Search

PURPOSE We assessed the safety and efficacy of cabozantinib and nivolumab (CaboNivo) and CaboNivo plus ipilimumab (CaboNivoIpi) in patients with metastatic urothelial carcinoma (mUC) and other genitourinary (GU) malignances. PATIENTS AND METHODS Patients received escalating doses of CaboNivo or CaboNivoIpi. The primary objective was to establish a recommended phase II dose (RP2D). Secondary objectives included objective response rate (ORR), progression-free survival (PFS), duration of response (DoR), and overall survival (OS). RESULTS Fifty-four patients were enrolled at eight dose levels with a median follow-up time of 44.6 months; data cutoff was January 20, 2020. Grade 3 or 4 treatment-related adverse events (AEs) occurred in 75% and 87% of patients treated with CaboNivo and CaboNivoIpi, respectively, and included fatigue (17% and 10%, respectively), diarrhea (4% and 7%, respectively), and hypertension (21% and 10%, respectively); grade 3 or 4 immune-related AEs included hepatitis (0% and 13%, respectively) and colitis (0% and 7%, respectively). The RP2D was cabozantinib 40 mg/d plus nivolumab 3 mg/kg for CaboNivo and cabozantinib 40 mg/d, nivolumab 3 mg/kg, and ipilimumab 1 mg/kg for CaboNivoIpi. ORR was 30.6% (95% CI, 20.0% to 47.5%) for all patients and 38.5% (95% CI, 13.9% to 68.4%) for patients with mUC. Median DoR was 21.0 months (95% CI, 5.4 to 24.1 months) for all patients and not reached for patients with mUC. Median PFS was 5.1 months (95% CI, 3.5 to 6.9 months) for all patients and 12.8 months (95% CI, 1.8 to 24.1 months) for patients with mUC. Median OS was 12.6 months (95% CI, 6.9 to 18.8 months) for all patients and 25.4 months (95% CI, 5.7 to 41.6 months) for patients with mUC. CONCLUSION CaboNivo and CaboNivoIpi demonstrated manageable toxicities with durable responses and encouraging survival in patients with mUC and other GU tumors. Multiple phase II and III trials are ongoing for these combinations.

PURPOSE Platinum-based chemotherapy for first-line treatment of metastatic urothelial cancer is typically administered for a fixed duration followed by observation until progression. “Switch maintenance” therapy with PD-1 blockade at the time of chemotherapy cessation may be attractive for mechanistic and pragmatic reasons. PATIENTS AND METHODS Patients with metastatic urothelial cancer achieving at least stable disease on first-line platinum-based chemotherapy were enrolled. Patients were randomly assigned double-blind 1:1 to switch maintenance pembrolizumab 200 mg intravenously once every 3 weeks versus placebo for up to 24 months. Patients with disease progression on placebo could cross over to pembrolizumab. The primary objective was to determine the progression-free survival. Secondary objectives included determining overall survival as well as treatment outcomes according to PD-L1 combined positive score (CPS). RESULTS Between December 2015 and November 2018, 108 patients were randomly assigned to pembrolizumab (n = 55) or placebo (n = 53). The objective response rate was 23% with pembrolizumab and 10% with placebo. Treatment-emergent grade 3-4 adverse events occurred in 59% receiving pembrolizumab and 38% of patients receiving placebo. Progression-free survival was significantly longer with maintenance pembrolizumab versus placebo (5.4 months [95% CI, 3.1 to 7.3 months] v 3.0 months [95% CI; 2.7 to 5.5 months]; hazard ratio, 0.65; log-rank P = .04; maximum efficiency robust test P = .039). Median overall survival was 22 months (95% CI, 12.9 months to not reached) with pembrolizumab and 18.7 months (95% CI, 11.4 months to not reached) with placebo. There was no significant interaction between PD-L1 CPS ≥ 10 and treatment arm for progression-free survival or overall survival. CONCLUSION Switch maintenance pembrolizumab leads to additional objective responses in patients achieving at least stable disease with first-line platinum-based chemotherapy and prolongs progression-free survival in patients with metastatic urothelial cancer.

467 Background: Neuroendocrine, small cell, or large cell carcinoma originating from the urothelium (uro-NE/SCC/LCC) is rare. Outcomes for metastatic disease are dismal. Treatment is extrapolated from small cell lung cancer, for which immune checkpoint inhibitors (ICIs) have modest activity. Preliminary activity has been reported with ICI for uro-NE. We aimed to compare real-world progression-free survival (PFS) and overall survival (OS) between ICI-containing and non-ICI-containing regimens in the first line (1L) metastatic setting for uro-NE/SCC/LCC. Methods: We performed a retrospective study at 11 cancer centers. Patients (pts) who received systemic therapy (2011-2021) for biopsy confirmed metastatic uro-NE/SCC/LCC were included. Pts with metastasis within 6 months of (neo)adjuvant chemotherapy (CT) (n=16) were excluded from 1L analyses. Results: 102 pts with metastatic uro-NE/SCC/LCC were evaluable. 17 (16.7%) had NE histology, 81 (79.4%) SCC, and 4 (3.9%) LCC. NE/SCC/LCC was mixed with urothelial histology in 19 (18.6%). Primary tumors were most often in the bladder (84.3%, n=86), less frequently upper tract (11.8%, n=12) or urethra (3.9%, n=4). 42 pts (41.2%) were previously treated for localized disease, the rest were de novo metastatic (n=60, 58.8%). Pts who received an ICI in any line (n=61) had significantly longer OS (p=0.038) than pts that never received an ICI (n=41). As shown in the table, in the 1L, ICI-containing regimens (n=33) resulted in significantly longer PFS, but not OS or ORR compared to non-ICI regimens (n=53). Subdividing 1L regimens into ICI without CT (n=14), CT without ICI (n=53), or ICI + CT (n=19), both PFS and OS were significantly different with similar ORR. ICI w/o CT had the longest median PFS and OS with an ORR 57.1% comparable to CT regimens. Of 61 pts that received ICI in any line, 14 (23.0%) had an immune-related adverse event of any grade; 11 (18.0%) received steroids. Conclusions: This is the largest ever report of ICI for metastatic uro-NE/SCC/LCC. ICIs were associated with improved outcomes with expected added toxicity. Further prospective investigation of ICI regimens is warranted. [Table: see text]

713 Background: The FOLH1 gene encodes prostate-specific membrane antigen (PSMA), a transmembrane glycoprotein that is highly expressed in prostate cancer cells and on endothelial cells in the neovasculature of solid tumors, including RCC. PSMA has been used as a target for diagnostic imaging and therapeutic radioligand therapy. We utilized a database of molecularly profiled RCC tumors to evaluate associations with FOLH1 expression. Methods: NextGen sequencing of DNA (592-gene/whole exome) and RNA (whole transcriptome) was performed for RCC patient specimens (n=1765) through Caris Life Sciences (Phoenix, AZ). FOLH1-High/Low expression were defined as ≥75th/

521 Background: Sacituzumab govitecan (SG) demonstrated an objective response rate (ORR) of 27% and median overall survival (OS) of 10.5 months (Mo) in metastatic urothelial carcinoma (mUC) patients (pts) progressing after platinum-based chemotherapy and PD1/L1 inhibitor, which led to accelerated US FDA approval in this setting. The combination of SG and pembrolizumab is safe and active following platinum-based chemotherapy. Nivolumab (NIVO) 1mg/kg plus Ipilimumab (IPI) 3mg/kg has shown promising activity in post-platinum mUC pts. Given the potential synergism between immunogenic cell death induced by SG and IPI-NIVO, we hypothesized that the combination of SG and IPI-NIVO would be safe and active as a frontline treatment for cisplatin ineligible mUC. Methods: 3+ 3 design was used for the phase I dose escalation of SG at 8 mg/kg and 10 mg/kg dose levels. IPI and NIVO were given at 3mg/kg and1mg/kg (I3+N1) intravenously (IV) every 3 weeks x 4 cycles followed by NIVO 360 mg IV day 1 every 3 weeks. SG was given IV at days 1,8 every 3 weeks The primary endpoint was safety and recommended phase 2 dose (RP2D) based on dose limiting toxicity (DLTs) observed in cycle 1; key secondary endpoints include ORR, DOR, PFS and OS. Key inclusion criteria were ECOG-PS 0-1, cisplatin-ineligibility, treatment naïve, no prior PD1/L1 inhibitor except >3 months earlier for non-metastatic disease. Results: The study has completed the phase I dose escalation after enrolling a total of 9 patients (8 men, 1-woman, median age: 74 years). 6 patients were enrolled at SG 8 mg/kg with 1 DLT, and 3 patients at 10 mg/kg with 2 DLTs. DLTs included grade 3 skin rash (n=2) and grade 3 pneumonitis (n=1). The RP2D of SG was determined to be 8 mg/kg with I3+N1. The most common treatment related adverse events (TRAE) included anemia (66.6%) neutropenia (66.6%), pruritus (66.6%), fatigue (66.6%), diarrhea (66.6%) and lymphopenia (55.5%). 2 patients developed grade 2 infusion reactions to SG. Other grade ≥ 3 TRAE included neutropenia (55.5%), anemia (33.3%), arthralgia (11.1%), and elevated amylase/lipase (11.1%). Both grade 2 pneumonitis and myositis were seen in 1 patient. Of the 9 pts, 6 pts (4 pts at SG dose of 8mg/kg and 2 pts at 10 mg/kg) were considered evaluable for response of whom 4 responded (ORR 66.6%) with 1 complete response and 3 partial responses. Median DOR was 9.2 Mo (range 4.6-12.0); mPFS was 8.8 Mo (95% CI 3.8-NR) and mOS was not reached. Conclusions: The RP2D of SG was identified as 8mg/kg in combination with Ipilimumab 3 mg/kg+ Nivolumab 1 mg/kg as first-line therapy for cisplatin-ineligible mUC. Early signals of promising activity were observed in a small cohort of evaluable pts. The Phase 2 trial is ongoing coupled with exploratory biomarker analyses. Clinical trial information: NCT04863885 .

650 Background: Management of metastatic renal cell carcinoma (mRCC) is highly individualized and often involves cytoreductive nephrectomy (CN) and systemic therapy (ST). The optimal sequencing of CN and ST is uncertain. A difference in perioperative outcomes based on sequence of CN and ST could influence decision-making. We conducted this NSQIP analysis to assess whether preoperative systemic therapy adversely impacted perioperative outcomes in patients receiving deferred CN. Methods: This analysis was conducted using the American College of Surgeons NSQIP Participant Use Data File for years 2019 and 2020. These years were selected because data on receipt of preoperative therapy is only available since 2019. Inclusion criteria were i) CPT code consistent with nephrectomy, ii) urologist operating surgeon & iii) presence of disseminated cancer. All cases with ICD-10 diagnosis codes not consistent with mRCC were excluded. Groups were stratified by their receipt of preoperative systemic therapy within 90 days before CN and we assessed 46 preoperative and perioperative outcomes. Results: The study cohort included 505 patients with 115 (23%) who received preoperative ST. No differences were noted in perioperative outcomes (Table). Patients receiving preoperative ST were more likely to be on steroids (23% vs. 7%, p0.05). Conclusions: Because preoperative ST did not have an appreciable impact on deferred CN perioperative outcomes, decision making for ST and CN sequencing should not be influenced by perioperative outcomes. Those who undergo deferred CN are unlikely to experience delayed time to surgery or perioperative complications from their ST. [Table: see text]

