KSMO 2026 — Seoul

[RAS inhibitors — mechanisms of action] What are learning - beyond KRAS G12C? 1. Many ways to skin a cat! RAS inhibitors: distinct mechanisms of action Mutant-selective off-state inhibitors: Adagrasib, sotorasib (KRAS G12C) · LY3962673, MRTX1133 (KRAS G12D) Pan-KRAS off/on-state inhibitors: BI 3706674, LY4066434 Mutant-selective on/off-state inhibitors: VS-7375/GFH375, INCB161734 KRAS-targeted proteolysis targeting chimeras (PROTACs): ASP3082, PT0253 Tri-complex mutant-selective on-state inhibitors: Elironrasib/RMC-6291 (KRAS G12C), Zoldonrasib/RMC-9805 (KRAS G12D) Tri-complex RAS-multi on-state inhibitor: RMC-7977, daraxonrasib/RMC-6236 (RAS WT + RAS Mut) Courtesy of C. Der · * = covalent modification · approved / terminated / clinical candidate / preclinical --- [What we know and need to know] What we know and need to know! What have we learned? (from KRAS G12C inhibitors) 1. Drug development is not easy! 2. First-in-class is not always best-in-class! 3. History never repeats but often rhymes! 4. Resistance is not simple! 5. Combinations are not straightforward What are we learning? (beyond KRAS G12C inhibitors) 1. Many ways to skin a cat! 2. Drug development is not easy! 3. Resistance is not easy! 4. The Final Frontier - frontline maybe near!

[The same shift across solid tumors] The Same Shift Is Happening Across Solid Tumors Immunotherapy BEFORE surgery — neoadjuvant / perioperative (with chemotherapy): Melanoma S1801 / NADINA · Lung (NSCLC) 6 phase 3 trials · Breast (TNBC) KEYNOTE-522 · Head & neck KEYNOTE-689 · Bladder (MIBC) NIAGARA · Gastric/GEJ MATTERHORN · KEYNOTE-585 · Colorectal (dMMR) NICHE-2 · dostarlimab Immunotherapy only AFTER surgery — adjuvant monotherapy (no chemotherapy, no neoadjuvant component): Renal cell (RCC) KEYNOTE-564 · Lung (NSCLC) IMpower010 · Esophageal/GEJ CheckMate 577 (post-CRT) · Melanoma (stage III) KEYNOTE-054 / CM238 Perioperative (chemo+) IO become standard of care across many cancer types. Patel SP et al., NEJM 2023; Schmid P et al., NEJM 2020; Uppaluri R et al., NEJM 2025; Powles T et al., NEJM 2024; Janjigian YY et al., NEJM 2025; Chalabi M et al., NEJM 2024; Choueiri TK et al., NEJM 2024; Kelly RJ et al., NEJM 2021; Eggermont AMM et al., NEJM 2018 --- [Adjuvant IO alone: inconsistent] Adjuvant Immunotherapy Alone: Inconsistent Results IMpower010 · Atezolizumab · DFS, PD-L1 ≥1%, stage II–IIIA · DFS HR 0.66 — significant · POSITIVE KEYNOTE-091 · Pembrolizumab · DFS, overall / PD-L1 ≥50% · Overall HR 0.76 (sig); ≥50% not met · MIXED BR.31 · Durvalumab · DFS, PD-L1 ≥25% (TC) · DFS HR ≈0.94 — not significant · NEGATIVE Adjuvant-only immunotherapy has produced inconsistent results across three phase 3 trials. Felip E et al., Lancet 2021; O'Brien M et al., Lancet Oncol 2022; Goss GD et al., Ann Oncol 2024 --- [Neoadjuvant & periop: uniformly positive] Neoadjuvant & Perioperative Trials: Uniformly Positive CheckMate 816 · Nivolumab + platinum-doublet ×3 · Neoadjuvant-only · EFS, pCR · Yes (EFS HR 0.63; OS improved) KEYNOTE-671 · Pembrolizumab + cisplatin-doublet ×4 (→ adj ×13) · Perioperative · EFS, OS · Yes (EFS HR 0.58; OS HR 0.72) AEGEAN · Durvalumab + platinum-doublet ×4 (→ adj ×12) · Perioperative · EFS, pCR · Yes (EFS HR 0.68) CheckMate 77T · Nivolumab + platinum-doublet ×4 (→ adj 1 yr) · Perioperative · EFS · Yes (EFS HR 0.58) RATIONALE-315 · Tislelizumab + platinum-doublet ×3–4 · Perioperative · EFS, MPR/pCR · Yes Neotorch · Toripalimab + platinum-doublet ×3 · Perioperative · EFS, MPR · Yes When give immunotherapy before surgery - neoadjuvant or perioperative - every phase three trial was positive. Forde P et al., NEJM 2022; Wakelee H et al., NEJM 2023; Heymach JV et al., NEJM 2023; Cascone T et al., NEJM 2024; Yue D et al., Lancet Respir Med 2025; Lu S et al., JAMA 2024 --- [Pre-op irAEs — Japanese RWD] Pre-operative irAEs from Japanese RWD Impact of pre-operative irAEs (CReGYT-04 subanalysis) Treatment completion: 58.3% with pre-op irAE vs 92.5% without Surgical cancellation: 21.7% with pre-op irAE vs 5.8% without ✓ irAEs reduced treatment completion and increased cancellation. ✓ However, among those reaching surgery, R0 resection and complication rates were preserved (94.4%). ✓ irAEs happen — but with early recognition and flexible scheduling, curative surgery is protected. Watanabe T et al., JTO Clin Res Rep, 2026

