Letter from Prof Rolf Stahel Dear Colleagues It is my pleasure to present this ETOP slide set which has been designed to highlight and summarise key findings in thoracic cancers from the major congresses in 2018. This slide set specifically focuses on the ESMO 2018 Congress and is available in 4 languages – English, French, Chinese and Japanese. The area of clinical research in oncology is a challenging and ever changing environment. Within this environment, we all value access to scientific data and research which helps to educate and inspire further advancements in our roles as scientists, clinicians and educators. I hope you find this review of the latest developments in thoracic cancers of benefit to you in your practice. If you would like to share your thoughts with us we would welcome your comments. Please send any correspondence to etop@etop.eu-org. I would like to thank our ETOP members Solange Peters and Martin Reck for their roles as Editors – for prioritising abstracts and reviewing slide content. The slide set you see before you would not be possible without their commitment and hard work. Finally, we are also very grateful to Lilly Oncology for their financial, administrative and logistical support in the realisation of this activity. Yours sincerely, Rolf Stahel President, ETOP Foundation Council

ETOP Medical Oncology Slide Deck Editors 2018 Focus: advanced NSCLC (not radically treatable stage III & stage IV) Dr Solange Peters Multidisciplinary Oncology Center, Lausanne Cancer Center, Lausanne, Switzerland Focus: other malignancies, SCLC, mesothelioma, rare tumours Dr Martin Reck Department of Thoracic Oncology, Hospital Grosshansdorf, Grosshansdorf, Germany

Contents Screening and biomarkers Early stage and locally advanced NSCLC – Stages I, II and III Advanced NSCLC – Not radically treatable stage III and stage IV First line Later lines Other malignancies SCLC, mesothelioma and thymic epithelial tumours

Screening and biomarkers

65PD: High tumor mutational burden (TMB) and PD-L1 have similar predictive utility in 2L+ NSCLC patients (pts) treated with anti-PD-L1 and anti-CTLA-4 – Higgs BW, et al Study objective To compare the utility of high TMB and PD-L1 to predict outcomes in patients with non-squamous NSCLC who were treated with durvalumab and tremelimumab Methods Immunotherapy-naïve patients (n=213) with non-squamous EGFR and ALK wild- type NSCLC who had progressed after one prior platinum-based therapy were enrolled to durvalumab 20 mg/kg q4w for up to 12 months with tremelimumab 1 mg/kg q4w for 4 cycles Tumour PD-L1 expression classified as ≥25% or <25% using the Ventana PD-L1 (SP263) assay DNA sequencing was performed on 106 of 200 (53%) available tumour biopsies and TMB high defined as the top tertile of values (11 mut/Mb) Higgs BW, et al. Ann Oncol 2018;29(suppl 5):Abstr 65PD

There was no appreciable correlation between TMB and PD-L1 65PD: High tumor mutational burden (TMB) and PD-L1 have similar predictive utility in 2L+ NSCLC patients (pts) treated with anti-PD-L1 and anti-CTLA-4 – Higgs BW, et al Key results There was no appreciable correlation between TMB and PD-L1 Low linear correlation between PD-L1 levels and TMB PD-L1 vs. TMB status overlap PD-L1 ≥25% n=16 16% TMB high n=17 17% n=18 PD-L1 TC, % TMB, mutations/Mb 30 40 50 60 70 80 90 100 20 10 5 25 35 55 45 15 rho=0.3 PD-L1 <25%/ TMB low n=52 50% Higgs BW, et al. Ann Oncol 2018;29(suppl 5):Abstr 65PD

Patients who were TMB high (≥11 mut/Mb) had improved ORR, PFS and OS 65PD: High tumor mutational burden (TMB) and PD-L1 have similar predictive utility in 2L+ NSCLC patients (pts) treated with anti-PD-L1 and anti-CTLA-4 – Higgs BW, et al Key results Conclusions In this study of patients with non-squamous NSCLC TMB and PD-L1 expression were not correlated Patients who were TMB high (≥11 mut/Mb) had improved ORR, PFS and OS TMB ≥11 mut/Mb (n=37) TMB <11 mut/Mb (n=69) PD-L1 ≥25% (n=57) PD-L1 <25% (n=136) Total* (N=213) PFS Median, months (95%CI) 7.1 (1.7, 9.1) 1.7 (1.6, 3.6) 7.1 (3.5, 9.2) 3.3 (1.7, 3.6) 3.5 (1.8, 4.0) 12-month PFS, % (95%CI) 26.8 (13.8, 41.6) 12.6 (5.6, 22.6) 34.2 (21.9, 46.8) 10.0 (4.8, 17.5) 17.6 (12.4, 23.6) OS NE (7.9, NE) 8.9 (4.8, NE) 15.5 (15.4, NE) 9.5 (6.7, NE) 15.4 (10.0, NE) 12-month OS, % (95%CI) 61.2 (42.4, 75.5) 43.0 (30.4, 55.0) 71.6 (57.3, 81.9) 47.3 (38.2, 55.9) 53.8 (46.4, 60.6) *20 patients had unknown PD-L1 expression; NE, not estimable Higgs BW, et al. Ann Oncol 2018;29(suppl 5):Abstr 65PD

1136PD: Discrepancy of tumor neoantigen burden between primary lesions and matched metastases in lung cancer – Jiang T, et al Study objective To compare neoantigen landscapes in primary tumours and metastases in patients with lung cancer Key result Conclusion Neoantigen landscapes were similar in primary tumours and metastases, however, only 20% of neoantigens were shared Comparison of neoantigens in primary tumours and metastases by baseline characteristics Primary Female Yes SCLC Metastasis Male No LUAD Neoantigen counts Age, years Gender Smoking Histology LM15 LM18 LM21 LM14 LM10 LM08 LM13 LM16 LM11 LM19 BM12 BM14 BM18 BM15 Gender Smoking Histology LUSC Jiang T, et al. Ann Oncol 2018;29(suppl 5):Abstr 1136PD

Early stage and locally advanced NSCLC – Stages I, II and III

LBA48_PR: CTONG 1103: Erlotinib versus gemcitabine plus cisplatin as neo-adjuvant treatment for stage IIIA–N2 EGFR-mutation non-small-cell lung cancer (EMERGING): a randomised study – Zhong W, et al Study objective To investigate the efficacy of erlotinib vs. gemcitabine + cisplatin as neoadjuvant/ adjuvant treatment in patients with locally advanced EGFR mutant NSCLC Erlotinib 150 mg/day for 42 days (n=37) Erlotinib 150 mg/day for 12 months Non-PD Key patient inclusion criteria Treatment naïve IIIA-N2 NSCLC EGFR activating mutation ECOG PS 0–1 (n=72) SURGERY Stratification Lymph node status Histology Smoking Sex R 1:1 Gemcitabine + cisplatin* q3w for 2 cycles (n=35) Gemcitabine + cisplatin* q3w for 2 cycles Primary endpoint ORR Secondary endpoints Downstaging rates of pathological lymph nodes, pCR, PFS, OS, safety *Gemcitabine 1250 mg/m2 D1, 8; cisplatin 75 mg/m2 D1 Zhong W, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA48_PR

Gemcitabine + cisplatin (n=35) LBA48_PR: CTONG 1103: Erlotinib versus gemcitabine plus cisplatin as neo-adjuvant treatment for stage IIIA–N2 EGFR-mutation non-small-cell lung cancer (EMERGING): a randomised study – Zhong W, et al Key results ORR in the ITT population was 54.1% (95%CI 37.2, 70.9) with erlotinib vs. 34.3% (95%CI 17.7, 50.8) with gemcitabine + cisplatin (OR 2.26 [95%CI 0.87, 5.84], p=0.092) Erlotinib (n=37) Gemcitabine + cisplatin (n=35) p-value Surgery, n (%) 31 (83.8) 24 (68.6) 0.129 Complete resection, n (%) R0 R1 R2 27 (73.0) 27 (73.0) 1 (2.7) 3 (8.1) 22 (62.9) 22 (62.9) 1 (2.9) 1 (2.9) 0.358 Lymph node downstage, n (%) N2pN0 N2pN1 N2pN2 4 (10.8) 3 (8.1) 1 (2.7) 27 (73.0) 1 (2.9) 1 (2.9) 0 (0) 23 (65.7) 0.185 Zhong W, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA48_PR

Progression-free survival, % LBA48_PR: CTONG 1103: Erlotinib versus gemcitabine plus cisplatin as neo-adjuvant treatment for stage IIIA–N2 EGFR-mutation non-small-cell lung cancer (EMERGING): a randomised study – Zhong W, et al Key results Conclusion Erlotinib in the neoadjuvant setting for EGFRm NSCLC improved ORR, MPR, R0 resection and lymph node downstaging and the safety profile was in line with previous reports PFS (ITT population) 100 80 60 40 20 Progression-free survival, % 37 35 6 28 12 21 13 18 16 4 24 7 3 30 2 1 36 42 Time, months 76.7% 49.0% 36.4% 14.4% Group Events mPFS, months (95%CI) Erlotinib (n=37) 20 21.5 (19.3, 23.6) Gemcitabine + cisplatin (n=35) 26 11.9 (9.1, 14.7) HR 0.42 (95%CI 0.23, 0.76), p=0.003 No.at risk Erlotinib Gemcitabine + cisplatin Zhong W, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA48_PR

