Exploring New Strategies in Bladder Cancer

Exploring New Strategies in Bladder Cancer Daniel P. Petrylak, MD Professor of Medicine and Urology Director, Genitourinary Translational Working Group Co-Director, Signal Transduction Program Smilow Cancer Hospital, Yale University

MVAC vs Cisplatin Phase III Long-Term Survival Cisplatin MVAC Evaluable 122 133 3 years 4 17 6 years* 2 9 *9 patients died: 6 of progressive disease/transitional-cell carcinoma (TCC), 1 of a second malignancy/colon cancer, 2 died of other causes (1 patient was lost to follow-up evaluation) MVAC, methotrexate, vinblastine, doxorubicin, and cisplatin Saxman SB, et al. J Clin Oncol. 1997;15(7):2564-2569.

Kaplan-Meier Curves for Overall Survival GC, gemcitabine/cisplatin von der Maase H, et al. J Clin Oncol. 2005;23(21):4602-4608. von der Maase, H. et al. J Clin Oncol; 23:4602-4608 2005

Salvage Chemotherapy for Advanced Urothelial Cancer Drug Author N RR (%) PFS (months) OS (months) Ifosfamide Witte 56 20 2.4 5.5 Gemcitabine Albers 30 11 4.9 8.7 Paclitaxel Vaughn 31 10 2.2 7.2 Docetaxel McCaffrey 30 13 NR 9.0 Pemetrexed Galsky 13 8 NR NR Pemetrexed Sweeney 47 27.7 2.9 9.6 Ixabepilone Dreicer 42 11.9 2.7 8.0 OS, overall survival; PFS, progression-free survival; RR, response rate

VFL European Phase III Trial Overall survival: Intent-to-treat population (N = 370) Survival Distribution Function VFL + BSC BSC VFL + BSC N = 253 BSC N = 117 No. of events 204 103 No. censored (%) 49 (19.4) 14 (12.0) Median OS (months) (95% Cl) Hazard ratio (95% CI) 6.9 (5.7, 8.0) 0.88 (0.69, 1.12) P value * 0.2868 4.6 (4.1, 7.0) BSC, best supportive care; VFL, vinflunine * Stratified log rank test Bellmunt J, et al. J Clin Oncol. 2009;27(27):4454-4461. Overall survival (months)

Atezolizumab (MPDL3280A): Phase Ia Study Ongoing dose-expansion phase UBC N = 92 a TNBC Melanoma NSCLC RCC Other tumor types 1. PD-L1 selected 2. Allcomers b 1. PD-L1 selected 2. Allcomers All-comers 1. Allcomers 2. PD-L1 selected 1. Allcomers 2. PD-L1 selected 1. PD-L1 selected 2. Allcomers Atezolizumab (MPDL3280A) administered IV Q3W 15 mg/kg or 1200 mg flat dose Key eligibility criteria: Measurable disease per RECIST v1.1 ECOG PS 0 or 1 a Safety-evaluable UBC population. b The UBC cohort originally enrolled patients with PD-L1 IC2/3 but was then expanded to include all-comers, primarily recruiting PD-L1 IC0/1 patients. NSCLC, non-small cell lung cancer; RCC, renal-cell carcinoma; TNBC, triple-negative breast cancer; UBC, urothelial bladder cancer Powles T, et al. J Clin Oncol. 2014;32(Suppl): Abstract 5011. Powles T, et al. Nature. 2014;515(7528):558-562. Bellmunt J, et al. Ann Oncol. 2014;25(Suppl 4): Abstract 8080. Xiao Y, et al. J Immunother Cancer. 2014; 2(Suppl 3): P131. Kim J, et al. J Clin Oncol. 2015;33(suppl): Abstract 297. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