703 Background: Bone-predominant metastatic renal cell carcinoma (mRCC) is associated with aggressive biology and poor prognosis. The optimal management of these patients (pts) remains not well-established. While contemporary trials showed the superiority of immune checkpoint inhibitors (ICI) compared to vascular endothelial growth factor tyrosine kinase inhibitors (TKI), these pts were often excluded due to non-measurable disease. mRCC bone metastases show an enhanced angiogenesis gene signature suggesting the potential for greater responses to TKI. We test this hypothesis in a retrospective study comparing treatment outcomes for the first line (1L) TKI versus ICI in pts with bone-predominant mRCC. Methods: Pts with mRCC who received care at the Ohio State University Comprehensive Cancer Center from 1/1/2008 to 6/1/2021 were identified through retrospective chart review. Bone-predominant metastasis was defined as the number of bone metastases > extra-osseous metastases. Pts who had ≥1 cycle of treatment and follow-up scans were evaluated for treatment responses per RECIST 1.1 criteria. Progression-free survival (PFS) and overall survival (OS) were estimated by the Kaplan-Meier method. The comparisons were made between TKI and ICI cohorts, and the significance was determined by logrank test. Pt who received ICI-TKI in 1L were excluded from the survival analysis due to limited N. Results: A total of 335 pts with mRCC were identified, of which 45 (13.4%) had bone-predominant metastasis. Most of the pts (44/45) had clear cell histology. TKI was the 1L intervention for 21 (47%) while 20 (44%) pts received ICI, and 4 (9%) pts received TKI-ICI. TKI and ICI cohorts were similar for International Metastatic RCC Database Consortium (IMDC) Risk (Table). In the TKI cohort, 19 pts had evaluable responses with ORR of 15.7% and SD of 42%. The mPFS was 5.9 months (95% CI 2.8–9), and mOS was 30.4 months (95% CI 3.3–57.6). In the ICI cohort, 16 had evaluable responses with ORR of 6% and SD of 43.8%. The mPFS was 3.4 months (95% CI 0–8.2), and mOS was 19.3 months (95% CI 14.9–23.7). The hazard ratio was 0.58 for PFS and 0.54 for OS, favoring TKI over ICI, although statistical significance was not reached due to the small sample size (p=0.10 and 0.12, respectively). Conclusions: In this single-center retrospective study, pts with bone-predominant mRCC who received TKI compared to ICI as 1L therapy showed improved ORR, PFS, and OS. These results warrant further investigation in larger studies comparing TKI-containing regimen vs ICI doublet. [Table: see text]

PURPOSE Patients with metastatic urothelial carcinoma are often ineligible for cisplatin-based treatments. A National Cancer Institute Cancer Therapy Evaluation Program–sponsored trial assessed the tolerability and efficacy of a gemcitabine-eribulin combination in this population. METHODS Patients with treatment-naïve advanced or recurrent metastatic urothelial carcinoma of the bladder, ureter, or urethra not amenable to curative surgery and not candidates for cisplatin-based therapy were eligible. Cisplatin ineligibility was defined as creatinine clearance less than 60 mL/min (but ≥ 30 mL/min), grade 2 neuropathy, or grade 2 hearing loss. Treatment was gemcitabine 1,000 mg/m2 intravenously followed by eribulin 1.4 mg/m2, both on days 1 and 8, repeated in 21-day cycles until progression or unacceptable toxicity. A Simon two-stage phase II trial design was used to distinguish between Response Evaluation Criteria in Solid Tumors, version 1.1 objective response rates of 20% versus 50%. RESULTS Between June 2015 and March 2017, 24 eligible patients with a median age of 73 years (range, 62 to 88 years) underwent therapy. Performance status of 0, 1, or 2 was seen in 11, 11, and two patients, respectively. Sites of disease included: lymph nodes, 16; lungs, nine; liver, seven; bladder, five; bones, two. Median number of cycles received was four (range, one to 16). Of 24 patients, 12 were confirmed responders; the observed objective response rate was 50% (95% CI, 29% to 71%). Median overall survival was 11.9 months (95% CI, 5.6 to 20.4 months), and median progression-free survival was 5.3 months (95% CI, 4.5 to 6.7 months). The most common treatment-related any-grade toxicities were fatigue (83% of patients), neutropenia (79%), anemia (63%), alopecia (50%), elevated AST (50%), and constipation, nausea, and thrombocytopenia (42% each). CONCLUSION Gemcitabine-eribulin treatment response and survival for cisplatin-ineligible patients compare favorably to other regimens. Additional research is needed.

4584 Background: We previously reported the clinical outcomes of a positive multi-center phase II trial of neoadjuvant gemcitabine (G) and cisplatin (C) plus atezolizumab (A) in patients with muscle-invasive bladder cancer (Funt, et al. JCO 2022). In this and another trial of neoadjuvant GC with pembrolizumab (Rose et al, JCO 2021), PD-L1 positivity by immunohistochemistry was not predictive of non–muscle-invasive downstaging ( < pT2N0). Therefore, we investigated other pre-treatment tissue-based genomic and gene expression biomarkers of response and resistance. Methods: 36 pts had pre-treatment tissue available for genomic analysis. We performed targeted hybridization capture DNA sequencing using the CLIA-certified MSK-IMPACT platform and whole transcriptome RNA sequencing. We examined genomic and gene expression biomarkers which have been previously investigated in the context of neoadjuvant cisplatin-based chemotherapy or anti-PD-1/L1 immunotherapy for MIBC, including tumor mutation burden (TMB), a DNA damage response (DDR) 9-gene panel (NCT03609216) associated with response to neoadjuvant chemotherapy, and an 8-gene cytotoxic T cell transcriptional signature associated with response to neoadjuvant A (tGE8; Powles et al, Nature Medicine 2019). We also evaluated TGF-β pathway activation, which was associated with resistance to A in pts with metastatic BC (Mariathasan et al, Nature 2018). Putative biomarkers were assessed for correlation with < pT2N0, the trial’s primary endpoint. Results: DNA was available from all 36 pts, and RNA met quality control metrics for 29 pts. TMB was significantly higher in pts with < pT2N0 (median 16 mut/Mb, IQR 12-25) versus ≥ pT2N0 (median 10 mut/ Mb, IQR 8-10; p < 0.01). A single patient had a TMB > 200 Mut/Mb with a POLE hotspot mutation and achieved pT0N0; TMB was still significantly higher in responders after omission of this patient (p < 0.01). Nine of 25 pts (36%) with < pT2N0 had a deleterious DDR mutation versus 1 of 10 pts (10%) with ≥ pT2N0 (p = 0.13). While tGE8 was significantly increased in patients with < pT2N0 compared to those without (p = 0.01), TGF-β pathway activation was not increased in pts with ≥ pT2N0 (p = 0.99). Conclusions: TMB and the tGE8 cytotoxic T cell transcriptional signature were associated with response to combination GC+A in muscle-invasive bladder cancer. More detailed molecular analyses will be reported. Clinical trial information: NCT02989584.

e17022 Background: De novo SCPC and tNEPC in metastatic castrate resistant prostate cancer (mCRPC) are rare, aggressive cancers with a poor prognosis. After first-line (1L) platinum chemotherapy (PLT), there is no consensus on 2L treatments. This multi-institutional, retrospective study examines practice patterns among clinicians and evaluates the outcomes of patients (pts) with SCPC or tNEPC in the 2L setting. Methods: After IRB approval, pts with a pathologic diagnosis of SCPC or tNEPC (any % of SC or NE histology defined by institutional review) were identified. Pts diagnosed between 2000-2020 who received 1L PLT and any 2L systemic therapy (tx) for SCPC or tNEPC were included. Standardized data collection templates containing demographic, clinical, and pathologic variables were collected. The primary endpoint was objective response rate (ORR) to 2L tx. Secondary endpoints included PSA response, time on 2L treatment, and overall survival (OS) from time of 2L tx. Data was analyzed using descriptive statistics. Results: Forty-two pts (21 SCPC, 21 tNEPC) from 6 institutions were included. At SCPC/tNEPC diagnosis, the overall cohort had a median age of 65.0 years (IQR 60.5, 68.0) and median PSA of 3.0 ng/dL (IQR 0.5, 21.8). The most common sites of metastasis included lymph node (78.6%), bone (54.8%), and liver (52.4%). For 1L tx, 37 pts (88.1%) received PLT and 5 (11.9%) had PLT + immunotherapy (IMM). For treatment to the prostate, 21 (51.2%) had none, 7 (17.1%) radiation (RT), 9 (22.0%) surgery, 3 (7.3%) RT + surgery, and 1 (2.4%) brachytherapy alone. Concurrent ADT was given to 32 pts (76.2%) in the 1L and 2L. At last follow-up, 34 pts (81.0%) were dead, 4 (9.5%) were alive, and 4 (9.5%) were lost to follow-up/censored. For 2L tx, 10 pts (23.8%) received PLT, 8 (19.0%) taxane monotherapy (TAX), 10 (23.8%) IMM, 8 (19.0%) other chemotherapy (CHX), and 6 (14.3%) other tx. Among 38 pts evaluable for response, the ORR was 15.8% (6 pts with PR: 4 PLT, 1 IMM, 1 CHX). PSA response ≥ 50% to 2L tx was seen in 5 of 29 pts with PSA data (17.2%). Other outcomes data are reported in Table. Conclusions: In this retrospective study, pts with SCPC or tNEPC who reached 2L tx received a wide variety of treatment regimens, reflecting the lack of consensus in this tx setting. ORR was low and overall prognosis was poor in the 2L regardless of tx choice. Pts on 2L immunotherapy seemed to have the shortest mOS. [Table: see text]

569 Background: Patients with metastatic genitourinary malignancies with neuroendocrine have limited therapeutic options following platinum therapy. Given encouraging results in initial cohort analysis for small cell urinary tract carcinoma, a cohort of any genitourinary malignancy with neuroendocrine differentiation was added to a multicenter, single arm, multi-cohort phase II trial to evaluate the efficacy of N and I in this setting. (NCT 03333616) Methods: Eligible patients had metastatic or locally advanced genitourinary malignancy with neuroendocrine differentiation with an ECOG performance status of 0-2; they may have received any line of prior therapy excluding prior immunotherapy. Patients underwent baseline biopsy and received treatment with N 3 mg/kg and I 1 mg/kg intravenously every 3 weeks for 4 cycles with continued maintenance of N 480 mg IV every 4 weeks. Imaging was performed at 12 weeks and then every 8 weeks through month 6 and then every 12 weeks thereafter. The primary endpoint was investigator assessed objective response rate (ORR) by RECIST 1.1. Results: A total of 27 patients were enrolled between 06/27/2018 and 06/21/2021, 10 (37%) had urinary tract cancer and 17 (63%) had prostate cancer (19 in expansion cohort, 3 urinary tract and 5 prostate cancer from earlier cohorts). The majority (n=25, 93%) patients received prior systemic therapy. Nine (33%) patients received all 4 doses of N and I during the induction period. Nine (33%) patients (7 of whom received 4 cycles N+I) received N maintenance (median number of cycles 9 (range, 2-37)). Median follow-up was 6.8 (range, 0.9-37.3) months. Objective response was achieved in 8 (30%, 80% CI 18%-44%) patients (Table). Median duration of response was not reached with 4 patients maintaining response >9 months. Median progression-free survival time was 2.6 (95% CI 1.8-6.5) months At time of analysis, 13 (48%) death events were reported due to progressive disease, in which 3 were bladder and 10 were prostate cancer. 8 (30%) patients developed treatment-related grade 3 or higher toxicities; one grade 5 toxicity was deemed treatment-unrelated. Conclusions: In this study we demonstrate N+I resulted in objective responses in patients with genitourinary malignancy with neuroendocrine differentiation. ORR of 50% in small cell carcinoma is bladder cancer is noteworthy and will be evaluated further in ongoing expansion cohort of bladder or upper tract carcinoma with variant histology. Clinical trial information: NCT03333616. [Table: see text]

