NIAGARA Trial

[Slide 1] Pre-Treatment and Presurgery ctDNA Predict Outcome KEYNOTE-B15 vs NIAGARA and Christensen 2019 KEYNOTE-B15 / EV-304 EFS Pre-chemotherapy ctDNA EV + pembro: curve labels at 24 months: [illegible, upper/ctDNA− curve]; 71.2% [ctDNA+] No. at risk (months 0 / 6 / 12 / 18 / 24 / 30 / 36 / 42 / 48 / 54): ctDNA+: 133 / 119 / 101 / 95 / 83 / 51 / 30 / 20 / 8 / 0 ctDNA−: 121 / 103 / 96 / 93 / 89 / 46 / 29 / 21 / 10 / 0 [faint] Cis + gem: curve labels at 24 months: 82.7% [ctDNA−]; 54.2% [ctDNA+] No. at risk: ctDNA+: 134 / 111 / 83 / 73 / 62 / 36 / 19 / 14 / 5 / 0 ctDNA−: 102 / 95 / 84 / 82 / 68 / 40 / 22 / 13 / 8 / 0 [faint] Pre-surgery ctDNA EV + pembro: Positive-Positive (n = 15); Negative-Negative (n = 79); curve labels at 24 months: 94.6%; 33.3% No. at risk: 15 / 12 / 8 / 6 / 5 / 3 / 2 / 1 / 1 / 0 79 / 73 / 71 / 69 / 62 / 32 / 21 / 16 / 8 / 0 Cis + gem: Positive-Positive (n = 34); Negative-Negative (n = 77); curve labels at 24 months: 87.8%; 35.3% No. at risk: 34 / 26 / 16 / 14 / 12 / 10 / 5 / 4 / 2 / 0 77 / 74 / 67 / 65 / 53 / 31 / 19 / 12 / 7 / 0 NIAGARA EFS, combined arms [top] EFS (combined arms): ctDNA− vs ctDNA+ HR, 0.42 (95% CI, 0.30-0.60) [bottom] EFS in the baseline ctDNA+ population that did or did not clear ctDNA pre-cystectomy (combined arms): ctDNA clearance vs persistent ctDNA+ HR, [illegible] (95% CI [illegible]) Christensen et al. 2019 RFS A ctDNA Before Chemotherapy — RFS (probability) vs Time Since Cystectomy (days); P < .001 No. at risk (0 / 300 / 600 / 900 days): Negative 35 / 32 / 23 / 1; Positive 24 / 16 / 4 / 0 B ctDNA Before Cystectomy — P < .001 No. at risk: Negative 55 / 49 / 29 / 1; Positive [illegible] Survival table Endpoint: EFS (KEYNOTE-B15/EV-304) by pre-chemotherapy and pre-surgery ctDNA status; NIAGARA EFS; Christensen 2019 RFS KEYNOTE-B15 EV + pembro pre-surgery Positive-Positive | 15 KEYNOTE-B15 EV + pembro pre-surgery Negative-Negative | 79 KEYNOTE-B15 Cis + gem pre-surgery Positive-Positive | 34 KEYNOTE-B15 Cis + gem pre-surgery Negative-Negative | 77 KEYNOTE-B15 EV + pembro pre-chemo ctDNA+ | 133 KEYNOTE-B15 EV + pembro pre-chemo ctDNA- | 121 KEYNOTE-B15 Cis + gem pre-chemo ctDNA+ | 134 KEYNOTE-B15 Cis + gem pre-chemo ctDNA- | 102 HR: NIAGARA EFS ctDNA- vs ctDNA+ 0.42 HR CI: 0.30-0.60 P: Christensen 2019: P < .001 (both panels) Landmark 24 mo | KEYNOTE-B15 EV + pembro pre-chemo ctDNA+ | 71.2% Landmark 24 mo | KEYNOTE-B15 Cis + gem pre-chemo ctDNA- | 82.7% Landmark 24 mo | KEYNOTE-B15 Cis + gem pre-chemo ctDNA+ | 54.2% Landmark 24 mo | KEYNOTE-B15 EV + pembro pre-surgery Negative-Negative | 94.6% Landmark 24 mo | KEYNOTE-B15 EV + pembro pre-surgery Positive-Positive | 33.3% Landmark 24 mo | KEYNOTE-B15 Cis + gem pre-surgery Negative-Negative | 87.8% Landmark 24 mo | KEYNOTE-B15 Cis + gem pre-surgery Positive-Positive | 