Great Debates & Updates in Solid Tumors 2026

[Slide 1] Landscape of Tumor-Agnostic Therapies Approved by US FDA and EMA Timeline of approvals (FDA/EMA): May 2017 - Pembrolizumab (dMMR/MSI-H) Nov 2018 - Larotrectinib (NTRK fusions) Aug 2019 (Jul 2019 EMA) - Entrectinib (NTRK fusions) Jun 2020 (May 2020 EMA) - Pembrolizumab (TMB-H) Aug 2021 - Dostarlimab (dMMR/MSI-H) Jun 2022 - Dabrafenib/Trametinib (BRAF V600E mutation) Sep 2022 - Selpercatinib (RET fusions) Apr 2024 (Mar 2024 EMA) - Trastuzumab deruxtecan (HER2 IHC3+) Jun 2024 - Repotrectinib (NTRK fusions) (Photo also shows two people posing in front of an Oncology Learning Network Great Debates Solid Tumors podium sign.)

[Slide 1] Every Waterfall Plot and CT Scan Has a Patient and a Family behind It [Photo of a group of physicians/staff in white coats posing with a patient] Book cover shown: "Like A Needle In A Haystack — My Survival from Stage 4 Pancreatic Cancer" by Allison Kuban (wristband reading #ALLYFORALL visible on cover) (Wide shot of ballroom screen — Oncology Learning Network Great Debates Solid Tumors branding visible)

[Slide 1] Response Rates with Dabrafenib + Trametinib across BRAF V600E-Mutated Cancers (ROAR Basket Trials and Others) ORR (%) by cancer type: ROAR cohort: ATC: 56.0 (n=36) Biliary tract cancer: 53.0 (n=43) High-grade glioma: 33.0 (n=45) Low-grade glioma: 54.0 (n=13) Adenocarcinoma small intestine: 67.0 (n=3) Hairy cell leukemia: 89.0 (n=55) Multiple myeloma: 50.0 (n=10) Adult Trials: Melanoma: 69.0 (n=211) NSCLC: 68.4 (n=93) Tumor cohorts in the NCI-MATCH study: 37.9 (n=29) DTC: 30.0 (n=27) CRC: 7.0 (n=43) Pediatric Trials: Langerhans cell histiocytosis, pediatric: 58.0 (n=12) Low-grade glioma, pediatric: 47.0 (n=?) Glioma, pediatric: 52.8 (n=?) ORR = objective response rate; ATC = anaplastic thyroid cancer; DTC = differentiated thyroid cancer; CRC = colorectal cancer. Subbiah V, et al. Nat Med. 2023;29(5):1102-1112. --- [Slide 2] Tumor Agnostic Activity of Selpercatinib in Patients with RET+ Cancers A. Response per independent review committee / B. Response per investigator assessment — waterfall plots of percentage change from baseline in sum of diameters, by tumor type (Pancreatic, Colon, Cholangiocarcinoma, Sarcoma, Rectal neuroendocrine, Small intestine, Unknown primary, Ovarian, Breast, Carcinoid, Salivary). - RET fusions are tissue-agnostic targets - FDA approval in RET fusion-positive NSCLC and thyroid cancer = May 8, 2020 - FDA approval in all RET fusion positive cancers = Sept 21, 2022 - FDA approval in all RET fusion positive pediatric cancers (>2 yrs) = May 29, 2024 Subbiah V, et al. Lancet Oncol. 2022;23(10):1261-1273. --- [Slide 3] Every Waterfall Plot and CT Scan Has a Patient and a Family behind It [Photo of a group of physicians/staff in white coats posing with a patient] Book cover shown: "Like A Needle In A Haystack — My Survival from Stage 4 Pancreatic Cancer" by Allison Kuban (wristband reading #ALLYFORALL visible on cover)

