Search

The incidence of metastatic renal cell carcinoma (mRCC) continues to rise. While treatment options have increased dramatically in the last few years, few patients achieve a cure. The standard of care for mRCC in cytokine-eligible candidates is nephrectomy followed by high-dose IL-2. High-dose IL-2 can induce durable complete remissions, but only select patients can enjoy its benefits owing to toxicities. While not curative, the newer targeted therapies offer a broader patient population the chance for treatment response and prolonged survival. This review highlights the historical background of cytoreductive nephrectomy in mRCC, discusses the available treatment options and considers alternative treatment paradigms, such as the integration of the targeted agents with nephrectomy and the use of systemic therapy as medical selection for determining appropriate nephrectomy candidates.

4559 Background: BKM120 is a pan-PI3K inhibitor with demonstrated activity in advanced solid tumors. The primary aim of this study was to determine the maximum tolerated dose (MTD) and dose limitin...

5037 Background: A and C (multi-tyrosine kinase inhibitor including MET and VEGFRs) have complementary mechanisms of action and activity in CRPC. In vivo work with LAPC4-CR (castration and A resistant cell line) has shown the combination has enhanced activity. Methods: Phase 1, 3 + 3 trial with A fixed at 1,000 mg/day and prednisone 5mg BID and escalating daily doses of C (20mg, 40mg, 60mg) in pts with mCRPC was conducted. Cycles were 28 days. Results: 27 patients were enrolled with a median follow-up of 9.2 months (Range: 1.6-29.4 months). 11 pts had prior docetaxel therapy. No DLTs in the first 28 day cycle of dose escalation cohorts. The 60 mg C cohort had Gr 2 adverse events (AEs) of myalgias (2), fatigue (2) and DVT (1) in cycles 2 and 3 necessitating a dose reduction to 40 mg (no change to A). Due to a preferable tolerability profile, the 20mg and 40mg C cohorts were expanded to a total of 12 pts each. See Table for activity data. Gr 3 treatment related AEs: 5 pts at 20mg C [skin infection (1); diar...

482 Background: Outcomes for patients with RCC are reported by pathologic stage; however, determination of therapy and patient counseling are based on clinical staging. Moreover, the development of neoadjuvant therapies that may requires knowledge of clinical stage-specific outcomes for trial design. Because of a paucity of reported outcomes, we characterized pathologic upstaging, recurrence-free survival (RFS) and cancer-specific survival (CSS) by stage at our institution. Methods: Our renal mass registry was queried for patients with localized RCC who underwent extirpative surgery with from 2003-2013 stratified by clinical stage. The proportion of cases upgraded by stage were determined. Survival outcomes were analyzed using the Kaplan-Meier method. Results: A total of 2,144 cases were captured during the study period. Median follow-up was 44.5 months. Clinical to pathologic upstaging is listed in the table. The highest degree of upstaging occured in cT2 cases, with 47% of cT2a and 48% of cT2b upstaged to ≥cT3a. RFS at 60 months for cT1a was 98.4%, cT1b 95.5%, cT2a 78.4%, cT2b 74.4%, cT3a 68.0%, cT3b 63.5%, cT3c 27.8%. CSS at 60 months for cT1a was 99.3%, cT1b 96.3%, cT2a 82.7%, cT2b 81.7%, cT3a 88.7%, cT3b 66.8%, and cT3c 76.9%. Harrell's C index for CSS was similar for both clinical (0.68) and pathologic stage (0.75). Conclusions: While the rate of upstaging increases markedly with increasing clinical stage, survival models based on clinical stage perform as well as those based on pathologic staging. The high rate of upstaging in cT2 disease supports enrollment of these patients in adjuvant clinical trials. Knowledge of these clinical stage-specific outcomes may be helpful for trial design. [Table: see text]

148 Background: Statin use has been associated with a decreased risk of recurrence after local therapy and a lower risk of prostate cancer mortality. There is little known about the association between statin use and prostate cancer progression among men newly initiating ADT. We undertook an analysis of the association between statin use at the time of ADT initiation and TTP. Methods: We queried our institutional clinical database (Prostate CRIS), which captures clinicodemographic data, treatment details, and clinical outcomes of our prostate cancer patients (pts), to identify pts treated with ADT for biochemical recurrence or metastatic disease. Date of progression and statin use at ADT initiation were collected by medical record review. Progression was defined as 3 PSA rises. Date of progression was defined as date of first rise (nadir + >0.02 ng/ml). The association between statin use and TTP on ADT was estimated from multivariable Cox regression, adjusting for known prognostic factors: biopsy Gleason score, primary therapy type, metastatic status, and PSA at ADT initiation. Results: Of the 926 patients eligible for analysis, 283 (31%) were taking a statin at ADT initiation. Statin users were less likely to have M1 (11 vs. 18%) or N1 (5 vs 10%) disease at diagnosis and more likely to have received local therapy +/-ADT compared to nonusers. They were also less likely to have metastases at ADT initiation (53 vs 63%). Median PSA at diagnosis and at ADT initiation were lower among statin users. Duration from diagnosis to ADT initiation was longer in users. After a median follow-up of 5.8 years, 644 pts (70%) had progressed on ADT and median TTP on ADT was 20.3 months (95% CI: 18,24). Men on statins had a longer median TTP on ADT compared to nonusers (27.5 vs 17.4 months). The association remained statistically significant after adjusting for known prognostic factors [adjusted HR=0.83 (p=.039)]. Conclusions: In our cohort, statin use at the time of ADT initiation was associated with a significant increase in TTP on ADT even after adjusting for established prognostic factors. Further investigation into the mechanisms behind this observation is planned.