Ipilimumab 1 mg/kg D1 + nivolumab 3 mg/kg D1, 15, 29 (n=18) LBA49: Neoadjuvant nivolumab (N) or nivolumab plus ipilimumab (NI) for resectable non-small cell lung cancer (NSCLC) – Cascone T, et al Study objective To investigate the efficacy and safety of nivolumab alone or in combination with ipilimumab in patients with untreated resectable NSCLC Nivolumab 3 mg/kg D 1, 15, 29 (n=18) Adjuvant therapy Surgery Key patient inclusion criteria Stage I–IIIA NSCLC Eligible for surgery (n=44) Stratification Stage R 1:1 Ipilimumab 1 mg/kg D1 + nivolumab 3 mg/kg D1, 15, 29 (n=18) Adjuvant therapy Surgery Primary endpoint Major pathological response (MPR) Secondary endpoints Safety, perioperative mortality and morbidity, ORR, RFS, OS, complete resection rate, pCR Cascone T, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA49

Nivolumab + ipilimumab LBA49: Neoadjuvant nivolumab (N) or nivolumab plus ipilimumab (NI) for resectable non-small cell lung cancer (NSCLC) – Cascone T, et al Key results Total Nivolumab Nivolumab + ipilimumab Evaluable (resected) n=26 n=14 n=12 MPR + pCR, n (%) 8 (31) 4 (28) 4 (33) 0% viable tumour cells (pCR), n (%) 5 (19) 2 (14) 3 (25) 1–10% viable tumour cells, n (%) 3 (11) 1 (8) Overall (resected + unresectable) n=31 n=16 n=15 8 (26) 4 (25) 4 (27) 5 (16) 2 (13) 3 (20) 3 (10) 1 (7) Cascone T, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA49

Nivolumab + ipilimumab (n=16) LBA49: Neoadjuvant nivolumab (N) or nivolumab plus ipilimumab (NI) for resectable non-small cell lung cancer (NSCLC) – Cascone T, et al Key results (cont.) Conclusions In resected patients with NSCLC, neoadjuvant nivolumab or nivolumab + ipilimumab induced an MPR rate of 31% Greater TIL proliferation and activation was demonstrated with neoadjuvant nivolumab and nivolumab + ipilimumab compared with untreated tumours Radiographic responses Evaluable Total (n=32) Nivolumab (n=16) Nivolumab + ipilimumab (n=16) Response, n (%) CR PR SD PD 1 (3) 6 (19) 19 (59) 0 (0) 5 (31) 8 (50) 3 (19) 1 (6) 11 (69) ORR, n/N (%) 7/32 (22) 5/16 (31) 2/16 (12) Nivolumab (evaluable) (n=16) Nivolumab + ipilimumab (evaluable) (n=16) 35 * 21 14 22 13 14 12 7 4 10 3 2 4 Overall % change in tumour size from baseline –2 –6 –12 –16 –17 –18 * –19 –20 –30 –21 –34 –31 –39 –39 –45 –100 *Considered SD in target lesion but overall PD due to new radiographic lesions Cascone T, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA49

1363O: Efficacy and safety evaluation based on time from completion of radiotherapy to randomization with durvalumab or placebo in pts from PACIFIC – Faivre-Finn C, et al Study objective To investigate the efficacy and safety outcomes of durvalumab in the PACIFIC trial based on PD-L1 expression and the components of chemoradiation received prior to durvalumab or placebo Key patient inclusion criteria Stage III NSCLC No progression after ≥2 cycles of platinum-CRT Pre-cCRT tumour tissue for PD-L1 testing if available WHO PS 0–1 (n=713) Durvalumab 10 mg/kg q2w (n=476) Stratification Age Sex Smoking history R 1:1 Placebo q2w (n=237) Primary endpoints PFS by BICR, OS Secondary endpoints ORR, DoR and TTDM by BICR, PFS2 by investigator, safety, PROs Faivre-Finn C, et al. Ann Oncol 2018;29(suppl 5):Abstr 1363O

1363O: Efficacy and safety evaluation based on time from completion of radiotherapy to randomization with durvalumab or placebo in pts from PACIFIC – Faivre-Finn C, et al Key results Median OS in the PD-L1 TC ≥1% was NR (95%CI NR, NR) with durvalumab and 29.1 months (95%CI 17.7, NR) for placebo; HR 0.53 (95%CI 0.36, 0.77) Median OS in the PD-L1 TC <1% was NR (95%CI 20.8, NR) with durvalumab and NR (95%CI 27.3, NR) for placebo; HR 1.36 (95%CI 0.79, 2.34) PFS (BICR) by PD-L1 TC ≥1% PFS (BICR) by PD-L1 TC <1% Events/ patients (%) Median PFS, months (95%CI) Durvalumab 84/212 (39.6) 17.8 (16.9, NR) Placebo 59/91 (64.8) 5.6 (3.6, 11.0) HR 0.46 (95%CI 0.33, 0.64) Events/ patients (%) Median PFS, months (95%CI) Durvalumab 49/90 (54.4) 10.7 (7.3, NR) Placebo 40/58 (69.0) 5.6 (3.7, 10.6) HR 0.73 (95%CI 0.48, 1.11) Probability of PFS 1.0 0.8 0.6 0.4 0.2 0.0 3 6 9 12 15 21 24 27 18 Time from randomisation, months 212 91 174 59 143 39 127 34 82 20 52 13 14 4 1 30 8 No. at risk Durvalumab Placebo Probability of PFS 1.0 0.8 0.6 0.4 0.2 0.0 3 6 9 12 15 21 24 27 18 Time from randomisation, months 90 58 70 45 51 25 42 23 14 8 1 4 5 No. at risk Durvalumab Placebo Faivre-Finn C, et al. Ann Oncol 2018;29(suppl 5):Abstr 1363O

1363O: Efficacy and safety evaluation based on time from completion of radiotherapy to randomization with durvalumab or placebo in pts from PACIFIC – Faivre-Finn C, et al Key results (cont.) Conclusions Durvalumab led to improvement in PFS and OS in the PD-L1 ≥1% subgroup and PFS in the <1% PD-L1 subgroup Durvalumab improved PFS and OS regardless of chemoradiation received prior to therapy PFS (BICR) Events / patients (%) HR (95%CI) Durvalumab Placebo Induction chemotherapy Yes No 59/123 (48.0) 155/353 (43.9) 49/68 (72.1) 108/169 (63.9) Platinum Cisplatin Carboplatin 115/266 (43.2) 91/199 (45.7) 87/129 (67.4) 65/102 (63.7) Taxane 97/207 (46.9) 117/269 (43.5) 72/112 (64.3) 85/125 (68.0) Etoposide 49/117 (41.9) 165/359 (46.0) 34/52 (65.4) 123/185 (66.5) Vinorelbine 58/124 (46.8) 156/352 (44.3) 42/59 (71.2) 115/178 (64.6) Dose of radiotherapy <60 Gy 60–66 Gy >66 Gy 16/38 (42.1) 187/407 (45.9) 10/30 (33.3) 11/15 (73.3) 130/202 (64.4) 15/19 (78.9) 0.25 0.5 1 2 Placebo better Durvalumab better Faivre-Finn C, et al. Ann Oncol 2018;29(suppl 5):Abstr 1363O

Chemotherapy: topotecan (oral or IV) q3w or carboplatin-etoposide q3w* 1664O: A randomized non-comparative phase II study of anti–PD-L1 atezolizumab or chemotherapy as second-line therapy in patients with small cell lung cancer: results from the IFCT-1603 Trial – Pujol J, et al Study objective To investigate the activity of atezolizumab as systemic therapy in patients with SCLC who have progressed after 1L platinum-based chemotherapy Atezolizumab 1200 mg q3w (n=49) PD Key patient inclusion criteria ED or LD SCLC Progressive disease after 1L chemotherapy No brain metastases PS 0–2 (n=73) Stratification Sensitive vs. refractory disease PS (0–1 vs. 2) Limited vs. extensive disease Gender R 2:1 Chemotherapy: topotecan (oral or IV) q3w or carboplatin-etoposide q3w* (n=24) PD Primary endpoint ORR at 6 weeks (investigator assessed) *Maximum 6 cycles Pujol J, et al. Ann Oncol 2018;29(suppl 5):Abstr 1664O

Progression, n (%) [95%CI] 30 (69.8) [56.0, 83.5] 6 (30) [9.9, 50.1] 1664O: A randomized non-comparative phase II study of anti–PD-L1 atezolizumab or chemotherapy as second-line therapy in patients with small cell lung cancer: results from the IFCT-1603 Trial – Pujol J, et al Key results Atezolizumab (n=43) Chemotherapy (n=20) Total (n=64) ORR at 6 weeks, n (%) [95%CI] 1 (2.3) [0.0, 6.8] 2 (10) [0.0, 23.1] 3 (4.8) [0.0, 9.9] DCR, n (%) [95%CI] 9 (20.9) [8.8, 33.1] 13 (65) [44.1, 85.9] 22 (34.9) [23.1, 46.7] Progression, n (%) [95%CI] 30 (69.8) [56.0, 83.5] 6 (30) [9.9, 50.1] 36 (57.1) [44.9, 69.4] Not done/not evaluable, n (%) [95%CI] 4 (9.3) [0.6, 18.0] 1 (5.0) [0.0, 14.6] 5 (7.9) [1.3, 14.6] Pujol J, et al. Ann Oncol 2018;29(suppl 5):Abstr 1664O

