All-Ireland Lung Cancer Conference 2026

[Slide 1] What is a paradigm shift? [Book cover: The Structure of Scientific Revolutions, Second Edition, Enlarged, Thomas S. Kuhn] Thomas S. Kuhn introduced the modern concept of paradigm shift in The Structure of Scientific Revolutions (1962), using the Copernican Revolution as one of his key examples. The planets did not change. The way we understood them did. [Two portraits: PTOLEMY (c. 100 - c. 170) The Earth is the center; COPERNICUS (1473 - 1543) The Sun is the center; orbit diagrams and quotations] Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. 3 GECP lung cancer research --- [Slide 2] What makes a true paradigm shift? > Changes outcomes. Improves clinically meaningful endpoints, ultimately survival. > Changes the standard of care. Not simply add another therapeutic option. > Changes how we think about the disease. Modifies the underlying therapeutic concept. > Creates new questions that did not previously exist. Once the paradigm changes, the old clinical questions are no longer sufficient. Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. GECP lung cancer research --- [Slide 3] Neoadjuvant Immunotherapy Creates new questions that did not previously exist A paradigm shift is not when a new treatment appears, it is when the new treatment changes the questions we ask. Neoadjuvant immunotherapy has done exactly that Once the paradigm changes, the old clinical questions are no longer enough BEFORE Should we give adjuvant chemotherapy? [arrow] NOW - Who needs adjuvant immunotherapy? - Can treatment be de-escalated after pCR/MPR negativity? - How should we escalate treatment in non-pCR/MPR positive patients? - And ultimately, does every exceptional responder need surgery? Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. 23 GECP lung cancer research

[Slide 1] Systemic therapy has become part of treatment Background and current situation in early stages of lung cancer IASLC | OS (60 mo) IB | 68% IIA | 60% IIIB | 53% IIIA | 36% Stages I, II -> Surgery and adjuvant QT Resectable stage IIIA -> Surgical multimodal treatment | Δ5 yr OS | HR (95% CI) | p value Neoadjuvant treatments | 5% | 0.87 (0.78-0.96) | 0.007 Adjuvant treatments | 4% | 0.86 (0.81-0.92) | < 0.0001 Both neoadjuvant and adjuvant therapies are treatment options that may improve outcomes in [patients under]going surgery. [Left KM-style plot: Survival (0-1.0) vs Time from randomization (years, 0-8); two curves labelled Surgery alone and Surgery + CT] NSCLC Collaborative Group, BMJ, 1995 [Right KM-style plot: Survival vs Time from randomization (years, 0-8); two curves labelled No preoperative chemotherapy and Preoperative chemotherapy] NSCLC Meta-analysis Collaborative Group, Lancet, 2014 Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. 8 GECP lung cancer research --- [Slide 2] Neoadjuvant Immunotherapy: A Paradigm Shift [Four tabs:] Changes outcomes. | Changes the standard of care. | Changes how we think about the disease. [highlighted] | Creates new questions that did not previously exist. [Ne]oadjuvant therapy turns the tumor into an in-vivo test of treatment sensitivity [Flow:] Tumor present -> Treatment -> Biological response -> Measurement of residual disease [Infor]mation that we did not have before surgery: pCR/MPR, ctDNA clearance, radiologic and [...]ics... Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. 17 GECP lung cancer research --- [Slide 3] Neoadjuvant treatment: Pathological Response > Nat Med. 2024 Jan;30(1):218-228. doi: 10.1038/s41591-023-02660-6. Epub 2023 Oct 30. Association between pathologic response and survival after neoadjuvant therapy in lung cancer Figure 1. Kaplan-Meier curves of EFS by %RVT-PT categories for patients treated with nivolumab plus chemotherapy. [KM plot: EFS (%) vs Months from randomization; four curves labelled 0-5% RVT-PT, >5-30% RVT-PT, >30-80% RVT-PT, >80% RVT-PT; landmark labels printed on plot: 94%, 90%, 79%, 74%, 72%, 60%, 57%, 39%] Median EFS, months (95% CI) 0-5% RVT-PT | NR (31.6-NR) [>5-30% RVT-PT] | NR (13.6-NR) [>30-80% RVT-PT] | 26.6 (11.6-NR) [>80% RVT-PT] | 18.9 (13.4-27.8) The deeper the pathologic response, the longer the EFS EFS by %RVT-PT category, nivolumab + chemotherapy arm Deutsch JS. et. al., Nat Med, 2024. Mariano Provencio MD, PhD; Puerta de Hierro University Hospital, Madrid, Spain. 20 GECP lung cancer research [Survival data captured as a table — no point estimates read from the curve] Endpoint: EFS by %RVT-PT category (nivolumab + chemotherapy arm; Deutsch JS et al., Nat Med 2024 - CheckMate 816 analysis as cited on slide) 0-5% RVT-PT | NR | 31.6-NR >5-30% RVT-PT | NR | 13.6-NR >30-80% RVT-PT | 26.6 | 11.6-NR >80% RVT-PT | 18.9 | 13.4-27.8 Printed on plot (category not unambiguously attached): 94%, 90% (upper curve, 0-5% RVT-PT); 79%, 60% (>5-30% RVT-PT); 74%, 57% (>30-80% RVT-PT); 72%, 39% (>80% RVT-PT) - at the ~12-month and ~24-month dashed reference lines Note: Median table: only the first row label (0-5% RVT-PT) is visible; row labels 2-4 are occluded by the speaker and are inferred from the legend order (0-5, >5-30, >30-80, >80). Landmark-to-curve mapping is by visual proximity only and should be verified against Deutsch et al. Nat Med 2024 Figure 1. No HR or p value printed.