365 Background: Neoadjuvant cisplatin-based chemo is the standard of care for muscle invasive UC. ddMVAC and GC are frequently used regimens but have not been directly compared. The choice is often based on physician preference and toxicity profile. We interrogated a pre-existing database of UC patients (pts) for differences in efficacy and toxicity among them. Methods: From 2007-2013, consecutive pts who had received presurgical chemo prior to primary tumor resection for muscle invasive, non-metastatic UC were identified. Tolerability, toxicity and efficacy were evaluated. Rates were calculated by regimen and compared using Fisher’s exact test. Disease-free survival (DFS) and overall survival (OS) were estimated using the Kaplan-Meier method and compared by regimen using logrank test. Cox regression estimated hazard ratios (HR) in univariate and adjusted models. Results: Of the96 patients eligible for analysis (GC: 40, ddMVAC: 56), 42% of GC pts had ≥cT3 and 23% had cN+ compared to 62% and 39% with ddMVAC. pCR rate was 18% for GC and 27% for ddMVAC (p=0.33). With a median follow-up of 28 mo., 2-yr OS probabilities were 59% [95% CI:(39-74)] on GC and 77%[95%CI:(60-87%)] on ddMVAC (p=0.1). Conclusions: Despite having more clinical ≥T3 and node positive disease at baseline, ddMVAC is at least as active as GC and achieved a numerically higher rate of pCRs/≤pT1 than GC in our cohort. No unexpected toxicities surfaced. Dose delays, discontinuations, and most selected toxicities appeared higher with GC. Neither DFS or OS significantly differed between the two regimens, however, there was a trend to greater benefit with ddMVAC. [Table: see text]

318 Background: VTE is common in cancer patients (incidence rate estimates 1-10%) and a leading cause of mortality in those receiving CTX. Little is known about the risk of VTE in patients with metastatic UTT. Understanding the frequency of VTE, whether CTX portends higher risk and impact on prognosis is of relevance. Methods: Data was collected via an electronic data capture platform from 25 centers. Patients diagnosed with metastatic UTT were eligible. Patients without a known date of diagnosis, last follow up, or VTE data were excluded. First-line treatment regimen was divided into 6 groups (Table). Cumulative and 90-day from metastatic disease diagnosis unadjusted VTE incidence rates were calculated. The impact of patient, therapy and tumor related factors were analyzed. A multivariate, competing risk regression was utilized to compute subdistribution hazard ratios (HR) of VTE risk for each treatment group. A VTE patient was matched to 4 non-VTE patients to carry out Kaplan-Meier survival analysis. Cases and controls were matched on survival from date of diagnosis of metastatic disease to date of VTE diagnosis and age group. Results: 1,762 patients were eligible. There were 144 (8.2%) and 60 (3.4%) events cumulative and within the first 90 days, respectively. Multivariate analysis revealed that non-urothelial histology and history of vascular disease were associated with increased risk. The HR for VTE based on treatment group, adjusted for covariates, is shown in the table. Overall survival was significantly worse in patients with VTE (p = 0.01). Conclusions: In metastatic UTT patients, there was no difference in VTE risk between CTX regimens. Development of VTE has negative impact on survival. [Table: see text]

e15616 Background: Outcomes with concurrent chemoradiation for penile squamous cell carcinoma (PSCC) are unclear and only anecdotal reports have been published. We performed a retrospective analysi...

4571 Background: mTOR inhibitors have clinical utility in VEGF-refractory RCC and may be useful in non-clear cell disease. In general, mTOR inhibition is inferior to targeting VEGF in the treatment...

4512 Background: GC has been adopted as a standard neoadjuvant regimen for MIBC, supported by practice guidelines, despite a lack of prospective (randomized) data in this setting. Given the histori...

5027^ Background: Abi (CYP17A1 and lyase inhibitor) and cabo (multi-tyrosine kinase inhibitor including MET and VEGFRs) have complementary mechanisms of action. Both have single agent activity in C...

