NorthStar Trial

[Slide 1] LBA72: Phase 2 NORTHSTAR Trial Does adding local consolidation therapy (LCT) to first-line osimertinib improve PFS in advanced EGFRm NSCLC? Primary Endpoint: mPFS 17.5 vs 25.3 mo (HR 0.66) 100 Points for Discussion: 82% 80 Osi +LCT Osi Which patients are best suited for LCT? 68% PFS % 60 40% When should LCT be delivered? 40 20 24% What type of LCT and to what sites? 0 0 12 24 36 48 60 72 What are optimal endpoints for LCT trials? Months No. at risk Osi LCT 56 44 25 19 11 2 0 Osi 63 41 21 13 5 1 1 Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use 2025 ESMO congress BERLIN --- [Slide 2] NORTHSTAR Patient population is heterogenous "Baseline" These subgroups reflect key differences in "Baseline" Oligometastatic Polymetastatic disease biology: Metastatic potential TKI sensitivity Tumor heterogeneity A "one size fits all" LCT approach will likely be inadequate Clearer definitions and better biomarkers are "Induced" needed to identify patients most likely to Oligometastatic benefit from LCT Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use congress BERLIN 2025 ESMO --- [Slide 3] LCT Target Selection: What does "residual disease" really mean? Baseline Disease Residual Disease NORTHSTAR LCT Requirements: Induction Therapy Oligometastatic disease: LCT to all sites Polymetastatic disease: LCT to as many 6-12 weeks TKI sites as feasible and safe CT or PET-CT, RECIST 1.1 1. What is the optimal duration of TKI treatment prior to LCT? Likely not the same for all patients. Need for dynamic biomarkers (MRD) 2. How can we accurately identify sites of disease that should be treated? Metabolic response by PET Radiomics Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use congress BERLIN 2025 ESMO --- [Slide 4] What is the best way to assess the benefit of LCT? PFS benefits have been observed in prior LCT trials, but the administration of LCT to RECIST target lesions in the experimental arm impacts likelihood of progression, and therefore PFS endpoint. Overall survival is needed to fully assess the impact of the LCT intervention What is the impact of local therapy "crossover" at later timepoints? Osimertinib LCT to Residual Disease Early VS. Late OS Osimertinib Local Therapy to Local Therapy Oligoprogression (or palliative RT) Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use BERLIN 2025 ESMO congre

[Slide 1] NCT NORTHSTAR: Osimertinib, Surgery, and Radiation Therapy in Treating Patients With 03410043 Stage IIIB or IV Non-small Cell Lung Cancer With EGFR Mutations RECRUITING Principal Investigators: Yasir Y. Elamin, MD- University of Texas M.D. Anderson Cancer Center Phase II Stage IIIB or IV non-small cell ARM 1 lung cancer with EGFR mutation Osimertinib and local consolidation therapy OUTCOMES and candidate for local consolidation therapy (LCT) PRIMARY to at least one disease site Osimertinib + and/or Progression free survival (up to 4 years) OD for 6-12 weeks prior to LCT followed by QD for 4 Surgical Rediation 00 5 days SECONDARY week cycles, as tolerated Intervention per week ( 8 weeks Overall survival Time to progression of 143 adults undergo ARM 2 target lesions randomization Time to apperance of new Osimertinib alone metastasis