IHC Characteristics of Prescreened Patients: PD-L1 Expression in the UBC Tumor Microenvironment The SP142 assay measures PD-L1 expression based on 4 IHC scoring levels a This assay is optimized for detection of PD-L1 expression in both tumor cells (TC) and tumor infiltrating immune cells (IC) In UBC, prevalence of PD-L1 on TC is low, 1,2 and patients who express PD-L1 on TC are captured within IC cutoffs In UBC, PD-L1 expression on IC is correlated with response, 1,2 while TC expression is not predictive PD-L1 expression is associated with atezolizumab clinical activity in other cancers, including NSCLC 3-6 PD-L1 prevalence in UBC b IHC level (N = 205) IC scored as PD-L1+, n (%) IC3 18 (9%) IC2 37 (18%) IC1 89 (43%) IC0 61 (30%) PD-L1 expression in IC a IC scoring criteria: IC3: 10% of IC PD-L1+; IC2: 5% but < 10% of IC PD-L1+; IC1: 1% but < 5% of IC PD-L1+; IC0: < 1% of IC PD-L1+. b Based on staining of archival tumor tissue from patients prescreened in atezolizumab Phase Ia study. IHC, immunehistochemical 1. Bellmunt J, et al. Ann Oncol. 2014;25(Suppl 4): Abstract 8080. 2. Powles T, et al. Nature. 2014;515(7528):558-562. 3. Spira A, et al. J Clin Oncol. 2015;33(suppl): Abstract 8010. 4. Horn L, et al. J Clin Oncol. 2015;33(suppl): Abstract 8029. 5. Liu S, et al. J Clin Oncol. 2015;33(suppl): Abstract 8030. 6. Spigel D, et al. J Clin Oncol. 2015;33(suppl): Abstract 8009. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Baseline Characteristics in UBC (Safety Evaluable) Characteristic Patients N = 92 a Median age, y (range) 66 (36-89) Male, n (%) 69 (75%) ECOG PS, n (%) 0 37 (40%) 1 55 (60%) Site of primary tumor Bladder 73 (79%) Renal pelvis 5 (5%) Ureter 9 (10%) Urethra 5 (5%) Site of metastases at baseline, n (%) Visceral b 73 (79%) Liver 34 (37%) Characteristic Poor prognostic factors included visceral mets, low hemoglobin levels, ECOG PS 1 and short time ( 3 months) from prior chemo 1,2 a Safety-evaluable patients received at least 1 dose of atezolizumab. b Includes lung, liver, non-lymph or soft tissue. c n = 89. Data cutoff, Dec 2, 2014. GFR, glomerular filtration rate; mets, metastases Patients N = 92 a Prior treatments, n (%) Cystectomy or nephroureterectomy 56 (61%) Platinum-based chemotherapy 86 (94%) Cisplatin-based 69 (75%) Carboplatin-based 35 (38%) 2 prior systemic therapies (metastatic) 66 (72%) 3 months from last chemotherapy 37 (42%) c Hemoglobin levels < 10 g/dl 16 (17%) GFR < 60 ml/min 38 (41%) 1. Bellmunt J, et al. J Clin Oncol. 2010;28(11):1850-1855. 2. Pond GR, et al. BJU Int. 2014;113(5b):E137-43. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Baseline Characteristics in UBC (Safety Evaluable) Characteristic Patients N = 92 a Median age, y (range) 66 (36-89) Male, n (%) 69 (75%) ECOG PS, n (%) 0 37 (40%) 1 55 (60%) Site of primary tumor Bladder 73 (79%) Renal pelvis 5 (5%) Ureter 9 (10%) Urethra 5 (5%) Site of metastases at baseline, n (%) Visceral b 73 (79%) Liver 34 (37%) Characteristic Patients N = 92 a Prior treatments, n (%) Cystectomy or nephroureterectomy 56 (61%) Platinum-based chemotherapy 86 (94%) Cisplatin-based 69 (75%) Carboplatin-based 35 (38%) 2 prior systemic therapies (metastatic) 66 (72%) 3 months from last chemotherapy 37 (42%) c Hemoglobin levels < 10 g/dl 16 (17%) GFR < 60 ml/min 38 (41%) Poor prognostic factors included visceral mets, low hemoglobin levels, ECOG PS 1 and short time ( 3 months) from prior chemo 1,2 a Safety-evaluable patients received at least 1 dose of atezolizumab. b Includes lung, liver, non-lymph or soft tissue. c n = 89. Data cutoff, Dec 2, 2014. GFR, glomerular filtration rate; mets, metastases 1. Bellmunt J, et al. J Clin Oncol. 2010;28(11):1850-1855. 2. Pond GR, et al. BJU Int. 2014;113(5b):E137-43. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Baseline Characteristics in UBC (Safety Evaluable) Characteristic Patients N = 92 a Median age, y (range) 66 (36-89) Male, n (%) 69 (75%) ECOG PS, n (%) 0 37 (40%) 1 55 (60%) Site of primary tumor Bladder 73 (79%) Renal pelvis 5 (5%) Ureter 9 (10%) Urethra 5 (5%) Site of metastases at baseline, n (%) Visceral b 73 (79%) Liver 34 (37%) Characteristic Prior treatments, n (%) Cystectomy or nephroureterectomy Patients N = 92 a 56 (61%) Platinum-based chemotherapy 86 (94%) Cisplatin-based 69 (75%) Carboplatin-based 35 (38%) 2 prior systemic therapies 66 (72%) (metastatic) 3 months from last chemotherapy 37 (42%) c Hemoglobin levels < 10 g/dl 16 (17%) GFR < 60 ml/min 38 (41%) Poor prognostic factors included visceral mets, low hemoglobin levels, ECOG PS 1 and short time ( 3 months) from prior chemo 1,2 a Safety-evaluable patients received at least 1 dose of atezolizumab. b Includes lung, liver, non-lymph or soft tissue. c n = 89. Data cutoff, Dec 2, 2014. GFR, glomerular filtration rate; mets, metastases 1. Bellmunt J, et al. J Clin Oncol. 2010;28(11):1850-1855. 2. Pond GR, et al. BJU Int. 2014;113(5b):E137-43. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Treatment- Related Adverse Events (AEs) in UBC Treatment-related AEs occurring in 5% of patients (all grade) or in 2 patients (grade 3-4) N = 92 a All grade Grade 3-4 b Any AE 60 (65%) 7 (8%) Fatigue 15 (16%) 0 Asthenia 12 (13%) 1 (1%) Nausea 10 (11%) 0 Decreased appetite 9 (10%) 0 Pruritus 9 (10%) 0 Pyrexia 6 (7%) 0 Rash 7 (8%) 0 Diarrhea 5 (5%) 0 Increased AST 2 (2%) 2 (2%) Atezolizumab was generally well tolerated Median safety follow-up was 16+ weeks (range, 3 to 86+ weeks) Median duration of treatment was 3 months (range, 0 to 19 months) No treatment-related deaths 1 discontinuation due to a treatmentrelated AE 5% of patients had a grade 3-4 immunemediated AE per investigator assessment Grade 3 AEs: increased AST (n = 3); increased ALT (n = 2); increased blood bilirubin (n = 1); hypophysitis (n = 1) 37 patients (40%) had a Grade 3-4 AE of any cause c a Safety-evaluable patients received at least 1 dose of atezolizumab. b Additional Gr 3-4 AEs (1% each) included anemia, confusional state, decreased blood phosphorus, hypophysitis, increased ALT, increased GGT and thrombocytopenia. c In addition, 2 grade 5 AEs not related to treatment were seen (acute respiratory failure and alcohol overdose). Data cutoff, Dec 2, 2014. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): ORR in UBC by IC Status PD-L1 IHC n = 87 b IC3 (n = 12) 67% (35%-90%) IC2 (n = 34) 44% (27%-62%) IC1 (n = 26) 19% (7%-39%) IC0 (n = 15) 13% (2%-40%) ORR (95% CI), % a 50% (35, 65) 17% (7, 32) 4 (33%) 5 (15%) CR, n (%) 9 (20%) 4 (33%) 10 (29%) - 5 (19%) - - 2 (13%) PR, n (%) 14 (30%) 7 (17%) Responses were observed in all PD-L1 subgroups, with higher ORRs associated with higher PD-L1 expression in IC Responders also included patients with visceral metastases at baseline: 38% ORR (95% CI, 21%-56%) in 32 IC2/3 patients and 14% (95% CI, 5%-30%) ORR in 36 IC0/1 patients a Efficacy-evaluable patients with measurable disease at baseline per RECIST v1.1. Responses are investigator assessed (unconfirmed); of 30 unconfirmed responses, 24 have been confirmed by the cutoff date. b 4 IC2/3 patients and 7 IC0/1 patients missing or unevaluable. Data cutoff, Dec 2, 2014. CR, complete response; PR, partial response; ORR, overall response rate Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Duration of Treatment and Response in UBC Patients with UBC and CR or PR as best response Median duration of response (DoR) has not yet been reached in either IC group (range, 0+ to 43 months) Median time to response was 62 days IC2/3 patients: range, 1+ to 10+ months IC0/1 patients: range, 1+ to 7+ months 20 of 30 responding patients had ongoing responses at the time of data cutoff 10 patients have been treated for over 1 year, including 3 retreated following protocol amendment a Discontinuation and ongoing response status markers have no timing implication. 4 patients discontinued treatment after cycle 16 prior to 1 year per original protocol. Responses plotted are investigator assessed and have not all been confirmed by the data cutoff (Dec 2, 2014). Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