Progression-free survival 1664O: A randomized non-comparative phase II study of anti–PD-L1 atezolizumab or chemotherapy as second-line therapy in patients with small cell lung cancer: results from the IFCT-1603 Trial – Pujol J, et al Key results (cont.) Conclusions In pre-treated patients with progressive SCLC, the efficacy of atezolizumab was inferior to chemotherapy ORR at 6 weeks with atezolizumab monotherapy was 2.3% and median PFS was 1.4 months Therefore, the planned phase 3 portion of the study was not activated PFS (ITT population) Median PFS, months (95%CI) 1.0 Chemotherapy (n=24): 4.3 (1.5, 5.9); 21 events, 3 censored Atezolizumab (n=49): 1.4 (1.2, 1.5); 46 events, 3 censored 0.8 Median follow-up: 13.7 months (95%CI 12.7, NR) 0.6 HR (adjusted) atezolizumab group = 2.26 (95%CI 1.30, 3.93); p=0.004 Progression-free survival 0.4 6-month PFS rate for atezolizumab group: 6.3% [0.0, 13.1] 0.2 1 2 3 4 5 6 7 8 9 Time, months Pujol J, et al. Ann Oncol 2018;29(suppl 5):Abstr 1664O

Advanced NSCLC Not radically treatable stage III and stage IV First line

LBA10: Primary results of ALESIA: A randomised, phase III, open-label study of alectinib vs crizotinib in Asian patients with treatment-naïve ALK+ advanced NSCLC – Zhou C, et al Study objective To investigate safety and efficacy of 1L alectinib vs. crizotinib in Asian patients with ALK+ NSCLC in the ALESIA study Key patient inclusion criteria Asian patients with stage IIIB/IV ALK+ NSCLC Treatment naïve ECOG PS 0–2 (n=187) Alectinib 600 mg bid (n=125) Stratification ECOG PS (0/1 vs. 2) Baseline brain metastases R 2:1 Crizotinib 250 mg bid (n=62) Primary endpoint PFS (investigator-assessed) Secondary endpoints PFS by IRC, time to CNS progression, ORR and DoR (investigator-assessed), OS, CNS ORR, safety, QoL, PK Zhou C, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA10

LBA10: Primary results of ALESIA: A randomised, phase III, open-label study of alectinib vs crizotinib in Asian patients with treatment-naïve ALK+ advanced NSCLC – Zhou C, et al Key result PFS (investigator-assessed) Time to CNS progression (IRC-assessed) Progression-free survival, % Time, months 100 80 60 40 20 3 6 9 12 15 18 21 Alectinib Crizotinib NE 11.1 months Cumulative incidence*, % Time, months 60 40 20 3 6 9 12 15 18 21 Alectinib Crizotinib 7.3% 35.5% (95%CI 23.5, 47.8) (95%CI 3.6, 12.8) Alectinib (n=125) Crizotinib (n=62) Patients with event, n (%) 26 (20.8) 37 (59.7) Median PFS, months (95%CI) NE (20.3, NE) 11.1 (9.1, 13.0) HR (95%CI); p-value (log-rank test) 0.22 (0.13, 0.38); p<0.0001 Cause-specific HR p-value (log-rank test) 0.14 (95%CI 0.06, 0.30) <0.0001 *Cumulative incidence of CNS progression without prior non-CNS progression or death Zhou C, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA10

LBA10: Primary results of ALESIA: A randomised, phase III, open-label study of alectinib vs crizotinib in Asian patients with treatment-naïve ALK+ advanced NSCLC – Zhou C, et al Key result (cont.) Conclusions Alectinib led to a significant improvement in PFS and showed benefits in regards to ORR, DoR and time to CNS progression compared with crizotinib The safety profile of alectinib in Asian patients was consistent with the previous findings DoR (investigator-assessed) Response, n (%) Alectinib (n=125) Crizotinib (n=62) ORR 114 (91.2) 48 (77.4) CR 5 (4.0) 3 (4.8) PR 109 (87.2) 45 (72.6) SD 7 (5.6) 8 (12.9) PD 2 (1.6) 4 (6.5) Missing or unevaluable 2 (3.2) DoR (unconfirmed) ,% Time, months 100 80 60 40 20 3 6 9 12 15 18 21 Alectinib Crizotinib NE 9.3 months Alectinib (n=114) Crizotinib (n=48) Median DoR, months (95%CI) NE (18.4, NE) 9.3 (7.4, NE) HR (95%CI), p-value (log-rank test) 0.22 (0.12, 0.40); p<0.0001 Zhou C, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA10

NGS was conducted using the Guardant360 assay or GuardantOMNI assay LBA50: Mechanisms of acquired resistance to first-line osimertinib: preliminary data from the phase III FLAURA study – Ramalingam SS, et al Study objective To investigate the mechanisms of acquired resistance to osimertinib and standard of care (gefitinib + erlotinib) in patients who progressed or discontinued treatment in the FLAURA study Methods Patients in the FLAURA study had EGFR+ (ex19del or L858R) locally advanced or metastatic NSCLC and were treated with osimertinib or gefitinib + erlotinib until progression Paired plasma samples were taken from patients at baseline and at progression/discontinuation for ctDNA genomic profiling NGS was conducted using the Guardant360 assay or GuardantOMNI assay Analysis set of valid paired NGS data were available for 272 patients Osimertinib: 113/279 (41%) Gefitinib + erlotinib: 159/277 (57%) Ramalingam SS, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA50

LBA50: Mechanisms of acquired resistance to first-line osimertinib: preliminary data from the phase III FLAURA study – Ramalingam SS, et al Key results The most common acquired resistance mechanisms in the osimertinib arm (n=91) were: MET amplification (15%) and EGFR C797S mutation (7%) There was no evidence of acquired T790M The most common acquired resistance mechanisms in the gefitinib + erlotinib arm (n=129) were: T790M (47%), MET amplification (4%) and HER2 amplification (2%) Conclusions There was no evidence of acquired resistance through T790M in the osimertinib-treated patients, while MET amplification and EGFR C797S mutations were the most common New mechanisms that may lead to aggressive disease biology were not identified Ramalingam SS, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA50

NGS was undertaken using the Guardant360 assay LBA51: Analysis of resistance mechanisms to osimertinib in patients with EGFR T790M advanced NSCLC from the AURA3 study – Papadimitrakopoulou VA, et al Study objective To investigate the mechanisms of acquired resistance to osimertinib in patients who progressed or discontinued treatment in the AURA3 study Methods Patients in the AURA3 study had T790M+ locally advanced or metastatic NSCLC and were treated with osimertinib or platinum-pemetrexed until progression Paired plasma samples were taken from patients at baseline and at progression/discontinuation for ctDNA genomic profiling NGS was undertaken using the Guardant360 assay Analysis set of valid paired NGS data were available for 113 patients Osimertinib: 83/279 (30%) Platinum-pemetrexed: 30/140 (21%) Papadimitrakopoulou VA, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA51

Loss of T790M was seen in 49% of patients LBA51: Analysis of resistance mechanisms to osimertinib in patients with EGFR T790M advanced NSCLC from the AURA3 study – Papadimitrakopoulou VA, et al Key results Loss of T790M was seen in 49% of patients Acquired EGFR mutations were observed in 21% of patients; the most common was C797S (14%) Other mutations included: MET amplification (19%); cell cycle gene alterations (12%); HER2 amplification (5%); PIK3CA amplification/mutation (5%); oncogenic fusion (4%), BRAF V600E (3%) All cases of C797X mutations were in the cis position when co-occurring with T790M Conclusion EGFR mutations and MET amplifications were the most common acquired resistance mechanisms to osimertinib among a diverse mixture, consistent with previous reports Acquired EGFR mutations with osimertinib (represents 21% of total patients) C797X: 15% (10 C797S, 1 C797G) L792H/F + C797S: 1% L792H: 1% G796S: 1% L718Q: 1% Ex20ins: 1% Papadimitrakopoulou VA, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA51

LBA52: Results of the GEOMETRY mono-1 phase II study for evaluation of the MET inhibitor capmatinib (INC280) in patients (pts) with METΔex14 mutated advanced non-small cell lung cancer (NSCLC) – Wolf J, et al Study objective To investigate the efficacy and safety of capmatinib, a selective MET inhibitor, in patients with METΔex14 mutated advanced NSCLC Cohort 4: Previously treated (1–2 prior lines) for advanced stage Capmatinib 400 mg bid (n=69) Key patient inclusion criteria Stage IIIB/IV NSCLC METΔex14 determined centrally EGFR wt and ALK negative ≥1 measureable lesion ECOG PS 0–1 Cohort 5b: Treatment-naïve for advanced stage Capmatinib 400 mg bid (n=28) Primary endpoint ORR by BIRC Secondary endpoints DoR by BIRC Wolf J, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA52

