Search

4503 Background: Transurethral resection of bladder tumor (TURBT) plus systemic therapy has been known for decades to achieve durable bladder-intact survival in a subset of patients with MIBC but efforts to advance this paradigm have been complicated by (a) lack of prospective studies exclusively testing cisplatin-based neoadjuvant chemotherapy, (b) lack of rigorous methods to define clinical complete response (cCR) and its association with long term outcomes and (c) limited understanding of the role of “salvage” cystectomy. Methods: Eligible patients were cisplatin-eligible with cT2-T4aN0M0 urothelial bladder cancer. Patients received 4 cycles of gemcitabine, cisplatin, plus nivolumab followed by clinical restaging including urine cytology, MRI/CT of the bladder, cystoscopy and bladder/prostatic urethral biopsies. Patients achieving a cCR (normal cytology, imaging, and cT0/Ta) were eligible to proceed without cystectomy and receive nivolumab q2 weeks x 8 followed by surveillance; otherwise, patients underwent cystectomy. Coprimary endpoints included (1) cCR rate and (2) ability of cCR to predict 2-year metastasis-free survival (MFS). The key secondary endpoint was the impact of genomic alterations in baseline TURBT (TMB, ERCC2, FANCC, RB1, ATM) on performance of cCR for predicting MFS. The cCR rate coprimary endpoint, and interim analysis of 1-year outcomes, are reported. Results: Between 8/2018-11/2020, 76 patients were enrolled at 7 sites (male 79%, median age 69; cT2 = 56%, cT3 = 32%, cT4 = 12%) and 64 (84%) have completed post-cycle 4 restaging; 31/64 achieved a cCR (48%; 95% CI 36%, 61%). The median follow-up of cCR patients is 13.7 months (range, 2.5-24 months). One cCR patient opted for immediate cystectomy (pTaN0M0). Outcomes for the entire cohort are summarized in the table below. Local recurrence has occurred in 8/31 cCR patients and 6 underwent cystectomy (pT0N0 = 1, pTaN0 = 1, pTisN0 =1, pT2N0 = 2, pT4N1 = 1). TMB ≥ 10 mut/Mb (p=0.02) or mutant ERCC2 (p=0.02) were associated with cCR or pT0. Conclusions: TURBT + gemcitabine, cisplatin, plus nivolumab achieves stringently defined cCR in a large subset of patients with MIBC. 1-year bladder intact survival is possible though the durability of responses, and role of genomic biomarkers in management algorithms, requires longer follow-up. Clinical trial information: NCT03558087. [Table: see text]

480 Background: Immune related adverse events (IRAEs) with immune checkpoint inhibitor (ICI) therapy are well recognized, but predictors for IRAEs are not well defined. We aim to characterize the type, timing, and clinical risk factors associated with (w/) IRAEs in ICI-treated, advanced urothelial carcinoma (UC) and renal cell carcinoma (RCC) patients (pts). Methods: We retrospectively reviewed charts of pts w/ advanced UC and RCC who received at least 2 ICI doses at our institution from 1/1/10 to 10/31/18. Patient baseline characteristics, treatment course, and clinical outcomes were collected. IRAEs were identified and graded (GR) based on CTCAE (v.4.0). Fisher’s exact test was used to study the differences between pts w/ versus without IRAE. Results: Of the 71 pts identified (UC n = 53; RCC n = 18), 27 pts (38%) developed IRAEs with 42 total events (38% GR1, 60% GR2, and 2% GR≥3) [table]. The majority of pts with dermatitis (70%) also developed a secondary, systemic IRAE(s). Systemic steroid (SS) was required in 17 events. The median time to any IRAE was 17.5 weeks (w, range 1-93). ECOG ≤ 1 predicted IRAE development (p < 0.05). No other characteristics (demographics, co-morbidities, metastatic sites, ICI type, line of therapy, and duration of ICI > 12w) were associated with IRAE. Conclusions: In our study, good function status is associated with the development of IRAE. Time to IRAE ranged from immediately to 93w after initiating ICI. Clinical validation with additional datasets will be needed to confirm these findings. [Table: see text]

482 Background: Despite the widespread use of immune checkpoint inhibitors (ICIs), patterns of disease progression (POD) are poorly characterized. We aim to define these characteristics in patients (pts) with advanced urothelial carcinoma (UC) and renal cell carcinoma (RCC) treated with ICIs. Methods: We retrospectively reviewed charts of pts with advanced UC and RCC who received at least 2 ICI doses at our institution from 12/1/10 – 10/31/18. Demographics, medical history, ICI course, toxicity, and outcomes were recorded. Characteristics at the time of radiographic POD including location of metastases (mets), symptoms (sx), and hospitalization details were collected. Fisher’s exact test was used to study differences in pts with and without hospitalization at POD. Results: Of the 71 pts identified (UC N=53; RCC N=18), 59 pts had POD. At POD, 19 (32.2%) pts had new sites of disease involvement, while the remaining pts (N=40, 67.8%) had progression only at previously known sites of disease. Fourty-six (78.0%) pts had sx at POD: 1 sx (N=19, 32.2%), 2 sx (N=13, 22.0%), 3+ sx (N=14, 23.7%). Pain was the most common sx at POD (N=32, 54.2%), followed by loss of energy (N=18, 30.5%), and loss of appetite (N=14, 23.7%). Twenty-five (42.4%) pts were hospitalized at POD, most commonly for sepsis (N=8, 32%). No clinical factors were identified to predict for pts being hospitalized at POD. Conclusions: In our review of GU cancer patients on ICIs, a large proportion of pts reported clinical sx at POD, pain being the most frequent. Furthermore, a substantial number of pts were hospitalized at POD, most commonly for sepsis. Thus, further studies are warranted to confirm these findings, and potentially identify strategies to optimize patients’ quality-of-life and reduce rates of hospitalizations at the time of POD on ICIs.