Atezolizumab (MPDL3280A): Survival in UBC Survival a N = 92 IC2/3 n = 48 IC0/1 n = 44 Median survival follow-up: 14 months (IC2/3) 12 months (IC0/1) PFS Median PFS (range) 6 months (0+ to 18) 1 months (0+ to 14+) Median OS not reached (95% CI, 9.0 mo-ne) 1-y PFS (95% CI) 39% (24-54) 10% (0-21) OS Median OS (range) Not reached (1 to 20+ months) 8 months (1 to 15+ months) IC2/3 IC0/1 + Censored Median OS 7.6 mo (95% CI, 4.7 mo-ne) 1-y survival (95% CI) 57% (41-73) 38% (19-56) PD-L1 IC status appeared to be predictive of benefit from atezolizumab treatment mpfs and 1-year PFS rates were higher in atezolizumab-treated patients with higher PD-L1 IC expression The same association was observed for 1-year OS rates, and median OS for IC2/3 patients was not yet reached Preliminary analysis using SP142 from an independent sample set (n = 110) suggests that PD-L1 IC status is not prognostic for OS in UBC 1 Data cutoff, Dec 2, 2014. Reference: 1. Genentech, unpublished data. Petrylak DP, et al. J Clin Oncol. 2015;33(suppl): Abstract 4501.