Cohort 5b (treatment naïve) LBA52: Results of the GEOMETRY mono-1 phase II study for evaluation of the MET inhibitor capmatinib (INC280) in patients (pts) with METΔex14 mutated advanced non-small cell lung cancer (NSCLC) – Wolf J, et al Key results ORR (BIRC) in Cohort 4 (pre-treated) was 39.1% (95%CI 27.6, 51.6) ORR (BIRC) in Cohort 5b (treatment naïve) was 72.0% (95%CI 50.6, 87.9) n=60†/69 * Cohort 4 (pre-treated) Best % change from baseline Cohort 5b (treatment naïve) * –25 –50 –75 –100 n=24†/25 PR SD PD NE Best % change from baseline *Patients still on treatment; †number of patients with measurable disease at baseline and ≥1 post-baseline assessment Wolf J, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA52

Cohort 5b (treatment naive) LBA52: Results of the GEOMETRY mono-1 phase II study for evaluation of the MET inhibitor capmatinib (INC280) in patients (pts) with METΔex14 mutated advanced non-small cell lung cancer (NSCLC) – Wolf J, et al Key results Conclusions In patients with METΔex14 advanced NSCLC, capmatinib shows promising results The differential benefit of 1L over later-line treatment highlights the need for prompt targeted treatment in this patient group AE, n (%) Cohort 4 (pre-treated) (n=69) Cohort 5b (treatment naive) (n=28) All patients (N=302) All grades Grade 3/4 Any 60 (87.0) 33 (47.8) 27 (96.4) 14 (50.0) 253 (83.8) 100 (33.1) Peripheral oedema 31 (44.9) 10 (14.5) 18 (64.3) 1 (3.6) 122 (40.4) 19 (6.3) Nausea 24 (34.8) 11 (39.3) 99 (32.8) 5 (1.7) Vomiting 13 (18.8) 4 (14.3) 58 (19.2) 6 (2.0) Blood creatinine increased 15 (21.7) 7 (25.0) Fatigue 9 (13.0) 4 (5.8) 2 (7.1) 40 (13.2) 10 (3.3) Decreased appetite 1 (1.4) 5 (17.9) 3 (1.0) Diarrhoea 8 (11.6) 3 (10.7) 35 (11.6) Wolf J, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA52

Induction treatment (4 or 6 cycles) Maintenance treatment LBA53: IMpower130: Progression-free survival (PFS) and safety analysis from a randomised phase 3 study of carboplatin + nab-paclitaxel (CnP) with or without atezolizumab (atezo) as first-line (1L) therapy in advanced non-squamous NSCLC – Cappuzzo F, et al Study objective To investigate the efficacy and safety of atezolizumab combined with carboplatin + nab-paclitaxel vs. carboplatin + nab-paclitaxel in chemotherapy-naïve patients with stage IV non-squamous NSCLC Induction treatment (4 or 6 cycles) Maintenance treatment Atezolizumab 1200 mg q3w + carboplatin AUC 6 q3w + nab-paclitaxel 100 mg/m2 q3w Loss of clinical benefit/toxicity Key patient inclusion criteria Stage IV non-squamous NSCLC Chemotherapy-naïve Pre-treated EGFR mutated or ALK translocation (TKI) (n=723; ITT-WT* n=679) Atezolizumab Stratification Sex Baseline liver metastases PD-L1 tumour expression R 2:1 Carboplatin AUC 6 q3w + nab-paclitaxel 100 mg/m2 q3w PD/ toxicity Best supportive care or pemetrexed q3w Co-primary endpoints Investigator-assessed PFS and OS (ITT-WT* population) Secondary endpoints OS and PFS (ITT population and by PD-L1 expression), ORR, safety *ITT-WT population, randomised patients excluding those with EGFR or ALK mutations Cappuzzo F, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA53

Investigator-assessed PFS (ITT-WT) LBA53: IMpower130: Progression-free survival (PFS) and safety analysis from a randomised phase 3 study of carboplatin + nab-paclitaxel (CnP) with or without atezolizumab (atezo) as first-line (1L) therapy in advanced non-squamous NSCLC – Cappuzzo F, et al Key results Investigator-assessed PFS (ITT-WT) OS (ITT-WT) PFS, % Months after randomisation 1.0 0.8 0.6 0.4 0.2 0.0 451 228 1 432 214 2 383 174 3 351 150 4 329 136 5 281 110 30 No. at risk Atezo + CnP Chemo 6 242 90 7 213 75 8 183 61 9 157 48 10 138 40 11 132 35 12 119 29 13 108 23 14 83 18 15 78 16 62 17 60 41 19 36 20 21 22 24 25 26 27 28 1.0 PFS rate, % 6-month 12-month Atezolizumab + CnP 56.1 29.1 CnP 42.5 14.1 0.8 OS rate, % 1-year 2-year Atezolizumab + CnP 63.1 39.6 CnP 55.5 30.0 0.6 Median follow-up: ~19 months HR 0.79 (95%CI 0.64, 0.98) p=0.033 HR 0.64 (95%CI 0.54, 0.77) p<0.0001 OS, % 0.4 0.2 Median: 5.5 mo (95%CI 4.4, 5.9) Median: 7.0 mo (95%CI 6.2, 7.3) 0.0 Median: 13.9 mo (95%CI 12.0, 18.7) Median: 18.6 mo (95%CI 16.0, 21.2) 451 228 1 435 218 2 422 206 3 400 190 4 384 176 5 365 167 6 351 161 7 333 154 8 315 147 9 305 136 10 294 132 11 284 124 12 268 119 13 253 109 14 217 96 15 194 90 16 167 75 17 147 65 18 129 58 19 103 49 20 88 39 21 75 31 2 22 59 24 32 1 23 17 33 24 40 13 34 25 29 9 26 19 8 27 12 3 28 10 1 6 30 4 Months after randomisation No. at risk Atezo + CnP Chemo Cappuzzo F, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA53

PD-L1-negative TC0 and IC0 LBA53: IMpower130: Progression-free survival (PFS) and safety analysis from a randomised phase 3 study of carboplatin + nab-paclitaxel (CnP) with or without atezolizumab (atezo) as first-line (1L) therapy in advanced non-squamous NSCLC – Cappuzzo F, et al Key results (cont.) Conclusions In the ITT-WT population atezolizumab + chemotherapy showed benefit in PFS and OS, which was maintained across all PD-L1 subgroups No new safety signals were observed with the combination of atezolizumab + chemotherapy PFS by baseline PD-L1 status (ITT-WT) PFS, % 100 80 60 40 20 88 42 2 76 32 4 64 23 No. at risk Atezo+CnP Chemo 6 43 18 8 34 11 10 28 9 12 22 14 17 1 16 24 3 26 PD-L1-high TC3 or IC3 Time, months 30 PD-L1-low TC1/2 or IC1/2 PFS, % 100 80 60 40 20 128 65 2 113 49 4 103 42 6 83 8 64 10 51 12 9 14 28 5 16 3 18 1 22 24 26 Time, months 30 PD-L1-negative TC0 and IC0 PFS, % 100 80 60 40 20 235 121 2 194 93 4 162 71 6 116 42 8 85 10 59 19 12 55 16 14 38 9 34 18 24 17 3 22 5 1 26 28 Time, months Atezo + CnP (n=88) CnP (n=42) Median PFS, mo (95%CI) 6.4 (5.49, 9.76) 4.6 (3.22, 7) HR (95%CI) 0.51 (0.34, 0.77) Atezo + CnP (n=128) CnP (n=65) Median PFS, mo (95%CI) 8.3 (7.16, 10.35) 6.0 (5.29, 6.93) HR (95%CI) 0.61 (0.43, 0.85) Atezo + CnP (n=235) CnP (n=121) Median PFS, mo (95%CI) 6.2 (5.52, 7.16) 4.7 (4.11, 5.72) HR (95%CI) 0.72 (0.56, 0.91) Cappuzzo F, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA53

LBA54: IMpower132: efficacy of atezolizumab (atezo) + carboplatin (carbo)/cisplatin (cis) + pemetrexed (pem) as 1L treatment in key subgroups with stage IV non-squamous non-small cell lung cancer (NSCLC) – Barlesi F, et al Study objective To investigate the PFS benefit of atezolizumab combined with carboplatin or cisplatin + pemetrexed in patients from IMpower132 in an exploratory analysis stratified by race, age, smoking history or liver metastasis at baseline Induction therapy (4 or 6 cycles) Maintenance therapy Atezolizumab + carboplatin or cisplatin + pemetrexed* PD/ loss of clinical benefit Atezolizumab + pemetrexed* Key patient inclusion criteria Stage IV non-squamous NSCLC Chemotherapy-naïve Without EGFR or ALK genetic alteration (n=578) Stratification Sex Smoking status ECOG PS Chemotherapy regimen R 1:1 Carboplatin or cisplatin + pemetrexed* PD/ loss of clinical benefit Pemetrexed* Exploratory endpoints Clinical and biomarker subgroup analysis *Atezolizumab 1200 mg IV q3w; carboplatin AUC 6 mg/mL/min IV q3w; cisplatin 75 mg/m2 IV q3w; pemetrexed 500 mg/m2 IV q3w Barlesi F, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA54