[Slide 1] CALL TO ACTION!!! Transformative acceleration of tissue-agnostic drug development from today's 10 approvals to 50-100 within 25 years Right Target + Biology + Patient -> Right Drug -> Clear Benefit -> Drug Approval "SOME DAY I'M GOING TO climb Everest" - Sir Edmund Hillary (image of hands holding a plant against a mountain backdrop) --- [Slide 2] 5 Strategic Imperatives for Tumor-Agnostic Precision Oncology 1. Genomic testing as standard of care, not last resort 2. Building molecular fluency in clinical practice 3. Revolutionizing clinical trial design for molecular medicine 4. Transforming regulatory frameworks for molecular precision 5. Real-world evidence as continuous learning engine --- [Slide 3] A Vision for Precision Oncology and Rethinking Drug Development [Illustration: world map puzzle being assembled by figures labeled Researchers, Clinicians, Clinical trialists, Payors, Funding agencies, Regulatory agencies, Other health care providers, Patients & caregivers, Governments, Patient advocates, Societies, Industry, Academia, Pharmaceutical companies] "We choose to go to the moon" "If not us, who? If not now, when?" - John F. Kennedy COLLABORATION IS KEY! CALL TO ACTION!!! Transformative acceleration of tissue-agnostic drug development from today's 10 approvals to 50-100 within 25 years Subbiah V, et al. Cancer Discov. 2024;14(4):579-584.

[Slide 1] Risk of Recurrence with Re-Challenge Is High Table — Adverse drug reaction / No. of cases / Recurrence rate (95% CI), %: Diabetes: 13 / 0 (0-27) Neurologic: 17 / 6 (0-29) Uveitis: 11 / 9 (0-40) Adrenal: 40 / 12 (5-27) Pancreatitis: 13 / 15 (3-43) Thyroiditis: 60 / 17 (9-28) Hypophysitis: 23 / 26 (12-47) Hepatitis: 31 / 29 (16-47) Hematologic: 10 / 30 (10-61) Pneumonitis: 101 / 34 (25-43) Colitis: 123 / 37 (29-45) Skin: 16 / 38 (18-61) Arthritis: 29 / 45 (28-62) - 30-35% of patients experienced recurrence of the same irAE - 68% of patients with G3+ irAE who were re-challenged had recurrence - 5% of patients experience a different irAE - 20-25% experience G3+ irAE - Shorter time to initial irAE associated with higher likelihood of recurrence Dolladille C, et al. JAMA Oncol. 2020;6(6):865-871. Mizuno K, et al. Jpn J Clin Oncol. 2026;56(2):117-129. Simonaggio A, et al. JAMA Oncol. 2019;5(9):1310-1317.

[Slide 1] Prioritize Efficacy of Osimertinib Alone (for Select Patients) as Front-Line Therapy Helena Yu, MD Thoracic Medical Oncologist and Early Drug Development Memorial Sloan Kettering Cancer Center New York City, New York --- [Slide 2] Options for First-Line Treatment for EGFR+ LC FLAURA (Osi mono): Progression-free survival — Osimertinib 18.9 mo, 1st-gen TKI 10.2 mo FLAURA 2 (Osi + chemo): Progression-free survival — Osi+chemo 25.5 mo, Osimertinib 16.7 mo MARIPOSA (Ami + Lazer): Progression-free survival — Ami+Laz 23.7 mo, Osimertinib 16.6 mo NSCLC = non-small cell lung cancer; TKI = tyrosine kinase inhibitor. Soria JC, et al. N Engl J Med. 2018;378(2):113-125. Planchard D, et al. N Engl J Med. 2023;389(21):1935-1948. Cho BC, et al. N Engl J Med. 2024;391(16):1486-1498. --- [Slide 3] Considerations for 1L Treatment Side effects / Time and effort / Financial cost / Quality of life <—> Progression-free survival / Overall survival / Ability to sequence treatments / CNS outcomes 1L = first-line; CNS = central nervous system. --- [Slide 4] Considerations for 1L Treatment Low-risk case: 76 yo, EGFR ex19 deletion only, Asymptomatic, Oligometastatic disease, Thoracic only disease, Slow growing, ctDNA neg, On osimertinib x 4 years High-risk case: 52yo, EGFR G719A, TP53, RB1, High symptom burden, Diffuse mets including brain, liver, bone, Large tumor burden, ctDNA pos at 3 weeks, Progression within 4 mo on osimertinib LOW RISK <-> HIGH RISK spectrum: Osimertinib alone -- Osimertinib + VEGF -- Osimertinib + chemotherapy / Amivantamab + lazertinib "When balancing toxicity, benefit and outcomes, I believe osimertinib monotherapy should be the default and standard for most. It aligns with many patients' desires." PFS chart: Osimertinib+Platinum-Pemetrexed 25.5 mo (24.7-NC) vs Osimertinib 16.7 mo (14.1-21.3), difference 8.8 mo, HR 0.62 (95% CI 0.49-0.79); combination therapy 57%/41% vs monotherapy at 24/33 months CtDNA clearance bar chart (Osimertinib n=238): Combination therapy 24% non-clearance; Osimertinib monotherapy 45% clearance, 32% baseline non-detectable VEGF = vascular endothelial growth factor. Planchard D, et al. N Engl J Med. 2023;389(21):1935-1948. Gray JE, et al. Clin Cancer Res. 2023;29(17):3340-3351.