Progression free survival Inclusion: in oligometastatic group EGFR exon 19 deletion/L858R Osimertinib mutation (must be TKI naive) Incidence of adverse events EGFR T790M mutation (must not have receivied 3rd gen EGFR TKI) I QD for 4 week cycles, as tolerated ANNALS OF SURGICAL ONCOLOGY --- [Slide 2] LBA72: Phase 2 NORTHSTAR Trial Does adding local consolidation therapy (LCT) to first-line osimertinib improve PFS in advanced EGFRm NSCLC? Primary Endpoint: mPFS 17.5 vs 25.3 mo (HR 0.66) 100 Points for Discussion: 82% 80 Osi +LCT Osi Which patients are best suited for LCT? 68% PFS % 60 40% When should LCT be delivered? 40 20 24% What type of LCT and to what sites? 0 0 12 24 36 48 60 72 What are optimal endpoints for LCT trials? Months No. at risk Osi . LCT 56 44 25 19 11 2 0 Osi 63 41 21 13 5 1 1 Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use 2025 ESMO congress BERLIN --- [Slide 3] What is the best way to assess the benefit of LCT? PFS benefits have been observed in prior LCT trials, but the administration of LCT to RECIST target lesions in the experimental arm impacts likelihood of progression, and therefore PFS endpoint. Overall survival is needed to fully assess the impact of the LCT intervention What is the impact of local therapy "crossover" at later timepoints? Osimertinib LCT to Residual Disease Early VS. Late OS Osimertinib Local Therapy to Local Therapy Oligoprogression (or palliative RT) Zosia Piotrowska, Boston, USA Content of this presentation is copyright and responsibility of the author. Permission is required for re-use BERLIN 2025 ESMO congre --- [Slide 4] Conclusions In patients with EGFR-mutant metastatic NSCLC, osimertinib + LCT improved PFS compared to osimertinib; 25.3 vs 17.5 months (HR 0.66; p = 0.025). The benefit of LCT was observed across disease burden groups including patients with polymetastatic disease; complete LCT was achieved in 59% of patients who presented with polymetastatic disease with median PFS of 27.9 months. The extent of consolidation (complete VS partial) may better predict LCT benefit than the initial metastatic burden (number of metastases). No excess toxicity was observed — in particular, there was no increased rate of pneumonitis beyond that expected with thoracic radiation. These findings support osimertinib + LCT as a safe and effective strategy to extend disease control and delay systemic progression in advanced EGFR-mutant NSCLC. Yasir Elamin, MD Content of this presentation is copyright and responsibility of the author. Permission is required for re-use BERLIN 2025 ESMO congress

[Slide 1] Higher Radiation BED to Primary Tumor Associated with Improved Outcom >75 <75 Gy Gy 10 Organ at Risk Median Range N (%) P 0.006 (1-sided) 0.8 Lung Rediation Modality V20Gy 17% (3%-31.6%) UMATO MRT 30 30 (71) 15 (36) Mean 8.2 Gy (0.14 Gy - 18.4 Gy) S887 30 Conformal or 0 Conformal 7 (17) PFS (Probability) 0.6 Esophagus 0.4 Mean 9.2 Gy (0.47 Gy - 36.7 Gy) Site Lung 38 (90) 0.2 Bone 9(21) Heart Brain 3 (7) Mean 6.0 Gy (0.02 Gy - 24.9 Gy) Gy 0.0 4 (10) 36 48 60 72 Other 0 12 24 Time (Months) a75 Gy Dose At risk 14 11 9 8 5 2 0 25 (60) Censored 0 1 1 1 3 4 6 4 5 5 8 8 50Gyin4-56x 9(21) Events 0 2 60-70Gyin9-20fx 3 (7) <75 Gy 30-40 