IMvigor210 Cohort 2: Study Design Basis for Accelerated Approval Locally advanced or metastatic urothelial carcinoma Predominantly TCC histology Tumor tissue for PD-L1 testing a Cohort 1 (N = 119) 1L Cisplatin Ineligible Cohort 2 (N = 310) Platinum-treated muc Cohort 1 presented at ASCO 2016 1 Atezolizumab 1200 mg IV q3w until loss of benefit Cohort 2: Specific inclusion criteria Progression during/following platinum (no restrictions on # prior lines of therapy) ECOG PS 0-1 CrCl 30 ml/min Median follow-up: 17.5 months (range, 0.2 to 21.1+ mo) a Patients and investigators blinded to PD-L1 IHC status. Trial Identifier: NCT02108652. Co-primary endpoints: ORR (confirmed) per RECIST v1.1 by central review ORR per immune-modified RECIST by investigator Key secondary endpoints: DOR, PFS, OS, safety Key exploratory endpoints: Intratumoral biomarkers (Presented at ASCO 2016 2 ) 1. Balar AV, et al. J Clin Oncol. 2016;34(suppl): Abstract LBA4500. 2. Rosenberg J, et al. J Clin Oncol. 2016;34(suppl): Abstract 104. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

IMvigor210 Cohort 2: Baseline Characteristics Representative of the Greater muc Population Characteristic (safety and efficacy evaluable patients) N = 310 Age, median (range) 66 years (32-91 years) Male 78% PD-L1 status on immune cells (IC): a IC2/3 IC1 IC0 32% 35% 33% Bladder primary tumor site 75% Metastatic sites: visceral b liver lymph node only 78% 31% 14% Creatinine clearance 30-60 ml/min 35% ECOG PS 1 62% Prior cystectomy or nephroureterectomy 66% Prior regimens (metastatic setting): 1 2 3 39% 21% 21% a PD-L1 expression on IC was evaluated (VENTANA SP142 IHC assay) based on 3 scoring levels: IC2/3 ( 5%), IC1 ( 1 but <5%), IC0 (<1%). b Defined as liver, lung, bone, or any non-lymph node or soft tissue metastasis. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