[Slide 1] CheckMate 816: 3-y efficacy/safety update and biomarker analys[is] Baseline 4-gene inflammatory signature scoreᵃ and EFS NIVO + chemo | Low 4-gene inflammatory score (n = 36) | High 4-gene inflammatory score (n = 41) Median EFS, mo (95% CI) | NR (10.5-NR) | NR (40.4-NR) HR (95% CI) | 0.65 (0.30-1.39) [KM plot: EFS (%) vs Months from randomization; curves labelled High and Low; landmark labels printed: 85%, 75%, 66%, 70%ᵇ, 59%, 55%ᶜ] Chemo | Low 4-gene inflammatory score (n = 40) | High 4-gene inflammatory score (n = 36) Median EFS, mo (95% CI) | 22.4 (13.9-NR) | 24.9 (9.8-NR) HR (95% CI) | 1.00 (0.52-1.92) [KM plot: EFS (%) vs Months from randomization; curves labelled High and Low; landmark labels printed: 72%, 68%, 51%, 50%, 47%ᵈ, 47%ᵉ] No. at risk [rows for Low and High under each plot] Composite signature of PD-L1, CD8A, STAT1 & LAG3 [Footnote, partially clipped at left:] ...months. ...ere grouped as high or low relative to the median z-score across the dataset. ᵇ⁻ᵉ95% CI for 3-year EFS rates: ᵇ56-86; ᶜ39-77; ᵈ32-69; ᵉ3[2-...] [Survival data captured as a table — no point estimates read from the curve] Endpoint: EFS by baseline 4-gene inflammatory signature score (CheckMate 816, 3-year update) NIVO + chemo, Low 4-gene inflammatory score | 36 | NR | 10.5-NR NIVO + chemo, High 4-gene inflammatory score | 41 | NR | 40.4-NR Chemo, Low 4-gene inflammatory score | 40 | 22.4 | 13.9-NR Chemo, High 4-gene inflammatory score | 36 | 24.9 | 9.8-NR HR: NIVO + chemo High vs Low: 0.65; Chemo High vs Low: 1.00 HR CI: 0.30-1.39 (NIVO + chemo); 0.52-1.92 (Chemo) NIVO + chemo 3-year EFS (per footnote b/c): High 70% (95% CI 56-86), Low 55% (95% CI 39-77) Chemo 3-year EFS (per footnote d/e): 47% (95% CI 32-69) and 47% (95% CI 3[2]-[clipped]) - high/low assignment of d vs e not legible Other printed landmarks (timepoint not labelled): NIVO + chemo 85%, 75%, 66%, 59%; Chemo 72%, 68%, 51%, 50% Note: Source slide is from the CheckMate 816 3-year update (header clipped at right). Footnote 'a' definition and first footnote lines are clipped at left edge. Last CI in footnote e is clipped ('3[2]-...'). --- [Slide 2] Conceptual framework to integrate pCR and ctDNA dynamics to guide adjuvant treatment escalation/de-escalation strategies a [Scatter schematic: y-axis RVT (%), x-axis VAF (%) / [Undetect]able ctDNA; three dashed regions labelled 'ctDNA adds value' (upper-left), 'ctDNA adds little value' (large diagonal cluster), 'ctDNA adds value' (lower-right)] b [2x2 schematic: y-axis RVT (%) vs MPR or pCR; x-axis ctDNA clearance -> VAF (%); quadrants labelled 'Adjuvant therapy or clinical trials', 'Adjuvant therapy or clinical trials (ctDNA probably does not provide additional value)', 'De-escalation strategies', 'Close surveillance'] [...]ents are uninformative in patients where both RVT and ctDNA are undetectable and in those with a [...]etectable ctDNA. [...] clinically informative in patients with MPR or pCR as it can provide early evidence of disease [...] also be informative when ctDNA is undetectable in patients with a substantial RVT. Garitaonaindia Y et al...Forde P, Cascone T, Spicer J, Provencio M.. Nat Rev Clin Oncol 2026 --- [Slide 3] Can neoadjuvant IO added to chemo benefit pts with tumor EGFR mutations? [red X icon] EFS using RECIST v1.1 (BICR) (EGFRm and mITT)* EGFRm subgroup | Durvalumab arm | Placebo arm No. events / no. patients (%) | 12/26 (46.2) | 9/25 (36.0) mEFS, months (95% CI) | 30.8 (11.4, NR) | 19.6 (14.3, NR) Unstratified HR† (95% CI) | 0.86 (0.35, 2.19) [Left KM plot: EFS vs Time from randomisation (months); landmark labels printed: 77.8%, 65.2%, 59.3%, 44.9%; Median follow-up (range) in censored patients: 16.6 months (0.0-36.4)] mITT population¹ | Durvalumab arm | Placebo arm No. events / no. patients (%) | 98/366 (26.8) | 138/374 (36.9) mEFS, months (95% CI) | NR (31.9, NR) | 25.9 (18.9, NR) Stratified HR† (95% CI) | 0.68 (0.53, 0.88) [Right KM plot: Probability of EFS vs Time from randomisation (months); landmark labels printed: 73.4%, 64.5%, 63.3%, 52.4%; Median follow-up (range) in censored patients: 11.7 months (0.0-46.1); Stratified log-rank P-value = 0.003902; + Censored] No. at risk: D arm | PBO arm [rows present] [Footnote, partially clipped at left:] ...defined as time from randomisation to the earliest of: (A) progressive disease (PD) that precludes surgery; (B) PD discovered and reported by the ...ompletion of surgery; (C) local/distant recurrence using BICR per RECIST v1.1; or (D) death from any cause. HR <1 favours the D arm versus the ...ulated using the Kaplan-Meier method. †HR for the EGFRm subgroup calculated from an unstratified Cox proportional hazards model; HR for the ...portional hazards model. CI, confidence interval; D, durvalumab; HR, hazard ratio; mEFS, median EFS; NR, not reached; PBO, placebo. ¹Heymach JV, et al. Cancer Res 2023;83 (8_Supplement):CT005 8 [Survival data captured as a table — no point estimates read from the curve] Endpoint: EFS by BICR per RECIST v1.1 (AEGEAN; durvalumab + chemo vs placebo + chemo), EGFRm subgroup and mITT EGFRm subgroup - Durvalumab arm | 26 | 30.8 | 11.4, NR EGFRm subgroup - Placebo arm | 25 | 19.6 | 14.3, NR mITT - Durvalumab arm | 366 | NR | 31.9, NR mITT - Placebo arm | 374 | 25.9 | 18.9, NR HR: EGFRm subgroup unstratified HR 0.86; mITT stratified HR 0.68 HR CI: 0.35, 2.19 (EGFRm); 0.53, 0.88 (mITT) P: mITT stratified log-rank P-value = 0.003902; EGFRm subgroup: no p value printed EGFRm plot printed labels at ~12 and ~24 month reference lines: 77.8% / 65.2% (12 mo), 59.3% / 44.9% (24 mo) - arm colour: blue = durvalumab, red = placebo; at 12 mo the red (placebo) label reads 77.8% and blue (durvalumab) 65.2%; at 24 mo blue 59.3%, red 44.9% mITT plot printed labels: 73.4% (D) / 64.5% (PBO) at 12 mo; 63.3% (D) / 52.4% (PBO) at 24 mo Note: Trial name is not printed on the visible slide; the cited reference (Heymach JV et al. Cancer Res 2023;83(8_Suppl):CT005) is the AEGEAN AACR 2023 presentation. Landmark-to-arm assignment is by label colour (blue = durvalumab, red = placebo); EGFRm 12-month assignment should be verified against the source. Events/patients as printed: EGFRm subgroup - Durvalumab arm: events 12/26 (46.2%); EGFRm subgroup - Placebo arm: events 9/25 (36.0%); mITT - Durvalumab arm: events 98/366 (26.8%); mITT - Placebo arm: events 138/374 (36.9%).