35.3% [Slide 2] Postsurgery ctDNA Findings Replicate NIAGARA KEYNOTE-B15 vs NIAGARA KEYNOTE-B15 / EV-304 study scheme [small schematic: Key eligibility: Adults with MIBC, cisplatin eligible, RC + PLND candidate; R 1:1; Neoadjuvant Enfortumab vedotin + Pembrolizumab N=405 → RC + PLND → Adjuvant Enfortumab vedotin + Pembrolizumab [highlighted]; Cisplatin + Gemcitabine N=403 → Observation; ctDNA assessed at baseline, presurgery, postsurgery and adjuvant] NIAGARA study scheme [small schematic: Durvalumab arm — Neoadjuvant Durvalumab + Gemcitabine + cisplatin → Radical cystectomy → Adjuvant Durvalumab [highlighted]; Comparator arm — Gemcitabine + cisplatin → No treatment; plasma samples analysed at baseline, pre-RC, C1D1; study population text too small to read] KEYNOTE-B15 / EV-304 EFS per BICR by postsurgery ctDNA status | Events, n (%) | HR (95% CI)a ctDNA+ EV + pembro (n = 10) | 9 (90.0) | 0.63 (0.30-1.35) Cis + gem (n = 30) | 27 (90.0) ctDNA− EV + pembro (n = 199) | 22 (11.1) | 0.48 (0.29-0.81) Cis + gem (n = 193) | 41 (21.2) Curve labels at 24 months: 89.9% [EV + pembro ctDNA−]; 82.2% [Cis + gem ctDNA−]; 10.0%; 10.0% [ctDNA+ arms] Y-axis: Event-Free Survival, % X-axis: Months No. at risk (months 0 / 6 / 12 / 18 / 24 / 30 / 36 / 42 / 48 / 54): EV + pembro: ctDNA+: 10 / 9 / 3 / 2 / 1 / 0 / 0 / 0 / 0 / 0 EV + pembro: ctDNA−: 199 / 197 / 186 / 180 / 162 / 91 / 58 / 38 / 16 / 0 Cis + gem: ctDNA+: 30 / 22 / 5 / 3 / 3 / 2 / 2 / 1 / 1 / 0 Cis + gem: ctDNA−: 193 / 191 / 173 / 162 / 131 / 75 / 43 / 26 / 12 / 0 NIAGARA7 DFS by post-RC ctDNA status, combined arms Curves: ctDNA−; ctDNA+ Y-axis: Probability of DFS X-axis: Time from radical cystectomy (months) [No. of patients at risk row too small to read] ctDNA− vs ctDNA+ HR, 0.09 (95% CI, 0.05–0.18) aUnstratified Cox proportional hazards model with treatment as a covariate. Data cutoff date: October 27, 2025. NIAGARA: DFS from radical cystectomy, ctDNA− vs ctDNA+ HR 0.09 (95% CI 0.05–0.18); Powles T, et al. J Clin Oncol. 2025;43(16 Suppl):4503. Survival table Endpoint: KEYNOTE-B15/EV-304 EFS per BICR by postsurgery ctDNA status; NIAGARA DFS by post-RC ctDNA status (combined arms) EV + pembro ctDNA+ | 10 Cis + gem ctDNA+ | 30 EV + pembro ctDNA- | 199 Cis + gem ctDNA- | 193 HR: ctDNA+: 0.63; ctDNA-: 0.48 (EV + pembro vs cis + gem); NIAGARA ctDNA- vs ctDNA+ DFS 0.09 HR CI: ctDNA+: 0.30-1.35; ctDNA-: 0.29-0.81; NIAGARA: 0.05-0.18 Landmark 24 mo | EV + pembro ctDNA- | 89.9% Landmark 24 mo | Cis + gem ctDNA- | 82.2% Landmark 24 mo | ctDNA+ (both arms) | 10.0% / 10.0%