452 Background: This study determined the association of functional imaging parameters obtained on FDG PET/CT and NaF PET/CT with OS for pts with mGU malignancies treated on a phase I study with CaboNivo +/- Ipi. Methods: Pts on this phase I study underwent sequential (1-hour apart) FDG PET/CT and NaF PET/CT imaging at baseline and at first-restaging (8 weeks follow up). Scan semi-quantitative parameters measures included: maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) for FDG and MTV for NaF. Total lesion number was captured for all scans. The association of imaging parameters and survival was determined with Kaplan-Meier curves. Baseline values and percent change values were calculated. Results: 81 pts were included in the analysis. 67 (83%) were males; Median age was 63 (range 25-86); Histologically, 30 pts had urothelial carcinoma, 15 clear cell renal cell carcinoma, 9 germ cell tumors, 8 urachal/adenocarcinoma, 8 prostate cancer, 3 penile cancer, 3 squamous cell carcinoma, 3 renal medullary carcinoma, and 2 small cell (1 bladder, 1 prostate). All 81 had a baseline FDG PET scan, 78 pts received baseline NaF PET scans; 66 received both FDG PET and NaF PET baseline and follow up scans. 957 total lesions were detected on FDG PET across all histologies, 87 liver (9%), 252 lung (26%), 152 bone (16%), 411 lymph node (43%), and 55 other visceral metastases (6%). 414 total lesions were detected on NaF imaging. Low vs high baseline FDG MTV (31 vs 11 months, p = 0.0002), TLG (30 vs 11 months, p = 0.0004), Lesion number (49 vs 15 months, p = 0.0005), and SUVmax (25 vs 12 months, p = 0.025), FDG lesion number decrease or no change vs increase (24 vs 12 months, p = 0.0068), and low vs high baseline NaF MTV (26 vs 16 months, p = 0.007), and lesion number (26 vs 16 months, p = 0.007) showed the strongest associations with OS. A multivariable Cox analysis demonstrated that baseline FDG MTV (HR = 2.87, 95% CI 1.62-5.08, p = 0.0003) and FDG lesion number percent change (HR = 2.71, 95% CI 1.40-5.24, p = 0.0031) were jointly associated with OS. Conclusions: Baseline functional imaging parameters and percent change seen on follow imaging with FDG PET and NaF PET are prognostic in mGU pts treated with CaboNivo +/- ipi. Additional parameters and histologic subsets will be presented.

e16567 Background: ICI/TKI combinations are a new standard of care for the initial treatment (tx) of mRCC. Efficacy and toxicity of such combination regimens beyond the first-line (1L) setting remain unknown. Methods: We retrospectively reviewed charts for adult patients (pts) receiving an ICI/TKI combination in any line of tx for mRCC of any histology at one of two academic centers as of May 1, 2020. ICIs included pembrolizumab (Pm), nivolumab (Ni), ipilimumab (Ip), or avelumab (Av); TKIs included sunitinib (Su), axitinib (Ax), pazopanib (Pz), lenvatinib (Ln), or cabozantinib (Ca). Clinical data including pt demographics, histology, International mRCC Database Consortium (IMDC) risk group, tx history, and ICI/TKI tx and toxicity details were recorded. Outcomes included objective response rate (ORR), median progression-free survival (mPFS), and safety, analyzed via descriptive statistics and the Kaplan-Meier method. Results: Of 85 pts, 69 (81%) were male and 67 (79%) had clear cell histology. IMDC risk was favorable (24%), intermediate (54%), poor (20%), and unknown (2%). 39% had ICI/TKI tx in the 1L setting. ICI/TKI regimens included Pm/Ax (33%), Ni/Ca (25%), Ni/Ax (20%), Av/Ax (11%), Ni/Ip/Ca (8%), Ni/Su (2%), and Ni/Ln (1%). ORR and mPFS stratified by line of tx and prior tx are shown in the table. Of 52 pts who received ICI/TKI tx as salvage (after 1L), 52% had a grade 3 or higher (≥G3) adverse event (AE), of which the most common were anorexia (13.5%), diarrhea and hypertension (11.5% each), and fatigue (9.6%). 65% of pts on salvage ICI/TKI tx stopped tx for progression/death, while 16% stopped tx for ≥G3 AE. ≥G3 AE rates by line of tx were 62.5% (2L), 50% (3L), and 45% (≥4L). Conclusions: ICI/TKI combination therapy is effective and safe beyond the 1L setting. Prior tx history appears to impact efficacy but has less of an effect on safety/tolerability. These observations will need to be confirmed in prospective studies.[Table: see text]

1025 Background: Analysis of circulating tumor DNA (ctDNA) over time allows non-invasive evaluation of tumor genomic evolution. We characterize changes in tumor fraction (TFx), somatic copy number alterations (SCNAs), and somatic mutations (muts) over time in patients (pts) with BRCA1/2 muts and metastatic breast cancer (mBC) who received a PARP inhibitor (PARPi) or platinum chemotherapy. Specifically, we seek to identify the frequency of BRCA1/2 reversion muts. Methods: Pts with mBC and germline or somatic BRCA1/2 muts were identified on a banking protocol of prospectively-collected serial samples of blood and plasma. Control pts without a BRCA1/2 mut were matched 2:1 by age and hormone receptor (HR) status. Ultra-low-pass whole genome sequencing (ULPWGS) with 0.1x depth was performed on all plasma samples (n = 103) and the ichorCNA algorithm was used to determine TFx and SCNAs. Targeted panel sequencing (TPS) of 402 cancer-related genes was performed at 10,000x depth on plasma samples, and one blood sample per pt. The panel includes BRCA1/2 and 38 other DNA damage repair (DDR) genes. Somatic muts were identified by joint calling with Mutect2 across plasma timepoints with paired pt normal blood. Germline variant calling from TPS on blood with HaplotypeCaller was used to confirm germline muts in BRCA1/2.Results: We identified 10 pts with mBC with a germline (n = 7) or somatic (n = 3) BRCA1 (n = 2) or BRCA2 (n = 8) mut and banked blood and plasma samples at 3-9 timepoints at a median of 8 weeks apart (range 1-43). The control cohort of 20 pts with mBC and wildtype BRCA1/2 was well matched by age and HR status. All pts with BRCA1/2 muts received a PARPi and/or platinum chemotherapy at some point during sample collection. Half of control pts received platinum chemotherapy. Germline BRCA1/2 muts were confirmed in all 7 pts with known germline muts. Among the BRCA1/2 mut cohort, median TFx was 0.04 (range 0-0.57) with 20% of samples having TFx > 0.10. A median of 1.5 (range 0-39) somatic muts per pt were found in DDR genes. Four pts (40%) had secondary non-reversion muts in BRCA1/2. A reversion mut of a germline BRCA2 mut, restoring the open reading frame of BRCA2, was discovered at the last timepoint from 1 pt while receiving carboplatin. A germline BRCA1/2 reversion mut in this cohort occurred in 2.3% of samples, 14.3% of pts. There was no significant difference in the percent of genome with a SCNA between the first and last time point, nor before and after PARPi/platinum. The somatic mut landscape and clonal evolution of TPS using PyClone will be presented. Conclusions: Evaluation of serial ctDNA samples for TFx, SCNAs, and somatic muts from banked plasma and blood from pts with mBC is feasible. The frequency of reversion muts in BRCA1/2 was low, suggesting that either their incidence is low or ctDNA TPS is not sensitive enough to detect them. Secondary non-reversion muts in BRCA1/2 and other somatic DDR muts were more common. SCNAs were stable over time.

4507 Background: Cisplatin with gemcitabine (CG) remains the standard upfront chemotherapy regimen for metastatic urothelial cancer (mUC). Preclinical synergy was noted between cisplatin and berzosertib, a selective ATR inhibitor. The current study sought to determine if the combination of berzosertib and CG could improve clinical outcomes in mUC. Methods: An open-label, randomized study was conducted across 23 centers in the United States through the Experimental Therapeutics Clinical Trials Network of the National Cancer Institute. Key eligibility criteria included confirmed mUC, no prior cytotoxic therapy for metastatic disease, ≥ 12 months since perioperative therapy and eligibility for cisplatin based on standard criteria. Patients (pts) were randomized to receive either CG alone (control arm) or CG plus berzosertib (experimental arm). In the control arm, 70 mg/m2 of cisplatin was given on day 1 and gemcitabine at 1000 mg/m2 on days 1 and 8 of a 21-day cycle. In the experimental arm, 60 mg/m2 of cisplatin was given on day 1, gemcitabine at 875 mg/m2 on days 1 and 8 and berzosertib at 90 mg/m2 on days 2 and 9 of a 21-day cycle. The primary endpoint of the study was progression-free survival (PFS), with secondary endpoints including response rate (RR), overall survival (OS) and toxicity. Results: A total of 87 pts (median age 67; M:F 68:19) were randomized; 41 pts received CG alone while 46 received CG with berzosertib. Visceral metastases were present in 49% of pts and 52%, 45% and 3% of pts were Bajorin risk 0, 1 and 2, respectively. Median PFS was 8.0 months for both arms (Bajorin risk adjusted hazard ratio [HR] 1.22, 95% confidence interval [CI] 0.72-2.08). RR was 54%(4 CR, 21 PR) in the CG with berzosertib arm and 63% (4 CR, 22 PR) in CG alone arm (P = 0.66). Median OS was shorter with CG with berzosertib as compared to CG alone (14.4 versus 19.8 months; Bajorin risk adjusted HR 1.42, 95%CI 0.76-2.68). Notably higher rates of grade 3/4 thrombocytopenia (59% vs 39%) and neutropenia (37% vs 27%) were observed with CG plus berzosertib compared to CG alone. Higher rates of toxicity-related discontinuation were seen in the experimental arm (24% vs 15%), and the median cumulative cisplatin dose in the experimental arm was 250 mg/m2, as compared to 370 mg/m2 in the control arm (P < 0.001). Conclusions: No improvement in PFS was observed with the addition of berzosertib to CG, and a trend towards inferior survival was observed. These results suggest caution in reducing the starting dose of cytotoxic therapy to accommodate addition of a myelosuppressive agent, as in the experimental arm of this study. Clinical trial information: NCT02567409.

4557 Background: Metastatic RCC (mRCC) involving the pancreas is distinct from RCC involving other metastatic sites and is characterized by an indolent clinical course, heightened angiogenesis, and an inflamed stroma (PMID: 32271170). We previously reported on outcomes of RCC patients (pts) with pancreatic oligometastasis (ASCO GU 2020). We now report on outcomes in pts with mRCC involving the pancreas in conjunction with other metastases (mets). Methods: We conducted a retrospective, multi-institutional study of mRCC pts with mets to the pancreas and other sites. Data on pt demographics, tumor characteristics, systemic therapy, and outcomes were collected. Pts were classified based on treatment category: immunotherapy (IO) or vascular endothelial growth factor/receptor inhibitors (VEGFI). Outcomes measured included objective response rates (ORR), time-on-treatment (TOT), and overall survival (OS). Results: The analysis included 229 pts from 9 institutions, diagnosed between 1985-2020. Of these, 211 (92%) had clear-cell histology; 131 (57%) had nephrectomy; 41 (18%) had local pancreas-directed therapy; 111 (48%) had synchronous presentation of disease in the pancreas and other sites at time of mets. IMDC risk was favorable in 33%, intermediate in 41%, poor in 11%, and unknown in 15% pts. Median lines of therapy was 2 (range 0-9). Of 219 pts who received first-line (1L) therapy, 151 (69%) had VEGFI therapy, 41 (19%) had IO, and 18 (8%) had VEGFI/IO combination (Table). The IO group included 21 pts on checkpoint inhibitor (CPI), 16 pts on HD-IL2, 4 pts on other IO. 1L ORR was 39.7% for VEGFI (95% CI 31.8-48.0) and 31.7% for IO (95% CI 18.1-48.1) and was not statistically significant (NS, OR 1.4, 95% CI 0.65-3.23, p = 0.371). Median TOT for 1L therapy was 11.6m for VEGFI and 6.5m for IO (p = 0.0106). With a median follow-up of 51.5m, the median OS (mOS) for all pts from time of metastatic disease was 7.7 years (y) (95% CI 6.3-10.3). The mOS for pts who received 1L VEGFI was 7.6y (95% CI 5.5-9.5) and was not reached (NR) for those who got 1L IO (95%CI 6.5-NR); this difference was significant with an unadjusted p-value of 0.029. The pair-wise comparison between mOS of the 1L CPI subgroup compared to that of the 1L VEGFI group was significant (p = 0.0148). Conclusions: Consistent with the literature, mRCC pts with involvement of the pancreas in this study have prolonged OS compared to historical OS for the standard mRCC population. Additionally, our findings suggest that the choice of first-line therapy may impact outcomes. Additional analyses will be presented.[Table: see text]