[Slide 1] What is Resectability? EORTC Consensus Statement [Table: rows x columns N0 | N1 | N2 SINGLE (non-bulky, non-invasiv[e])] T1-2 | Not stage III disease | Not stage III disease | Resectable [T]3 size, satellite, and invasion | Not stage III disease | Resectable | Resectable [T]4 size or satellite | Resectable | Reseactable [sic] | Resectable [T4 in]vasion | Potentially resectable‡ | Potentially resectable‡ | Potentially resectable‡ [Right table: structure | Unresectable | Potentially resectable*] Pulmonary artery in the pericardium | | ✓ Superior vena cava | | ✓ Diaphragm | | ✓ Heart | ✓* | Carina | | ✓ Trachea | ✓* | Oesophagus | ✓* | Spinal cord | ✓ | Vertebral body | | ✓ Recurrent laryngeal nerve | | ✓ Mediastinal fat | | ✓ Great vessels: aorta, inferior vena cava, and pulmonary vein | | ✓ Figure 3: Resectability consideration for major structures involved in T4 definition [Logos: Trinity St James's Cancer Institute / Institiúid Ailse Choláiste na Tríonóide/San Séamas; partial second logo] [Podium banner: 12TH ALL IRELAND LUNG CANCER CONFERENCE ... LUNG CANCER RAPID FIRE]

[Slide 1] A Paradigm Shift for Locally Advanced NSCLC Past Max single station N2 No pneumonectomy in N2 disease Equipoise neoadjuvant vs adjuvant chemotherapy Current Resectability and operation determined upfront Tumour profile guides treatment Resource and timing limits Possible Near Future Deescalate extent of surgery Response-informed risk Neoadjuvant oncogene TT Expanding driver list Beyond Unresectable -> resectable Salvage surgery N3 disease [Logos: St James's Hospital; Trinity St James's Cancer Institute / Institiúid Ailse Choláiste na Tríonóide/San Séamas; ECI Comprehensive Cancer Centre; European Society of Thoracic Surgeons; 12TH ALL IRELAND LUNG CANCER CONFERENCE]

[Slide 2] BOTTOM LINE KRAS in NSCLC: where we are and what comes next Where we are - [...]gen G12C (OFF) [inhibit]ors poised to become [se]cond-line standard, [pendin]g full ESMO data - [...] for every patient: [...] variant plus STK11 Coming soon - First-line G12C inhibitor + IO combinations - Early-stage disease - New KRAS targets arriving fast Still to solve - Toxicity is improving but still a challenge - Resistance limits the durability - Sequencing and combinations need further research --- [Slide 3] PART ONE · PLASMA AND TISSUE Complementary Techniques 3 [Table — columns: Alteration class | Plasma ctDNA NGS | Tissue NGS | IHC / FISH] EGFR C797S, ALK and ROS1 resistance SNVs | Best analyte | Yes | — Fusions (ALK, ROS1, RET, NRG1, NTRK) | Under-calls — 2–7 of 14 ALK fusions detected | Yes — RNA preferred | IHC screen only MET amplification (Bontoux et al) | Partial — but found 12 that tissue NGS missed | Yes | FISH is the reference CDKN2A / MTAP loss and other deletions | Poor | Yes | MTAP IHC surrogate Small-cell or squamous transformation | Structurally invisible (but see below) | — | Histology only 20–43% Incremental actionable yield from adding the second analyte Plasma turnaround 10 days vs 29.9 days for tissue in matched series [...]cer Discov 2026;16:1074–1086, PMID 41790042 (ELIOS) · Bontoux C et al. J Exp Clin Cancer Res 2025, PMID 40765058 (matched tissue/plasma, n=102) · Choudhury Y et al. Front Med 2022 · ESMO PMWG, [...] 750–768.

[Slide 2] The Royal Marsden OMD: Nine Distinct Clinical Scenarios Classification De-Novo OMD Synchronous Oligo-Metastatic Disease Metachronous Oligo-Recurrence Metachronous Oligo-Progression Repeat OMD: With history of oligo-metastatic but not poly-metastatic disease Repeat Oligo-Recurrence Repeat Oligo-Progression Repeat Oligo-Persistence Induced OMD: With history of poly-metastatic disease Induced Oligo-Recurrence Induced Oligo-Progression Induced Oligo-Persistence [Right margin: nine small lung-schematic icons, one per scenario, each labelled with T-X / T0 timepoints and 'Active systemic therapy' arrows — labels too small to read reliably] Guckenberger Lancet Oncol 2020; --- [Slide 3] HALT Randomised Ph II Advanced NSCLC EGFR / ALK + with response to TKI TKI Progression Oligo-Progression Randomise (2:1) Widespread Progression Primary Endpoint: Progression Free Survival Continue TKI [Map of Europe with site dots:] UK Sites · Paris, France · Zurich, Switzerland · Toulouse, France · Turin, Italy · Rome, Italy · Barcelona, Spain Primary Analysis 2027 [Logos:] EORTC — The future of cancer therapy · British Thoracic Oncology Group (BTOG) · CTRad · SABR UK Consortium · Cancer Research UK --- [Slide 4] The Royal Marsden NORTHSTAR: PFS Metastases ≤3 PFS Median PFS 33.1 mths Median PFS 22.4 mths Metastases >3 Median PFS 19.7 mths Median PFS 15.9 mths Complete LCT Median PFS 27.9 mths Partial LCT Median PFS 14.5 mths Time (Months) Time (Months) Elamin ESMO 2025; [Survival data captured as a table — no point estimates read from the curve] Endpoint: PFS (four Kaplan-Meier panels: metastases ≤3; metastases >3; complete LCT; partial LCT) Metastases ≤3 — green curve (arm label not printed) | 33.1 mths Metastases ≤3 — red curve (arm label not printed) | 22.4 mths Metastases >3 — green curve (arm label not printed) | 19.7 mths Metastases >3 — red curve (arm label not printed) | 15.9 mths Complete LCT | 27.9 mths Partial LCT | 14.5 mths Note: Medians are colour-coded to the curves (green vs red) but the slide does NOT print which treatment arm each colour represents (NORTHSTAR randomised osimertinib ± local consolidative therapy — arm assignment must be confirmed from the ESMO 2025 presentation, not inferred here). No HR, CI, p-value, N or landmark rates printed. Dashed vertical reference lines at 36 months on the lower panels are unlabelled.