Former/current smokers LBA54: IMpower132: efficacy of atezolizumab (atezo) + carboplatin (carbo)/cisplatin (cis) + pemetrexed (pem) as 1L treatment in key subgroups with stage IV non-squamous non-small cell lung cancer (NSCLC) – Barlesi F, et al Key results Conclusion Atezolizumab added to carboplatin/cisplatin + pemetrexed improved PFS across several subgroups including patients from Asia, never smokers, those who were older and those without liver metastases at baseline, but currently has not shown improvement in OS PFS, % Time, months 1.0 0.8 0.6 0.4 0.2 0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20 19 21 22 Median PFS (95%CI), mos HR (95%CI) Non-Asian patients Atezolizumab + PP 6.9 (5.9, 8.1) PP 5.0 (4.2, 5.7) 0.65 (0.53, 0.81) Atezolizumab + PP (n=221) PP (n=221) Asian patients 10.2 (8.3, 15.3) 5.3 (4.3, 6.7) 0.42 (0.28, 0.63) Atezolizumab + PP (n=71) PP (n=65) PFS, % Time, months 1.0 0.8 0.6 0.4 0.2 0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20 19 21 22 Median PFS (95%CI), mos HR (95%CI) Former/current smokers Atezolizumab + PP 7.5 (6.3, 8.4) PP 5.1 (4.3, 5.6) 0.61 (0.50, 0.74) Atezolizumab + PP (n=225) PP (n=226) Never smokers 8.6 (6.5, 15.4) 5.5 (4.0, 8.3) 0.49 (0.28, 0.87) Atezolizumab + PP (n=37) PP (n=30) Barlesi F, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA54

LBA55: Primary efficacy results from B-F1RST, a prospective phase II trial evaluating blood-based tumour mutational burden (bTMB) as a predictive biomarker for atezolizumab (atezo) in 1L nonsmall cell lung cancer (NSCLC) – Kim ES, et al Study objective To investigate the use of bTMB as a predictive biomarker for atezolizumab monotherapy in 1L NSCLC Key patient inclusion criteria Stage IIIB/IVA locally advanced or metastatic NSCLC Immunotherapy naïve PD-L1 unselected ECOG PS 0–1 (n=152) Atezolizumab 1200 mg q3w (biomarker evaluable population n=119) PD/toxicity/ loss of benefit Co-primary endpoints Investigator-assessed ORR and PFS (using pre-specified bTMB cut-off of 16) Secondary endpoints Safety, investigator-assessed DoR and OS Kim ES, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA55

Overall response rate, % LBA55: Primary efficacy results from B-F1RST, a prospective phase II trial evaluating blood-based tumour mutational burden (bTMB) as a predictive biomarker for atezolizumab (atezo) in 1L nonsmall cell lung cancer (NSCLC) – Kim ES, et al Key results ORR per RECIST v1.1 bTMB subgroups Overall response rate, % ITT (N=152) BEP (n=119) High (n=49) Low (n=70) (n=28) (n=91) (n=19) (n=100) PR CR ≥10 cut-off ≥16 cut-off ≥20 cut-off p<0.0001 p=0.0002 p=0.0595 Kim ES, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA55

PFS in bTMB high (≥16) vs. low (<16) subgroups LBA55: Primary efficacy results from B-F1RST, a prospective phase II trial evaluating blood-based tumour mutational burden (bTMB) as a predictive biomarker for atezolizumab (atezo) in 1L nonsmall cell lung cancer (NSCLC) – Kim ES, et al Key results (cont.) Conclusion In patients with NSCLC treated with atezolizumab monotherapy a numerical improvement in outcomes was seen in those with a bTMB cut-off of ≥16 PFS in bTMB high (≥16) vs. low (<16) subgroups Progression-free survival, % Time, months 100 80 60 40 20 28 91 1 27 86 2 17 56 3 14 43 No. at risk High (≥16) Low (<16) High, ≥16 (n=28) Low, <16 (n=91) 6-month PFS 41.6% vs. 32.8% 9-month PFS 37.4% vs. 9.7% 4 39 5 13 6 11 21 7 8 9 10 12 bTMB high (n=28) bTMB low (n=91) Median PFS, months (90%CI) 4.6 (1.6, 11.0) 3.7 (2.6, 4.3) HR (90%CI) 0.66 (0.42, 1.02) p-value 0.12 Kim ES, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA55

LBA58: Intracranial efficacy of brigatinib (BRG) vs crizotinib (CRZ) in the phase 3 ALTA-1L trial – Popat S, et al Study objective To investigate the intracranial efficacy of brigatinib vs. crizotinib in ALK inhibitor- naïve patients with advanced ALK+ NSCLC in the ATLA-1L study Key patient inclusion criteria Stage IIIb/IV NSCLC ALK+ (based on local ALK testing) No prior ALK inhibitor ≤1 prior systemic therapy Brain metastases were allowed (n=275) Brigatinib 180 mg/day (90 mg/day for 7 day lead-in) (n=137) PD/toxicity/ discontinuation Stratification Brain metastases at baseline Prior chemotherapy R 1:1 Crizotinib 250 mg bid (n=138) PD*/toxicity/ discontinuation Primary endpoint PFS (BIRC-assessed) Secondary endpoints ORR, intracranial ORR, intracranial PFS, OS, safety *Crossover to brigatinib permitted at PD Popat S, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA58

LBA58: Intracranial efficacy of brigatinib (BRG) vs crizotinib (CRZ) in the phase 3 ALTA-1L trial – Popat S, et al Key results Conclusion In patients with advanced ALK+ NSCLC, brigatinib demonstrated superior intracranial efficacy vs. crizotinib, with significantly higher intracranial response and improved intracranial PFS in those with brain metastases Whole body BIRC-assessed PFS by brain metastases at baseline With brain metastases Without brain metastases PFS, % of patients Time, months 100 80 60 40 20 3 6 9 12 15 18 HR 0.20 (95%CI 0.09, 0.46) p<0.0001 by log-rank test Brigatinib (n=40) Crizotinib (n=41) PFS, % of patients Time, months 100 80 60 40 20 3 6 9 12 15 18 HR 0.72 (95%CI 0.44, 1.18) p=0.20 by log-rank test Brigatinib (n=97) Crizotinib (n=97) Treatment Pts with events, n (%) Median PFS, months (95%CI) 1-year PFS rate, % (95%CI) Brigatinib (n=40) 8 (20) NR 75 (56, 87) Crizotinib (n=41) 24 (59) 5.6 (3.8, 11.1) 25 (8, 46) Treatment Pts with events, n (%) Median PFS, months (95%CI) 1-year PFS rate, % (95%CI) Brigatinib (n=97) 28 (29) NR 63 (50, 74) Crizotinib (n=41) 39 (40) 11.1 (9.2, NR) 49 (36, 61) Popat S, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA58

Tepotinib 500 mg/day + gefitinib 250 mg/day 1377O: Phase 2 study of tepotinib + gefitinib (TEP+GEF) in MET-positive (MET+)/epidermal growth factor receptor (EGFR)-mutant (MT) non-small lung cancer (NSCLC) – Wu Y, et al Study objective To investigate the efficacy and safety of tepotinib + gefitinib vs. chemotherapy in Asian patients with advanced MET+/EGFR+T790M- NSCLC Tepotinib 500 mg/day + gefitinib 250 mg/day (n=31) Key patient inclusion criteria Locally-advanced/metastatic IV NSCLC EGFR+, T790M–, MET+ Asian Resistance to prior EGFR TKI No prior HGF/MET pathway-directed therapy (n=55) PD/ toxicity Stratification MET+ type (IHC2+, IHC3+ or MET amplification) Prior EGFR-TKI therapy R* Chemotherapy: Pemetrexed 500 mg/m2 q3w+ cisplatin 75 mg/m2 or carboplatin AUC 5 or 6 q3w (n=24) PD/ toxicity Primary endpoint PFS (investigator-assessed) Secondary endpoints ORR, safety *Initially 1:1 and changed to 2:1 at protocol amendment (30 Sept 2016) Wu Y, et al. Ann Oncol 2018;29(suppl 5):Abstr 1377O

Tepotinib + gefitinib (n=31) 1377O: Phase 2 study of tepotinib + gefitinib (TEP+GEF) in MET-positive (MET+)/epidermal growth factor receptor (EGFR)-mutant (MT) non-small lung cancer (NSCLC) – Wu Y, et al Key results In the MET IHC3+ subgroup (n=34) median PFS was 8.3 months with tepotinib + gefitinib and 4.4 months with chemotherapy (HR 0.35 [95%CI 0.17, 0.74]) In the MET amplification subgroup (n=19) median PFS was 21.2 months with tepotinib + gefitinib and 4.2 months with chemotherapy (HR 0.17 [95%CI 0.05, 0.57]) Investigator-assessed PFS (ITT population) Tepotinib + gefitinib (n=31) Chemotherapy (n=24) Events 23 19 Median PFS, months (90%CI) 4.9 (3.9, 6.9 4.4 (4.2, 6.8) Stratified HR (90%CI) 0.71 (0.36, 1.39) 1.0 0.8 0.6 Tepotinib + gefitinib Chemotherapy KM estimate for PFS 0.4 0.2 0.0 31 24 3 20 16 6 11 7 12 1 18 1 24 30 No. at risk Tep + gef Chemo Time, months Wu Y, et al. Ann Oncol 2018;29(suppl 5):Abstr 1377O