Gy in 10 fx 8 (29) At risk 16 11 4 3 2 0 0 Censored 0 0 3 (7) 3 3 3 3 3 Events 0 $ 9 10 11 13 13 19-20Gra1b 3 (7) 50Gyin 25 6x 1 (2) Relevant treatment related AEs in radiation cohort G1-3 Pneumonitis Event Grade 1 Grade 2 BED ≥75 Gy: median PFS = 49.1 mo (95% CI 21.6-NA) Grade 3 Lung V20 Gy BED <75 Gy: median PFS = 22.3 mo (95% CI 11.6-39.7) Pneumonitis 1 (2.3) 5(11.9) 1 (2.3) > 25% < 25% HR 0.31 (90% CI 0.14-0.70), p = 0.006 Esophagitis 3(7.1) 1 (2.3) Dyspnea 12 (28.5) 1 (2.3) 4/9 2/29 p<0.007 Saumil (Dudhi, MD PhD Content of this copyright responsibility of the author Permission is required for re-use Two presentation RT "levers": is and aim for thoracic BED ≥75 Gy when feasible. Keep lung V20 <25% to mitigate pneumonitis risk. ESMO ASLC ASSOCIATION INTERNATIONAL FOR THE STUDY or LUNG CANCER ESTRO Increase I 2 ETOP-IBCSG lings - 1 --- [Slide 2] Conclusions Addition of LCT to osimertinib significantly prolonged PFS compared with osimertinib alone Clearance of thoracic nodal disease and pleural effusion following induction osimertinib are potential predictors of benefit from LCT Consolidative radiotherapy results in excellent local control rates with failures occurring outside the RT field at new distant metastatic sites outcomes Thoracic BED and ≥ 75Gy and lung V20Gy < 25% are associated with improved minimal toxicity Saumil Gandhi, MD PhD Content of his presentation IS copyright and responsibility of the author Permission is required for re-use Organisers ESMO IASLC Partners INTERNATIONAL FOR THE STUDY OF LING CANCER ESTRO I 1 I i ETOP-IBCSG --- [Slide 3] Approach Tested in Unselected Real-World Population Oligo Poly Baseline 84 (71%) Unanswered Questions 35 (29%) Oligo Induced Oligo Poly 15 Post Induction 35 (29%) (13%) 69 (58%) 1. How do we deliver safe and effective RT? Randomized to LCT 56 2. Which patients derive the largest benefit Cige induced Cligo Poly Complete LCT 100% from LCT after induction osimertinib? 56% - ESMD 3. Where do patients fail after LCT Saumil Gandhi, MD PhD Content of a this presentation is copyright and responsibility of the author Permission is required for re-use Organisers ESMO IASLC Partners ASSOCIATIONAL OF LUNG CANCER ESTRO Empensions 1 L ETOP-IBCSG ling / --- [Slide 4] Higher Radiation BED to Primary Tumor Associated with Improved Outcom >75 <75 Gy Gy 10 Organ at Risk Median Range N (%) P 0.006 (1-sided) 0.8 Lung Rediation Modality V20Gy 17% (3%-31.6%) UMAT MRT 30 (71) 15 (36) Mean 8.2 Gy (0.14 Gy - 18.4 Gy) S887 30 Conformal or 0 Conformal 7 (17) PFS (Probability) 0.6 Esophagus 0.4 Mean 9.2 Gy (0.47 Gy - 36.7 Gy) Site Lung 38 (90) 0.2 Bone 9(21) Heart Brain 3 (7) Mean 6.0 Gy 0.02 Gy 24.9 Gy) 0.0 Other 0 12 36 48 60 72 4 (10) 24 Time (Months) a75 Gy Dose At risk 14 11 9 8 5 2 0 25 (60) Censored 0 1 1 1 3 4 6 5 8 8 50Gyin4-56x 9(21) Events 0 2 4 5 60-70Gyin9-20fx 3 (7) <75 Gy 30-40 Gy in 10 fx 8 (29) At risk 16 11 4 3 2 0 0 Censored 0 0 3(7) 3 3 3 3 3 Events 0 $ 9 10 11 13 13 19-20Gra1b 3 (7) 50Gyin 25 6x 1 (2) Relevant treatment related AEs in radiation cohort G1-3 Pneumonitis Event Grade 1 Grade 2 BED ≥75 Gy: median PFS = 49.1 mo (95% CI 21.6-NA) Grade 3 Lung V20 Gy BED <75 Gy: median PFS = 22.3 mo (95% CI 11.6-39.7) Pneumonitis 