[Slide 1] Taiwan national screening: not a pilot anymore Population-scale implementation in smokers and family-history enriched never/light smokers [Table — column groups: LCFH (Non-smokers | Light smokers | Overall) | Heavy Smokers | Both | Total] Screened | 124,043 | 12,278 | 136,321 | 111,407 | 10,229 | 257,957 Lung cancer | 1,905 | 115 | 2,020 | 991 | 128 | 3,139 Detection rate (%) | 1.5 | 0.9 | 1.5 | 0.9 | 1.3 | 1.2 Stage 0 (%) | 10[.?] | 8.7 | 10.5 | 4.4 | 8.6 | 8.5 Stage 1 (%) | [occluded] | [occluded] | 78.5 | 62.1 | 68.8 | 72.9 Stage 0-1 (%) | [occluded] | [occluded, ends in 8] | 89.0 | 66.5 | 77.4 | 81.4 Key numbers 2022-2025 257,957 screened 3,139 lung cancers 1.2% detection rate 81.4% stage 0–I Note: There were 30,003 screening case[s ...] [occluded ...]me. Health Promotion Administration. Min[istry ...] [occluded] https://www.hpa.gov.tw/EngPages/In[...] [occluded] Stage shift is co[...occluded...]benefit in never-smokers remains the key evidentiary question Taiwan Health Promotion Administration. [...occluded...]kshop 2026. --- [Slide 2] Next IO #1 — targeting the suppressive microenvironment The next checkpoint may not be another checkpoint. TREG / CCR8 • Tumor-selective Treg depletion • CD8 activation dem[onstrated — partially occluded] • Responses across s[olid tumo]rs [partially occluded] • NSCLC effica[cy ...]hed [partially occluded] TCE and CAR-T • Overcoming the immune desert • Bypassing AG presentation • Effective in chemo/ICI resistance • Infinite technological opportunities IMMUNE–TME BISPECIFICS • PD-(L)1×VEGF • PD-(L)1×4-1BB • PD-(L)1×TGFβ / IL-2 / IL-15 • Spatial co-localization as drug design Wang Y et al. Nat Rev Clin Oncol. 20[...] [occluded] [...]. 2025;22:806–830 • Cho BC et al. Nat Med. 2026;32:2245–2253 --- [Slide 3] The next decade: scientific abundance, implementation scarcity SCIENCE • Risk replaces pack-years as the screening currency • MRD turns recurrence into a measurable biological state • AI extracts latent phe[no]types from CT and H&E • Therapy becomes [molecu]larly and immunologically engineered • Long surviva[l makes toxicity,] time and survivorship first-order en[dpoints] DELIVERY • Testing access remains geographically unequal • Novel drugs reach countries at different speeds and prices - joint HTA bodies are desirable • Complex modalities require infrastructure, not only approval • MFN-type pricing policies will alter global launch and pricing incentives – meaning access • An innovation that cannot reach patients is not an innovation – time to enter politics ? Smeltzer MP et al. J Thorac Oncol[...] [occluded] [...]ng TJ, Clemans-Cope L. N Engl J Med. 2026 • J Clin Oncol. 2026;44(16_suppl):11056 --- [Slide 4] The perioperative revolution • CheckMate 816 established neoadjuvant nivolumab + chemotherapy and shows OS benefit • KEYNOTE-671 demonstrated EFS and OS benefit with perioperative pembrolizumab. AEGEAN, CheckMate 77T, NEOTORCH and RATIONALE-315 reinforce the perioperative paradigm [Figure — schematic table: Trial | Perioperative Strategy: Neoadjuvant CT IO (C1 C2 C3 C4) | SURGERY | Adjuvant treatment] CheckMate 816: Nivolumab → [3 neoadjuvant cycles] → Surgery → (no adjuvant IO row) Che[ckMate 77T]: Nivolumab → [4 cycles] → Surgery → Adjuvant IO: Nivolumab 12 months KE[YNOTE-671: P]embrolizumab → [4 cycles] → Surgery → Pembrolizumab 9 months A[EGEAN: Durva]lumab → [4 cycles] → Surgery → Durvalumab 12 months [NEOTORCH: Toripali]mab → [3 cycles] → Surgery → Toripalimab + CT x 1 → Toripalimab 9 months [RATIONALE-315: Tislel]izumab → [3-4 cycles, 4th bracketed] → Surgery → Tislelizumab 11 months Adjuvant CT Atezolizumab 12 months Pembrolizumab 12 months Durvalumab 12 months Nivolumab 12 months [Right margin: tiny flag icons with PD-L1 eligibility labels per regimen, e.g. 'All PDL1', 'PD-L1 ≥1%', 'IB-IIIA PDL1≥1% / PDL1≥50%', 'All PDL1 POST-CT' — too small to transcribe reliably] Adapted from Remon, (...) Mountzios and[...]

[Slide 2] NHS Galleri is the largest and first randomised controlled clinical trial evaluating utility of a Multi Cancer Early Detection test Approximately 1.5 million individuals invited NHS England 142,826 consented participants aged 50–77 years with no cancer diagnosis or treatment in the past 3 years Blood sample collected Randomised 1:1 BLINDED Intervention (71,122) MCED test performed BLINDED Control (71,128) Sample stored Positive — Results are reported back to participant Negative UNBLINDED Referred to Diagnostic Pathway Through NHS (Outside of Trial) NHS England Cancer Diagnosed No Cancer Diagnosed BLINDED Return for 1- and 2-year blood sample visit[b] BLINDED Primary Endpoint Significant reduction in incidence rate of late-stage cancer (stage III/IV) after [annu]al screening rounds [in] a prespecified group of [12] primary cancer types[a], [in] all routinely staged cancer types [(]excluding and including [p]rostate cancer) Secondary Endpoints Prespecified secondary endpoints of clinical interest • Reduction in stage IV cancer incidence rate • Proportion stage I/II cancers • Safety (based on MCED test only) • Mortality (after longer follow-up) • Stage distribution by cancer type Exploratory/Adhoc Endpoints • Route to diagnosis --- [Slide 3] Three doors for AI in lung cancer • Radiology: detection, segmentation, risk prediction and radiomics • Pathology: diagnosis, biomarker prediction and spatial phenotyping • Decision support: multimodal synthesis, longitudinal prediction and trial matching • The challenge shifts from model performance to calibration, workflow and accountability [...]PG et al. J Clin Oncol. 2023 • Vorontsov E et al. Nat Med. 2024 --- [Slide 4] The next decade: scientific abundance, implementation scarcity SCIENCE • Risk replaces pack-years as the screening currency • MRD turns recurrence into a measurable biological state • AI extracts latent phenotypes from CT and H&E • Therapy becomes molecularly and immunologically engineered • [L]ong survival makes toxicity, time and survivorship [firs]t-order endpoints DELIVERY • Testing access remains geographically unequal • Novel drugs reach countries at different speeds and prices - joint HTA bodies are desirable • Complex modalities require infrastructure, not only approval • MFN-type pricing policies will alter global launch and pricing incentives – meaning access • An innovation that cannot reach patients is not an innovation – time to enter politics ? [Smeltzer MP et] al. J Thorac Oncol. 2025 • Cherny NI et al. Ann Oncol. 2025 • Hwang TJ, Clemans-Cope L. N Engl J Med. 2026 • J Clin Oncol. 2026;44(16_suppl):11056