1377O: Phase 2 study of tepotinib + gefitinib (TEP+GEF) in MET-positive (MET+)/epidermal growth factor receptor (EGFR)-mutant (MT) non-small lung cancer (NSCLC) – Wu Y, et al Key results Tepotinib + gefitinib was generally well tolerated, with TEAEs leading to discontinuation in 9.7% of patients compared with 4.3% in the chemotherapy group Conclusion In patients with advanced NSCLC and MET amplifications, improvements in PFS were observed with tepotinib + gefitinib, indicating that MET may be considered as a biomarker for treatment with tepotinib Tepotinib + gefitinib Chemotherapy Odds ratio (90%CI) Responders* n/N ORR, % (90%CI) Responders* n/N ORR, % (90%CI) Overall (n=55) 14/31 45.2 (29.7, 61.3) 8/24 33.3 (17.8, 52.1) 1.99 (0.56, 6.87) MET IHC3+ (n=34) 13/19 68.4 (47.0, 85.3) 5/15 33.3 (14.2, 57.7) 4.33 (1.03, 18.33) MET amplification (n=19) 8/12 66.7 (39.1, 87.7) 3/7 42.9 (12.9, 77.5) 2.67 (0.37, 19.56) *No confirmation required Wu Y, et al. Ann Oncol 2018;29(suppl 5):Abstr 1377O

1379PD: Impact of the EML4-ALK variant on the efficacy of alectinib (ALC) in untreated ALK+ advanced NSCLC (aNSCLC) in the global phase III ALEX study – Dziadziuszko R, et al Study objective To assess efficacy of alectinib vs. crizotinib in the ALEX trial in patients with ALK+ advanced NSCLC stratified by EML4-ALK fusion variants Methods FOUNDATIONACT and FOUNDATIONONE NGS platforms were used to determine the ALK fusion variants in plasma and tissue BEP subgroups Alectinib 600 mg bid Key patient inclusion criteria ALK+ NSCLC No prior treatment for NSCLC ECOG PS 0–2 (n=303) PD/death/ withdrawal R 1:1 Crizotinib 250 mg bid PD/death/ withdrawal Dziadziuszko R, et al. Ann Oncol 2018;29(suppl 5):Abstr 1379PD

PFS (investigator-assessed) 1379PD: Impact of the EML4-ALK variant on the efficacy of alectinib (ALC) in untreated ALK+ advanced NSCLC (aNSCLC) in the global phase III ALEX study – Dziadziuszko R, et al Key results Conclusion In patients with ALK+ NSCLC alectinib demonstrated greater efficacy than crizotinib, regardless of EML4-ALK variant PFS (investigator-assessed) Plasma BEP subgroup Tissue BEP subgroup Crizotinib Alectinib Crizotinib Alectinib EML4-ALK V1: 7.4 EML4-ALK V2: 8.8 EML4-ALK V3a/b: 9.1 Median PFS, months Median PFS, months EML4-ALK V1: 12.9 EML4-ALK V2: 8.8 EML4-ALK V3a/b: 14.6 Median PFS, months Median PFS, months EML4-ALK V1: NE EML4-ALK V2: 14.9 EML4-ALK V3a/b: NE EML4-ALK V1: NE EML4-ALK V2: 11.5 EML4-ALK V3a/b: NE 1.0 1.0 1.0 1.0 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 Survival probability 0.4 0.4 0.4 0.4 0.2 0.2 0.2 0.2 0.0 0.0 0.0 0.0 3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24 27 30 3 6 9 12 15 18 21 24 Time, months 3 6 12 15 18 21 24 9 27 30 Time, months Time, months Time, months ALK V1 ALK V2 ALK V3a/b Dziadziuszko R, et al. Ann Oncol 2018;29(suppl 5):Abstr 1379PD

Molecular profiling of tumour tissue and blood was performed 1380PD: Efficacy of lorlatinib in patients (pts) with ROS1-positive advanced non-small cell lung cancer (NSCLC) and ROS1 kinase domain mutations – Solomon BJ, et al Study objective To investigate molecular profiling of patients with ROS1+ advanced NSCLC in the B7461001 study of lorlatinib Methods Patients with locally advanced or metastatic NSCLC with ROS1 rearrangements were enrolled in the phase 1/2 study Patients were treatment-naïve in the advanced setting of PD after ≥1 prior ROS1 inhibitor (phase 1) or any number of prior therapies (phase 2) Lorlatinib was orally administered in continuous 21-day cycles with escalating doses (10 mg/day to 100 mg bid) in phase 1 and 100 mg/day in phase 2 Molecular profiling of tumour tissue and blood was performed All patients had tissue samples collected before enrolment and tissue collection was encouraged on PD; tumour tissue was analysed with a ROS1 kinase domain mutation- focused NGS panel Blood samples were collected at screening, at the beginning of cycle 3, and at the end- of-treatment visit for circulating-free DNA (cfDNA) analysis using an NGS panel or digital PCR Solomon BJ, et al. Ann Oncol 2018;29(suppl 5):Abstr 1380PD

ROS1 kinase domain mutations occurred in 10–14% of samples 1380PD: Efficacy of lorlatinib in patients (pts) with ROS1-positive advanced non-small cell lung cancer (NSCLC) and ROS1 kinase domain mutations – Solomon BJ, et al Key results Conclusions ROS1 kinase domain mutations occurred in 10–14% of samples In patients with ROS1 kinase domain resistance mutations, lorlatinib exhibited some anti-tumour activity Percentage change in tumour size from baseline in patients with ≥1 ROS1 kinase domain mutation ORR by presence of ROS1 mutations Best percentage change from baseline G2032R Partial response Stable disease Progressive disease Best overall response ROS1 kinase domain mutation in: cfDNA Tumour tissue Both cfDNA and tumour tissue G2032R/ L2026M S1986F K1991E ROS1 TKI-naïve (n=17) Prior crizotinib (n=36)* Prior non-crizotinib TKI or ≥2 TKIs (n=5) No mutation (n=17) ≥1 mutation (n=0) No mutation (n=27) ≥1 mutation (n=8) No mutation (n=4) ≥1 mutation (n=1) Best overall response, n (%) CR 2 (11.8) – 1 (3.7) PR 9 (52.9) 8 (29.6) 2 ( 25.0) SD 5 (29.4) 5 (62.5) 3 (75.0) 1 (100) PD 1 (5.9) 3 (11.1) 1 (12.5) 1 (25.0) Indeterminate 7 (25.9) Responders, n (%) 11 (64.7) 9 (33.3) 2 (25.0) *One patient had a non-analysable or uninformative sample Solomon BJ, et al. Ann Oncol 2018;29(suppl 5):Abstr 1380PD

Pemetrexed 500 mg/m2 q3w + gefitinib 250 mg/day (n=126) 1381PD Gefitinib with or without pemetrexed in nonsquamous non-small cell lung cancer with EGFR mutation: Final overall survival results from a randomized phase II study – Chih-Hsin Yang J, et al Study objective To investigate the efficacy and safety of pemetrexed + gefitinib vs. gefitinib in non-squamous NSCLC with EGFR mutations Key patient inclusion criteria Stage IV or recurrent non-squamous NSCLC Activating EGFR mutations East Asian ≥18 years (≥20 years in Japan and Taiwan) No prior systemic therapy for stage IV or recurrent NSCLC ECOG PS 0–1 (n=191) Pemetrexed 500 mg/m2 q3w + gefitinib 250 mg/day (n=126) R 2:1 Gefitinib 250 mg/day (n=65) Primary endpoint PFS Secondary endpoints OS, time to progressive disease, tumour response rates, DoR, QoL, biomarker analysis, safety Chih-Hsin Yang J et al. Ann Oncol 2018;29(suppl 5):abstr 1381PD

Median PFS, months (95%CI) 1381PD Gefitinib with or without pemetrexed in nonsquamous non-small cell lung cancer with EGFR mutation: Final overall survival results from a randomized phase II study – Chih-Hsin Yang J, et al Key results 36 (28.6%) patients had TEAEs in the pemetrexed + gefitinib group and 7 (10.8%) in gefitinib group Conclusion In patients with EGFR+ advanced NSCLC pemetrexed + gefitinib significantly prolonged PFS, with a numerically longer, but not statistically significant, improvement in OS OS Updated PFS Time, months Probability 1.0 0.8 0.6 0.4 0.2 0.0 126 65 6 118 64 12 106 53 18 92 45 24 79 36 30 66 27 58 26 42 51 19 48 43 15 54 8 60 3 2 1 72 No.at risk P+G G Median OS, months (95%CI) 43.4 months (33.4, 50.8) 36.8 months (26.7, 42.6) Pemetrexed + gefitinib Gefitinib Adjusted HR 0.77 (95%CI 0.5, 1.2) 1-sided p=0.105 Time, months Probability 1.0 0.8 0.6 0.4 0.2 0.0 126 65 6 98 51 12 70 29 18 50 24 9 30 19 36 14 4 42 2 48 7 54 5 1 60 No.at risk P+G G Median PFS, months (95%CI) 16.2 months (12.6, 18.7) 11.1 months (9.7, 13.8) Pemetrexed + gefitinib Gefitinib Adjusted HR 0.67 (95%CI 0.5, 0.9) 1-sided p=0.009 Chih-Hsin Yang J et al. Ann Oncol 2018;29(suppl 5):abstr 1381PD