Efficacy Responses to Atezolizumab by PD-L1 IC Subgroup IC2/3 n = 100 IC1/2/3 n = 207 All a N = 310 IC1 n = 107 IC0 n = 103 ORR: confirmed IRF RECIST v1.1 (95% CI) 28% (19, 38) 19% (14, 25) 16% (12, 20) 11% (6, 19) 9% (4, 16) CR rate: confirmed IRF RECIST v1.1 (95% CI) 15% (9, 24) 9% (6, 14) 7% (4, 10) 4% (1, 9) 2% (0, 7) Responses were seen in all IC subgroups, but ORR was enriched with higher PD-L1 status Complete responses accounted for nearly half of the observed responses CRs were observed in all PD-L1 subgroups, with the highest rate in IC2/3 patients ORRs per immune-modified RECIST were concordant a Includes 46 patients with missing/unevaluable responses. b CR + PR + SD 24-wk rate per IRF RECIST v1.1. Treated patients had measurable disease at baseline per investigator-assessed RECIST v1.1. Data cutoff: March 14, 2016. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

Efficacy Time to Response Patients With CR or PR as Best Response 1 st responses Median follow-up: 17.5 months (range, 0.2 to 21.1+ mo) 0 2 4 6 8 10 12 14 16 18 20 Months Median time to first response was 2.1 months Range: 1.6 to 8.3 months Median time to CR was 4.2 months Range 2.0 to 13.2 months Additional CRs and PRs were observed with longer follow up 5 PR to CR and 2 SD to PR conversions CR as best response PR as best response First CR/PR Treatment discontinuation b Ongoing response c a Per IRF RECIST v1.1 b Discontinuation symbol does not indicating timing. c No PD or death only. Data cutoff: March 14, 2016. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

Evidence of Clinical Benefit In Patients Treated Beyond Progression Subsequent responses to atezolizumab were seen across PD-L1 subgroups, highlighting the potential for benefit through non-classical responses SLD Changed from PD, % 100 0-100 Patients Treated Beyond PD With Reduction in Tumor Burden IC2/3 IC1 IC0 Discontinued New lesion In patients treated beyond PD 19% (26/134) had a reduction in tumor burden ( 30% decrease in target lesions) 8% (11/134) achieved PR, CR or SD per IRF RECIST v1.1 post PD Median OS was 11.4 months (all patients) and 16.5 months (IC2/3 patients) 12-month OS was 50% (all patients) and 57% (IC2/3 patients) The safety of atezolizumab was consistent with that in the ITT population 0 18 36 54 72 90 Time on Study, weeks Patients without post-pd baseline tumor assessments (n = 29) are not included in Plot. Data cutoff: March 14, 2016. PD, progressive disease Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

Efficacy Overall Survival Overall Survival 100 80 60 40 20 0 All patients + Censored Time, months All patients 0 2 4 6 8 10 12 14 16 18 20 # at Risk: All pts: 310 265 203 176 146 126 110 97 82 35 5 Longer OS observed in patients with higher PD-L1 IC status mpfs (2.1 mo per RECIST v1.1; 2.6 mo per imrecist) underscores a disconnect between PFS and OS Subgroup All pts (N = 310) Subgroup All pts (N = 310) 12-mo OS (95% CI) IC2/3 IC0/1 All 50% (40, 60) Median OS (95% CI) IC2/3 IC0/1 All 11.9 mo (9.0, 17.9) 6.7 mo (5.4, 8.0) 31% (24, 37) Median follow-up (range): All Pts: 17.5 mo (0.2 to 21.1+ mo) 7.9 mo (6.7, 9.3) 37% (31, 42) NE, not estimable. Data cutoff: March 14, 2016. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