[Slide 1] [Discussion questions slide, styled as a checklist with a Devil Wears Prada-themed image captioned "I have four questions."] 1. Did FLAURA2 improve OS vs osimertinib? 2. Did MARIPOSA improve OS vs osimertinib? 3. Can you guarantee every patient receiving osimertinib alone will remain fit enough to receive every treatment you're saving for later? 4. So why should monotherapy automatically remain the default?

[Slide 1] First-Line Strategies in EGFR+ Non-Small Cell Lung Cancer: Combination Therapy Eric K. Singhi, MD Medical Oncology Lead, Young-Onset Lung Cancer Program Assistant Professor UT MD Anderson Cancer Center @lungoncdoc --- [Slide 2] Or... since we are in NYC [Image: mock movie poster "EGFR WEARS PRADA 2" styled after The Devil Wears Prada, cast labeled Meryl Streep, Anne Hathaway, Emily Blunt, Stanley Tucci] Why monotherapy may be so last season In theaters May 1 --- [Slide 3] FLAURA2: Overall Survival CTx was the most common FST (74%) after osi + CTx — 44% of FSTs were rechallenge with platinum CTx OS benefit with osi + CTx was observed despite SoC CTx being the most common FST after osi mono Osi + CTx (n=127): No FST vs FST — post-discontinuation therapy breakdown: 14% other, 5%, 8%, 44% platinum-based CTx Osi mono (n=185): 14%, 3%, 7%, 3%; 72% of those who received FST (n=143) got platinum-based chemo "And upfront chemotherapy STILL improved OS." "But why not just save chemotherapy for later?" Legend: Platinum-based CTx / Non-platinum-based CTx / EGFR targeted therapy (other than osi), mono or combo / Osi + targeted agent/investigational drug (no CTx) / Other [Janne PA, et al.] N Engl J Med. 2026;394(1):27-38.