[Beyond genotyping — clonal phylogenetics] Beyond Genotyping: Spatial & Temporal Clonal Phylogenetics Use of ctDNA in DREAM NSCLC study JAMA Original Investigation: Lung Transplant for Refractory Lung-Limited Stage IV Non-Small Cell Lung Cancer Ankit Bharat, MD; ... G. R. Scott Budinger, MD; Young Kwang Chae, MD (Bharat/Chae et al. JAMA 2026) Pre-DLT vs Post-DLT CT chest; overall survival after transplant. Chae et al. ASCO 2026; Bharat (Chae) et al. JAMA 2026; Abbosh et al. Nature 2023 --- [ctDNA take-home messages] Take-Home Messages 01 The ctDNA application landscape and testing platforms are evolving rapidly. It is our essential tool in cancer treatment. 02 ctDNA testing for tumor genetic alterations may be used as first assessment, replacement, or as additional testing to tissue-based testing (ASCO guideline). ctDNA can be used for detection of emergent resistant clones and to evaluate treatment response. 03 ctDNA use for MRD surveillance is now a standard practice in bladder cancer. Other tumor types will follow. 04 The future of cancer treatment will increasingly involve ctDNA-adaptive strategies across both targeted therapy and immunotherapy settings. X (twitter) @youngkwangchae --- [SWOG MRD Restart design] Multi-histology Adjuvant therapy Reinitiation Study (SWOG MRD Restart) [Design schema, headline values as legible on the slide:] Total N = 224 · 8 baskets · n=28 per basket NeXT MRD LOD 3.45 PPM at 95% confidence SWOG ETRC trial; Chae & Moon PIs (as shown on slide footer)

[Next neoantigen classes] Future plans — Incorporation of additional classes of Neoantigens • Endogenous human retroviruses (ERV): Jiang et al. HIF regulates multiple translated endogenous retroviruses: Implications for cancer immunotherapy. Cell 2025 • nuORFs: Ouspenskaia et al. Unannotated proteins expand the MHC-I-restricted immunopeptidome in cancer. Nat Biotechnology 2022 Feb • Fusion antigens such as in NUT carcinoma • Viral cancer antigens such as Merkel Cell Polyoma Virus • MS detected epitopes --- [Vaccine conclusions] Conclusions • Current Epitope selection criteria at DFCI generated immune responses to many tumors. • Immune response to the vaccine improved overall survival and progression free survival in patients in these small cohorts.

[Why LA EGFR-mutant disease is distinct] Why EGFR-mutant locally advanced disease is distinct 40–50% EGFR mutation prevalence in Asian NSCLC (vs ~15% Western) CNS: dominant relapse pattern — shorter PFS, distant & brain recurrence Low benefit from immunotherapy — oncogene-addicted, low-TMB "cold" tumors Two clinically distinct populations under one label: (1) Resectable / operable: early-stage & selected Stage III → surgery-centered pathway (neoadjuvant / adjuvant) (2) Unresectable Stage III: definitive chemoradiotherapy (CRT) pathway --- [LAURA — PFS and CNS benefit] LAURA: PFS and CNS disease-free survival benefit Median PFS 39.1 months vs 5.6 with placebo Hazard ratio, PFS: 0.16 — 84% risk reduction ORR 57% vs 33% placebo New brain lesions: 8% vs 29% with placebo • Pneumonitis overlaps with radiation — any-grade radiation pneumonitis 48%; grade 3 pneumonitis 2% • Indefinite therapy until progression — QoL burden and cost • A PFS gain is not yet an OS gain Lu S, et al. N Engl J Med 2024;391:585–597; CNS and distant progression analyses: Ann Oncol 2024; Updated OS: Ramalingam SS, et al. ELCC 2025 LBA4. --- [ADAURA — is three years arbitrary?] Is three years arbitrary? (ADAURA) DFS by timepoint (stage II–IIIA), osimertinib vs placebo: 24 mo: 90% vs 46% (+44 pt) 36 mo: 84% vs 34% (+50 pt) 48 mo: 70% vs 29% (+41 pt) • The slope bends at 36 mo — the 36→48 mo fall is far steeper than 24→36. Placebo has already plateaued; the closing gap is driven by late events in the osimertinib arm. MRD data prove it directly (ctDNA): • MRD event: osimertinib 13% (15/112) vs placebo 49% (53/108) • 68% (19/28) of osimertinib-arm events occurred after stopping the drug — 58% (11/19) of those within 12 months of stopping • event-free rate 80%, 66% at 12, 24 months after stopping Updated analysis (Herbst JCO 2023): Stage II–IIIA DFS HR 0.23 (95% CI 0.18–0.30); Stage IB–IIIA at 48 mo 70% vs 38%, HR 0.27. Herbst JCO 2023 / Tsuboi NEJM 2020 (ADAURA updated DFS); ADAURA ctDNA MRD analysis, Nat Med 2025 (n=220). DFS values read from the updated KM curve shown. --- [NeoADAURA — major outcomes] NeoADAURA: Major outcomes Major pathologic response (MPR): Osi + CTx 26% · Osi mono 25% · PBO + CTx 2% pathCR: 4% · 9% · 0% (osi + CTx / osi mono / PBO + CTx) EFS HR vs PBO + CTx — interim EFS (15% maturity): Osi + CTx: 0.50 (99.8% CI 0.17–1.41) Osi mono: 0.73 (95% CI 0.40–1.35) • Resectable Stage II–IIIB EGFRm; ≥9 weeks neoadjuvant osimertinib ± chemo vs chemo • First randomized proof that neoadjuvant TKI markedly improves pathologic response • Not FDA-approved in the neoadjuvant setting and not in NCCN He J, Tsuboi M, et al. NeoADAURA. J Clin Oncol 2025;43:2875–2887. Blakely JCO 2024 (single-arm neoadjuvant osimertinib).