Efficacy Overall Survival Overall Survival 100 80 60 40 20 0 All Patients All 2L Patients only + Censored 0 2 4 6 8 10 12 14 16 18 20 Time, months All patients and 2L only # at Risk: All pts: 310 265 203 176 146 126 110 97 82 35 5 2L only: 120 101 83 72 61 54 44 34 29 11 2 Longer OS observed in patients with higher PD-L1 IC status 12-mo OS compares favorably with historic estimates of 20% 1 Subgroup All pts (N = 310) 2L only (n = 120) Subgroup All pts (N = 310) 2L only (n = 120) 12-mo OS (95% CI) IC2/3 IC0/1 All 50% (40, 60) 61% (44, 77) Median OS (95% CI) IC2/3 IC0/1 All 11.9 mo (9.0, 17.9) NE (10.9, NE) 6.7 mo (5.4, 8.0) 7.1 mo (5.0, 9.2) 31% (24, 37) 29% (19, 39) Median follow-up (range): All pts: 17.5 mo (0.2 to 21.1+ mo) 2L only: 17.3 mo (0.5 to 21.1+ mo) 7.9 mo (6.7, 9.3) 9.0 mo (7.2, 11.3) 37% (31, 42) 38% (29, 47) NE, not estimable. a 1 prior line of therapy for muc and no (neo)adjuvant therapy. Data cutoff: March 14, 2016. 1. Aggarwal R, et al. Clin Genitourin Cancer. 2014;12(5):e167-172. Dreicer R, et al. J Clin Oncol. 2016;34(suppl): Abstract 4515.

ORR, % 100 75 50 25 TCGA Subtype II Is Associated With Higher ORR n = 73 n = 52 n = 38 n = 36 RECIST v1.1 response PD SD PR CR Gene expression data used to classify IMvigor210 tumor samples recapitulated TCGA subtypes 1,2 Responses occurred in all subtypes, but ORR was significantly higher in luminal II vs other subtypes (P =.0072) Urothelium Luminal 0 I II III IV Luminal Basal Basal TCGA, The Cancer Genome Atlas. Data cutoff: March 14, 2016. 1. Cancer Genome Atlas Research Network. Nature. 2014;507(7492):315-322. 2. Rosenberg JE, et al. Lancet. 2016;387(10031):1909-1920. Choi W, et al. Nat Rev Urol. 2014;11(7):400-410. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

TCGA Subtype II Is Associated With Higher ORR ORR, % 100 75 50 25 n = 73 n = 52 n = 38 n = 36 RECIST v1.1 response PD SD PR CR Gene expression data used to classify IMvigor210 tumor samples recapitulated TCGA subtypes 1,2 Responses occurred in all subtypes, but ORR was significantly higher in luminal II vs other subtypes (P =.0072) What might be the drivers of this subtype-specific response? 0 I II III IV Luminal Basal TCGA, The Cancer Genome Atlas. Data cutoff: March 14, 2016. 1. Cancer Genome Atlas Research Network. Nature. 2014;507(7492):315-322. 2. Rosenberg JE, et al. Lancet. 2016;387(10031):1909-1920. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CDKN2A, cyclin-dependent kinase inhibitor 2A; EGFR, epidermal growth factor receptor; FGFR3, fibroblast growth factor receptor 3; GATA3, GATA-binding protein 3. Data cutoff: March 14, 2016. Choi W, et al. Nat Rev Urol. 2014;11(7):400-410. Aine M, et al. Sci Rep. 2015;5:10957. Cancer Genome Atlas Research Network. Nature. 2014;507(7492):315-322. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 Urothelium Luminal Basal Luminal and basal tumors are distinct subtypes with unique histology and gene expression 1-3 CDKN2A, cyclin-dependent kinase inhibitor 2A; EGFR, epidermal growth factor receptor; FGFR3, fibroblast growth factor receptor 3; GATA3, GATA-binding protein 3. Data cutoff: March 14, 2016. Choi W, et al. Nat Rev Urol. 2014;11(7):400-410. Aine M, et al. Sci Rep. 2015;5:10957. Cancer Genome Atlas Research Network. Nature. 2014;507(7492):315-322. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype T eff Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CD8A GZMA GZMB IFNG CXCL9 CXCL10 PRF1 TBX21 Urothelium Luminal Basal Luminal I patients have low T eff gene expression Data cutoff: March 14, 2016. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype T eff Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CD8A GZMA GZMB IFNG CXCL9 CXCL10 PRF1 TBX21 Urothelium Luminal Basal Luminal I patients have low T eff gene expression In Luminal II patients with disease progression, a trend toward lower baseline T eff expression was seen, compared with Luminal II responders Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype T eff Stromal Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CD8A GZMA GZMB IFNG CXCL9 CXCL10 PRF1 TBX21 COL4A1 COL4A2 PDGFRB BGN NUAK1 Urothelium Stromal genes were also analyzed Luminal Basal COL4A, collagen type IV α; PDGFRB, platelet-derived growth factor β. Data cutoff: March 14, 2016. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype T eff Stromal Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CD8A GZMA GZMB IFNG CXCL9 CXCL10 PRF1 TBX21 COL4A1 COL4A2 PDGFRB BGN NUAK1 Urothelium Luminal Basal Subtype II patients tended to have lower stromal gene expression Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc TCGA Subtype T eff Stromal Luminal Basal I II III IV Papillary like Squamous Mesenchymal Response IC status TC status FGFR3 CDKN2A KRT5 KRT14 EGFR GATA3 FOXA1 ERBB2 CD8A GZMA GZMB IFNG CXCL9 CXCL10 PRF1 TBX21 COL4A1 COL4A2 PDGFRB BGN NUAK1 Urothelium Luminal Basal Luminal I tumors have low T eff expression Luminal II tumors have high T eff and low stromal gene expression Basal tumors have high T eff and high stromal gene expression Data cutoff: March 14, 2016. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