3 Background: CaboNivo and CaboNivoIpi showed promising efficacy and safety in a dose-escalation phase I study in patients (pts) with metastatic genitourinary (mGU) tumors. We now report the final results from a pooled analysis of the phase I dose-finding and 7 subsequent expansion cohorts. Methods: Pts with mGU tumors in the phase I cohort received 8 escalating doses of CaboNivo or CaboNivoIpi. In the 7 expansion cohorts, pts received the recommended phase II dose for CaboNivo (cabo 40mg qd + nivo 3mg/kg q2wks in 28-day cycles) and for CaboNivoIpi (CaboNivo same dose + ipi 1mg/kg q3wks in 21-day cycles x 4 cycles followed by CaboNivo). The CaboNivo expansion cohorts included pts with urothelial carcinoma (UC); clear cell renal cell carcinoma (RCC), bladder adenocarcinoma (BlaAdeno), and other rare mGU tumors. The CaboNivoIpi expansion cohorts included, UC, RCC, and penile carcinoma (penile). The objectives of the study were to determine the clinical activity, safety and tolerability of both combinations. A secondary objective was the detection of EpCAM+ circulating tumor cells (CTCs). Biomarker correlatives of plasma VEGFA/VEGFR2, PIGF, and inflammatory cytokines will be presented. Results: A total of 120 pts (median age 59; range 20-82) were enrolled: 54 in the phase I and 66 in the dose expansion cohorts. 64 pts received CaboNivo and 56 CaboNivoIpi. Median follow-up was 40.4 months (range 2.2-62.2 months). The ORR for 108 evaluable pts was 38% (95% CI: 28.8-47.8%) with 12 complete responses (CRs) (11.1%) and 29 partial responses (26.9%). The ORRs for the following mGU tumors were: UC 42.4% (n=33) with CR=21.2%; RCC 62.5% (n=16); prostate cancer 11.1% (n=9); germ cell tumor no responses (n=6); BlaAdeno 20% (n=15); penile 44.4% (n=9); bladder squamous 85.7 (n=7); renal medullary 50% (n=2); bladder small cell 33.1 (n=3). The median overall survival for the entire population was 15.9 months (95% CI: 11.6-23.9); 24.9 months (95% CI: 11.8-41.6) for pts with UC (n=39); and 38.6 months (95% CI: 19.4-not estimable) for RCC (n=16). Median duration of response was 22.8 months (95% CI: 18.3-40.1 months) for all pts, 32.1 months [95% CI: 20.3-NE)] for the UC pts and 20.1 months (95% CI: 5.8-NE) for RCC pts. Grade 3 or 4 treatment related adverse events (AEs) occurred in 84% and 80% of pts treated with CaboNivo and CaboNivoIpi pts respectively, and included hypophosphatemia (25% and 16%), lipase elevation (20% and 20%), fatigue (20% and 18%), ALT elevation (5% and 14%), AST elevation (9% and 11%), diarrhea (9% and 11%), and thromboembolic event (11% and 4%). One pt had Grade 5 pneumonitis on CaboNivoIpi. Baseline EpCAM+ CTC count of < 5 vs. > 5, was associated with longer median OS (24.3 vs. 12.3 months p=0.037). Conclusions: CaboNivo and CaboNivoIpi demonstrated promising clinical activity and manageable safety in many mGU histologies including rare tumors. Clinical trial information: NCT02496208.

5037 Background: Previous treatment with ICI is more common in clinical practice since recent FDA-approval of 5 ICIs in second-line and 2 in first-line for mUC. There is lack of data regarding the use of ICI after progression on a prior ICI. Cabozantinib has been shown to have immunomodulatory properties and may have synergistic effect with ICI. Methods: This is a phase I expansion cohort of mUC pts, who received prior ICI, treated with Cabozantinib 40mg daily and Nivolumab 3mg/kg every 2 weeks until disease progression/unacceptable toxicity. The primary objective was to determine the efficacy and tolerability of CaboNivo. Results: Twenty-nine mUC pts were treated. Median follow-up was 14.1 months (mo). The majority of pts were male (75.8%); 27 were White (93.1%), and 2 were Asian (6.9%). Primary tumor was bladder in 21 pts (72.4%) and upper tract in 8 (27.6%). Twenty-two pts (75.9%) had visceral metastasis (mets), 4 (13.8%) had lymph node only mets and 13 (44.8%) had liver mets. The median number of prior lines of treatment for mUC was 2 (range 0-8) with 17 pts (58.6%) receiving 2 prior lines of treatment. The majority of pts (86.2%) received prior chemotherapy for mUC and all pts received prior ICI. The median number of cycles of prior ICI was 7 (range 1-20) and median time between previous ICI and CaboNivo was 2.5 mo (range 1-18). The best response to previous ICI was partial response (PR) in 1 pt (3.4%), stable disease (SD) in 13 (44.9%), progressive disease (PD) in 14 (48.3%) and one (3.4%) was not evaluable (NE). The overall response rate for CaboNivo was 13.8% with 4 pts achieving PR (13.8%), 15 SD (51.7%), 7 PD (24.2%) and 3 NE (10.3%). Responses were seen in the liver, lung, and lymph nodes. Among 4 pts with PR, 2 were primary refractory to previous ICI and 2 had SD. At cutoff date the median duration of response was not reached and 3 PR were still ongoing: 1 had just began and the other 2 were ongoing at 12.3 and 26.4 mo. Among 15 pts with SD, 4 had SD for more than 6 mo and 2 were still ongoing at 8.1 and 25.1 mo. Median progression-free survival was 3.6 mo (95% CI: 2.1 – 5.3 mo) and median overall survival was 10 mo (95% CI: 5.8 – 16.7 mo). Grade 1/ 2 treatment related adverse events (AEs) occurred in 28 pts (97%) and >Grade 3 (G>3) AEs occurred in 14 pts (48%). The most common G>3 AEs were fatigue (14%), hypophosphatemia (14%), lymphocyte count decrease (14%), hypertension (7%) and hyponatremia (7%). Conclusions: CaboNivo is clinically active and safe in heavily pretreated pts with progressive mUC following ICI. Clinical trial information: NCT02496208 .

5039 Background: We investigated challenging/re-challenging pts with ipilimumab (ipi) after progression on CaboNivo or CaboNivoIpi. Methods: In a phase I expansion study, patients with mUC post-platinum chemotherapy and other GU tumors patients who progressed on Cabo 40 mg daily plus nivolumab, 3 mg/kg every 21 days (CaboNivo) alone or with ipi, 1 mg/kg every 21 days for 4 cycles (CaboNivoIpi)-and achieved a PR or SD≥6 mo, were challenged/re-challenged with ipi, 1 mg/kg every 21 days for up to 4 cycles. Restaging scans were done every 6 wks for the first 12 wks, then every 8 wks and evaluated by RECIST 1.1. Results: In total, 24 patients were evaluated: 18 pts (8 UC (5 bladder and 3 upper tract), 4 clear cell renal cell carcinoma (RCC), 3 urachal adenocarcinoma (adeno), 2 bladder adeno, and 1 sarcomatoid clear cell RCC) who progressed on CaboNivo were challenged with ipi. In the challenge group, median (m) follow-up was 21.2 months. One pt achieved a PR in the LNs, but was found to have brain metastases before the next restaging, 13 had SD and 4 had PD. Median duration of PR or SD was 3.6 months (95% CI: 1.4 – 7.8 months). The mOS from start of ipi challenge was 13.9 months (95% CI: 5.8 months- not estimable); mPFS was 4.6 months (95% CI: 1.9 – 8.7 months). Grade 1/2 treatment related adverse events (AEs) occurred in all 18 pts (100%) and ≥Grade 3 (G≥3) AEs occurred in 11 pts (61%). The most common G≥3 AEs were hypophosphatemia (22%), hypertension (6%), adrenal insufficiency (6%), increased AST (6%), and ALT (6%). Six patients (3 bladder UC, 1 penile squamous cell (SCC) carcinoma, 1 urethral SCC, and 1 clear cell RCC with sarcomatoid features) who progressed on CaboNivoIpi were re-challenged with Ipi. On re-challenge, mfollow-up was 20.9 months. There were no PRs, 3 SDs and 3 PDs. mOS from start of re-challenge was 4.0 months (95% CI: 2.2 – 23.3 months) and mPFS was 1.9 months (95% CI: 1.1 – 2.6 months). Grade 1/2 treatment related AEs occurred in all 6 pts (100%) and ≥Grade 3 (G≥3) AEs occurred in 2pts (33%). G≥3 AEs included 1 hypertension (17%) and 1 hyperphosphatemia (17%). Conclusions: Ipi challenge/re-challenge showed low response rates in pts previously treated with CaboNivo or CaboNivoIpi. However, pts treated with CaboNivo who were challenged with ipi had a better OS than patients who had progressed on CaboNivoIpi and were re-challenged with ipi. Larger trials are warranted testing the ipi challenge in pts progressing on CaboNivo. Clinical trial information: NCT02496208 .

e15142 Background: Immune checkpoint inhibitors (ICIs) are increasingly utilized across cancer types. Risk factors for emergency room (ER) and inpatient (IP) care in this patient (pt) population remain poorly defined. Methods: We retrospectively reviewed charts for pts with (w/) advanced solid tumors who received more than 1 ICI dose at Mount Sinai from 1/1/11 – 4/28/17. Demographics, medical history, cancer diagnosis and therapy (tx) details, and outcomes were recorded. Descriptive data detailing ER/IP care during ICI tx (from first dose to 3 months after last dose) were collected. Fisher’s exact test and multivariable regression analyses were used to study differences between pts w/ versus w/out ER/IP care during ICI tx. Significance level was set at 0.05 for all tests. Results: 345 patients met the inclusion criteria for this study. Mean age was 64 years. Common tumor types were non-small cell lung cancer (n = 87), melanoma (n = 78), hepatocellular carcinoma (n = 53), and urothelial carcinoma (n = 38). 83% of patients had metastatic disease. Median duration of ICI tx was 2 months. During ICI tx, 50% received ER care and 43% received IP care, resulting in 369 total episodes of ER and/or IP care. 12% of episodes were attributed to ICI toxicity; 6% required intensive care. Median time to ER/IP care was 111 days. Median IP length of stay was 2 days. Several factors were found to be significant predictors of risk for ER or IP care on multivariable analysis (Table). African-American race and Hispanic ethnicity were each predictive of increased risk for both ER and IP care. Conclusions: ER/IP care is common among pts w/ advanced solid tumors on ICI tx. Understanding demographic and clinical risk factors can help prospectively identify higher-risk pts to inform preventive programs aimed at reducing such care. [Table: see text]

5018 Background: Patients with rare genitourinary malignancies (Bladder cancer of variant histologies (BCVH), adrenal tumors, chemotherapy refractory germ cell tumors (CRGCT), penile carcinoma and prostate cancer of variant histology (PCVH)) have poor outcomes. Nivolumab and ipilimumab has demonstrated safety and efficacy in genitourinary malignancies. In this multicenter, single arm, multi-cohort phase II trial we evaluated the efficacy of nivolumab and ipilimumab in pts with advanced rare genitourinary cancers (NCT 03333616). Herein, we report the results of the fully accrued BCVH, adrenal and other cohorts. Methods: Eligible pts had a metastatic rare genitourinary malignancy, ECOG performance status of 0-2 and no prior immune checkpoint inhibitor exposure; aside from CRGCT pts could be treatment naïve. Eligible pts received nivolumab 3 mg/kg and ipilimumab 1 mg/kg intravenously every 3 weeks for 4 doses with continued nivolumab 480 mg IV every 4 weeks. The primary endpoint was overall response rate (ORR) by RECIST 1.1. Results: 56 pts were enrolled at 6 institutions between 4/2018 and 7/2019 in 3 cohorts: BCVH (N = 19), adrenal tumors (n = 18) and other tumors (n = 19). Median follow-up was 9.9 (range < 1~21) months. 28(50%) pts received all 4 doses of nivolumab and ipilimumab. 25 pts received nivolumab maintenance at a median of 4 (range 1-18) cycles. ORR in entire cohort was 16% (n = 9: 2 CR 7 PR, Table); 67% of responders (6/9) maintained response greater than 9 months. Median PFS was 2.8 (2.7-5.2) months. 22 pts (39%) developed treatment-related ≥ grade 3 toxicity; 23% required high dose steroids (≥40 mg prednisone or equivalent) and 15 (27%) pts discontinued treatment due to an AE. Grade 5 toxicity occurred in 3 pts; 1 death was treatment related. Conclusions: Nivolumab and ipilimumab resulted in objective responses in a subset of pts with rare GU malignancies, especially in BCVH. Biomarkers are being evaluated and an additional cohort exploring the activity in genitourinary tumors with neuroendocrine differentiation is ongoing. This combination warrants further investigation in these pts with substantial unmet needs. Clinical trial information: NCT 03333616 . [Table: see text]