299 Background: Currently, there are no standard treatments for relapsed or refractory urothelial carcinoma (UC). Discouraging results have been achieved in trials evaluating drugs as single agents or in combination regimens. Paclitaxel has activity when used alone and in combination in urothelial cancer, and pazopanib is active in solid tumors secondary to its potent anti-angiogenic effect. We report the results of an ongoing multicenter phase II study evaluating the combination of paclitaxel with pazopanib in refractory UC. Methods: Patients (pts) eligible for the study have histologically confirmed UC, with relapse after receiving up to 2 chemotherapeutic regimens. Pazopanib (800 mg) is administered daily, with weekly paclitaxel (80mg/m2) for 3 weeks in a 28 day cycle. Treatment is continued until disease progression or unacceptable toxicity. Primary endpoint of the study is the response-rate (RR) based on RECIST criteria. Secondary endpoints include safety, and progression free-survival (PFS). For designing the study, Simon’s two-stage method was applied, and 9 pts were recruited in the first stage. After having ≥1 response in the first group, a full enrollment of 32 pts has been initiated. Results: 25 pts were enrolled from April 2010 to September 2013. Their median age was 67 years (47-89), with a median ECOG performance status of 1 (0-2). 10 pts (40%) had UC of the upper urinary tract and 15 had primary bladder/ureter tumors. All pts had multiple metastatic sites, including 11 (44%) with liver metastases. Median number of prior cytotoxic regimens was 2, and 50% were considered cisplatin responsive. Based on RECIST (v1.1) criteria,13 pts (52%) had partial response (PR), 5 (20%) had stable disease (SD), and 2 (8%) had complete response (CR) (80% clinical benefit). The side effects included grade 3 hypertension (n=2), grade 3 fatigue (n=4), grade 3 thrombosis (n=2) and grade 4 neutropenia (n=2). 14 pts required growth factor support. Conclusions: Our phase II study combining paclitaxel and pazopanib demonstratedsignificant anti-tumor activity of combining paclitaxel in relapsed/refractory UC. This combination is safe, and is worthy of evaluation in larger studies. Clinical trial information: 1108055.

470 Background: The neutrophil to lymphocyte ratio (NLR) is a marker of host inflammation and appears to have prognostic value in many solid tumors. We have found in a pilot RCC study that a NLR > 2.5 was predictive of a lower likelihood of response to TT on a multivariable analysis. Here we aim to explore the added value of the NLR to improve the prognostic value of the established IMDC criteria (Heng et al JCO 2009). Methods: We included patients from 7 consortium sites where NLR data was available for patients treated with TT and compared NLR cutoff 2.5 (i.e. low vs. high NLR) and adjusted using proportional hazards regression for the known poor prognostic criteria (listed in Table). Results: Data from 859 patients were included. NLR values were: Mean 4.98, Median 3.51, Mode 2.5, 95%CI 1.42 – 14.0. Using Cutoff 2.5 Median overall survival (OS) is 30.4 months (95%CI 24.9-37.0, n= 237) vs. 15.7 months (95%CI 13.0-17.2, n=622); log-rank p value

398 Background: Prior prognostic models for 2nd-line systemic therapy have not been studied in the setting of contemporary sequential targeted therapy (TT). We sought to validate the IMDC prognostic model in patients with mRCC receiving next-line TT after progression on 1st-line TT. Methods: Patients who received 2nd-line TT after progressing on 1st-line TT for mRCC at 19 centres were analyzed. For the patients who had immunotherapy (22%) prior to their 1st TT, we examined their second TT (ie 3rd-line therapy). The endpoint was median overall survival (OS) since the initiation of 2nd-line therapy. Additionally, we compared the IMDC model with the 3-factor-MSKCC model (Motzer et al JCO 2004) used for previously-treated patients. Results: 1,021 patients treated with a second TT were included. Median time on 2nd-line TT was 3.9 months (range 0-76+). 871 (85%) of patients had stopped 2nd-line TT by the time of analysis. Median OS since 2nd-line TT was 12.5 months (95% CI: 11.3-14.3 months), with 369 (36.1%) of patients remaining alive. 5 out of 6 pre-defined factors in IMDC model (anemia, thrombocytosis, neutrophilia, KPS

e15518 Background: FDA approval of the mTOR inhibitors (mTORi) in mRCC was based on efficacy in poor risk patients (pts) in the first line setting for temsirolimus (T) and in VEGF inhibitor-refractory pts for everolimus (E). Little is known about T’s effectiveness in good and intermediate risk patients and E’s outcomes in the first line setting. Methods: We interrogated the IMDC for the outcomes of pts who received mTORi as first-line targeted therapy. Results: 127 pts received a first line mTORi; the majority received T (93 T, 34 E). The main reasons for T administration were poor risk (38%), non-clear cell histology (27%), and clinical trial (15%) whereas clinical trial (82%) and non-cc histology (6%) drove E use. Of the T and E pts, 58% and 32% were poor risk, respectively. Median age was 61 years and median KPS was 80%. 68% had prior nephrectomy (62% T vs. 82% E). Median progression-free survival (PFS) and overall survival (OS) are detailed below. In the 97 pts with response data, 5% and 53% for T and 8% and 58% for E achieved partial responses and stable disease, respectively. Progressive disease as best response occurred in 41% for T and 33% for E. Second line therapy was captured in 52 pts (41%), of whom 48 received VEGF inhibitors. Conclusions: Given the different populations in which they were administered, direct comparisons of the frontline efficacy of T vs. E cannot be made. The majority of T pts were poor risk, which their dismal PFS and OS reflect. The better outcomes in the E pts highlight that the majority were not poor risk and were healthy enough for clinical trials. While limited by small numbers, this data characterizes a real world experience of mTORi in the first line setting. [Table: see text]