[Slide 1] Oligometastatic vs Oligoremnant vs Oligoprogressive 1. Oligometastatic Disease (De Novo / Baseline) — Limited metastatic burden from the time of diagnosis. 2. Oligoremnant Disease (Induced State) — Polymetastatic at baseline; Systemic therapy cleared most, leaving isolated remnants. 3. Oligoprogressive Disease (Resistant Subclones) — Systemic treatment controls most, but an isolated subclone escapes and grows. Imagine created with Gemini. --- [Slide 2] Local Consolidation Therapy in EGFR-Mutant Lung Cancer (NorthStar) Randomized phase 2 trial Key Eligibility: Locally advanced or metastatic NSCLC; ECOG PS 0 or 1; Measurable disease; TKI naive EGFR Ex19del or L858R OR Acquired T790 no prior 3rd gen TKI Induction osimertinib 6-12 weeks -> Non-PD, R 1:1, N=120 -> Osimertinib vs Osimertinib plus LCT Primary Endpoint: PFS LCT modality n=56: Radiation 33 (58.9%), Surgery 17 (32.1%), Surgery and radiation 6 (9%) Surgical procedure n=23: Lobectomy 18 (78.3%), Lobectomy and wedge resection 1 (4.3%), Wedge resection 2 (8.7%), Segmentectomy 1 (4.3%), Adrenalectomy 1 (4.3%) Radiation modality n=50, 39 patients: VMAT or IMRT 28 (56%), SBRT 15 (30%), 3D conformal or 2D conformal 7 (14%) PFS = progression-free survival; IMRT = intensity-modulated radiotherapy; VMAT = volumetric-modulated arc therapy; LCT = local consolidative therapy. Elamin YY, et al. Presented at: ESMO Congress; Oct 17, 2025; Berlin, DE. LBA72. --- [Slide 3] How Do These Trials Compare with Systemic Therapy Alone in the Modern Era of Combination Therapy? Clinical Trial / Phase & Design / Interventions Evaluated / Median PFS (95% CI) / Median OS (95% CI): FLAURA2 — Ph III, global, open-label — Osimertinib+Chemo vs Osimertinib mono — 29.4 mo (25.1-NE) [BICR] / 25.5 mo (24.7-NE) [Inv] — 47.5 mo (43.4-NE) vs 37.6 mo (33.9-43.1) MARIPOSA — Ph III, global, open-label — Amivantamab+Lazertinib vs Osimertinib mono — 23.7 mo (19.1-27.7) vs 16.6 mo (14.8-18.5) — Not Yet Reached (42.9-NE) vs 36.7 mo (33.4-41.0) Sampath — Ph II, multicenter, single-arm — Osimertinib induction + Consolidative RT at 8-10 wk — 32.3 mo (21.9-51.7) — 45.0 mo (39.3-56.4) NORTHSTAR — Ph II, multicenter, randomized — Osimertinib+LCT at 6-12 wk vs Osimertinib mono — 25.3 mo (19.4-45.0) vs 17.5 mo (14.5-24.3) — N/A Janne PA, et al. N Engl J Med. 2026;394(1):27-38. Yang JC, et al. N Engl J Med. 2025;393(17):1681-1693. Sampath S, et al. EClinicalMedicine. 2025;87:103435. Elamin YY, et al. Presented at: ESMO Congress; Oct 17, 2025; Berlin, DE. LBA72.