[Slide 2] Three curative-intent pathways Potentially resectable Selected IIIA / limited N2 [partially occluded] Induction sy[stemic therapy] [occluded] ([...]) [occluded] then [...] [occluded] Surgery + m[...] [occluded] (with RT [...]) [occluded] Unresectable but fit Bulky/multistation N2, N3, IIIB–IIIC Concurrent CRT then durvalumab Radiation oncology coordinates local curative treatment Not suitable for cCRT Frailty or major comorbidity Individualised sequential, RT-alone or symptom-directed care Supportive MDT care and shared decisions --- [Slide 3] Stage III – Heterogenous group Stage IIIA • cT1a-c N2b • cT2a-b N2a • cT3 N1 / N2a • cT4 N0 / N1 T3 - Invasion of Azygos vein, phrenic nerve, pericardium T4 – Invasion of Trachea, Carina, SVC, Diaphragm, Vertebral body, Mediastinum, Heart, Vagus nerve, RLN, Oesphagus, Great vessels, Thymus Pancoasts – Invasion of subclavian vessels, vertebral body, lamina, spinal canal, cervical nerve roots, or [brachial pl]exus [partially occluded] 9th Edition TNM Descriptors and Stages [Table — T/M | Categories and Descriptors | N0 | N1 | N2a | N2b | N3] T1 | T1a ≤1 cm | IA1 | IIA | IIB | IIIA | IIIB T1 | T1b >1 to ≤2 cm | IA2 | IIA | IIB | IIIA | IIIB T1 | T1c >2 to ≤3 cm | IA3 | IIA | IIB | IIIA | IIIB T2 | T2a Visceral pleura / central invasion | IB | IIB | IIIA | IIIB | IIIB T2 | T2a >3 to ≤4 cm | IB | IIB | IIIA | IIIB | IIIB T2 | T2b >4 to ≤5 cm | IIA | IIB | IIIA | IIIB | IIIB T3 | T3 >5 to ≤7 cm | IIB | IIIA | IIIA | IIIB | IIIC T3 | T3 Invasion | IIB | IIIA | IIIA | IIIB | IIIC T3 | T3 Same lobe separate tumor nodules | IIB | IIIA | IIIA | IIIB | IIIC T4 | T4 >7 cm | IIIA | IIIA | IIIB | IIIB | IIIC T4 | T4 Invasion | IIIA | IIIA | IIIB | IIIB | IIIC T4 | T4 Ipsilateral separate tumor nodules | IIIA | IIIA | IIIB | IIIB | IIIC M1 | M1a Contralateral tumor nodules | IVA | IVA | IVA | IVA | IVA M1 | M1a Pleural / pericardial effusion, nodules | IVA | IVA | IVA | IVA | IVA M1 | M1b Single extrathoracic metastasis | IVA | IVA | IVA | IVA | IVA M1 | M1c1 Multiple metastases in 1 organ system | IVB | IVB | IVB | IVB | IVB M1 | M1c2 Multiple metastases in >1 organ systems | IVB | IVB | IVB | IVB | IVB [Logo:] Institiúid Ailse Choláiste na Tríonóide/San Séamas · Trinity St James's Cancer Institute --- [Slide 4] When do Medical Oncology get involved Management approach: Stage I-III NSCLC [Table — Stage | Tumour / nodal status | Local therapy | Systemic therapy | Key trials] Stage IA | ≤3 cm, N0 | Surgery (lobectomy or segmentectomy); SBRT if inoperable | None | JCOG0802; CALGB 140503; RTOG 0236 Stage IB | >3–4 cm[, N0 — occluded] | Resection; SBRT if inoperable | Consider adjuvant targeted therapy for eligible mutations | ADAURA; ALINA Stage II–III resectable | >4[... occluded ...]ed / re[... occluded] | Resection | Platinum doublet + ICI (neoadjuvant/perioperative); selected adjuvant ICI or targeted therapy | CM816; KN671; AEGEAN; CM77T; NEOTORCH; IMpower010; PEARLS (KN091) Stage III unresectable | L[... occluded] / un[... occluded] | [... occluded ...]therapy (cCRT) | Durvalumab consolidation if eligible; osimertinib following CRT for EGFR-mutated disease | PACIFIC; LAURA [Logos:] UHW · Ospidéal Ollscoile Phort Láirge / University Hospital Waterford · HSE

[Slide 1] Lung SABR in Northern Ireland [Stacked bar chart — y-axis 0 to 350 (ticks 50, 100, 150, 200, 250, 300, 350); x-axis years 20[08?], 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025; three stacked series per year] Legend: [blue] NSCLC SCRT · [green] NSCLC RT Alone · [red] NSCLC SABR [Figure description: NSCLC SABR (red) segment grows from near zero in the earliest years to the dominant component by 2022–2025, with total bar height rising from roughly 50 to roughly 290 per year; no bar values are printed on the chart] --- [Slide 3] Tricks of the Trade Restrict Dmax ≤120Gy Increase Fractions Compromise PTV Coverage Decrease Total Dose [...occluded...] to 60Gy/15# or 55Gy/20#