4586 Background: Limited data exists on outcomes for mRCC patients treated with multiple lines of therapy. Benchmarks for survival are required for patient counseling and clinical trial design. Methods: Outcomes of mRCC patients from the IMDC treated with 1, 2, or 3+ lines of targeted therapy (TT) were compared and adjusted by proportional hazards regression. Overall survival (OS) and progression-free survival (PFS) benchmarks were calculated using different population inclusion criteria. OS and PFS are calculated from the line of therapy under consideration unless otherwise specified. Results: 2,705 patients were treated with TT of which 1,533 (57%) received only 1st-line TT, 734 (27%) received 2 lines of TT, and 438 (16%) received 3+ lines of TT. The median OS of patients that received 1, 2 or 3+ lines of TT starting from initial TT was 14.9, 21.0, and 39.2 months, respectively (p

4525 Background: Most studies reporting outcomes of pts with metastatic UC are derived from clinical trial data, potentially limiting the breadth/generalizability of the findings. To explore patterns of care/outcomes in “real world” pts, we initiated an international retrospective cohort study. Methods: Data were collected via an electronic data capture platform from 23 centers. Eligible pts had UC (at least muscle-invasive) and were initially evaluated from 1/1/2006-1/1/2011. Parameters were subjected to regression analysis to identify prognostic variables. Results: By 12/18/12, 1905 pts were enrolled. Among 1077 with metastatic UC, median age was 67 (IQR 60-75), 80% were male, 87% had bladder primary tumors, and 33% received perioperative chemotherapy. Only 758 (70%) received 1st-line chemotherapy for metastatic UC: cisplatin-based (51%), carboplatin-based (29%), non-platinum single-agent (16%), and non-platinum multi-agent (4%). The median survival from date of diagnosis of metastatic UC was 5.2 months (95% CI 4.4-6.5) and 16.1 months (95% CI 15.1-17.5) for pts who did and did not receive 1st-line chemotherapy, respectively [13.9 months (95% CI 12.78-14.98) from start of chemotherapy for latter group]. Among pts receiving 1st-line chemotherapy, univariable analysis revealed gender, primary tumor site, removal of primary, creatinine clearance, LDH, and hgb were not significantly associated with survival whereas smoking status, performance status, perioperative chemotherapy, metastatic sites, study site and chemotherapy regimen were. The multivariable analysis is shown in the Table. Conclusions: The current analysis identifies previously unrecognized prognostic factors in an international cohort of “real world” pts with metastatic UC treated with 1st-line chemotherapy. A large subset of pts with metastatic UC receives no chemotherapy. [Table: see text]

Novel agents that inhibit targets in the hypoxia and MTOR pathways can achieve response rates of 30-60% in renal cell carcinoma (RCC); however, biomarkers have not yet been identified to measure degree of target inhibition and mechanism of biologic response in individual patients. We developed the use of a high-throughput, nanoscale immunoassay (NIA, Nanopro1000 Instrument, Protein Simple) to profile hypoxia pathways and downstream signaling in serially sampled clinical specimens from patients with RCC. NIA uses charge-based separation to distinguish multiple charged modifications of individual proteins, and measure relative ratios of individual unphosphorylated and phosphorylated isoforms. We first developed and optimized assays to measure therapeutic targets and proteins in hypoxia, proliferation and metabolic pathways (including CaIX, HIF, VEGFR, ERK, AKT, MYC, glutaminase) to analyze frozen surgical specimens, fine needle aspirates (FNA's) and blood PBMC specimens from patients with RCC. Next, we analyzed more than 200 FNA's from solid tumors, comparing RCC with paired adjacent non-tumor tissue and other epithelial malignancies. Basal MAPK signaling across the samples ranged across 3 logs for both normal and tumor tissue, demonstrating the difficulty in setting a threshold predictive of response based on a single pre-treatment measurement. To test whether analysis of serial specimens could detect changes, we prospectively collected and analyzed blood peripheral mononuclear cells from 20 RCC patients before and during treatment with standard targeted therapies including MTOR and tyrosine kinase inhibitors. We report for the first time that targeted inhibitors can preferentially inhibit or activate specific protein phospho-isoforms in patients with RCC. Our studies demonstrate that rapid and quantitative nanoproteomic profiling in very small amounts of serially sampled clinical specimen may accelerate translational studies for predictive proteomic biomarkers of response to targeted therapy. Citation Format: Alice C. Fan, John Leppert, Joanna E. Liliental, Liwen Xu, Thomas J. Metzner, Emelyn Shroff, Alia Yaghi, Christine Yost, James D. Brooks, Lauren C. Harshman, Chiara Sabatti, Sandhya Srinivas, Dean W. Felsher. Nano-scale proteomic profiles of response to targeted therapy in patients with RCC. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2283. doi:10.1158/1538-7445.AM2013-2283