1382PD: Phase III study of gefitinib (G) versus gefitinib+carboplatin+pemetrexed (GCP) as 1st-line treatment for patients (pts) with advanced non-small cell lung cancer (NSCLC) with EGFR mutations (NEJ009) – Seike M, et al Study objective To investigate the efficacy and safety of gefitinib + carboplatin + pemetrexed vs. gefitinib in patients with non-squamous NSCLC Induction phase Maintenance phase Gefitinib 250 mg/day + carboplatin AUC6 + pemetrexed 500 mg/m2 Gefitinib + pemetrexed q3w Key patient inclusion criteria Treatment naïve patients with non-squamous IIIB/IV NSCLC EGFR+ PS 0–1 (n=345) Stratification Sex EGFR mutation Smoking history R 1:1 Gefitinib 250 mg/day Platinum-based regimen Primary endpoints PFS, PFS2, OS Secondary endpoints ORR, safety, QoL Seike M, et al. Ann Oncol 2018;29(suppl 5):Abstr 1382PD

Progression-free survival, % 1382PD: Phase III study of gefitinib (G) versus gefitinib+carboplatin+pemetrexed (GCP) as 1st-line treatment for patients (pts) with advanced non-small cell lung cancer (NSCLC) with EGFR mutations (NEJ009) – Seike M, et al Results Conclusions PFS and OS were significantly improved in patients treated with combination therapy of gefitinib + carboplatin + pemetrexed There was no difference in PFS2 PFS Updated OS Overall survival, % Time, months 100 80 60 40 20 172 170 12 153 162 24 115 131 36 86 106 48 62 77 26 29 No. at risk Gefitinib Gef + carb + pem Progression-free survival, % Time, months 100 80 60 40 20 172 169 12 78 123 24 29 68 36 11 37 48 2 10 No. at risk Gefitinib Gef + carb + pem Gefitinib Gefitinib + carb + pem Median PFS, months (95%CI) 11.2 (9.0, 13.4) 20.9 (18.0, 24.0) HR (95%CI); p-value 0.490 (0.388, 0.620); <0.001 Gefitinib Gefitinib + carb + pem mOS, months (95% CI) 38.8 (31.1, 47.3) 50.9 (41.8, 62.5) HR (95%CI); p-value 0.72 (0.58, 0.95); 0.02 Seike M, et al. Ann Oncol 2018;29(suppl 5):Abstr 1382PD

1385PD: A randomised phase III trial evaluating the addition of denosumab to standard first-line treatment in advanced NSCLC – the ETOP and EORTC SPLENDOUR trial – Peters S, et al Study objective To investigate the efficacy of denosumab as an add-on to standard 1L doublet chemotherapy + BSC in patients with NSCLC in the SPLENDOUR study Denosumab 120 mg every 3–4 weeks + chemotherapy (4–6 cycles) (n=255) Key patient inclusion criteria Stage IV NSCLC (n=514) Stratification Bone metastases PS Histology Region R 1:1 Chemotherapy (4–6 cycles) + BSC* (n=259) Primary endpoint OS Secondary endpoints PFS, OS by bone metastases, safety *Zoledronic acid in case of skeletal involvement Peters S, et al. Ann Oncol 2018;29(suppl 5):Abstr 1385PD

1385PD: A randomised phase III trial evaluating the addition of denosumab to standard first-line treatment in advanced NSCLC – the ETOP and EORTC SPLENDOUR trial – Peters S, et al Key results OS PFS OS, % 100 80 60 40 20 2 3 4 1 29 23 83 82 259 255 Years PFS, % 100 80 60 40 20 28 36 12 6 5 4 2 29 19 259 254 Months 24 32 16 1 13 10 140 138 8 55 50 Denosumab BSC Denosumab BSC No. at risk Denosumab BSC No. at risk Denosumab BSC Non-parametric Cox model Treatment Events/ patients Median, years (95%CI) 1 year, % (95%CI) HR (95%CI) p-value (Score test) Denosumab 177/258 0.68 (0.62, 0.87) 40.2 (33.8, 46.5) 0.96 (0.78, 1.18) BSC 178/255 0.73 (0.63, 0.92) 1.00 0.689 Non-parametric Cox model Treatment Events/ patient Median, years (95%CI) 1 year, % (95%CI) HR (95%CI) p-value (Score test) Denosumab 228/259 0.39 (0.35, 0.44) 13.1 (9.1, 17.9) 0.97 (0.81, 1.17) BSC 227/254 0.39 (0.34, 0.44) 9.3 (6.0, 13.6) 1.00 0.777 Peters S, et al. Ann Oncol 2018;29(suppl 5):Abstr 1385PD

Denosumab added to 1L platinum-based chemotherapy did not improve OS 1385PD: A randomised phase III trial evaluating the addition of denosumab to standard first-line treatment in advanced NSCLC – the ETOP and EORTC SPLENDOUR trial – Peters S, et al Key results (cont.) Conclusions Denosumab added to 1L platinum-based chemotherapy did not improve OS There were no safety concerns with denosumab OS Bone metastases OS No bone metastases OS, % 100 80 60 40 20 28 12 15 14 7 36 41 N 138 137 Months 4 24 32 16 17 13 5 3 27 25 87 95 8 50 53 2 OS, % 100 80 60 40 20 28 36 12 19 14 7 6 47 41 N 121 118 Months 4 24 32 16 10 2 35 95 87 8 65 66 Denosumab BSC Denosumab BSC No. at risk Denosumab BSC No. at risk Denosumab BSC Non-parametric Cox model Treatment Events/ patients Median, years (95%CI) 1 year, % (95%CI) HR (95%CI) p value (Score test) Denosumab 98/138 0.62 (0.46, 0.73) 34.4 (26.1, 43.0) 1.03 (0.78, 1.37) BSC 98/137 0.61 (0.51, 0.82) 35.8 (27.5, 44.2) 1.00 0.816 Non-parametric Cox model Treatment Events/ patients Median (years) (95%CI) 1 year, % (95%CI) Hazard Ratio (95%CI) P-Value (Score test) Denosumab 79/121 0.89 (0.67, 1.35) 46.9 (37.2, 55.9) 0.89 (0.65, 1.22) BSC 80/118 0.92 (0.71, 1.05) 45.4 (35.7, 54.6) 1.00 0.483 Peters S, et al. Ann Oncol 2018;29(suppl 5):Abstr 1385PD

Advanced NSCLC Not radically treatable stage III and stage IV Later lines

LBA63: Long-term survival in patients (pts) with advanced NSCLC in the KEYNOTE-010 study overall and in pts who completed 2 years of pembrolizumab (pembro) – Herbst RS, et al Study objective Long-term follow-up of KEYNOTE-010 investigating pembrolizumab vs. docetaxel in patients with advanced NSCLC and PD-L1 TPS ≥1% who had progressed after platinum-containing chemotherapy Pembrolizumab 2 mg/kg q3w for 24 months Key patient inclusion criteria Advanced NSCLC PD-L1 TPS ≥1% Progression after ≥1 line of chemotherapy No active brain metastases Pembrolizumab 10 mg/kg q3w for 24 months R 1:1:1 Docetaxel 75 mg/m2 q3w Stratification ECOG PS (0 vs. 1) Region (East Asia vs. non-East Asia) PD-L1 status (TPS ≥50% vs. 1–49%) Endpoints in TPS ≥50% and ≥1% populations OS, PFS ORR, DoR, safety Herbst RS, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA63

LBA63: Long-term survival in patients (pts) with advanced NSCLC in the KEYNOTE-010 study overall and in pts who completed 2 years of pembrolizumab (pembro) – Herbst RS, et al Key results OS PD-L1 TPS ≥50% population PD-L1 TPS ≥1% population N Events, n (%) Median OS, mo (95%CI) HR (95%CI); p-value Pembrolizumab 290 199 (69) 16.9 (12.3, 21.4) 0.53 (0.42, 0.66); <0.00001 Docetaxel 152 127 (84) 8.2 (6.4, 9.8) N Events, n (%) Median OS, mo (95%CI) HR (95%CI); p-value Pembrolizumab 690 548 (79) 11.8 (10.4, 13.1) 0.69 (0.60, 0.80); <0.00001 Docetaxel 343 295 (86) 8.4 (7.6, 9.5) Overall survival, % 100 80 60 40 20 290 152 5 229 97 10 178 58 15 149 39 131 29 25 115 23 35 94 18 50 45 26 8 1 55 30 101 21 Time, months 35% 13% Overall survival, % 100 80 60 40 20 690 343 5 523 226 10 374 135 15 295 90 248 57 25 211 44 35 151 86 45 13 50 2 55 30 171 Time, months 23% 11% No. at risk Pembro Docetaxel No. at risk Pembro Docetaxel Herbst RS, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA63

LBA63: Long-term survival in patients (pts) with advanced NSCLC in the KEYNOTE-010 study overall and in pts who completed 2 years of pembrolizumab (pembro) – Herbst RS, et al Key results (cont.) Overall, median treatment duration was 3.5 months (range 0.03–31.7) in the pembrolizumab group (n=682) and 2.0 months (range 0.03–26.4) in the docetaxel group (n=309) 79 patients completed 35 cycles or 2 years of pembrolizumab with a median follow-up of 43.4 months (range 35.7–49.8) 75/79 (95%) patients had a CR or PR 48/75 (64%) patients had ongoing response, median DoR NR (range 4–46+ months) Median OS was NR (95%CI NR, NR) The Kaplan-Meier estimate of 36-month OS rate was 98.7% (95%CI 91.2, 99.8) 25/79 (32%) patients had PD after stopping 35 cycles or 2 years of pembrolizumab In 14 patients who received a second course of pembrolizumab after 35 cycles or 2 years of treatment and subsequent PD, 6 (43%) had PR and 5 (36%) had SD Conclusions In patients with PD-L1-expressing advanced NSCLC, pembrolizumab continued to prolong OS compared with docetaxel In patients who completed 2 years of pembrolizumab AEs were manageable and responses were durable Herbst RS, et al. Ann Oncol 2018;29(suppl 5):Abstr LBA63