[Slide 1] 13:30 - 15:00 Proffered paper 1: Genitourinary tumour, renal and urothelial CHAIRS: AXEL BEX, ANJA LORCH Is the combination hypothesis incorrect? pCR outcomes were consistent across trials KEYNOTE-866 [Bar chart; y-axis: pCR Rate, % (95% CI)] Pembro + cis + gem vs Placebo + cis + gem Estimated difference:b 8.9% (95% CI, 2.5-15.1)* CA017-078 pCR overall (co-primary endpoint)a 9.5% (95% CI 2.3-16.7) (99.91% CI −2.6 to 21.6)b [Bars; y-axis: Patients (%)] 39.1 | 29.6 Overall ITT NIAGARA Re-analysis (Apr 2024) Odds ratio 1.60 (95% CI, 1.23-2.08) nominal P=0.0005 Durvalumab arm N=533: 37.3% 199/533 95% CI, 33.2-41.6 Comparator arm N=530: 27.5% 146/530 95% CI, 23.8-31.6 [y-axis: pCR rate (%)] Robert Jones Invited Discussant 47350, 47360 and 47420 BARCELONA AUDITORIUM - HALL 9 [Slide 2] 13:30 - 15:00 Proffered paper 1: Genitourinary tumour, renal and urothelial CHAIRS: AXEL BEX, ANJA LORCH Were populations different? EFS outcomes were good in control arms of these studies [Thumbnail 1] Primary End Point: EFS by Investigator Assessment, ITT Population — Pembro + cis + gem n = 453 vs Placebo + cis + gem n = 454; Events, n (%) 157 (34.7) vs 168 (37.0); Median (95% CI), mo 74.1 (72.4-NR) vs NR (67.7-NR); HR 0.88 (95% CI, 0.71-1.10), 1-sided P = 0.1366*; landmark labels 76.7%/70.0%, 69.2%/66.3%, 66.0%/62.3%, 63.5%/61.9%; No. at risk Pembro 453 401 366 338 329 304 296 273 224 175 114 51 11 0; Placebo 454 400 345 310 300 291 283 257 208 168 106 51 12 0 [Thumbnail 2] Event-free survival by treatment arm (co-primary endpoint) — NIVO+GC (N = 325) vs GC (N = 328); events 104/325 (32.0) vs 104/328 (31.7); median NR vs 67.6 (63.0-NE); HR (95% CI) 0.94 (0.72-1.24); HR (95.74% CI) 0.94 (0.71-1.25); P 0.6697; landmark labels 82.6%/78.6%, 72.3%/71.6%, 68.1%/68.5%, 67.3%/65.8%; No. at risk NIVO+GC 325 283 243 227 209 197 189 146 100 60 25 5 0 0; GC 328 271 219 199 187 176 168 131 90 61 23 4 1 0 KEYNOTE-866 | CA017-078 | NIAGARA T2 (%) 15 | 76 | 40 N1 (%) 0.7 | Excluded | 5 ECOG 0 (%) 76 | - | 78 24 month EFS in control arm (%) 70.0 | 71.6 | 59.8 Robert Jones Invited Discussant 47350, 47360 and 47420 BARCELONA AUDITORIUM - HALL 9 [Slide 3] 13:30 - 15:00 Proffered paper 1: Genitourinary tumour, renal and urothelial CHAIRS: AXEL BEX, ANJA LORCH How I explain these two studies • More similarities than differences with NIAGARA • Patient mix had better than expected prognosis - this makes it harder to demonstrate modest improvement • Small imbalances, unseen at baseline, may explain different findings between the trials Robert Jones Invited Discussant 47350, 47360 and 47420 BARCELONA AUDITORIUM - HALL 9 [Slide 4] 13:30 - 15:00 Proffered paper 1: Genitourinary tumour, renal and urothelial CHAIRS: AXEL BEX, ANJA LORCH Were studies designed differently? Callouts: Only KN-866 used placebo | Differences in cisplatin eligibility | Only CA017-078 excluded N1 | Subtle differences in primary endpoint definitions | Some differences in statistical design [Panel] Phase 3 KEYNOTE-866 Study Design (NCT03924856) — Adults with MIBC; clinical stage T2-T4aN0M0 or T1-T4aN1M0; urothelial histology ≥50%; eligible for RC + PLND; did not meet Galsky criteria for cisplatin ineligibility incl. GFR ≥60 mL/min; ECOG PS 0 or 1. R 1:1: n = 453 Pembrolizumab 200 mg + cisplatin 70 mg/m2 + gemcitabine 1000 mg/m2 x4 → RC + PLND → Pembrolizumab 200 mg Q3W 13 cycles; n = 454 Placebo + cisplatin + gemcitabine → Placebo Q3W 13 cycles. Primary end point: EFS by investigator assessment. Key secondary: OS and pCR (pT0N0) by central pathologist review. Other secondary: Safety. [Panel] Study design: CA017-078 — Phase 3 randomized trial of neoadjuvant NIVO plus GC, followed by adjuvant NIVO, vs neoadjuvant GC alone for previously untreated MIBC (NCT03661320). Key inclusion: MIBC (predominant urothelial component); cT2-T4a, N0, M0; GFR ≥ 50 mL/min; ECOG PS 0-1. R 1:1: GC Q3W x 4 cycles + NIVO 360 mg Q3W x 4 cycles (N = 325) → radical cystectomy → NIVO 480 mg Q4W x 9 cycles; GC Q3W x 4 cycles (N = 328) → radical cystectomy → no further treatment. [Greyed arm: GC + NIVO 360 mg + BMS986205 (blinded) 100 mg QD → NIVO 480 mg Q4W + BMS986205 — TERMINATED (Dec 2021), excluded from formal testing of NIVO+GC vs GC alone.] Stratification: clinical stage (T2 vs T3/T4a); GFR (50 to <60 mL/min vs ≥60 mL/min); PD-L1 status (≥1% vs <1%/indeterminate/NE) by central review. Co-primary endpoints: pCR and EFS per BICR. Secondary: OS, safety. Exploratory: ctDNA analyses. Median follow-up (range): 54.5 months (39.4-77.0). [Panel] NIAGARA: Study Design — Adults; cisplatin-eligible MIBC (cT2-T4aN0/1M0); UC or UC with divergent differentiation or histologic subtypes; evaluated and confirmed for RC; CrCl of ≥40 mL/min. R 1:1: Durvalumab arm N=533: neoadjuvant Durvalumab 1500 mg IV Q3W + Gemcitabine + cisplatin (4 cycles) → radical cystectomy → adjuvant Durvalumab 1500 mg IV Q4W (8 cycles); Comparator arm N=530: Gemcitabine + cisplatin → radical cystectomy → no treatment. Dual primary endpoints: EFS, pCR. Key secondary endpoint: OS. Safety. EFS was defined as: progressive disease that precluded RC; recurrence after RC; date of expected surgery in patients who did not undergo RC; death from any cause. Placebo, primary endpoints Robert Jones Invited Discussant 47350, 47360 and 47420 BARCELONA AUDITORIUM - HALL 9

[Header row] ENERGYZE (n=646) Nivolumab | KN866(n=907) Pembrolizumab | NIAGARA (n=1063) Durvalumab | B15/EV304 (n=808) EV + pembro Study arm | GCN | GC | GCP | GC | GCD | GC | EVP | GC Medan FU | 54 months | 61 months | 46 months | 33 months pCR/ctDNA clearance | 39% 82% | 29% 58% | 44% | 36% | 37% | 27% | 56% | 33% 2 yr EFS | 72% | 72% | 76% | 70% | 68% | 60% | 79% | 66% 2 yr OS | 85% | 85% | 83% | 82% | 74% | 68% | 87% | 81% EFS HR | 0.94 | 0.88 | 0.68 | 0.53 OS HR | 0.88 | 1.00 | 0.75 | 0.6, Competed IO | 56% | 46% | 54% | 51% surgery | 90% | 88% | 91% | 87% | 88% | 85% | 87% | 91% T2 at | 77% | 15% | 40% | 20%