681 Background: Patients (pts) with RCC and oligometastatic pancreas metastases are treated with pancreatectomy, stereotactic body radiation therapy (SBRT), or systemic therapy. The optimal approach is not clear. We aimed to evaluate the comparative efficacy of the modalities in terms of progression-free survival (PFS) and overall survival (OS). Methods: This IRB-approved, multi-institutional, retrospective study evaluated pts with pancreatic-only RCC metastasis without concurrent metastases elsewhere. Data on pt demographics, tumor characteristics, treatment, and outcomes were collected. PFS and OS in pts treated with pancreatectomy vs. systemic therapy were compared by log rank tests. Results: Fifty-one pts from 9 institutions were included. All had clear cell RCC; 50 pts had nephrectomy; 30 pts (58.8%) and 18 pts (35.3%) had IMDC favorable and intermediate risk, respectively. Median time from RCC diagnosis to oligometastatic disease was 120 months (mo) (range: 0, 175). As initial treatment, 23 (45%) pts had pancreatectomy (mostly partial); 25 (49%) had systemic therapy (VEGFR TKI and/or immunotherapy); 1 had SBRT; 2 had other treatments. Too few pts had SBRT for comparison. With a median follow-up of 25 mo (2, 68), median PFS for the population was 25 mo (17, 42 95% CI). Median PFS was 36 mo (8, 43 95% CI) for surgery pts and 22 mo (17, NR 95% CI) for systemic therapy pts; this was not statistically significant (NS), p = 0.3. Median OS for the population was 121 mo (100, NR 95% CI). With a median follow-up of 51 mo (2, 217), OS was 121 mo (100, NR 95% CI) for surgery pts and not reached (64, NR 95% CI) for systemic therapy pts; NS, p = 0.52. Conclusions: In this retrospective series, RCC pts with oligometastatic pancreatic-only disease had similar PFS and OS outcomes from initial pancreatectomy or systemic therapy. RCC pts with pancreas-only metastases represent a unique patient population and studies informing the underlying biology are needed to optimize clinical management.

4544 Background: NaF+FDG PET imaging are used to assess soft tissue and bone metastases. The prognostic value of NaF+FDG PET in GU cancer pts was assessed as a secondary endpoint within a phase I trial of combination CaboNivoIpi. Methods: NaF+FDG PET scans were collected at baseline and cycle (C)3 day (D)1 for 50pts. Up to 50 lesions/patient were analyzed at baseline, up to 10 lesions/organ in case of extensive disease. Lesion number and whole-body metabolic tumor volume (wMTV) were recorded at baseline, C3D1 and percent change for FDG and NaF scans. Whole-body total lesion glycolysis (wTLG) and its percent change was obtained for FDG scans. Parameters were evaluated with respect to OS using Kaplan-Meier and log-rank, with quartiles, then refined to show strongest distinction, adjusting p-values to account for this exploration. Parameters with strongest OS association were also analyzed for associations with OS in urothelial carcinoma (UC) pts. Results: 50 pts, (UC (n = 20); others (renal cell, prostate, urachal/adenocarcinoma, germ cell, penile, bladder squamous cell (n = 2-7 each)). Median (m) overall survival (OS) was 23.9 months (mo) (95% CI: 13.7mo – NE) with 29.7mo m potential follow up and mOS of 24.7mo (95% CI: 13.7mo-NE) for UC. For FDG in all pts, wMTV: baseline ≤ vs > 51.6, mOS (NR vs 10mo, p = .0006), C3D1 ≤ vs > 85 mOS (25.9mo vs 5.1mo, p = .0001), percent change (0/increase vs decrease, mOS 14mo vs 25.9mo, p = .015); wTLG: baseline ≤ vs > 178, mOS (NR vs 11.5mo, p = .011), C3D1 ≤ vs > 300, mOS (25.9mo vs 8.3mo, p < .0001), percent change decrease vs increase, mOS (25.9mo vs 14.0mo, p = .016); and lesion number: baseline ≤ vs > 13, mOS (25.9mo vs 9.9mo, p = .0090), C3D1 ≤ vs > 13 mOS (25.9mo vs 9.9mo, p < .0001) significantly predicted OS. In UC pts, wMTV percent change (0/increase vs decrease, mOS 14mo vs. 25.9mo, p = .057), wTLG percent change decrease vs increase, mOS (NR vs 8.4mo, p = .0015), and lesion number C3D1 ≤ vs > 13 mOS (25.9mo vs 2.8mo, p = .022) significantly predicted OS. NaF parameters failed to do so. Conclusions: FDG wMTV and wTLG at baseline, C3D1, and percent change and lesion number at baseline and C3D1, predicted OS in GU cancer pts on CaboNivoIpi. Clinical trial information: NCT02496208.

4571 Background: RCC and UC pts with clinically significant AD were generally excluded from CPI trials due to potential AD exacerbation. Thus, the safety and clinical activity of CPI in AD pts are not well characterized. Methods: We retrospectively collected data from RCC and UC pts with AD treated with CPI at 9 centers. Adverse events (AEs) were assessed using CTCAEv5 criteria. Objective response rate (ORR) was assessed by RECIST principles. Overall survival (OS) was estimated by Kaplan Meier. Results: Of 103 pts (57 RCC & 46 UC) with a broad spectrum of AD such as psoriasis (22%), thyroiditis (20%), rheumatoid arthritis (13%), polymyalgia rheumatica (8%), inflammatory bowel disease (6%), multiple sclerosis (3%) and lupus (3%), most received CPI as 1st or 2nd line (77% RCC, 93% UC) and anti-PD-1/L1 monotherapy (65% RCC, 98% UC). At CPI start, 36 had clinically active AD (all grade 1-2) and 4(11%) requiring systemic immunosuppression. AD exacerbations occurred in 37% (n = 38); most frequent: arthritis (12% RCC, 24% UC), rash (11% RCC, 9% UC). New onset immune related (ir) AEs occurred in 36% (n = 37); most frequent: colitis (12% RCC, 4% UC), rash (11% RCC, 9% UC), hypothyroid (each 7%), nephritis (7% UC). Table details timing and management. Median followup was 12.5 (1-52) mos for RCC and 14.5 (1-53) mos for UC. Median time on CPI was 6 mos (1-36) RCC and 4 mos (0.5-40) UC. At data cutoff, 39 RCC & 36 UC had discontinued CPI; 16% for toxicity. ORR was 31% for RCC and 35% for UC. 1 yr-OS rate was 74% (95%CI 58-84) for RCC and 60% (95%CI 43-74) for UC. Conclusions: AD exacerbations or new irAEs occurred in 37% and 36% respectively and were generally manageable. Only a minority required CPI cessation due to toxicity. [Table: see text]

e16602 Background: Radical prostatectomy (RP) and definitive radiation with androgen deprivation therapy (ADT) are both standard of care options for intermediate and high-risk localized prostate cancer (PCa). It remains controversial if the two modalities produce equivalent outcomes because the PROTECT trial (2016) and three recent retrospective studies (2018) drew inconsistent conclusions. Methods: The Surveillance, Epidemiology and End Results (SEER) database was queried from 2004 to 2015 for localized PCa (TxN0M0) who received upfront surgery or who were recommended for surgery but instead received external beam radiation (EBRT) with or without brachytherapy boost to control selection bias. The Kaplan-Meier method was used to generate overall survival (OS) curves and the Cox regression analysis was performed to compare survival outcomes. Results: A total of 175,349 patients were eligible with a median age of 62 years old (age range: 26 to 99). 167,234 patients had upfront surgery and 8,115 patients had upfront radiotherapy, including 7,099 EBRT and 1,016 EBRT with brachytherapy. Overall, upfront surgery was associated with a better OS, compared with patients who received radiotherapy (all patients: adjusted HR [adjHR], 0.79; 95% CI, 0.74–0.83, P < 0.001). The survival benefit is more pronounced in patients of intermediate-risk group (adjHR, 0.61; 95% CI, 0.57–0.66, P < 0.001; N = 112,816), and remains significant in patients of high-risk group (adjHR, 0.91; 95% CI, 0.84–0.98, P = 0.02; N = 62,533). However, in elderly high-risk PCa patients (age ≥ 70), surgery was associated with a significantly inferior OS, compared with radiotherapy (adjHR, 1.16; 95% CI, 1.06–1.28, P = 0.002). Conclusions: Compared with upfront radiation therapy, definitive surgery was associated with overall survival benefit in intermediate- and high-risk localized PCa, except patients with old age and high-risk features, who should be considered for radiation therapy.

4503 Background: The combination of gemcitabine (G) and cisplatin (C) is a standard therapy for metastatic urothelial carcinoma (mUC). Based on data that angiogenesis plays a role in UC growth and progression, a randomized placebo-controlled trial was performed. Methods: Patients mUC, no prior chemotherapy for metastatic disease and >12 months from prior (neo)adjuvant chemotherapy and ECOG PS 0-1 were randomized 1:1 to G 1000 mg/m2 IV days 1 and 8 and C IV 70 mg/m2 day 1 with bevacizumab (GCB) 15 mg/kg IV or placebo (GCP) day 1 every 21 days. Randomization was stratified by the presence of visceral metastases and prior chemotherapy. The primary endpoint was overall survival (OS) defined as the time from randomization to death or last follow-up (FU). Secondary endpoints included progression-free survival (PFS), objective response rate (ORR), and ≥ grade 3 toxicity. With 445 deaths, the log-rank test had an 87% power to detect a hazard ratio (HR) of 0.74 with a 2-sided α=0.05. The primary analysis was based on the stratified log-rank test adjusting on stratification factors. Alliance Data Safety and Monitoring Board approved the final OS analysis be performed at 420 events due to lower than expected event rates. Results: 506 patients were randomly assigned (252 GCB, 254 GCP) stratified by the presence of visceral disease and prior chemotherapy for UC. The median FU for patients still alive was 46.2 months. Median OS was 14.5 months for patients treated with GCB and 14.3 months for patients treated with GCP with a HR of 0.87 (95%CI 0.72-1.06; 2-sided Wald p=0.17). The HR for PFS was 0.77 (95%CI 0.63-0.93) in favor of GCB (p=0.0074). Grade 3 or greater adverse event rate was 83.5% with GCB compared to 80.7% with GCP. Conclusions: The addition of bevacizumab to GC chemotherapy did not result in improved OS (primary endpoint) in patients with mUC but there was a PFS improvement. The observed median OS of about 14 months is consistent with prior phase III trials of cisplatin-based chemotherapy. Support: U10CA180821, U10CA180882, U10CA180820, U10CA180853, U10CA180888, Genentech https://acknowledgments.alliancefound.org. Clinical trial information: NCT00942331.