432 Background: Novel agents that inhibit kinases in the hypoxia and MTOR pathways can achieve response rates of 30-60% in renal cell carcinoma (RCC); however, protein biomarkers to diagnose RCC and guide treatment have not been identified. Methods: We developed the use of two high-throughput, nanoscale immunoassays to profile hypoxia and downstream signaling in clinical specimens from patients with RCC. These automated technologies minimize errors due to manual variability, and make 96 independent measurements per overnight experiment. We used size-based protein separation (Sally Instrument, Protein Simple) to quantify proteins using antibody-specific detection, normalized to loading control. Next, we used a nano-immunoassay with charge-based separation (NIA, Nanopro1000 Instrument, Protein Simple) to distinguish multiple charged modifications of individual proteins, and measure relative ratios of individual unphosphorylated and phosphorylated isoforms. Results: We first optimized assays for proteins in MAPK, PI3K and STAT pathways, and loading controls, to analyze frozen surgical specimens and ex-vivo fine needle aspirates (FNA’s) performed immediately following nephrectomy. We analyzed more than 200 FNA’s from solid tumors, comparing RCC with paired adjacent non-tumor tissue, and other epithelial malignancies. Basal MAPK signaling of tumors varied across 3 logs of expression. Unique to NIA, we analyzed percent phosphorylation and resolved differences in single phosphorylations. Subtle differences in basal MAPK signaling between tumor calls and adjacent non-tumor tissue can distinguish RCC from other malignancies. Lastly, we prospectively collected and analyzed blood peripheral mononuclear cells from 20 RCC patients before and during treatment with standard targeted therapies. Kinase inhibitors can preferentially inhibit or activate specific protein phospho-isoforms. Conclusions: Our studies demonstrate that rapid and quantitative nanoproteomic profiling in very small amounts of clinical specimen may accelerate translational studies for novel diagnostic and predictive biomarkers.

10513 Background: Novel agents that inhibit kinases in the hypoxia pathway [VEGF, PDGF] can achieve response rates of 30-57% in renal cell carcinoma (RCC); however, threshold levels of targets and downstream signaling proteins in patient cells have not been identified as biomarkers to guide treatment. Rapid and quantitative measurements of overall levels and % phosphorylation of proteins in the hypoxia pathway from patient specimens might better enable the development of proteomic biomarkers. Methods: We have developed the use of an automated nanoscale immuno assay (NIA) to profile hypoxia proteins in solid tumors, and now present NIA assays to measure EGFR, VEGF-R, AKT2, GSK3β, CA-9, MEK1 and ERK1/2 proteins in RCC clinical specimens. Furthermore, to decrease the amount of tissue and the invasiveness of the procedure required to obtain specimens for analysis, we have optimized assays to profile RCC specimens acquired by fine needle aspiration (FNA). Results: Unique to NIA technology, we have analyzed percent phosphorylation for phosphorylated proteins and can distinguish differences in even a single phosphorylation in the FNA specimens, allowing us to group tumors based upon different patterns of phosphorylation. In addition, we compared 10-parameter hypoxia profiles in a panel of RCC paired with adjacent non-tumor tissue and will present the proteins that can distinguish tumor from non-tumor tissue. Importantly, we have determined that we can present protein profiles to the clinical team within 8 hours of receiving the specimen (1 hour preparation time, 6 hours of unattended automated instrument time, 1 hour of analysis). Conclusions: NIA is an extremely efficient, quantitative and high-throughput platform that has the potential to be used to rapidly generate proteomic profiles for clinical specimens for the development of novel diagnostic and predictive biomarkers.

4536 Background: Clinical trials have strict eligibility criteria that exclude many patients to whom the trial results are later extrapolated to in clinical practice. Methods: mRCC patients treated with VEGF targeted therapy were retrospectively deemed ineligible for clinical trials (according to commonly used inclusion/exclusion criteria) if they had a Karnofsky Performance Status (KPS) 2x the upper limit of normal, platelet count of 3/uL, neutrophil count 3 or corrected calcium>=12 mg/dL. Results: 894/2076 (43%) patients were deemed ineligible for clinical trials by the above criteria. Between ineligible versus eligible patients, the response rate, median progression free survival (PFS) and median overall survival of first-line targeted therapy were 21% vs 29%, 5.2 vs 8.8 months and 14.5 vs 28.8 months (all p=12, brain metastases, and non-clear cell histology, had a hazard ratio (HR) for death of 2.8 (95%CI 2.4-3.4), 1.8 (95%CI 1.4-2.2), 1.8 (95%CI 1.2-2.7), 1.4 (95%CI 1.1-1.8), and 1.4 (95%CI 1.1-1.7), respectively (all p