[Slide 1] AI and Healthcare Cycle [Hype-cycle style curve: y-axis EXPECTATION, x-axis TIME; labelled points along curve, each marked with a circle whose fill indicates expected time to standard of care] Cohort Refinement Spatial Omics [Radio]mics [partially hidden by speaker] [Biomarker discovery - hidden by speaker] [Simulated Clinical] Trial [partially hidden] Clinical Decision Support Models Patient - ChatBot Interface Restructuring Free Text Data Security Challenges Wearable devices Real World Data Integration Telehealth Decentralized Trials Digital Twin Creation Patient-Trial Matching Adverse Event Automated Reporting Ambient AI Scribe Note Generation [Phase labels along x-axis:] [Inspiration for Innovation - partially hidden] | Rise of Inflated Expectation | Fall of Disillusionment | Ascent of Enlightenment | Stabilization of Productivity TIME [Legend:] [Expected time to s]tandard of care: O <2 yr. [light] 2-5 yrs. [mid] 5-10 yrs. [dark] >10 yrs. [Az]enkot, Patel ASCO Ed Book 2025 UC San Dieg[o] Moores Cancer C[enter] --- [Slide 2] 2024 World Conference on Lung Cancer | SEPTEMBER 7-10, 2024 SAN DIEGO, CA USA | #WCLC24 wclc2024.iaslc.org NSQ/non-AGA BEP: Efficacy by TROP2 QCS-NMR Status TROP2 QCS-NMR positivity is predictive for longer PFS with Dato-DXd in the NSQ/non-AGA biomarker-evaluable population NSQ/non-AGA BEP, n=221 [Kaplan-Meier plot: y-axis PFS probability, %; x-axis Time from randomization, months; four curves] [Table:] | | TROP2 QCS-NMR+ | | TROP2 QCS-NMR- | | | | Dato-DXd n=68 | Docetaxel n=72 | Dato-DXd n=40 | Docetaxel n=41 | ORR, % | 36.8 | 15.3 | 22.5 | 12.2 Median PFS, months | 7.2 | 4.1 | 4.0 | 4.4 PFS HR (95% CI) | 0.52 (0.35-0.78) | 1.22 (0.74-2.00) Treatment by biomarker status interaction: p=0.0098 [Legend:] Dato-DXd, QCS-NMR+ / Dato-DXd, QCS-NMR- / Docetaxel, QCS-NMR+ / Docetaxel, QCS-NMR- Dr Marina Chiara Garassino | Normalized M[embrane ratio ...] Continuous Scoring is Predictive of Clinical [...] [footer clipped by podium] Data cutoff: March 29 2023 PFS HR (95% CI) by TROP2 QCS-NMR status (+ vs -) within treatment: Dato-DXd: 0.40 [0.25-0.64]; Docetaxel: [1.0?]0-1.49] [partially occluded] [Survival data captured as a table — no point estimates read from the curve] Endpoint: PFS (NSQ/non-AGA biomarker-evaluable population, n=221; TROPION-Lung01 Dato-DXd vs docetaxel, by TROP2 QCS-NMR status) Dato-DXd, TROP2 QCS-NMR+ | 68 | 7.2 months Docetaxel, TROP2 QCS-NMR+ | 72 | 4.1 months Dato-DXd, TROP2 QCS-NMR- | 40 | 4.0 months Docetaxel, TROP2 QCS-NMR- | 41 | 4.4 months HR: QCS-NMR+: 0.52; QCS-NMR-: 1.22 HR CI: QCS-NMR+: 0.35-0.78; QCS-NMR-: 0.74-2.00 P: treatment-by-biomarker interaction p=0.0098 Note: ORR %: 36.8 / 15.3 (QCS-NMR+ Dato-DXd / docetaxel); 22.5 / 12.2 (QCS-NMR-). Data cutoff March 29 2023. Footnote within-treatment HR (+ vs -): Dato-DXd 0.40 [0.25-0.64]; docetaxel value partially occluded, reads '[..]0-1.49]'. No landmark % printed. --- [Slide 3] Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum [A] Quality ratings [Kernel density plot; y-axis Density; curves labelled Physicians and Chatbot; x-axis Response options: [Very poor / Poor / Accepta]ble, Good, Very good - left labels hidden by speaker] [B] Empathy ratings [Kernel density plot; y-axis Density; curves labelled Physicians and Chatbot; x-axis Response options: Not empathetic, Slightly empathetic, Moderately empathetic, Empathetic, Very empathetic] Dist[ribution of Qualit]y Ratings for Chatbot and Physician Responses to Patient Questions Kernel density plot[s ... in]dependent licensed health care professional evaluators using principles of crowd eva[luation ... s]hown. B, The overall empathy metric is shown. 11 [JAMA] Intern Med. 2023;183(6):589-596. doi:10.1001/jamainternmed.2023.1838 UC San Diego Moores Cancer Center

[Slide 1] The Consequences - From Research to Equitable Access & Care [Five-step flow with icons and arrows:] Basic science - Research questions Translational research - [Wha]t moves forward Clinical trials - Design and access Clinical practice - Treatment and services Real-world outcomes - Impact on people's lives Trinity Colle[ge Dublin ...] [footer partially hidden] 5

[Slide 2] AI and Healthcare Cycle [Hype-cycle style curve: y-axis EXPECTATION, x-axis TIME; labelled points along curve, each marked with a circle whose fill indicates expected time to standard of care] Cohort Refinement Spatial Omics [Radio]mics [partially hidden by speaker] [Biomarker discovery - hidden by speaker] [Simulated Clinical] Trial [partially hidden] Clinical Decision Support Models Patient - ChatBot Interface Restructuring Free Text Data Security Challenges Wearable devices Real World Data Integration Telehealth Decentralized Trials Digital Twin Creation Patient-Trial Matching Adverse Event Automated Reporting Ambient AI Scribe Note Generation [Phase labels along x-axis:] [Inspiration for Innovation - partially hidden] | Rise of Inflated Expectation | Fall of Disillusionment | Ascent of Enlightenment | Stabilization of Productivity TIME [Legend:] [Expected time to s]tandard of care: O <2 yr. [light] 2-5 yrs. [mid] 5-10 yrs. [dark] >10 yrs. [Az]enkot, Patel ASCO Ed Book 2025 UC San Dieg[o] Moores Cancer C[enter] --- [Slide 3] The Consequences - From Research to Equitable Access & Care [Five-step flow with icons and arrows:] Basic science - Research questions Translational research - [Wha]t moves forward Clinical trials - Design and access Clinical practice - Treatment and services Real-world outcomes - Impact on people's lives Trinity Colle[ge Dublin ...] [footer partially hidden] 5