4504 Background: Platinum-based chemotherapy for 1st-line treatment of pts with metastatic urothelial cancer (mUC) is typically administered for a fixed duration followed by observation until recurrence. PD-1 blockade with pembro improves survival of pts with mUC progressing despite platinum-based chemotherapy. We explored the potential benefit of earlier use of PD-1 blockade using a "switch maintenance" approach. Methods: Pts with mUC achieving at least stable disease after up to 8 cycles of 1st-line platinum-based chemotherapy were enrolled. Pts were randomized 1:1 to pembro 200 mg IV q3 weeks versus placebo for up to 24 months; pts progressing on placebo could cross over to pembro. Randomization was stratified based on pre-chemotherapy visceral metastases (Y/N) and response to 1st-line chemotherapy (CR/PR vs. SD). The primary objective was to determine the progression-free survival (PFS) as per irRECIST among pts treated with pembro versus placebo. Results: Between 12/2015 and 11/2018, 107 pts were randomized to placebo (n=52) versus pembro (n=55). The baseline pt characteristics are shown in the Table. Pts randomized to placebo and pembro received a median of 6 and 8 cycles, respectively. Excluding patients with baseline CRs, the objective response rate was 12% (5/42) on placebo and 22% (10/46) on pembro. Grade 3-4 treatment emergent adverse events occurred in 48% of pts on placebo and 56% on pembro. At a median follow-up of 14.7 months, 41 pts have died and 26/52 pts randomized to placebo have crossed over to pembro. PFS was significantly longer in patients randomized to pembro vs. placebo (Maximum Efficiency Robust Test p=0.036; log-rank p = 0.038). The 18-month restricted mean progression-free survival time was 5.6 months with placebo and 8.2 months with pembro (p=0.023). Conclusions: Switch maintenance pembro may “deepen” responses achieved with 1st-line chemotherapy. Switch maintenance pembro prolongs PFS in pts with mUC completing 1st-line platinum-based chemotherapy. Clinical trial information: NCT02500121. [Table: see text]

4555 Background: CTCs may serve as biomarkers for clinical outcomes in GU tumor pts. We examined the association between baseline CTC enumeration, CTC heterogeneity, CTC morphologic subtypes, and progression-free-survival, overall survival and response to therapy with combination CaboNivo or CaboNivoIpi. Methods: 123 samples from 52 pts with mGU tumors treated with CaboNivo (38 pts) or CaboNivoIpi (14 pts) drawn at Baseline, Cycle (C) 2 Day (D) 1, and C3D1 were processed using the Epic Sciences platform. CTCs were defined as cytokeratin (CK)+, CD45-, distinct morphology, intact nucleus. PD-L1 expression was also assessed. Results: From 07/20/2016-09/01/2018, 52 pts [urothelial carcinoma (UC) N = 33; plasmacytoid N = 1; Clear cell renal cell carcinoma N = 4; bladder adenocarcinoma N = 8; bladder squamous cell carcinoma N = 2; bladder small cell N = 2; renal medullary N = 2] were treated. Median age was 61.5 years (range 20-82); 35 (67%) were male. N = 37 (71%) had visceral involvement, N = 15 (29%) with liver involvement, N = 11 (21%) with bone involvement. CTCs were found in the peripheral blood of 26/40 (65%) pts at baseline. 1/40 pts (bladder adenocarcinoma) had PDL1+ CTCs at Baseline. Median CTC/mL at Baseline, C2D1, C3D1, were not significantly different. CTC counts of > 2 and > 4 at C2D1 were potentially associated with shorter OS (p = 0.071 and p = 0.045 without adjustment for multiple cutoffs for evaluation). PFS results exhibited similar trends. Unsupervised clustering of CTC images identified 5 main CTC subtypes, of which the presence of one was significantly associated with shorter OS (p = 0.0014, not adjusted for multiple testing). Additionally, we observed a trend towards patients with higher CTC heterogeneity at C2D1 having longer OS, when adjusting for the risk associated with CTC enumeration (p = 0.084). Conclusions: CTC were detected in pts with mGU tumors treated with CaboNivo and CaboNivoIpi. CTC values were somewhat but not statistically lower in responders vs. non-responders. On treatment lower CTCs and the absence of aggressive CTC subtypes were associated with better clinical outcomes. Ongoing analyses include single cell genomics, and analysis of T-cell populations. Clinical trial information: NCT02496208.

Purpose Vandetanib is an oral once-daily tyrosine kinase inhibitor with activity against vascular endothelial growth factor receptor 2 and epidermal growth factor receptor. Vandetanib in combination with docetaxel was assessed in patients with advanced urothelial cancer (UC) who progressed on prior platinum-based chemotherapy. Patients and Methods The primary objective was to determine whether vandetanib 100 mg plus docetaxel 75 mg/m2 intravenously every 21 days prolonged progression-free survival (PFS) versus placebo plus docetaxel. The study was designed to detect a 60% improvement in median PFS with 80% power and one-sided α at 5%. Patients receiving docetaxel plus placebo had the option to cross over to single-agent vandetanib at progression. Overall survival (OS), overall response rate (ORR), and safety were secondary objectives. Results In all, 142 patients were randomly assigned and received at least one dose of therapy. Median PFS was 2.56 months for the docetaxel plus vandetanib arm versus 1.58 months for the docetaxel plus placebo arm, and the hazard ratio for PFS was 1.02 (95% CI, 0.69 to 1.49; P = .9). ORR and OS were not different between both arms. Grade 3 or higher toxicities were more commonly seen in the docetaxel plus vandetanib arm and included rash/photosensitivity (11% v 0%) and diarrhea (7% v 0%). Among 37 patients who crossed over to single-agent vandetanib, ORR was 3% and OS was 5.2 months. Conclusion In this platinum-pretreated population of advanced UC, the addition of vandetanib to docetaxel did not result in a significant improvement in PFS, ORR, or OS. The toxicity of vandetanib plus docetaxel was greater than that for vendetanib plus placebo. Single-agent vandetanib activity was minimal.

379 Background: CTCs may serve as biomarkers for clinical outcomes in GU tumor pts. We examined the association between baseline CTC enumeration, change in CTC enumeration post treatment, progression-free-survival, overall survival and response to therapy with combination CaboNivo or CaboNivoIpi. Methods: 128 samples from 52 pts with mGU tumors treated with CaboNivo (38 pts) or CaboNivoIpi (14 pts) were drawn at baseline, cycle (C) 2 Day (D) 1, and C3D1 and processed with Epic Sciences platform. CTCs were defined as cytokeratin (CK)+, CD45-, distinct morphology, intact nucleus. PD-L1 expression was assessed. Results: From 07/20/2016-09/01/2018, 52 pts were treated for metastatic [urothelial carcinoma (UC) N = 33; UC plasmacytoid N = 1; clear cell renal cell carcinoma N = 4; bladder adenocarcinoma (BAC) N = 8; bladder squamous cell carcinoma N = 2; bladder small cell carcinoma N = 2; renal medullary carcinoma N = 2]. Median age was 61.5 years (range 20-82); 35 (67%) were male; N = 37(71%) had visceral involvement, N = 15(29%) had liver involvement, N = 11(21%) had bone involvement. CTCs were found in the peripheral blood of 39 (75%) pts, 1 pt (BAC) had PDL1+ CTCs at baseline. Median CTC/mL at C1D1, C2D1, C3D1, were 0.5 (0-357.1), 0.4 (0-71.8), and 0.75 (0-18), respectively. Median (lower and upper quartile) CTCs in pts with complete or partial responses (responders) vs stable or progressive disease (non-responders) were: 0.0 (0 to 5.40) vs 0.7 (0 to 2.5) at baseline; -0.35 (-3.65 to 0) vs 0 (-2 to 1.7) change from baseline to C2D1; and 0 (-5.4 to 0.5) vs 0 (-0.9 to 0.65) change from baseline to C3D1 (all p > 0.15). CTC counts of > 2 and > 4 at C2D1 were potentially associated with shorter OS (p = 0.071 and p = 0.045 without adjustment for multiple cutoffs for evaluation). PFS results exhibited similar trends. Conclusions: CTC were detected in pts with mGU tumors treated with CaboNivo and CaboNivoIpi. CTC values were lower in responders than in non-responders. On treatment lower CTCs were associated with better clinical outcomes. Ongoing analyses with the CTCs include cell morphology and heterogeneity, single cell genomics, and T-cell populations.

370 Background: Somatic mutations of ATM are frequently found in UC and have been associated with a better response to cisplatin-based neoadjuvant chemotherapy. However, we previously showed ATM mutations were associated with a short survival in UC (PMID: 29682192 ). In this study, we focused on prognostic values of mutations in ATM in tumors of patients with relapsed or advanced (TxN2-3M0-1) UC through three independent datasets. Methods: 81 UC pts who underwent FoundationOne genomic sequencing (315 cancer-related genes) were used as a discovery dataset. Results were then validated in additional 91 pts with UC who received FoundationOne test (collected separately) and 129 relapsed/advanced UC patients selected from 412 TCGA bladder cohort. Fisher Exact test was used to determine difference of ATM mutation rates. Logistic regression analysis was performed to calculate odds ratio (OR) and 95% confidence interval (CI). Overall survival (OS) was measured from time of initial diagnosis and Cox proportional hazard regression analysis was performed to calculate the hazard ratio (HR) and 95% CI. Results: The median ages of the 3 cohorts were 65 (44–84), 65 (21–91) and 67 (45–-90). The majority of pts were Caucasians (86.4%, 75.8% and 81.4%) and ever smokers (77.8%, 67% and 76.7%). ATM mutations were present in 14.8% (12/81), 11% (10/91) and 6.2% (8/129) of the three cohorts (Fisher Exact, p = 0.118). In all three groups of pts, ATM mutations consistently correlated with a significantly shorter OS (Discovery: HR = 2.25, 95% CI, 1.03–4.89, p = 0.041; Validation 1: HR = 3.15, 95% CI, 1.17–8.44, p = 0.023; and Validation 2: HR = 2.17, 95% CI, 0.99–4.75, p = 0.051). In 144 pts treated with cisplatin or carboplatin pooled from the three cohorts, ATM mutations correlated with a non-significantly higher objective response rate (OR = 1.54, 95% CI 0.44–5.35, p = 0.5), but were still associated with a poorer survival (HR = 1.95, 95% CI 1.00–3.83, p = 0.05). Conclusions: These results suggest that ATM mutations may be considered as a negative prognostic biomarker in relapsed/advanced UC pts. Further studies are required to determine the underlying mechanisms.

TPS504 Background: Patients with high-risk MIBC have a poor prognosis. Radical cystectomy remains the standard treatment. Yet despite substantial improvements in surgical techniques, mortality from metastatic recurrence remains high. A large number of MIBC patients are ineligible for cisplatin-based chemotherapy or have persistent muscle-invasive disease despite neoadjuvant chemotherapy (NAC). Pembrolizumab, a PD-1 inhibitor, has demonstrated significant activity and is FDA-approved for patients with advanced/chemotherapy-refractory metastatic urothelial carcinoma. We hypothesize that pembrolizumab given post-cystectomy will improve overall survival (OS) and disease-free survival (DFS) in patients with high-risk MIBC. Methods: Patients must have histologically confirmed muscle-invasive urothelial carcinoma of the bladder or upper tract, have received NAC and have ≥ pT2 and/or pN+ at surgery or be cisplatin-ineligible and have ≥ pT3 and/or pN+ at surgery or have declined adjuvant cisplatin-based therapy and have ≥ pT3 and/or pN+ at surgery. Surgery could be radical cystectomy, nephrectomy, or ureterectomy. Patients are stratified by pathologic stage, central PD-L1 status, and prior NAC. Patients are randomized to receive pembrolizumab 240 mg every 3 weeks for 1 year, or observation. The dual primary objectives are to determine DFS and OS. Secondary objectives are to determine DFS and OS in PD-L1-positive and -negative patients and assess safety. Correlative objectives are to determine whether 12 immune gene signatures, tumor molecular subtypes, diversity of T-cell receptor (TCR) clonotypes, persistence of TCR clonotypes, tumor and neoantigen burden, HLA subtypes, and plasma HGF and VEGF levels with IL-10 and IL-17 are associated with OS and DFS. Quality of Life Correlative objectives are to compare health-related quality of life as assessed by the EORTC QLQ-C30. Clinical trial information: NCT03244384.

438 Background: DDR gene alterations may contribute to higher tumor mutational burden (TMB) via genomic instability in addition to APOBEC mutagenesis. We previously showed that ATM mutations correlated with shorter OS in UC, while Teo et al. showed patients (pts) with DDR alterations benefited from PD-1/PD-L1 blockade in advanced UC. Here, we aimed to validate those findings and further explore the prognostic role of ATM mutations in advanced UC treated with anti-PD-1/PD-L1 agents. Methods: The study included 53 pts who had FoundationOne tumor tissue genomic sequencing and anti-PD-1/PD-L1 therapy. Fisher exact test was used to test difference in objective response rate (ORR). OS was measured from time of initial UC diagnosis and Cox proportional hazard regression analysis was performed to calculate the hazard ratio (HR) and 95% confidence interval (CI). Results: The cohort had a median age of 66 (range 21–81) with 34% females and 64.2% platinum-based chemotherapy. DDR alterations (including ATM) were present in 49.1% pts (26/53) and favored a higher ORR (37.5% vs. 23.1%, p = 0.26). Compared with those without DDR alterations, pts with DDR alterations (excluding ATM) seemed to have longer OS, although significance was not reached likely due to a short follow-up time (HR = 0.53, 95% CI 0.20–1.38, p = 0.19). ATM alterations seemed to favor higher response rate to PD-1/PD-L1 blockade (ORR, 40% vs. 28.9%, p = 0.6), but was associated with significantly shorter OS (HR = 5.7, 95% CI 1.65–19.74, p = 0.006) in overall pts and in subgroups with/without platinum-based chemotherapy (data not shown). Pts with ≥ 3 DDR alterations (including ATM) had substantial higher TMB (13.9–72.2 perMb, median 22.6) and benefited the most from PD-1/PD-L1 blockade with 80% ORR vs. 24.4% ORR in pts with < 3 DDR alterations. Conclusions: Our study supported that DDR alterations are associated with higher response rate and prolonged OS in advanced UC pts receiving anti-PD-1/PD-L1 agents, likely from impact on TMB. However, ATM alterations correlated with poor prognosis also in those pts. Further studies are needed to assess the clinical utility of DDR alterations in directing therapies in UC.