e15005 Background: Adjuvant (adj) treatment of UC generally entails platinum (plat)- and gemcitabine (gem)-based chemotherapy (chemo). Two candidate prognostic and predictive biomarkers are the excision repair cross-complementation group 1 gene (E1) and the regulatory subunit of ribonucleotide reductase (R1). Studies in advanced UC patients (pts) suggest that high E1 and R1 expression confers chemoresistance to cisplatin (cis) and gem respectively. We identified UC pts treated with adj chemo to assess for differences in time to recurrence (TTR) and overall survival (OS) based on tumoral E1/R1 expression and chemotherapy regimen. Methods: 184 pts, who received adj plat or gem for T2-4NxM0 UC from 1984-2009, were identified. Tissue microarrays were created from their primary bladder tumor and involved nodal tissue. E1 and R1 protein expression was assessed using AQUA (HistoRx). The median (med) values were established a priori as the cutoff for high and low expression. Medical records were retrospectively reviewed for pathology, treatment, and outcomes data. Results: 154 pts who had both clinical and gene expression data were eligible for analysis. Pts received cis (134), carboplatin (carbo, 20) and gem (62) based regimens. Med age was 65 yrs. The majority were male (79%) and had pT3 (48.7%) or nodal disease (53.2%). At med followup of 3.9 yrs, 53.4% have recurred. Med TTR and OS were 1.2 and 2.1 yrs. Primary tumor E1/R1 Spearman correlation was moderate at 0.49 but less between primary and nodal disease (E1 0.34, R1 0.23). Irrespective of chemo, E1 and R1 expression were not prognostic for TTR (E1 p=0.8, R1 p=0.3) or OS (E1 p=0.25, R1 p=0.83). Degree of E1 expression did not impact TTR or OS in plat-treated pts nor did R1 expression in gem-treated pts. Stage predicted increased TTR and OS (

4610 Background: Almost all patients with advanced RCC are treated with therapies targeting the hypoxia inducible factor axis. The potential for these agents to cause cardiovascular (CV) toxicity has been increasingly recognized but the overall incidence and extent have not been well described. We sought to identify and characterize the incidence and severity of CV toxicity among RCC patients treated with targeted therapies. Methods: Between 2004 and 2011, all RCC patients treated with targeted therapies directed against the VEGF or mTOR axes were identified at our institution. The incidence of hypertension, left ventricular (LV) dysfunction, changes in serum markers of CV toxicity (e.g., troponin I, NT-proBNP) and heart failure were assessed and graded according to the Common Terminology Criteria for Adverse Events (CTCAE v.4.0). Results: Cardiovascular toxicity developed in 116 of 159 patients (73%). Excluding hypertension, 52 of 159 (33%) developed cardiovascular toxicity. Cardiac toxicity ranged from asymptomatic drops in LV ejection fraction (LVEF) to rises in NT-proBNP to severe heart failure. Asymptomatic cardiotoxicity as defined by decrease in LVEF or increase in NT-proBNP occurred in 43 patients (27%). Symptomatic heart failure (grade 2-3) and grade 3-4 decrease in LVEF each occurred in 4%. Sunitinib was the most frequently used and most common offending agent, with 66 of 101 sunitinib-treated patients (65%) developing a form of CV toxicity, or 32 of 101 (32%) excluding hypertension. However, it was notable that CV toxicity was observed in 68%, 66%, and 51% of patients treated with bevacizumab, sorafenib, and pazopanib as well. The mTOR inhibitors elicited significantly less CV toxicity, but sample sizes were small. Conclusions: Cardiovascular toxicity is an important adverse event related to targeted therapy administration. Close monitoring for the development of CV toxicity with the use of these agents should become standard of care as early detection of asymptomatic patients could preempt symptomatic toxicity and reduce treatment-related morbidity and mortality.

358 Background: Survival estimates for patients with mRCC are traditionally reported from the time of TT initiation. These survival projections, however, may not be applicable to patients who have already survived a period of time after initiating therapy. CS accounts for elapsed time since starting therapy, providing more relevant prognostic information. Methods: Data on 1673 patients treated with first-line VEGF TT between 4/7/2003 and 10/12/2010 was analyzed. Median follow up for patients still alive is 20.1 months. Conditioned survival was calculated on the set of patients alive or on TT at 3 months and using 3 months increments for up to 18 months. Results: The 2-year CS probability tends to slightly improve from 44 to 51% when conditioned on having already survived 0 to18 months since initiation of TT, respectively. The Heng et al (JCO 2009) risk criteria (defined at therapy initiation) retains prognostic ability over time independent of previous survival time or previous time on TT up to 18 months (p Conclusions: Conditional survival may be a more relevant measure of prognosis for those who have already survived or have been on TT for a period of time. The largest improvement was seen in patients in the poor risk group. [Table: see text]

e15103 Background: Obesity is strongly associated with an increased risk of RCC. Patients with higher BMI have improved outcomes with nephrectomy. No data exists evaluating the impact of BMI on (PF...