IMvigor210: TCGA Subtype in muc Luminal Basal I II III IV Papillary like Squamous Mesenchymal Immune desert Inflamed Immune suppressed a Response IC status TC status IMvigor210 subtypes have distinct tumor-immune landscapes that reflect responsiveness to atezolizumab Tumor stroma Tumor cells TIL/immune cells Increased responses TIL, tumor-infiltrating lymphocyte. a High myeloid, inflammatory, activated stromal/fibroblast markers. Data cutoff: March 14, 2016. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

Mutation Load by FoundationOne and Response muc has a high mutation load and thus potential for neoantigen generation and recognition by the immune system 1-3 Median load was significantly higher in responders vs non-responders This relationship was statistically independent of other predictors of response RECIST v1.1 response Responder Non-responder Mutation Load/MB 40 30 20 10 0 All (n = 150) I II Luminal III Basal IV Mutation Load/MB 50 40 30 20 10 0 IC0/1 IC2/3 1. Lawrence MS, et al. Nature. 2013;499(7457):214-218. 2. Cancer Genome Atlas Research Network. Nature. 2014;507(7492):315-322. 3. Kandoth C, et al. Nature. 2013;502(7471):333-339. Data cutoff: March 14, 2016. Rosenberg JE, et al. J Clin Oncol. 2016;34(suppl): Abstract 104.

Front-Line Therapy for Urothelial Cancer: Cisplatin-Ineligible Balar AV, et al. J Clin Oncol. 2016;34(suppl): Abstract LBA4500.

Efficacy: Change in Tumor Burden Balar AV, et al. J Clin Oncol. 2016;34(suppl): Abstract LBA4500.

Study 1108: Dose-Escalation and Dose-Expansion Study in Patients With Advanced Solid Tumors Massard C, et al. J Clin Oncol. 2016;34(suppl): Abstract 4502.

Confirmed ORR and DCR12 by PD-L1 Localization Massard C, et al. J Clin Oncol. 2016;34(suppl): Abstract 4502.

Study Design Sharma P, et al. J Clin Oncol. 2016;34(suppl): Abstract 4501.

Antitumor Activity Sharma P, et al. J Clin Oncol. 2016;34(suppl): Abstract 4501.

Progression-Free Survival Sharma P, et al. J Clin Oncol. 2016;34(suppl): Abstract 4501.

Overall Survival Sharma P, et al. J Clin Oncol. 2016;34(suppl): Abstract 4501.

ASG-22CE (ASG-22ME; Enfortumab Vedotin) in a Phase I Dose Escalation Trial in Patients (Pts) With Metastatic Urothelial Cancer CONCLUSIONS Enfortumab vedotin is well tolerated and shows encouraging anti-tumor activity with 10 PRs (ORR = 30.3%) Enfortumab vedotin shows promising anti-tumor activity in subjects with liver metastases (4/10 = 40%) and those previously treated with checkpoint inhibitors (3/12 = 25%) The ocular AEs experienced with enfortumab vedotin are of low grade and reversible with appropriate management. Preliminary PK profile for the antibody drug conjugate (ADC) reveals a short half-life of 1.5 days with no accumulation between the 1 st and 3 rd doses of cycle 1 FUTURE PLANS Enrollment at 1 and 1.25 mg/kg is ongoing to identify the maximum tolerated dose and a recommended dose for future studies Petrylak DP, et al. J Clin Oncol. 2016;34(suppl): Abstract 4532.