1 (2.3) 5(11.9) 1 (2.3) Esophagitis > 25% < 25% HR 0.31 (90% CI 0.14-0.70), p = 0.006 3(7.1) 1 (2.3) Dyspnea 12 (28.5) 1 (2.3) 4/9 2/29 p<0.007 Content of this copyright responsibility of the author Permission is required for re-use Two presentation RT "levers": is and aim for thoracic BED ≥75 Gy when feasible. Keep lung V20 <25% to mitigate pneumonitis risk. Saumil (Dudhi, MD PhD ESMO ASLC ASSOCIATION INTERNATIONAL FOR THE STUDY or LUNG CANCER ESTRO Increase 2 ETOP-IBCSG lings - 1 --- [Slide 5] Nodal Clearance after Induction Osimertinib PREDICTIVE of LCT Benefit Thoracic Nodes Absent at Randomization PFS Thoracic Nodes Present at Randomization PFS Osimertinib + LCT 1.0 13 Osimertinib LCT Osimertinib Oximertinib 08 0.8 P - 0.008 (1-sided) p . 0.388 (1-sided) È / 04 04 PFS (Probability) 0.6 0.4 02 0.2 & 0 2 24 38 & 0.0 60 72 Time (Months) 0 12 24 LCT 36 48 60 72 At mk 34 34 * Censored 0 7 $ Osimertinb LCT Time (Months) 2 3 0 4 0 Events 0 $ $ At nisk 22 5 20 22 23 5 Censored 0 16 12 24 0 6 29 Demerting 29 Events 0 2 2 2 0 2 5 At nok 34 4 7 8 9 Censored 0 12 21 1 11 11 8 Osimertinib 13 4 13 Eventy 0 1 1 1 22 3 At risk 28 25 4 27 4 19 Censored 0 10 29 5 29 1 3 1 Events 0 4 0 8 8 0 15 19 8 Osi Osi + LCT: median PFS = 19.0 mo (95% CI 12.6-23.2) 19 8 20 20 alone: HR median PFS = 15.9 mo (95% CI 10.9-23.9) 0.93, 90% CI 0.60-1.43, p = 0.388 Osi Osi alone: + LCT: median PFS = 41.5 mo (95% CI Saumil Gandhi, MD PhD HR median PFS = 19.6 mo (95% CI 31.2-NA) 9.1-31.5) 0.43, 90% CI 0.23-0.78, p = 0.008 Content of this presentation is copyright and responsibility of the author Permission is required for re-use Organisers ESMO ASLC Partners INTERNATIONAL STUDY OF LUNG CANCER ESTRO --- [Slide 6] Conclusions Addition of LCT to osimertinib significantly prolonged PFS compared with osimertinib alone Clearance of thoracic nodal disease and pleural effusion following induction osimertinib are potential predictors of benefit from LCT Consolidative radiotherapy results in excellent local control rates with failures occurring outside the RT field at new distant metastatic sites outcomes Thoracic BED and ≥ 75Gy and lung V20Gy < 25% are associated with improved minimal toxicity Saumil Gandhi, MD PhD Content of his presentation S copyright and responsibility of the author Permission is required for re-use Organisers ESMO IASLC Partners INTERNATIONAL FOR THE STUDY OF LUNG CANCER ESTRO I I i ETOP-IBCSG

[Slide 1] Patterns of Failure after Osimertinib + LCT Site of First Recurrence New VS Prior Involvement 100 100 80 80 Percent 60 Percent 60 40 40 20 20 0 0 New Site Site Present at Both Locoregional Distant Both Baseline Recurrence In Relation to RT Field Organ Site of First Recurrence (RT Alone Cohort) 25 100 20 80 Percent 60 Percent 15 40 10 20 5 0 0 In RT Field Out of RT Field Lung/Pleura Bone Spine Brain Thoracic Liver Adrenal Other node Saumil Gandhi, MD PhD Organizers Partners Content of this presentation n copyright and responsibility of the author Permission is required for re-une ESMO IASLC ESTRO L ETOP-IBCSG --- [Slide 2] Nodal Clearance after Induction Osimertinib PREDICTIVE of LCT Benefit Thoracic Nodes Present at Randomization PFS Thoracic Nodes Absent at Randomization PFS Osimertinib + LCT 1.0 12 Osimertinib LCT Osimertinib Osimertinib as 0.8 p 0.388 (1-sided) P - 0.008 (1-sided) F / 04 04 PFS (Probability) 0.6 0.4 a 0.2 00 0 2 24 36 & 0.0 60 72 Time (Months) 0 12 Chined CT 24 36 48 60 At nat 34 72 34 * Censored 0 7 5 Osimertinb LCT Time (Months) 2 3 0 Events 0 4 0 $ At risk 22 8 5 20 22 23 5 Censored 0 16 24 12 29 0 6 Ownerting 29 2 2 Events 0 2 0 2 5 At nok 34 4 7 27 8 Cansored 0 11 9 11 1 8 4 Osimertinib 13 Eventy 0 1 1 1 13 1 2 22 3 25 I At nisk 28 27 4 19 Censored 0 10 29 29 1 5 Events 0 3 1 4 0 8 8 0 15 19 8 Osi Osi + LCT: median PFS = 19.0 mo (95% CI 12.6-23.2) 19 8 20 20 alone: HR median PFS = 15.9 mo (95% CI 10.9-23.9) 0.93, 90% CI 0.60-1.43, p = 0.388 Osi Osi alone: + LCT: median PFS = 41.5 mo (95% CI Saumil Gandhi, MD PhD HR median PFS = 19.6 mo (95% CI 31.2-NA) 9.1-31.5) 0.43, 90% CI 0.23-0.78, p = 0.008 Content of this presentation is copyright and responsibility of the author Permission is required for re-use Organisers ESMO ASLC Partners INTERNATIONAL FOR STUDY OF LUNG CANCER ESTRO ETOP

[Slide 1] Background The NorthStar trial demonstrated improved progression free survival (PFS) with local consolidative therapy (LCT; radiotherapy [RT] or surgery) added to Osimertinib (Osi) in metastatic EGFR-mutant NSCLC (25.4 VS 17.5 months [mo]; HR 0.66, 90% CI 0.50-0.87; p=0.025). Evaluation of RT techniques, patterns of failure, and predictors of LCT benefit are of immense interest for patient (pt) selection. Methods This secondary analysis assessed RT dosimetry, outcomes, and failure patterns after LCT. Cox proportional- hazards modeling evaluated factors associated with PFS. Results Among 119 pts without progression after 6-12 weeks of Osi, 63 were randomized to Osi alone and 56 to Osi + LCT. LCT modalities were RT alone, 33 (59%); RT + surgery, 6 (11%); surgery alone, 17 (30%). 76% of pts continued Osi during RT. Common RT regimens were 52.5 Gy/15 fx, 50 Gy/4 fx, 60 Gy/15 fx, and 45 Gy/15 fx. Pts without residual thoracic nodal involvement on post-induction diagnostic CT scan derived the greatest benefit, with median PFS of 41.5 mo with Osi + LCT versus 19.6 mo with Osi alone (HR 0.43, 90% CI 0.23-0.78; p=0.008). Similarly, among patients without persistent pleural effusion on post-induction diagnostic CT scan, median PFS was 32.7 versus 22.3 mo (HR 0.63, 90% CI 0.39-1.02; p=0.057). In contrast, no PFS benefit was observed in pts with persistent thoracic nodal disease (19.0 VS 17.5 months; HR 0.94, 90% CI 0.61-1.46; p=0.41) or persistent pleural effusion (15.3 VS 12.9 months; HR 0.90, 90% CI 0.52-1.55; p=0.37) after induction Osi. Among 37 recurrences after LCT, 7 (19%) were locoregional, 23 (62%) were distant, and 7 (19%) were both; 23 (62%) occurred at new metastatic sites. Among the subset who had RT, there were 24 recurrences, only 2 (8%) were in-field, whereas 22 (92%) were out-of-field failures. Conclusions Addition of LCT to Osi significantly prolonged PFS compared with Osi alone. Clearance of thoracic nodal disease and pleural effusion following induction Osi emerged as potential predictors of benefit from LCT. Failures after LCT were predominantly distant, occurring outside the RT field and most commonly at new metastatic sites.