[Slide 1] The Spectrum of EGFR Mutations Pie chart: Classical (incl. T790M) 69.2%, PACC (incl. compound) 12.5%, Other uncommon (incl. L861Q) 9.2%, Exon 20 insertions 9.1% Exon map: Exon 18 (E709_T710delinsD 4%, E709X 10.3%, G719X 40.7%, G724S* 1.6%), Exon 19 (I740_K745dup 1.7%, L747X 3.1%), Exon 20 (S768I 20.7%, V769X 4%, V774M 2.8%, R776X 3.4%, G779F 1.7%), Exon 21 (C797S* 3%) PACC Ex G719X: Afatinib alone; Off-label Osimertinib +/- chemotherapy; Off-label amivantamab + lazertinib; Clinical trials Classical-Like Ex: L861Q: Reasonable to use similar strategy to L858R/del19; Eg, osimertinib monotherapy or 1L combinations 1L = first-line. Le X, et al. Presented at IASLC WCLC 2024. Abstract PL04.07. Robichaux JP, et al. Nature. 2021;597(7878):732-737. Yang JCH, et al. Lancet Oncol. 2015;16(7):830-838. Yang G, et al. Lung Cancer. 2020:145:186-194. Bar J, et al. J Oncol. 2021;2021:7836264. Cho BC, et al. J Clin Oncol. 2024;42(16 Suppl):8516. --- [Slide 2] The Spectrum of KRAS Inhibitors 1. OFF-State Inhibitors (G12C Selective) — Bind inactive GDP-bound KRAS G12C — Switch-II pocket 2. ON-State Inhibitors (Tri-Complex) — Bind active GTP-bound KRAS using cyclophilin A (CypA) — Tri-complex formation blocks effector binding 3. Dual-State Inhibitors (Non-Tri-Complex) — Bind KRAS in both GDP and GTP states directly 4. Pan-KRAS / Pan-RAS Inhibitors — Inhibit multiple KRAS mutations or RAS isoforms — Broad binding across KRAS variants or RAS isoforms 5. Mutation-Specific (Non-G12C) Inhibitors — Selective for specific KRAS mutations (G12D, G12V, G13C) 6. KRAS Degraders (PROTAC / Molecular Glue) — Induce degradation of KRAS protein — Degrader brings KRAS to E3 ligase -> ubiquitination and proteasomal degradation Image generated with AI, referenced from Singhal A, et al. Nat Med. 2024;30(4):969-983. --- [Slide 3] Significant and Increasing Impact on [Outcomes] for Those Patients with Actionable Driver Oncogenes Bar chart of PFS (months) by driver oncogene: ALK ~59, ROS1 ~46, NTRK ~35, BRAF ~30, EGFR ~25, RET ~18, HER2 ~13, MET ~12, KRAS ~7, Chemo ~4 Covering today: EGFR uncommon; EGFR exon 20ins; KRAS; ALK; RET PFS = progression-free survival. Soria JC, et al. N Engl J Med. 2018;378(2):113-125. Wolf J, et al. J Clin Oncol. 2021;39(Suppl 15):9020. Mok TSK, et al. Ann Oncol. 2020;31(8):1056-1064. Drilon A, et al. N Engl J Med. 2020;383(9):813-824. Drilon A, et al. JTO Clin Res Rep. 2022;3(6):100332. Drilon A, et al. J Clin Oncol. 2023;41(2):385-394. Li F, et al. Front Oncol. 2021;27(4):913-915. Drilon A, et al. N Engl J Med. 2024;390(2):118-131. Riely GJ, et al. Ann Oncol. 2024;35(Suppl 2):S1246-S1247. Planchard D, et al. N Engl J Med. 2023;389(21):1935-1948.

[Slide 1] A Tale of Two Climates NEW ENGLAND — Local ablation / radiation: "Chip ice off one window at a time, spot by spot. North Easter - keeps coming." MIAMI — Systemic therapy: "Warm the whole house at once, room and the attic (the brain). D[...]" [text cut off at right edge of frame] "Residency in New England, fellowship in Miami. This debate is New E[ngland vs Miami:] two ways to warm a cold house." [quote partially cut off]

[Slide 1] [Top half of image: selfie of two attendees at the conference, not a data slide] Is There a Role for Adjuvant HAI Following Liver Resection vs Adjuvant Chemotherapy? Maybe, in selected patients ... still to be determined [Illustration: boxing-themed cartoon, referee raising a fighter's glove, fighter resting on a stool]

[Slide 1] Are RAS Inhibitors Ready for Prime-Time Use as a Standard of Care in Pancreatic Cancer?: Yes! Olatunji Alese, MD FWACS FASCO Professor & Director of GI Oncology Department of Hematology and Medical Oncology Lead, Winship GI Disease Team Winship Cancer Institute of Emory University --- [Slide 2] Audience live-poll results displayed on two side screens: "Session 7: Are RAS Inhibitors Ready for Prime-Time Use as a Standard of Care in Pancreatic Cancer?" — Bar chart: No 75%, Yes 25% (shown on both screens, panelists seated at dais below).

