR was associated with a significantly increased risk of cancer-specific (hazard ratio [HR] 1.66 [95% confidence interval 1.41-1.96], p < 0.0001) and all-cause mortality (HR 1.48 [1.23-1.71], p < 0.0001). Conclusions: SR after RC is associated with worse oncologic outcomes than post-RC recurrence detected by routine surveillance. As such, continued surveillance is warranted, while further study is needed to determine the optimal follow-up regimen balancing patient and disease risks.

416 Background: Urethral recurrence (UR) has been reported to occur in up to 5% of patients following radical cystectomy (RC). Debate continues regarding the utility of screening for UR after RC. Moreover, oncologic outcomes of patients with UR remain incompletely described, and reports have been limited by small cohort sizes. Herein, we evaluated risk factors for UR as well as cancer-specific survival (CSS) and overall survival (OS) among patients with UR. Methods: We reviewed our institutional RC registry to identify patients with UR following RC. Logistic regression was used to assess risk factors for UR. Kaplan-Meier and Cox proportional hazard regression were used to compare outcomes in those with UR diagnosis prompted by symptoms versus those with asymptomatic UR diagnosed on surveillance cytology. Results: Overall, 2930 patients underwent RC from 1980-2018, with a median post-operative follow-up of 4.9 years (IQR 1.6-11.0), of whom 144 (4.9%) were subsequently diagnosed with UR. Prostatic urethral involvement at RC (odds ratio [OR] 5.75 [3.67-9.01], p < 0.0001) and higher pT-stage (OR 3.57 [2.07-6.14], p < 0.0001) were associated with an increased risk of UR, whereas receipt of continent urinary diversion (OR 0.34 [0.20-0.58], p < 0.0001) was associated with a decreased risk of UR. A total of 72/144 (50%) patients were diagnosed with UR based on symptoms, and 104/144 (72.2%) patients with UR underwent subsequent urethrectomy. Patients with symptomatic recurrence had higher tumor stage on urethrectomy than those with asymptomatic recurrence (≥pT2 in 13.1% vs 3.1%, p = 0.007). At a median follow-up of 2.6 years (IQR 1.0-7.1) after UR, 68 of the patients with UR died of urothelial carcinoma. Kaplan-Meier analyses suggest longer median overall and cancer-specific survival for patients with UR detected by cytology than those presenting with symptoms (p = 0.05 for both). On multivariable analyses, patients who experienced UR had significantly increased risk of cancer-specific (hazard ratio [HR] 1.93 [95% confidence interval 1.50-2.50), p < 0.0001) and all-cause mortality (HR 1.34 [1.13-1.65], p = 0.001). Conclusions: Approximately 5% of patient undergoing RC experience UR, with higher pT-stage and prostatic urethral involvement increasing the risk of UR. Asymptomatic detection of UR was associated with lower pathologic stage at urethrectomy as well as longer cancer-specific and overall survival, supporting urethral surveillance after RC.

405 Background: The diagnosis of upper tract urothelial carcinoma (UTUC) is frequently obtained via diagnostic ureteroscopy (URS) with biopsy. In a recent meta-analysis of five retrospective series comparing UTUC RNU patients with and without prior diagnostic URS, prior URS was associated with intravesical recurrence of urothelial carcinoma. However, there remains an incomplete understanding of the risks of URS on intravesical recurrence after RNU, and existing studies have not examined the risks of URS without endoscopic biopsy or percutaneous biopsy alone. Here, we query our institutional RNU registry to determine if UTUC diagnostic modality influences post-RNU intravesical recurrence risk. Methods: Years 1995 to 2019 of our institutions RNU registry were queried. Patients with UTUC were divided into four groups: 1) URS with endoscopic biopsy; 2) URS without endoscopic biopsy (visual confirmation alone); 3) percutaneous biopsy without any URS; 4) no URS and no percutaneous biopsy. Exclusion criteria included pure non-urothelial histology, RNU for benign indications, and prior/concomitant cystectomy. The primary outcome was intravesical recurrence of urothelial carcinoma compared across the four groups using Kaplan-Meier log-rank analyses and Cox-proportional hazard modeling (hazard ratio = HR). Results: In a cohort of 878 patients (mean post-RNU follow-up 43 months), 461 (53%) had URS with biopsy, 130 (15%) had URS without biopsy, 229 (26%) had no URS or percutaneous biopsy, and 58 (7%) had percutaneous biopsy alone. The 3-year intravesical recurrence rate was 27%, 22%, 18%, 12% for groups 1-4, respectively. Covariates associated with intravesical recurrence on univariable analysis included age (HR 1.03, p < 0.01), female gender (HR 0.76, p = 0.06), history of bladder cancer (HR 1.39, p = 0.01), current smoker (HR 1.49, p = 0.04), and multifocality (HR 1.34, p = 0.03). After adjusting for these covariates, multivariable analysis found that group 1 (URS with biopsy) was associated with increased intravesical recurrence (HR 1.41, p = 0.03) relative to group 3 (no URS or percutaneous biopsy). Compared to group 1, the hazard ratio for intravesical recurrence was not significantly different for group 2 (HR 1.18, p = 0.45) or group 4 (HR 1.11, p = 0.79). Conclusions: Patients undergoing ureteroscopic biopsy to diagnose UTUC prior to RNU had a higher risk of intravesical recurrence after RNU compared to patients who did not undergo any URS or percutaneous biopsy. This agrees with prior literature and suggests that aggressive ureteroscopic manipulation of UTUC tumors via biopsy may promote shedding of urothelial carcinoma cells capable of seeding the bladder. Our study does support the initiation of a randomized trial to determine if intravesical chemotherapy after ureteroscopic biopsy can reduce intravesical urothelial carcinoma recurrences.