1378O: ARCTIC: durvalumab + tremelimumab and durvalumab monotherapy vs SoC in ≥3L advanced NSCLC treatment – Kowalski DM, et al Study objective To investigate the efficacy and safety of durvalumab vs. SoC and the combination of durvalumab + tremelimumab vs. SoC in subgroups based on PD-L1 expression (≥25% or <25%) in the ARCTIC study Durvalumab 10 mg/kg q2w for up to 12 mo (n=62) Study A PD-L1 TC ≥25% (n=126) R 1:1 Key patient inclusion criteria Stage IIIB/IV NSCLC ≥2 prior treatments Immunotherapy-naïve EGFR/ALK wild-type WHO PS 0/1 SoC* (n=64) Durvalumab 20 mg/kg + tremelimumab 1 mg/kg q4w for 12 weeks then durvalumab 10 mg/kg for 34 weeks (n=174) Study B PD-L1 TC <25% (n=469) SoC* (n=118) Stratification Planned SoC, histology R 3:2:2:1 Durvalumab 10 mg/kg q2w for up to 12 mo (n=117) Tremelimumab 10 mg/kg q4w for 24 weeks then q12w for 24 weeks (n=60) Primary endpoint OS, PFS (investigator assessed Secondary endpoints 1-year OS and PFS rates, ORR, safety *Erlotinib, gemcitabine, or vinorelbine Kowalski DM, et al. Ann Oncol 2018;29(suppl 5):Abstr 1378O

1378O: ARCTIC: durvalumab + tremelimumab and durvalumab monotherapy vs SoC in ≥3L advanced NSCLC treatment – Kowalski DM, et al Key results Study A Probability of OS Time from randomisation, months 1.0 0.8 0.6 0.4 0.2 0.0 62 64 3 53 45 6 35 9 37 27 No. at risk Durvalumab SoC 12 30 19 15 25 18 21 10 24 16 8 4 33 OS (PD-L1 TC ≥25%) 31.3% 49.3% Durvalumab(n=62) SoC (n=64) Events, n (%) 48 (77.4) 55 (85.9) Median OS, months (95%CI) 11.7 (8.2, 17.4) 6.8 (4.9, 10.2) HR (95%CI) 0.63 (0.42, 0.93) Durvalumab(n=62) SoC (n=64) Events, n (%) 58 (93.5) 58 (90.6) Median PFS, months (95%CI) 3.8 (1.9, 5.6) 2.2 (1.9, 3.7) HR (95%CI) 0.71 (0.49, 1.04) Probability of PFS Time from randomisation, months 1.0 0.8 0.6 0.4 0.2 0.0 62 64 3 34 25 6 22 14 9 18 10 No. at risk Durvalumab SoC 12 5 15 2 7 21 24 1 27 30 33 PFS (PD-L1 TC ≥25%) 9.9% 19.4% Kowalski DM, et al. Ann Oncol 2018;29(suppl 5):Abstr 1378O

1378O: ARCTIC: durvalumab + tremelimumab and durvalumab monotherapy vs SoC in ≥3L advanced NSCLC treatment – Kowalski DM, et al Key results (cont.) Conclusions Clinically meaningful improvement in OS was seen with durvalumab vs. SoC in the PD-L1 TC ≥25% subgroup Combination therapy led to non-significant improvement in OS vs. SoC in PD-L1 TC <25% subgroup Study B Probability of OS Time from randomisation, months 1.0 0.8 0.6 0.4 0.2 0.0 174 118 3 137 97 6 111 70 9 93 55 No. at risk D + T SoC Durvalumab + tremelimumab 12 82 44 15 67 35 18 56 27 21 33 14 24 7 1 30 OS (PD-L1 TC <25%) 38.8% 49.5% PFS (PD-L1 TC <25%) Probability of OS Time from randomisation, months 0.8 0.6 0.4 0.2 0.0 174 118 3 90 51 6 49 25 9 41 13 No. at risk D + T SoC 12 31 15 23 18 17 21 11 1 24 27 30 8.0% 20.6% 1.0 Durvalumab + tremelimumab SoC D + T (n=174) SoC (n=118) Events, n (%) 118 (67.8) 90 (76.3) mOS, months (95%CI) 11.5 (8.7, 14.1) 8.7 (6.5, 11.7) HR (95%CI) 0.80 (0.61, 1.05), p=0.109 D + T (n=174) SoC (n=118) Events, n (%) 146 (83.9) 92 (78.0) mPFS, months (95%CI) 3.5 (2.3, 4.6) 3.5 (1.9, 3.9) HR (95%CI) 0.77 (0.59, 1.01), p=0.056 Kowalski DM, et al. Ann Oncol 2018;29(suppl 5):Abstr 1378O

SCLC, mesothelioma and thymic epithelial tumours Other malignancies SCLC, mesothelioma and thymic epithelial tumours

1665PD: Preliminary efficacy of durvalumab plus tremelimumab in platinum-refractory/resistant EDSCLC from arm A of the phase II BALTIC study – Bondarenko I, et al Study objective To determine the efficacy and safety of durvalumab + tremelimumab in patients from the BALTIC study with extensive-disease (ED) SCLC who have platinum- refractory or -resistant disease Key patient inclusion criteria ED-SCLC Refractory or resistant to 1L chemotherapy Life expectancy ≥8 weeks No prior exposure to immunotherapy WHO/ECOG PS 0–1 Cohort A: Durvalumab 1500 mg + tremelimumab 75 mg q4w for up to 4 months, then durvalumab 1500 mg q4w (n=10) Interim analysis* Primary endpoint ORR Secondary endpoints DCR at 12 weeks, DoR, TTR, PFS, OS, safety *Interim ORR analysis followed by expansion stage with n=20 Bondarenko I, et al. Ann Oncol 2018;29(suppl 5):Abstr 1665PD

Maximum percentage change in target lesion size 1665PD: Preliminary efficacy of durvalumab plus tremelimumab in platinum-refractory/resistant EDSCLC from arm A of the phase II BALTIC study – Bondarenko I, et al Key results Confirmed ORR was 9.5% (95%CI 1.2, 30.4) with PR in 2 patients Median PFS was 1.9 months (95%CI 1.8, 4.3) Median OS was 6.0 months (95%CI 1.9, 12.0) Conclusion In patients with platinum-refractory ED-SCLC, durvalumab + tremelimumab showed promising activity, with a safety profile consistent with previous reports Maximum percentage change in target lesion size Best change from baseline in target lesion size, % 100 80 60 40 20 –20 –40 –60 –80 –100 Non-response Response Bondarenko I, et al. Ann Oncol 2018;29(suppl 5):Abstr 1665PD

NCF was tested prior to PCI and at weeks 11, 29 and 53 1667PD: Impact of early prophylactic cranial irradiation with hippocampal avoidance on neurocognitive function in patients with limited disease small-cell lung cancer. A multicenter phase II trial (SAKK 15/12) – Vees H, et al Study objective To investigate the neurocognitive function (NCF) in patients with limited-disease SCLC treated with prophylactic cranial irradiation (PCI) concomitant with chemotherapy and thoracic radiotherapy (tRT) in the SAKK 15/12 study Methods Patients received hippocampal avoidance PCI at the start of the 2nd cycle of chemotherapy (cisplatin or carboplatin and etoposide) and tRT (total of 6 cycles of chemotherapy) NCF was tested prior to PCI and at weeks 11, 29 and 53 Memory was assessed using Hopkins Verbal Learning Test Revised (HVLT-R); language and verbal fluency was assessed using Controlled Oral Word Association Test (COWAT); visual search, scanning, speed of processing and executive function was assessed using Trail Making Tests A and B (TMT A&B) NCF decline was defined as a decrease of 1 SE of measurement in any test A rate of ≤30% in patients with no NCF decline was assumed unpromising and a rate of ≥50% in patients with no NCF decline was assumed promising Vees H, et al. Ann Oncol 2018;29(suppl 5):Abstr 1667PD

Deterioration of NCF by cognitive test 1667PD: Impact of early prophylactic cranial irradiation with hippocampal avoidance on neurocognitive function in patients with limited disease small-cell lung cancer. A multicenter phase II trial (SAKK 15/12) – Vees H, et al Key result No brain metastases were observed during 6 months OS was 87% (95%CI 72, 94) Overall 7 patients died: 4 due to disease progression, 1 due to respiratory failure, 1 due to haemorrhage and 1 for unknown reason The most common grade ≥3 AEs were: anaemia (21%), febrile neutropenia (19%) and fatigue (14%) Conclusion The proportion of patients with no decline in NCF at 6 months is similar to that in those receiving sequential PCI Deterioration of NCF by cognitive test Deterioration of NCF at 6 months Deterioration of NCF n (%) 90%CI No 13 (34.2) 21.6, 48.8 Yes 25 (65.8) Number of patients Vees H, et al. Ann Oncol 2018;29(suppl 5):Abstr 1667PD