[Slide 1] DeLLphi-308 - Subcutaneous tarlatamab formulation - Will lowering the Cmax to reduce CRS - 15mg SC had similar exposure to 10mg IV Key Treatment-Related Adverse Events by Grade 15 mg Q2W SC-treated patients (n = 40) TRAE, n (%) | G1 | G2 | >= G3 | Overall CRS | 11 (28) | 4 (10) | 0 | 15 (38) ICANS | 0 | 0 | 0 | 0 Injection site reactions | 18 (45) | 2 (5) | 0 | 20 (50) Dysgeusia | 14 (35) | 5 (13) | 0 | 19 (48) Neutropenia | 2 (5) | 2 (5) | 0 | 4 (10) CRS - IV, historical DeLLphi-304 (n = 252)1 | 107 (42) | 32 (13) | 3 (1) | 142 (56) [CRS and ICANS rows outlined in red box] [Swimmer plot: Duration of Treatment (Weeks) and Best Overall Response; x-axis 0-70; legend: Partial Response, Stable Disease, Progressive Disease, Not Evaluable, First PD, Ongoing Treatment] Rocha, WCLC 2026 Georgetown | Lombardi --- [Slide 2] Tambotatug Pelitecan (Tam-Peli, YL201) - TAISHAN-302 - Phase III trial, 2L SCLC, China only, tam-peli vs topotecan Key eligibility criteria: - Histologically or cytologically confirmed SCLC - Progressed after 1 prior line of platinum-based therapy - At least one measurable lesion per RECIST v1.1 - ECOG PS 0-1 R 1:1 N=451 Treatment until disease progression per RECIST v1.1 or intolerable toxicity. Crossover between arms is not permitted. Tam-Peli 2.0 mg/kg, D1 Q3W, maximum dose of 200 mg Topotecan 1.2 mg/m2, D1-5 Q3W* Stratification factors - Disease status (local** or systemic disease) - Brain metastasis (yes or no) - Chemotherapy-free interval (<90 or >=90 days) Primary endpoint: - OS Key secondary endpoints: - PFS by investigators - ORR by investigators Other secondary endpoints: - DCR, DoR, TTR by investigators - Safety, PK, and immunogenicity Exploratory endpoints: - Intracranial efficacy Zhang, WCLC 2026; Zhao, NEJM 2026 Georgetown | Lombardi --- [Slide 3] Where We Are Today - Standard of care for ES-SCLC is evolving [Three chevron sequences:] Chemotherapy + Immunotherapy -> Maintenance Lurbinectedin -> Tarlatamab -> ADC Chemotherapy + Immunotherapy -> Maintenance Tarlatamab -> ADCs -> ? ADC + Bispecifics + Tarlatamab -> ? -> ? -> ? Georgetown | Lombardi

[Slide 1] A. Neoadjuvant Arm NSCLC-OVA+ cells [Schematic: mouse injected with cells -> mouse with tumor -> R (randomization) -> Treatments:] Treatments: 1. IgG 2. Anti-PD-1 3. Anti-CTLA-4 4. Anti-PD-1+anti-C[TLA-4] [clipped at right edge] [Timeline:] D1 D2 D[3...] [clipped] B. Adj[uvant Arm] [mostly hidden by speaker] [... c]ells [R] [The University of Texas] MD An[derson] Cancer Center [logo partially visible at lower left] --- [Slide 2] Neoadjuvant combined ICIs are superior to adjuvant therapy at reducing lung metastases in NSCLC-OVA+ model A. [Photographs of excised mouse lungs; rows: Neoadjuvant, Adjuvant; columns: IgG, anti-PD-1, anti-CTLA4, anti-PD-1+anti-CTLA4; blue arrows mark the combination column] B. [Violin plot; y-axis Microscopic lung metastases, 0-80; x-axis Neoad[juvant Arm] / [Adjuvant Arm - hidden by speaker]; legend: IgG, anti-PD-1, anti-CTLA-4, anti-CTLA-4+anti-PD-1; asterisks * mark significance] C. [Violin plot; y-axis Macroscopic lung metastases, 0-50; x-axis Neoadjuvant Arm, Adjuvant Arm; legend: IgG, anti-PD-1, anti-CTLA-4, anti-PD-1+anti-CTLA-4; asterisks * and *** mark significance] Cascone T et al, Cancer Res (2018) 78 (13_Supplement[):] 1719 The University of Texas MD Anderson Cancer Center --- [Slide 3] Responses to neoadjuvant Nivo/Ipi in stage II-IIIA resectable NSCLC (JHU & MSKCC) - Terminated early after 9/15 pts enrolled: 67% (6/9) TRAEs and 33% (3/9) grade >=3 TRAEs. - 3 of 9 patients (33%) had PD and no definitive surgery; 6 pts underwent resection (67%): 2 pCR (33%) - 6 pts (67%) had tumor STK11 mutations with or without KRAS or KEAP1 co-mutations [Left: colour-coded heatmap grid, columns Patient 1, Patient 5, Patient 6, Patient 3, Patient 7, Patient 8, Patient 2, Patient 4, Patient 9; rows pCR, Res Tumor (%), RECIST#, Prim Pro[gression?], Recurren[ce], [further rows hidden by speaker]; legend partially hidden: PD, PR, No, ...ained] Table 3 Radiographic, pathologic and molecular response characteristics Patient number | Radiographic response* | Residual tumor (%) | Pre-treatment PD-L1 (%) | Normalized tumor mutation burden | Driver genes with sequence alterations 1 | PD | 100 | 0 | 344 | KRAS, STK11 2 | PD+ | N/A | 1 | 109 | KRAS, KEAP1, STK11 3 | SD | 90 | 10 | 147 | KRAS, STK11, TP53 4 | PD | N/A | 1 | 63 | KRAS, KEAP1, STK11 5 | PR | 0 (pCR) | 75 | 554 | KRAS, STK11 6 | SD | 0 (pCR) | 95 | Undeterminable++ | Undeterminable++ [7] | SD | 70 | 0 | 914 | TP53 [8] | PD | N/A | 0 | 78 | BRAF, STK11, TP53 [9] | SD | 20 | 30 | 99 | TP53 [Driver genes column outlined in red box] Reuss JE, et al, JITC 2020 The University of Texas MD Anderson Cancer Center --- [Slide 4] Apparent radiographic nodal disease progression after ICIs with evidence of granulomas and no tumor: Nodal Immune Flare (NIF) Neoadjuvant ICIs Baseline [PET/CT axial image and coronal image at baseline] P[ost-treatment?] [label hidden by speaker] [PET/CT axial image and coronal image after ICIs showing increased nodal uptake] Neoadjuvant ICIs Baseline FNA - No granulomas At resection - Non-caseating granulomas 16% of patients developed NIF post ICIs (7/44) with non-caseating granulomas and no tumor Neoadjuvant Chemotherapy Baseline FNA - No granulomas At resection - No non-caseating granulomas No cases of NIF after chemotherapy (0/28) Cascone, T et al. Nat Commun 2021 The University of Texas MD Anderson Cancer Center