4528Background: Preliminaryclinical activity of CaboNivo and CaboNivoIpi has been previously reported for mUC pts and other genitourinary tumors. Here, we report longer follow-up data of CaboNivo a...

515 Background: Tolerability and efficacy of CaboNivo and CaboNivoIpi were demonstrated in the initial phase I cohort, prompting longer follow-up and the addition of expansion cohorts to further evaluate both combinations Methods: Phase I cohort had 7 dose levels (DL). Recommended phase 2 doses for CaboNivo=Cabo 40mg/Nivo 3mg/kg (DL2) & CaboNivoIpi=Cabo 40mg/Nivo 3mg/kg/Ipi 1mg/kg (DL6) (Nadal ESMO 2017).In expansion cohorts, pts were treated: [1] DL8: Cabo 40mg/Nivo 1mg/kg/Ipi 3mg/kg; [2]DL2: mUC, metastatic renal cell carcinoma (mRCC), adenocarcinoma bladder (AcB): post-PD-1-PDL1 inhibitor mUC (p-mUC), and [3] DL6 mRCC. Objectives: safety, objective response rate (ORR), duration of response (DOR), progression-free survival (PFS) & overall survival (OS). Results: 75 pts enrolled. Median age 59 yo. 83% male, 17% female. 47 treated with CaboNivo (mUC n=24; AcB n= 9; germ cell tumor (GCT) n= 5; castrate-resistant prostate cancer (CRPC) n= 4; bladder squamous cell carcinoma (SCC) n= 2; penile n=1; mRCC n= 7). 28 treated with CaboNivoIpi: (mUC n=8; penile n=3; CRPC n=7; Sertoli n=1; mRCC n=6; bladder small cell carcinoma n=1; renal medullary carcinoma (RMC) n=2). Any grade (G) adverse events (AEs) (CaboNivo 96% & CaboNivoIpi 96%)/G3–4 AEs (CaboNivo 62% & CaboNivoIpi 71%). Most common any G, CaboNivo: fatigue 70%, diarrhea 60%, AST/ALT 60%, hypophosphatemia (hypoP) 45% & CaboNivoIpi: fatigue 71%, diarrhea 68%, hypoP 50%, AST/ALT 43%. Most common G3-4, CaboNivo: lipase 17%, hypoP 15%, fatigue 6% & CaboNivoIpi: hypoP 21%, AST/ALT 14%, lipase 14%. Selected irAs: colitis (2.6%), hepatitis (2.6%), pneumonitis (2.6%), aseptic meningitis (1.3%). ORR: 36%; 3CR (2mUC, 1SCC) & 20PR (5mUC, 2SCC, 7mRCC, 2Penile, 2AcB, 1CRPC, 1RMC). mDOR: 24.1 mo [95% CI: 14.7-NR], mPFS: 7.2 mo [95%CI: 5.1-18.4], mOS:18.8 mo [95%CI: 10.6-NR]. OS 6/12 mo: 83%/61%. Cohort of p-mUC pts(n=5):1PR/3SD/1PD Conclusions: Updated results from phase I and expanded cohorts confirm initial safety findings and promising antitumor activity for both combinations in mUC, mRCC, and rare GU malignancies Clinical trial information: NCT02496208.

4562 Background: We report the safety and clinical activity of the combination of CaboNivo and CaboNivoIpi in pts with mUC and other mGU tumors (NCT02496208). Methods: In this phase I trial 30 pts were treated in 4 dose levels (DL) for part 1 (CaboNivo) and 18 pts were treated in 3 DL for part 2 (CaboNivoIpi). Pts received Cabo PO daily and Nivo IV (part 1) with Ipi 1mg/kg x 4 doses q3wks (part 2). A mUC and a renal cell carcinoma (RCC) expansion cohort of CaboNivo has initiated enrollment. Tumors were assessed for overall response rate (ORR) q8wks (RECIST 1.1). Adverse events (AEs) were graded (G) by NCI-CTCAE v4.0. Results: From 7/22/15 to 12/31/2016, 48pts (CaboNivo N = 30; CaboNivoIpi N = 18) (mUC N = 19; bladder urachal N = 4; bladder squamous cell carcinoma (bSCC) N = 2; germ cell tumor (GCT) N = 4; castrate-resistant prostate cancer (CRPC) N = 9; RCC N = 2, sarcomatoid RCC N = 2, Sertoli cell N = 1, and trophoblastic tumor N = 1 were treated. Median age was 58 (range 35-77), 41 (85%) were male. Common treatment-related G1/2 AEs for CaboNivo: ALT increase (67%), fatigue (63%), diarrhea (60%), hypothyroidism (57%); CaboNivoIpi: fatigue (72%), diarrhea (61%), anorexia (61%); Grade 3 AEs for CaboNivo: hypertension (23%), neutropenia (17%), hypophosphatemia (13%), lipase increase (10%), fatigue (7%), aseptic meningitis (3%); CaboNivoIpi: hypophosphatemia (19%), hypertension (19%), fatigue (13%), hyponatremia (13%), nausea (13%), lipase increase (11%), colitis (6%); G4 CaboNivo: lipase increase (7%) thrombocytopenia (3%); CaboNivoIpi lipase increase (6%) There were no G5 toxicities, no DLTs. 43 pts were evaluable for response: ORR was 30% 13/43 [3 CR (2 mUC, 1 bSCC); 10 PRs (4 mUC, 2 penile, 1 sarcomatoid RCC, 1 urachal, 1 CRPC, 1 bSCC)]. ORR for CaboNivo 39% (mUC 44%); CaboNivoIpi 18% (mUC 29%). 11/13 (85%) of responses were ongoing at cutoff. Conclusions: CaboNivo and CaboNivoIpi combinations were well tolerated with no DLTs and have durable efficacy in mGU tumors particularly mUC. Rare tumors such as bSCC, urachal, and penile cancers demonstrated response to the combination. Larger cohorts of mUC and rare GU tumors are ongoing. Clinical trial information: NCT02496208.

e16017 Background: The primary objective of the study is to assess the feasibility of obtaining combined FDG and NaF PET/CT in detecting soft-tissue and bone metastatic disease in mGU pts treated with CaboNivo+/-Ipi. Methods: Pts in this single-arm, multicenter, phase I trial had FDG PET/CT followed by NaF PET/CT within 1 hour, at baseline and 8 weeks. The number and location of metastatic lesions on FDG PET/CT and on NaF PET/CT were captured. Lesions were considered positive when there was focal abnormal radiotracer uptake with CT correlation. We analyzed soft tissue and bone lesions distribution and the concordance of baseline metastatic bone disease. Results: From 7/14/15 to 9/9/16, 27 pts (urothelial carcinoma N = 11, bladder squamous cell carcinoma N = 2, sarcomatoid renal cancer N = 1, sertoli cell N = 1, germ cell tumor N = 4, trophoblastic testicular cancer N = 1, urachal N = 4, castrate-resistant prostate cancer N = 2, prostate neuroendocrine N = 1) had scans. Median age was 55 (range 35-75 yr), 24 (89%) were male. Pts completed the dual PET/CTs on the same day (imaging time 3-4 hours) without complications. On FDG, the distribution of 340 lesions were lymph node 41%, bone 35%, lung 18%, liver 11%, soft tissue 4%, kidney 0.6%, adrenal 0.3%, spleen 0.3%, pancreas 0.3%. On NaF, the distribution of 130 lesions was spine 33%, rib 16%, pelvis 13%, femur 10%, scapula 9%, sacrum 8%, skull 5%, humerus 4%, sternum 3%, manubrium 2%, clavicle 2%. Of the 130 lesions on FDG/NaF PET/CT, 38% were not seen on initial FDG. The combination of the two tracers did not affect the quality of the PET/CTs given distinct radiotracer uptake in the anatomical locations with no distortions. Residual FDG activity on combined FDG-NaF images required background intensity to be adjusted to delineate focal radiotracer uptake. There was no difficulty in identifying metastatic lesions, and no instances of prior FDG uptake affecting the determination of metastatic disease. Conclusions: Combination FDG-NaF PET/CT is a feasible imaging modality that offers convenience to pts without compromising detection of metastatic disease in mGU pts. FDG alone detected fewer bone lesions than combined FDG-NaF PET/CT. Clinical trial information: NCT02496208.

293 Background: We report the safety and clinical activity of CaboNivo and CaboNivoIpi in pts with mUC and other GU tumors. Methods: Part I included 4 dose levels (DLs) (Cabo PO daily and Nivo IV q2wk): DL1 Cabo40/Nivo1, DL2 Cabo40/Nivo3, DL3 Cabo60/Nivo1, DL4 Cabo 60/Nivo3. Part II included 3 DLs (Cabo PO daily, plus NivoIpi IV q3 wk x 4 doses then Nivo q2wk): DL5 Cabo 40/Nivo1/Ipi1, DL6 Cabo40/Nivo3/Ipi1, DL7 Cabo 60/Nivo3/Ipi 1. Tumors were assessed for overall response rate (ORR) by RECIST 1.1. Adverse events (AEs) were graded (G) by NCI-CTCAE v4.0. Results: From 7/22/15-10/14/16, 40pts [mUC N=14/(plasmacytoid N=1); bladder urachal N=4; bladder squamous cell carcinoma (bSCC) N=2; germ cell tumor (GCT) N=4; castrate-resistant prostate cancer (CRPC) N=9/(neuroendocrine prostate N=1); sarcomatoid renal cell carcinoma (sRCC) N=1; trophoblastic tumor N=1); sertoli cell tumor N=1; and penile SCC N=4] were treated. Median age was 58 (range 31-77); 36 (90%) were male. AEs related to study drugs with [1] CaboNivo included G3 hyponatremia 4/24 (17%), hypophosphatemia 4/24 (17%), lipase increase 3/24 (13%), dehydration 2/24 (8%), diarrhea 2/24 (8%), fatigue 2/24 (8%), HTN 2/24 (8%), thromboembolic event 1/24 (4%), rash 1/24 (4%), chest pain 1/24 (4%), amylase increase 1/24 (4%), hyperthyroid 1/24 (4%), proteinuria 1/24 (4%), thrombocytopenia 1/24 (4%); and G4 pyelonephritis 1/24 (4%); [2] CaboNivoIpi included G3 hypophosphatemia 2/16 (13%), lipase increase 2/16 (13%), fatigue 1/16 (6%), ALT increase 1/16 (6%), and HTN 1/16 (6%). There were 2/40 (5%) G3 immune-related AEs: 1 aseptic meningitis/CaboNivo; and 1 colitis/CaboNivoIpi. There were no G5 toxicities, or DLTs. 38 pts were evaluable for response. ORR was 12/38 (32%): 1 CR (bSCC); 11 PRs (5 mUC, 1 sRCC, 1 urachal, 1 bSCC, 1 CRPC, 2 penile). SD 20/38 (53%); 9/11 responses were ongoing and 26/39 (67%) pts remain on study. Conclusions: CaboNivo and CaboNivoIpi were well tolerated with no DLTs. Responses were seen at all DLs. The recommended dose for Part I is Cabo40/Nivo3 and for Part II is Cabo40/Nivo3/Ipi1. Rare tumors such as bSCC, urachal, and penile cancer demonstrated responses. Clinical trial information: NCT02496208.