4560 Background: The Heng et al JCO 2009 prognostic model was developed in the age of VEGF-targeted therapy and this study serves as an external validation while comparing its performance to other models. Methods: 1,028 previously unanalyzed patients were used to externally validate the prognostic criteria for overall survival (OS). The model’s discriminatory value was compared to four other prognostic models using C-indices and Net Reclassification Improvement (NRI) (Pencina et al 2008). NRI indicates how the Heng et al model improves upon other models by reclassifying a percentage of patients (pts) into the more correct risk group based on the observed OS at 2 years. Results: The median OS of all pts was 18.8 mons. On multivariable analysis, the pre-identified Heng’s risk factors (anemia, thrombocytosis, neutrophilia, hypercalcemia, Karnofsky performance status

e15035 Background: ERCC1 (E1), the excision repair cross-complementation group 1 gene, and RRM1 (R1), the regulatory subunit of ribonucleotide reductase are DNA repair enzymes that are being invest...

4555 Background: In mRCC pts who fail first-line VEGF-targeted therapy, there are no available randomized data comparing active drugs. Currently, many clinicians choose a second-line VEGF-targeted therapy based on the type of response to first-line VEGF therapy, although no data exists to support this practice. Methods: Pts treated with targeted therapy in the International mRCC Database Consortium were examined to identify pts who received a second-line VEGF-targeted therapy after failure of a different first-line VEGF-targeted therapy. Response rates and progression-free survival (PFS) were compared in the first- and second-line settings. Results: Of 1,602 total database pts, 464 pts received second-line VEGF-targeted therapy (sunitinib 37%, sorafenib 51%, axitinib 2%, bevacizumab 7%, pazopanib 3%) after failure of first-line VEGF-targeted therapy (sunitinib 54%, sorafenib 33%, bevacizumab 13%). The median overall survival from initiation of first-line therapy for pts who received second-line therapy wa...

e15112 Background: While independent of androgens, the growth and proliferation of CRPC cells remain dependent on the presence of androgen receptors (ARs). Estrogens are known to down regulate ARs ...

e16021 Background: RRM1, the regulatory subunit of ribonucleotide reductase, plays a role in DNA repair after chemotherapy damage and in gemcitabine metabolism. Prior studies demonstrated a survival benefit to high expression in resected early stage lung cancer and a trend toward longer time to progression with low expression in advanced bladder cancers treated with gemcitabine and cisplatin. We hypothesized that patients with resected locally advanced (T2–4NxM0) urothelial transitional cell carcinoma (TCC) whose tumors had higher RRM1 expression would have longer overall survival (OS). Methods: 84 radical cystectomy specimens with muscle invasive TCC were identified from existing tissue microarrays (TMAs) containing 343 specimens. The medical records of these patients were retrospectively reviewed to confirm pathology and stage. Presence of muscle invasion was required. Specimens were analyzed for RRM1 expression using AQUA. The median value of RRM1 was established a priori as the cutoff for high and low expression. Results: Median age of the patients was 69.3 years. There was near equal distribution of stages: 30%, 38%, and 32% for stage II, III, and IV respectively. The majority were high grade (99%) with no nodal involvement (69%). Median OS was 2.0 years (0–13.1). Median RRM1 expression was 1493.3. Degree of RRM1 expression did not correlate with OS, but when adjusted for age, adding an interaction term, high RRM1 expression in younger patients correlated with increased OS (p = 0.0278). Median OS for high expressors age [Table: see text]

e16003 Background: Tyrosine kinase inhibitors (TKI) can lead to cardiotoxicity either directly by causing left ventricular dysfunction, or indirectly by increasing blood pressure. In light of prior studies at our institution documenting a high rate of symptomatic heart failure in patients with renal cell carcinoma undergoing treatment with sunitinib, we instituted a prospective screening protocol to characterize the incidence and natural history of TKI-associated cardiotoxicity. Methods: From March-December 2008, patients receiving TKI therapy for renal cell carcinoma received cardiac biomarker screening (NT- BNP and Troponin I at baseline and after week 4 of each cycle) and transthoracic echocardiography (baseline, 1 month, 3 months, and every 3 months thereafter). If biomarkers were elevated or a decline in ejection fraction was observed, patients were referred for cardiology evaluation. Results: Twenty-six patients have been included since the protocol's initiation. No elevations in cardiac troponin I have been observed to date. Eight patients (31%) had elevations in NT-BNP (a sensitive marker for heart failure). The TKIs involved included sunitinib (5 patients), sorafenib (2 patients), and bevacizumab (1 patient). One patient who was treated with sunitinib had frank left ventricular systolic dysfunction. Seven of the eight patients with elevated NT-BNP values had baseline hypertension, and five patients had significant increases in blood pressure during TKI treatment. In all patients with follow-up biomarkers, NT-BNP levels fell after initiation of heart failure therapy. TKI treatment appeared to ‘unmask’ previously subclinical cardiac injury, including prior silent myocardial infarction (one patient), left ventricular hypertrophy (four patients), and valvular disease (three patients). Updated data from the ongoing screening protocol will be presented. Conclusions: The Stanford TKI screening protocol identified a high rate of subclinical cardiotoxicity and allowed for early initiation of heart failure therapy. Further studies are needed to determine if this approach can decrease cardiac morbidity and improve oncologic outcomes by preventing discontinuation or dose interruption of TKI therapy. [Table: see text]