[Slide 1] Nearly all patients in the US with EGFRm mNSCLC treated with EGFR-TKIs present with known risk factors for mortality1 [Pictogram grid of coloured person icons, colours keyed to the risk factors at right] Age >=65 EGFR p.L858R* ECOG PS >= 2 Brain metastases Liver metastases TP53 co-mutation ctDNA positive+ *Compared with exon 19 d[eletion ...] [clipped] ctDNA, circulating tumour D[NA ...] [clipped] 1. Sabari J, et al. J Thorac [Oncol ...] [clipped] [...detec]ted in ctDNA at baseline by NGS [...] epidermal growth factor receptor; NGS, next-generation sequencing; TP53, tumour protein p53 [AstraZeneca logo] --- [Slide 2] 1L treatment for EGFRm mNSCLC is evolving, with emerging combination approaches1-3 Should all patients with EGFRm mNSCLC receive escalated 1L treatment? In the FLAURA trial, 28% of patients were continuing to receive osimertinib at 3 years1 FLAURA2 Osimertinib + platinum-pemetrexed2 - Patients with untreated locally advanced / metastatic EGFRm NSCLC - Ex19del / L858R - No prior systemic / EGFR-TKI therapy for advanced NSCLC - [Sta]ble CNS metastases allowed R 1:1 N=557 Osimertinib 80 mg QD + pemetrexed 500 mg/m2 + carboplatin AUC5 or cisplatin 75 mg/m2 Q3W for 4 cycles (n=279) -> Maintenance osimertinib 80 mg QD + pemetrexed 500 mg/m2 Q3W Osimertinib monotherapy 80 mg QD (n=278) Primary endpoint: PFS by investigator assessment Secondary endpoints: OS, ORR, DoR, DCR, HRQoL, safety, PFS2, TFST, TSST MARIPOSA Amivantamab + lazertinib3 [Patients] with untreated locally [advanced /] metastatic EGFRm NSCLC [... EGFR-T]KI therapy for [...] [partially hidden by speaker] R 2:2:1 N=1,074 Amivantamab 1,050 mg (1,400 mg if >=80 kg) Q1W for 1 cycle, then Q2W + lazertinib 240 mg QD (n=429) Osimertinib monotherapy 80 mg QD (n=429) Lazertinib 240 mg QD (n=216) Primary endpoint: PFS by BICR (amivantamab + lazertinib vs. osimertinib) Secondary endpoints (amivantamab + lazertinib vs. osimertinib): OS, ORR, DoR, PFS2, symptomatic PFS, IC PFS, safety NOTE: Cross tria[l comparisons ... may b]e due to differences such as patient populations, trial designs/endpoints and follow-up MARIPOSA: lazertinib monoth[erapy ...] 1L, first-line; AUC, area under [the curve ...] EGFR(m), epidermal growth f[actor receptor ...] free survival; PFS2, time from ra[ndomisation ...] [...] system; DCR, disease control rate; DoR, duration of response; ECOG, Eastern Cooperative Oncology Group performance status; [...] progression- [...] quality of life; IC, intracranial; (m)NSCLC, (metastatic) non-small cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression- [...]; R, randomisation; TFST, time to first subsequent therapy; TKI, tyrosine kinase inhibitor; TSST, time to second subsequent therapy. 1. Ramalingam SS, et al. N Engl [J Med ...] [...] 1935-48; 3. Cho B, et al. N Engl J Med. 2024;391:1486-98 [AstraZeneca logo]

[Slide 1] ADRIATIC: Consolidation durvalumab significantly improved OS and PFS vs. placebo in patients with LS-SCLC1,2 OS1 [Kaplan-Meier plot: y-axis Probability of OS; x-axis Time from randomisation (months) 0-63; durvalumab (magenta) vs placebo (grey); printed landmark labels 68.0% and 58.5% at first dashed line, 56.5% and 47.6% at second dashed line] Results shown are from the prespecified primary analysis (data cut-off: 15 January 2024). No. at risk D 264 261 [...occluded...] 68 51 39 27 19 11 5 1 0 PBO 266 26[...occluded...] 62 44 [3?] 23 19 8 5 1 0 [OS table, left column partially hidden by speaker:] | [Durvalumab] | Placebo [mOS, mo (95% CI)] | [hidden] | 33.4 (25.5, 39.9) [3-year OS, %] | [hidden] | 47.6 HR (98.32[1]% CI); [p-value] | [hidden] [...]8); p=0.01 PFS1 [Kaplan-Meier plot: y-axis Probability of PFS; x-axis Time from randomisation (months) 0-63; printed landmark labels 48.8% and 36.1% at first dashed line, 46.2% and 34.2% at second dashed line] No. at risk D 264 212 161 135 113 105 101 98 84 78 51 51 33 21 19 10 10 4 4 0 0 [0] PBO 266 208 146 122 100 88 79 76 71 69 47 47 34 23 22 15 14 5 5 0 0 [0] | Durvalumab | Placebo mPFS, mo (95% CI) | 16.6 (10.2, 28.2) | 9.2 (7.4, 12.9) 2-year PFS, % | 46.2 | 34.2 HR (99.816% CI); p-value | 0.76 (0.53, 1.08); p=0.02 Data cut-off: 15 January 202[4 ...] [...]ancer; (m)OS, (median) overall survival; (m)PFS, median progression-free survival; NR, not reached; PBO, placebo. CI, confidence interval; D, durva[lumab ...] 1. Cheng Y, et al. N Engl J Med. [...] [...]es.ie/medicines/imfinzi-50-mg-ml-concentrate-for-solution-for-infusion-A0002311/spc (Accessed October 2026) [AstraZeneca logo] [Survival data captured as a table — no point estimates read from the curve] Endpoint: OS and PFS, consolidation durvalumab vs placebo, LS-SCLC (ADRIATIC, prespecified primary analysis, data cut-off 15 January 2024) Durvalumab (OS) | 264 | [hidden by speaker] | [hidden by speaker] Placebo (OS) | 266 | 33.4 months | 25.5, 39.9 Durvalumab (PFS) | 264 | 16.6 months | 10.2, 28.2 Placebo (PFS) | 266 | 9.2 months | 7.4, 12.9 HR: OS: [value hidden by speaker]; PFS: 0.76 HR CI: OS: 98.32[1]% CI, upper bound ends '...8)' (rest hidden); PFS: 99.816% CI 0.53, 1.08 P: OS p=0.01; PFS p=0.02 OS printed labels: 68.0% / 58.5% (first dashed line) and 56.5% / 47.6% (second dashed line); table row for placebo reads 47.6 (3-year OS row label hidden) PFS printed labels: 48.8% / 36.1% (first dashed line) and 46.2% / 34.2% (second dashed line); table: 2-year PFS 46.2% durvalumab vs 34.2% placebo Note: Durvalumab mOS, OS HR point estimate and CI lower bound, and the OS table row labels are all hidden behind the speaker (Gerry Hanna) - do NOT fill from memory; obtain from Cheng et al NEJM 2024 if needed. Landmark time points for the dashed lines are not printed as labels on the curves; the table rows confirm 2-year PFS; the OS row label is hidden. At-risk rows for OS partially occluded; one PBO at-risk value uncertain '[3?]'.