[Slide 1] Adverse Events with DLL3-Targeting T-Cell Engagers Cytokine Release Syndrome (CRS): Occurs early in the majority of patients (~60%), mostly low grade; Step-in dosing; Admission for initial infusions; Anti-IL6 drugs; No grade >=3 CRS in combination with chemo in DAREON-7 Immune effector cell-associated neurotoxicity syndrome (ICANS): Occurs in ~10%; Supportive care, steroids if grade 2; No >=2 ICANS in combination with chemo in DAREON-7 Other: Fever; Dysgeusia; Decreased appetite; Asthenia, fatigue; Tumor flare

[Slide 1] [Top half] DLL3-Targeting ADCs in NEC Table — Agent / Payload / Linker / Indication / Phase / Sponsor / NCT: Zocilurtatug pelitecan (ZL-1310) — DNA topoisomerase I inhibitor (YL0010014) — Cleavable — NECs excluding SCLC — Ib/II — Zai Lab — NCT06885281 BL-M14D1 — DNA topoisomerase I inhibitor (Ed-04) — Cleavable — SCLC, NEC — I — SystImmune Inc. — NCT07080242 IDE849 — DNA topoisomerase I inhibitor — Cleavable — SCLC, High-grade NEC, Other DLL3+ tumors — I/II — IDEAYA Biosciences — NCT07174583 [Bottom half] Early Results from DLL3 Targeting ADCs in EP-NEC Zocilurtatug pelitecan (ZL-1310) Phase 1b: Total (n=34): ORR 38.2%; GEP-NEC (n=18): ORR 33%; Other DLL3+NEC (n=16): ORR 43.8%; No clear association between response and DLL-3 expression; Gr >=3 TEAEs in 30.4% (mostly hematologic); Study is ongoing, expansion into phase 2; FDA fast-track designation for treatment of EP-NEC following progression on 1st-line Tx BL-M14D1 — Phase 1 in SCLC, NEC: Enrolled 40 pts with NEC, 87 with SCLC; Efficacy at 4.0 and 4.5 mg/kg presented; In 22 evaluable pts with NEC: ORR 40.9%; cORR 27.3%; Gr >=3 TRAEs: thrombocytopenia (54.3%), neutropenia (53.5%), leukopenia (46.5%), anemia (32.3%); Gr 3 ILD in 1; 1 treatment-related death TEAE = treatment-emergent AE; Tx = treatment; cORR = confirmed ORR; ILD = interstitial lung disease. Thummalapalli R, et al. Presented at: American Association for Cancer Research (AACR) Annual Meeting; 2026.

[Slide 1] [Top half] Clinicopathologic and Imaging Features of NEN Well-differentiated Gastroenteropancreatic NET (Grade G1 low / G2 intermediate / G3 high) vs Poorly differentiated NEC (High grade by definition): Mitotic count: G1 <2/2mm2, G2 2-20/2mm2, G3 >20/2mm2; NEC high grade by definition Ki-67: G1 <3%, G2 3-20%, G3 >20%; NEC often >70% Clinical course: G1 Indolent, G2 Intermediate, G3 Intermediate-Rapid; NEC Rapid Nuclear Imaging Features: Somatostatin receptor (SSTR)+ dotatate PET CT (well-differentiated) trending to FDG-PET avid (+) (poorly differentiated) Mutations: DAXX/ATRX, MEN-1, mTOR pathway (pNET), CDKN1B (SI-NET) [well-differentiated] vs TP53, RB1 [NEC] [Bottom half] Emerging Role of DLL3 in EP-NEC - Notch signaling is involved in cell fate and differentiation, proliferation, angiogenesis - Dysregulation of Notch signaling is implicated in various cancers, particularly neuroendocrine malignancies — Notch inhibition -> ASCL1 expression - DLL3 is an inhibitory ligand of Notch [Diagram of Notch Ligand/Receptor signaling pathway showing DLL3, ADAM10/17, gamma secretase, Canonical Notch signaling, HES1/HEY1/ASCL1/DLL3 gene regulation]