e14564 Background: In a phase I dose finding study, the maximum tolerated dose of sorafenib was determined to be 400 mg bid. In subsequent practice, these doses of sorafenib have been associated with minimal or reversible toxicity, which presented an opportunity to explore more intensive dosing. A single institution trial of 44 patients demonstrated complete responses in 26% and partial responses in 36%. 93% of patients were able to be dose escalated. The objective of our study was to confirm these findings. The primary endpoints for this study were safety, feasibility, and toxicity. Secondary endpoints included overall response rate (ORR) and progression- free survival (PFS). Methods: Patients with treatment-naïve metastatic renal cell with clear cell histology were enrolled at Stanford University and the University of Nebraska Medical Center. Cycles consisted of 28 days. Dose limiting toxicity (DLT) was defined as grade 3/ 4 toxicity. Patients received sorafenib 400mg bid for 28 days. If no DLT's were reported after the first cycle, patients were dose escalated to sorafenib 600 mg bid for cycle 2. Patients without DLTs at the end of the second cycle were dose escalated to the highest dose, sorafenib 800 mg bid. Patients continued on therapy until progression or unacceptable toxicity. Results: To date, thirteen patients have been treated. Median age was 63 (47–84) years. The majority (84%) were men. Ten patients had nephrectomy prior to trial initiation. Sites of metastases included lung, nodes, liver, adrenal glands, and bone. Dose escalation was possible in 70% of the patients. Median number of cycles was 7 (3–12). Hand-foot syndrome (skin toxicity) limited escalation in the other 30% of patients. One pathologic complete response was obtained after a patient underwent metastectomy of adrenal gland after 4 cycles. One patient achieved a partial response and seven experienced disease stabilization. Conclusions: This dose-escalation trial is ongoing. Dose escalation has been feasible in 70% of patients. The main toxicity limiting escalation was hand-foot syndrome. Thus far, dose escalation did not result in an increased number of responses as compared to the standard 800mg/d dosing. PFS data will be presented as data matures. [Table: see text]

16103 Background: Although commonly used as second-line treatment for patients with progressing prostate cancer, hormonal therapy is associated with unpleasant side effects, and has not unequivocal...

16096 Background: Bevacizumab (Bev), a monoclonal vascular endothelial growth factor (VEGF) antibody, has activity in mRCC. A surrogate biomarker that would accurately predict therapeutic VEGF inhi...

Abiraterone acetate is the first second-line hormonal agent proven to improve survival in metastatic castration-resistant prostate cancer. It selectively inhibits cytochrome P450 17 (CYP17) α-hydroxylase and cytochrome17,20 (C17,20)-lyase, which are enzymes critical for androgen synthesis. Abiraterone acetate was initially approved in the United States in 2011 after demonstrating a 4-month survival benefit in docetaxel-refractory metastatic prostate cancer. The FDA recently expanded its indication for use in the pre-chemotherapy setting after it elicited significant delays in disease progression and a strong trend for increased overall survival in phase III studies. Ongoing investigations of abiraterone are evaluating its efficacy in earlier disease states, exploring its synergy in combination with other therapeutic agents, and assessing the necessity for administration of concurrent steroids and gonadal suppression. The identification and development of predictive biomarkers will optimize the incorporation of abiraterone into the management of advanced prostate cancer.

481 Background: Immune checkpoint inhibitors (ICIs) are associated with immune-related adverse events (irAEs). While the incidence and prevalence of irAEs have been well characterized in the literature, much less is known about the cumulative incidence (CI) rate of irAEs. We sought to evaluate the CI of irAEs in metastatic urothelial carcinoma (mUC) and metastatic renal cell carcinoma (mRCC) patients (pts) treated with ICIs. Methods: We identified a cohort of mUC and mRCC pts who received ICIs at DFCI. irAEs were classified using CTCAE v.5.0 guidelines. The CI rate was a defined measure that accounted for elapsed time since treatment initiation and estimated the risk of irAE development conditioned on time elapsed without experiencing an irAE, accounting for the competing risk of death. Incidence and CI of irAEs at each monthly landmark time was calculated. Prognostic factors of irAE were assessed using the Fine and Gray method. Results: A total of 470 pts was treated with ICIs between July 2013 and October 2018 [mUC: 199 (42.3%); mRCC: 271 (57.7%)]. 341 (72.6%) pts received ICI monotherapy, 86 (18.3%) received ICIs in combination with targeted therapies, and 43 (9.2%) received a combination of two ICIs. Overall, 186 pts (39.5%) experienced any irAE at any time point. Common irAEs included hypothyroidism (n=42 [22.6%]), skin (n=36 [19.4%]), colitis (n=35 [18.8%]), transaminitis (n=32 [17.2%]), and pneumonitis (n=14 [7.5%]). The risk of developing an irAE over time was as follows: 33.5% if no irAE within the 1st month(mo), 27.3% if no irAE in 3mo, 18.8% if no irAE in 6mo, and 16.4% if no irAE by 12mo. No difference was observed in CI based on type of cancer (mUC vs mRCC) or agent (PD1 vs. PD-L1). Multivariable analysis showed that ICI combined with ICI or other agents vs. ICI monotherapy (p