[4G EGFR-TKIs targeting C797S] 4G EGFR-TKIs targeting C797S under clinical development BLU-945 · Blueprint Medicines · EGFR+/T790M, EGFR+/T790M/C797S · Discontinued BBT-176 · Bridge Biotherapeutics · EGFR+/T790M/C797S · Discontinued · NCT04862780 JIN-A02 · J Ints Bio · EGFR+/T790M/C797S · Phase I/II · NCT04820023 Silevertinib · Black Diamond Therapeutics · Del 19/T790M/C797S · Phase I/II · NCT05394831 TQB3804 · Chia Tai Tianqing · EGFR+/C797S · Phase I · NCT05256290 HS-10375 · Jiangsu Hansoh Pharmaceutical · EGFR+/T790M/C797S · Phase I · NCT04128085 BAY2927088 · Bayer · EGFR+/T790M/C797S · Phase I, ORR: 4.8% · NCT05435248 BH-30643 · Blossom Hill · EGFR+/del 19/C797S · Developed as HER2 TKI · NCT05099172 DZ6008 · Dezal · EGFR+/T790M/C797S/Ex20ins · Phase I/II · NCT06706076 VRN110755 · Voronoi · EGFR+/T790M/C797X · Phase I · NCT06905197, NCT06813365 · Phase I/II · NCT07699328 Lim ASCO 2025, Zhan J Transl Med 2025, Spira AACR-NCI-EORTC 2025, Wang ASCO 2026, Hong ASCO 2026 --- [MET inhibitor + osimertinib] MET inhibitor + osimertinib Selected trials of MET inhibitors + osimertinib in patients with advanced EGFR-mutated NSCLC following first-line EGFR TKI therapy INSIGHT 2 · n=128 · Phase 2 · plasma GCN ≥2.3; tumor GCN ≥5 and/or MET/CEP7 ≥2 · osimertinib + tepotinib · ORR 50% · PFS 5.6 m INSIGHT 2 (tepotinib mono) · n=12 · tepotinib · ORR 8.3% · PFS 2.7 m TATTON · n=69 · Phase 1b · GCN ≥5 or MET/CEP7 ≥2 · savolitinib + osimertinib · ORR 33% · PFS 5.5 m SAVANNAH · n=80 · Phase 2 · OE IHC3+/≥90% and/or FISH10+ · savolitinib + osimertinib · ORR 56.3% · PFS 7.4 m Arm E (NCT02099058) · n=38 · Phase 1b · IHC3+/≥25% · Teliso-V + osimertinib · ORR 50% · PFS 7.4 m SACHI · n=211 · Phase 3 · GCN ≥5 and/or MET/CEP7 ≥2 · savolitinib + osimertinib vs chemo · ORR 58% v 34% · PFS 9.8 v 5.4 m SAFFRON · n=324 · Phase 3 · OE IHC3+/≥90% or FISH10+ · savolitinib + osimertinib vs chemo · NA Arter, Soo Med 2026 --- [ADCs in pre-treated EGFR+ NSCLC] Studies of ADCs in pre-treated EGFR + NSCLC Some hits, some misses and some promising data Trop2: ORCHARD module 10 (n=34) · OptiTROP-Lung03 (n=137) · pooled TROPION-Lung01/05 (n=117) · Phase III OptiTROP-Lung04 (n=376) HER3: HERTHENA-Lung01 (n=225) · Phase III HERTHENA-Lung02 (n=586) HER3xEGFR: BL-B01D1-101 (n=40) · BL-B01D1-203 cMET: Teliso V + osimertinib (n=38) · M21-404 (n=40) Payloads: deruxtecan · KL610023 · ED-04 · vedotin · adizutecan