2558Background: We have shown that AR-42, an oral pan-histone deacetylase inhibitor, has clinical activity in some hematologic malignancies and preclinical activity in certain solid tumors, including neurofibromatosis type 2 (NF2)-associated meningiomas and schwannomas. Therefore, this phase 1 trial was initiated to investigate the dose limiting toxicities (DLTs) and the maximum tolerated dose (MTD) in patients with advanced solid tumors. Methods: This is a single-agent, standard 3+3, phase 1 dose-escalation study, with an additional expansion cohort at the MTD; mainly enrolling patients with malignancies of interest, based on our preclinical data. AR-42 was given in a 28-day cycle (3 times weekly, 3-week on, 1-week off). Plasma samples were collected following the first dose to assess pharmacokinetics in each patient. Results: Seventeen patients were enrolled in this study. Two patients were replaced and 15 patients were evaluable for safety and response. Patients had a median age of 49 (range: 20-80) an...

4517Background: FGFR3 alterations are frequent in UC (≈ 21%) and have a putative role in pathogenesis and progression. Antitumor activity observed in pts with UC in a phase 1 trial of the FGFR inhi...

TPS4575 Background: Platinum-based chemotherapy is standard treatment for advanced UC, but no standard exists after first-line therapy. Mocetinostat (MGCD0103) is a spectrum selective inhibitor of ...

146 Background: Tivantinib is a putative non-ATP competitive inhibitor of c-MET receptor tyrosine kinase that has additional cytotoxic mechanisms including tubulin inhibition. Prostate cancer demonstrates higher c-MET expression as the disease progresses to more advanced stages and to castration resistance. Methods: 80 patients (pts) with asymptomatic or minimally symptomatic mCRPC were assigned (2:1) to either tivantinib 360 mg PO BID or placebo (P). The primary endpoint was progression free survival. PCWG2 guidelines were utilized for determining eligibility and progression. Results: Of the 80 pts enrolled, 78 (52 tivantinib, 26 P) received treatment and were evaluated. Median age was 67 yrs (range: 43 to 85). Baseline characteristics were balanced between arms for ECOG PS, Gleason score, PSA, LDH, hemoglobin, Alk Phos, prior treatment, bone and organ involvement. More African Americans and those with lymph node involvement were randomly assigned to placebo. Median follow up is 8.2 months (range: 1.4 to 27.6). To date 59 patients have progressed. Patients treated with tivantinib had significantly better PFS vs. those treated with placebo (medians: 5.6 mo vs 3.8 mo, respectively; HR = 0.53, 95% CI: 0.32 to 0.89; p=0.015). Toxicity was mild overall. Grade 3 febrile neutropenia was seen in 1 patient on tivantinib while grade 3 and 4 neutropenia were recorded in 1 patient each on tivantinib and placebo. Grade 3 sinus bradycardia was recorded in two men on the tivantinib arm. 8 deaths (3 P and 5 tivantinib) have been recorded and were all considered unrelated to therapy. Conclusions: Tivantinib significantly improved PFS in men with asymptomatic or minimally symptomatic mCRPC. Given the favorable toxicity profile and evidence of anti-tumor activity, investigation of tivantinib with other agents may be a rational strategy. Clinical trial information: NCT01519414.

TPS2628 Background: BRAF inhibitor therapy is associated with an impressive albeit short-lived response rate in the treatment of malignant melanoma. Mechanisms of resistance include upregulation of...

e15158 Background: PRCC accounts for 10-15% of all renal cancer. Historically metastatic disease is associated with a short median survival of 5.5 months. There are anecdotal accounts of activity w...

337 Background: HD-IL2 has demonstrated activity for immune therapy of RCC and melanoma. Sorafenib is a multitargeted kinase inhibitor with antiangiogenic activity and may also modulate antitumor immunity. This ongoing phase I study assesses the safety, tolerability, dose-limiting toxicity (DLT), and maximum tolerated dose of the combination of HD-IL2 and sorafenib in patients with unresectable or metastatic clear cell RCC and melanoma. Methods: Eligible patients with ECOG PS 0-1, adequate organ function, and favorable or intermediate Motzer status (RCC patients) received up to two series of therapy. Each 82-day series consisted of two cycles of bolus HD-IL2 and 8 weeks of sorafenib. HD-IL2 (600,000 IU/kg IV q8h x 8-12 doses) was given on days 1-5 (cycle 1) and 15-20 (cycle 2). Sorafenib was given on days 29-82 and discontinued 48-72 hours prior to the next series. Dosing levels for sorafenib were (1) 200mg daily, (2) 200mg BID, (3) 200mg qAM and 400mg qPM, and (4) 400mg BID. Patients who achieved stable disease (SD), partial response (PR), or complete response (CR) after Series 1 proceeded to Series 2. Results: Fourteen patients (8 RCC, 6 melanoma) have been treated. Nine patients completed Series 1 (5 patients came off study due to poor tolerance of HD-IL2, prior to receiving sorafenib). Six patients had SD or better (3 PR, 3 SD) after Series 1. Of these, five have completed Series 2 and one patient progressed during sorafenib treatment in the second series. No DLTs were observed at the first 3 dose levels. All observed grade 3 and 4 AEs were transient and attributable to the known toxicities of HD-IL2 or sorafenib, and were not considered DLT. Adverse events occurring more frequently in Series 2 (HD-IL2 after sorafenib) included proteinuria and dry skin. One death occurred in a patient with melanoma who received only the first cycle of HD- IL2 (never received sorafenib) and one week later experienced sudden death at home, an event reported as possibly related to the HD-IL2. Conclusions: No DLT was seen in this study of HD-IL2 and sorafenib and the toxicity was predictable and manageable. Both RCC and melanoma patients experienced PR and SD, but not CR. Enrollment to cohort 4 is ongoing. [Table: see text]

280 Background: Compared to MVAC, HD-MVAC achieves significantly higher complete response rates in patients (pts) with metastatic bladder cancer. Based on current literature, 7%-38% of pts with muscle-invasive bladder cancer achieve pathological down-staging (pT0) with neoadjuvant chemotherapy (NC), which correlates with improved disease-free and overall survival. The role of HD-MVAC has not been evaluated in the neoadjuvant setting. In this retrospective study, we present our data in pts who received NC with HD-MVAC or GC followed by radical cystectomy (RC). Methods: From July 2008 to August 2010, 38 (pts) received 4 cycles of NC with either HD-MVAC or GC for at least T2 bladder cancer followed by RC. The endpoints of interest of this study were the complete pathologic response (pT0) or down-staging to < pT2 (pT0, pTis, and pT1) at RC; and median interval to RC from the time of diagnosis of muscle-invasive bladder cancer and start of NC. Results: Median age at the time of diagnosis was 66 years (35-80 years). Fifteen pts received neoadjuvant HD-MVAC, and 23 received neoadjuvant GC. Clinical T stage at the time of diagnosis was T2 in 29 (76%), T3 in 3 (8%), and T4 in 6 (16%) pts. Down-staging to < pT2 was achieved in 8 (53%) of HD-MVAC pts and 10 (43%) of GC pts. pT0 was achieved in 5 (33%) of HD-MVAC pts and 9 (39%) of GC pts. The median interval from time of diagnosis to RC was 129 days (range 84-154) for the HD-MVAC pts, and 145 days (range 108-252) for the GC pts. The median interval from initiation of NC to RC was 85 days (range 53-122) for the HD-MVAC pts and 107 days (range 60-126) for the GC pts. Overall, NC was well tolerated with 80% of HD-MVAC pts and 78% of GC pts completing the planned chemotherapy. To this date, none of the pT0 pts had recurrence. Conclusions: Both neoadjuvant HD-MVAC and GC appear to be well tolerated, with very promising rate of pathological down-staging. Longer follow-up is needed for the survival outcomes of these patients. A shorter interval from diagnosis and initiation of NC to RC might be responsible for our better outcomes comparing to some historical data. No significant financial relationships to disclose.

232 Background: Histone deacetylase inhibitors (HDACIs) modulate gene expression and induce cellular differentiation, growth inhibition and apoptotic cell death by chromatin hyperacetylation. Developmental arrest of germ cell differentiation earlier in the life is responsible for the pathogenesis of germ cell tumors (GCT). With current treatment nearly 95% of patients with GCT can be cured. Yet, effective agents with less toxicity are desired. In addition, those with relapsed/refractory disease have a dismal prognosis, indicating a clear need for new, more effective agents. Here we assess the antitumor effects of AR-42, a novel HDACI in in vitro and in vivo models of embryonal carcinoma. Methods: In vitro effects of AR-42 and suberoylanilide hydroxamic acid (SAHA) were evaluated in NTERA-2, an embryonal carcinoma (EC) cell line derived from a human testicular cancer. Cell viability (MTS assay), apoptosis (caspase 3/7 activity and PARP cleavage), cell cycle analysis (flow cytometry) and HDAC inhibition (immunoblotting) were assessed. The in vivo efficacy of AR-42 was assessed in a NTERA-2 xenograft tumor model in male athymic nude mice. Mice were fed control diet and diet containing AR-42 at an average dose of 25 mg/kg/day. Tumor volumes and weights were used as in vivo endpoints. Results: Treatment of NTERA-2 cells with both agents at 0.1-10 μM concentrations showed a time- and dose-dependent reduction in cell viability. Both agents significantly induced apoptosis, cell cycle inhibition and hyperacetylation of histones H-3 and H-4 in a dose-dependent manner. In vitro studies showed that AR-42 was more potent than SAHA. In our rodent model, AR-42-containing diet resulted in a significant reduction in tumor volumes and weights (50% and 56%, respectively). The results for intratumoral changes of proliferation and apoptosis are pending. There were no significant toxicities associated with AR-42, except for testicular atrophy, known to be reversible. Conclusions: AR-42 appears to be a potent inhibitor of EC through different mechanisms, orally bioavailable and well tolerated in our rodent model. Our data indicates that AR- 42 may have clinical value in the treatment of GCT and requires further investigation in clinical trials. No significant financial relationships to disclose.

278 Background: Sorafenib is an oral tyrosine kinase inhibitor with anti-angiogenic properties that may enhance the antitumor activity of GC and when administered as maintenance therapy after GC may delay disease recurrence. This study investigated the safety and clinical outcome of adding S to GC for pts with advanced TCC and the role of S maintenance after GC. Methods: Metastatic/unresectable chemotherapy naive TCC pts ECOG PS 0-1 received G 1,000 mg/m2 d1, 8 and C AUC = 5 d1 with S 400 mg bid d2-19 q21 d for 6 cycles. Pts with stable disease or PR/CR continued on 400 mg bid of S. The primary endpoint was proportion of pts free of progression at 5 months with secondary endpoint of proportion of pts with PR/CR. A Simon two-stage design would enroll 30 pts; if 8 of the initial 17 followed for 5 months or more post treatment showed disease progression, the trial would terminate. Results: 17 pts enrolled, 14 male, 3 female; median age 64 (47-86); ECOG 0 (7)/1 (10). The median number of cycles of GC + S given was 4.4 with multiple dose reductions in 71%. 7 pts (47%) experienced grade (gr) 3/4 neutropenia and 4 (24%) gr 3/4 anemia during GC + S; no neutropenic fever was seen. Gr 3/4 thrombocytopenia occurred in 59% with 2 events of hematuria, 1 epistaxis, and 1 bleeding colitis. Gr 3 nonhematologic toxicities: hypophosphotemia (29%), diarrhea (6%), fatigue (24%), anorexia (6%), dehydration (6%), hand foot syndrome (6%), rash (12%), hypertension (6%). Responses: 1 CR (6%), 4 PR (24 %), 8 SD (47%); 3 pts with evaluable disease showed tumor shrinkage, including 1 with resolution of bone metastases. 9 (53 %) were progression free at 5 months; 7 (41%) came off trial due to > 3 dose reductions (1 prior to first restaging) and 1 withdrew consent. 7 (41%) that achieved CR, PR or SD after a minimum of 4 cycles GC + S received maintenance S for a median of 8 cycles. Conclusions: GC + S showed clinical activity in advanced TCC similar to GC alone but prolonged progression-free intervals were observed in patients treated with S maintenance. GC + S was associated with increased hematologic toxicity although S maintenance was well tolerated. Further studies exploring the role of S maintenance after chemotherapy are warranted. [Table: see text]

e13022 Background: ACUP remains a very challenging diagnosis with very little advances over the last two decades. A principal determinant of the therapeutic index with capecitabine-based treatment ...

2602 Background: Triapine (T), an inhibitor of ribonucleotide reductase, enhances the activity of gemcitabine (G), and the synergism is most evident when the pre-exposure time to T is prolonged. In...