Pembrolizumab (MK-3475) Humanized IgG4, High-Affinity, Anti-PD-1 Antibody Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Baseline Characteristics Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Treatment Related Adverse Events Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

AEs of Interest Based On Immune Etiology Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Antitumor Activity Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Treatment Exposure and Response Duration Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Survival (Central Radiology Assessment) Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

PD-L1 IHC Staining in Bladder Cancer Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Exploratory Predictive Value of PD-L1 Scoring Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Association Between Immune-Related Gene Expression Signatures and Clinical Outcome Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Immune-Related Gene Expression Signatures Identified in Melanoma Patients Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Exploratory Analysis of the Association Between Immune-Related Gene Expression Signatures and Response Plimack ER, et al. J Clin Oncol. 2015;33(suppl): Abstract 4502.

Rational for VEGF Blockade in Bladder Cancer Antiangiogenic agents, particularly anti-vegfr-2 monoclonal antibodies (mabs), may be capable of acting as chemosensitizing agents when given in combination with docetaxel, since this effect was demonstrated in mice when an anti-vegfr-2 mab, DC101, was combined with paclitaxel Anti-VEGFR-1 mabs may inhibit metastasis, based on the observed impact of the anti-vegfr-1 mab, MF1, on VEGFR-1- positive circulating hematopoietic progenitor cells in mice

Binding of Ramucirumab to VEGFR-2 and Icrucumab to VEGFR-1 Inhibits Subsequent Signaling

Expression of VEGFR-2 in Bladder Cancer But Not in Normal Urothelium

JCDC: Study Design S C R E E N 1:1:1 R A N D O M I Z E Docetaxel 75 mg/m 2 day 1 of a 21-day cycle N = 44 Docetaxel 75 mg/m 2 day 1 + Ramucirumab 10 mg/kg day 1 of a 21-day cycle N = 46 Docetaxel 75 mg/m 2 day 1 + Icrucumab 12 mg/kg days 1 and 8 of a 21-day cycle N = 49 Treat until disease progression or intolerable toxicity Survival and safety follow-up Primary Endpoint: Progression-free survival (PFS) Secondary Endpoints: Overall survival, objective response rate, duration of response, safety, PK/PD and immunogenicity profile Stratification factors: Visceral metastasis (yes vs. no) Prior antiangiogenic therapy (yes vs. no)

Progression-Free Survival Interim Analysis

RANGE (Trial I4T-MC-JVDC): Study Design S C R E E N 1:1 R A N D O M I Z E Docetaxel 75 mg/m 2 + Placebo 10 mg/kg I.V. on day 1 of a 21-day cycle N = 262 Docetaxel 75 mg/m 2 + Ramucirumab 10 mg/kg I.V. day 1 of a 21-day cycle N = 262 Oversight by an IDMC Treat until disease progression or intolerable toxicity Primary Objective PFS Key Secondary Objectives OS and ORR Important Inclusion Criteria: Locally advanced or unresectable or metastatic UC and ECOG PS 0 or 1 Progression on or after first-line platinum-based chemotherapy ( 14 months; or 24 months if prior treatment with one immune checkpoint inhibitor) Key Exclusion Criteria: Hemoglobin < 9 g/dl Uncontrolled bleeding or thrombotic disorder Known untreated brain metastasis Petrylak DP, et al. J Clin Oncol. 2015;33(suppl 7): Abstract 295.

Phase I Trial Pembrolizumab + Ramucirumab in Metastatic Cancer PI Overall Roy Herbst Bladder Petrylak Fuchs Gastric

Conclusions Checkpoint inhibition therapy demonstrates significant antitumor activity in cisplatin treated metastatic urothelial carcinoma Phase II and III trials are ongoing to confirm initial observations of anti PD-1 and PDL-1 in metastatic urothelial carcinoma A thorough understanding of the markers of resistance and response will help to designing future trials in earlier disease