[Slide 1] T-cell engagers / bispecifics - Tarlatamab (Imdelltra): DLL3 x CD3 BiTE - Obrixtamig (BI 764532): DLL3 x CD3 TCE - Alveltamig (ZG006): DLL3 x DLL3 x CD3 trispecific TCE - Gocatamig (MK-6070 / HPN328): DLL3 x CD3 x albumin TriTAC - Peluntamig (PT217): DLL3 x CD47 bispecific - RO7616789: DLL3 x CD3 x 4-1BB trispecific TCE - QLS31904: DLL3 x CD3 TCE - IBI115: DLL3 x CD3 TCE - SHR-7787: DLL3 x CD3 TCE ADCs - SHR-4849 / IDE849: DLL3 ADC (TOP1 inhibitor) - ZL-1310 / zocilurtatug pelitecan: DLL3 ADC (TOP1 inhibitor) - IBI3009 / RG6810: DLL3 ADC - FZ-AD005: DLL3 ADC (DXd / TOP1 inhibitor)

[Slide 1] Summary of Phase 3 Trials for cc-RCC CLEAR (Pembrolizumab + Lenvatinib, N=1069): mPFS 23.9 mo (v 9.2); HR 0.47 (0.38-0.57); mOS 53.7 mo, HR 0.79 (0.63-0.99); ORR/CR 71.3/18.3%; Sarcomatoid features 7.9%; %TEAEs leading to DC IO or TKI 37.2%; IMDC/MSKCC Risk F/I/P 27/63.9/9; Median follow-up 49.8 mo CHECKMATE-9ER (Nivolumab + Cabozantinib, N=651): mPFS 16.4 mo (v 8.3); HR 0.58 (0.49-0.70); mOS 46.5 mo, HR 0.79 (0.65-0.96); ORR/CR 55.7/13.9%; Sarcomatoid features 11%; %TEAEs leading to DC IO or TKI 28%; IMDC/MSKCC Risk F/I/P 23/58/19; Median follow-up 67.6 mo CHECKMATE-214 (Ipilimumab + Nivolumab, N=1096, I/P only): mPFS 12.4 mo (v 8.5); HR 0.73 (0.61-0.87); mOS 52.7 mo, HR 0.72 (0.62-0.83); ORR/CR 42/12%; Sarcomatoid features 16.4%; %TEAEs leading to DC IO or TKI 24%; IMDC/MSKCC Risk F/I/P 23/61/17; Median follow-up 99.1 mo KEYNOTE-426 (Pembrolizumab + Axitinib, N=861): mPFS 15.7 mo (v 11.1); HR 0.69 (0.59-0.81); mOS 47.2 mo, HR 0.84 (0.71-0.99); ORR/CR 60.6/11.6%; Sarcomatoid features 11.8%; %TEAEs leading to DC IO or TKI 33.3%; IMDC/MSKCC Risk F/I/P 31.9/55.1/13; Median follow-up 67 mo DC = discontinue; MSKCC = Memorial Sloan Kettering Cancer Center; F/I/P = favorable/intermediate/poor. 1. Motzer RJ, et al. J Clin Oncol. 2024;42(11):1222-1228. 2. Motzer RJ, et al. J Clin Oncol. 2025;43(5 Suppl):439. 3. Bourlon MT, et al. J Clin Oncol. 2024;42(4 Suppl):362. 4. Rini BI, et al. J Clin Oncol. 2023;41(17 Suppl):LBA4501. --- [Slide 2] [Bar/line chart of clinical event timelines since start of ICPI (months), by treatment status on/off ICPI, with events marked progressive disease / other death / cancer death, CR label at top] - Decruyenaere 2025 retrospectively showed favorable outcomes among patients with mRCC treated in any line with IPI/NIVO or NIVO alone, who received treatment for 21+ months, and who did not stop treatment for toxicity - A separate study retrospectively showed that immunotherapy (IO) rechallenge in second line setting after CR in first line setting, with 1L IO-IO or IO-VEGF TKI therapy, resulted in 75% ORR among 8 patients (Zarba 2026) Decruyenaere A, et al. Acta Oncol. 2025;64:979-988. Zarba M, et al. J Clin Oncol. 2026;44(7_suppl):446.

[Slide 1] Not Everyone Received Adjuvant Therapy Table — Study / Study arms / Adjuvant treatment / Key inclusion criteria / Patients / Median follow-up / Started Adjuvant Therapy / Completed Adjuvant Therapy: NIAGARA (N=1063): Durva+GemCis (n=533) vs GemCis (n=530); Durva Q4W x 8 cycles; Cisplatin-eligible MIBC (cT2-T4aN0/1M0), UC or UC with divergent differentiation, CrCl >=40 mL/min; T2N0 40%, CrCl>=40 to <60 19%, Divergent differentiation 14%; Median follow-up 46 mo; Started 71%; Completed 54% KEYNOTE-B15/EV-304 (N=808): EV+Pembro (n=405) vs GemCis (n=403); EV D1/D8 Q3W x5 cycles + Pembro Q3W x13 cycles; Does not meet any Galsky criteria for cisplatin ineligibility, Stage cT2-T4aN0M0 or cT1-T4aN1M0 MIBC, Urothelial histology >=50%; T2N0 20%, CrCl>=30 to <60 1%, Divergent differentiation 9%; Median follow-up 33 mo; Started 66%; Completed ~25% KEYNOTE-905/EV-303 (N=344): EV+Pembro (n=170) vs observation (n=174); EV D1/D8 Q3W x6 cycles + Pembro Q3W x14 cycles; Cisplatin ineligible or declined cisplatin for MIBC, Stage cT2-T4aN0M0 or cT1-T4aN1M0; T2N0 18%, CrCl>=30 to <60 60%, Divergent differentiation 11%; Median follow-up 25.6 mo; Started ~56%; Completed ~51% VOLGA (N=712): Durva+Treme+EV vs Durva+EV; Durva x9 cycles +/- Treme C1D1 only; Cisplatin ineligible or declined cisplatin for MIBC, Stage cT2-T4aN0-N1M0; Stage cT2-T4aN0-N1M0; Primary completion estimated Q2 2026; Started N/A; Completed N/A UC = urothelial carcinoma; CrCl = creatinine clearance. 1. Powles T, et al. N Engl J Med. 2024;391(19):1773-1786. 2. Galsky M, et al. N Engl J Med. 2026;395(4):338-348. 3. Vulsteke C, et al. N Engl J Med. 2026;394(13):1257-1269. 4. Doherty K [www.onclive.com]. Last updated May 14, 2026. --- [Slide 2] Role of pCR in Determining Adjuvant Treatment NIAGARA: Event-free survival by pCR status (pCR vs non-pCR), and overall survival by pCR status curves. Pathological Responses bar chart (Path CR vs Path Downstaging): ddMVAC 42%/63%, CG 36%, CG+Durva 37%, EV+P (KN905) 57%/66%, EV+P (KN B15) 56%/63% KN-905: EFS by pCR Status, ITT Population — EV+pembro (pCR) n=97: events 16 (16.5%), median NR (NR-NR); Control (pCR) n=15: events 5 (33.3%), median NR (12.7-NR); HR 0.43 (0.18-1.16). No pCR status — EV+pembro n=73: events 32 (43.8%), median 26.1 (10.1-41.2); Control n=168: events 96 (56.6%), median 14.2 (10.1-19.5); HR 0.76 (0.51-1.14) KN-B15: Median Disease-free Survival — EV-Pembro (pCR) NR (NR to NR) vs Cis-Gem (pCR) 31.2 (19.1 to 43.2); pCR HR 0.54 (95% CI 0.27 to 1.07); No pCR HR 0.90 (95% CI 0.68 to 1.21) CG = cisplatin gemcitabine. Vulsteke C, et al. Presented at: European Society for Medical Oncology (ESMO) Congress; 2025. Meeks J, et al. Presented at: AUA Annual Meeting; 2025.