SPOTLIGHT Trial PDF: Zolbetuximab & mFOLFOX6 for Gastric Cancer (2023) AQA

Articles Zolbetuximab plus mFOLFOX6 in patients with CLDN18.2-positive, HER2-negative, untreated, locally advanced unresectable or metastatic gastric or gastro- oesophageal junction adenocarcinoma (SPOTLIGHT): a multicentre, randomised, double-blind, phase 3 trial Kohei Shitara, Florian Lordick, Yu...

Articles Zolbetuximab plus mFOLFOX6 in patients with CLDN18.2-positive, HER2-negative, untreated, locally advanced unresectable or metastatic gastric or gastro- oesophageal junction adenocarcinoma (SPOTLIGHT): a multicentre, randomised, double-blind, phase 3 trial Kohei Shitara, Florian Lordick, Yung-Jue Bang, Peter Enzinger, David Ilson, Manish A Shah, Eric Van Cutsem, Rui-Hua Xu, Giuseppe Aprile, Jianming Xu, Joseph Chao, Roberto Pazo-Cid, Yoon-Koo Kang, Jianning Yang, Diarmuid Moran, Pranob Bhattacharya, Ahsan Arozullah, Jung Wook Park, Mok Oh, Jaffer A Ajani Summary Background Zolbetuximab, a monoclonal antibody targeting claudin-18 isoform 2 (CLDN18.2), has shown efficacy in Lancet 2023; 401: 1655–68 patients with CLDN18.2-positive, human epidermal growth factor receptor 2 (HER2)-negative, locally advanced Published Online unresectable or metastatic gastric or gastro-oesophageal junction adenocarcinoma. We report the results of the April 14, 2023 https://doi.org/10.1016/ SPOTLIGHT trial, which investigated the efficacy and safety of first-line zolbetuximab plus mFOLFOX6 (modified S0140-6736(23)00620-7 folinic acid [or levofolinate], fluorouracil, and oxaliplatin regimen) versus placebo plus mFOLFOX6 in patients with This online publication has been CLDN18.2-positive, HER2-negative, locally advanced unresectable or metastatic gastric or gastro-oesophageal junction corrected. The corrected version adenocarcinoma. first appeared at thelancet.com on December 12, 2023 Methods SPOTLIGHT is a global, randomised, placebo-controlled, double-blind, phase 3 trial that enrolled patients See Comment page 1630 from 215 centres in 20 countries. Eligible patients were aged 18 years or older with CLDN18.2-positive (defined as Department of Gastrointestinal ≥75% of tumour cells showing moderate-to-strong membranous CLDN18 staining), HER2-negative (based on local or Oncology, National Cancer Center Hospital East, central evaluation), previously untreated, locally advanced unresectable or metastatic gastric or gastro-oesophageal Kashiwa City, Japan junction adenocarcinoma, with radiologically evaluable disease (measurable or non-measurable) according to (K Shitara MD); Department of Response Evaluation Criteria in Solid Tumors version 1.1; an Eastern Cooperative Oncology Group performance Medicine and University Cancer status score of 0 or 1; and adequate organ function. Patients were randomly assigned (1:1) via interactive response Center Leipzig, University of Leipzig Medical Center, technology and stratified according to region, number of organs with metastases, and previous gastrectomy. Patients Leipzig, Germany received zolbetuximab (800 mg/m² loading dose followed by 600 mg/m² every 3 weeks) plus mFOLFOX6 (Prof F Lordick MD); (every 2 weeks) or placebo plus mFOLFOX6. The primary endpoint was progression-free survival assessed by Department of Internal independent review committee in all randomly assigned patients. Safety was assessed in all treated patients. The Medicine, Seoul National University College of Medicine, study is registered with ClinicalTrials.gov, NCT03504397, and is closed to new participants. Seoul, South Korea (Prof Y-J Bang MD); Center for Findings Between June 21, 2018, and April 1, 2022, 565 patients were randomly assigned to receive either zolbetuximab Esophageal and Gastric Cancer, Dana-Farber Cancer Institute, plus mFOLFOX6 (283 patients; the zolbetuximab group) or placebo plus mFOLFOX6 (282 patients; the placebo Boston, MA, USA group). At least one dose of treatment was administered to 279 (99%) of 283 patients in the zolbetuximab group and (P Enzinger MD); Memorial 278 (99%) of 282 patients in the placebo group. In the zolbetuximab group, 176 (62%) patients were male and Sloan Kettering Cancer Center, 107 (38%) were female. In the placebo group, 175 (62%) patients were male and 107 (38%) were female. The median New York City, NY, USA (Prof D Ilson MD); follow-up duration for progression-free survival was 12·94 months in the zolbetuximab group versus 12·65 months Weill Cornell Medical College, in the placebo group. Zolbetuximab treatment showed a significant reduction in the risk of disease progression or New York City, NY, USA death compared with placebo (hazard ratio [HR] 0·75, 95% CI 0·60–0·94; p=0·0066). The median progression-free (Prof M A Shah MD); survival was 10·61 months (95% CI 8·90–12·48) in the zolbetuximab group versus 8·67 months (8·21–10·28) in the Department of Digestive Oncology, University Hospitals placebo group. Zolbetuximab treatment also showed a significant reduction in the risk of death versus placebo Gasthuisberg, Leuven, and (HR 0·75, 95% CI 0·60–0·94; p=0·0053). Treatment-emergent grade 3 or worse adverse events occurred in 242 (87%) KULeuven, Leuven, Belgium of 279 patients in the zolbetuximab group versus 216 (78%) of 278 patients in the placebo group. The most common (Prof E Van Cutsem MD); all-grade adverse events with zolbetuximab plus chemotherapy were nausea, vomiting, and decreased appetite. Sun Yat-Sen University Cancer Center, Guangzhou, China Treatment-related deaths occurred in five (2%) patients in the zolbetuximab group versus four (1%) patients in the (Prof R-H Xu MD); Department placebo group. No new safety signals were identified. of Oncology, Azienda ULSS 8 Berica, Veneto, Italy Interpretation Targeting CLDN18.2 with zolbetuximab significantly prolonged progression-free survival and overall (G Aprile MD); Department of Gastrointestinal Oncology, survival when combined with mFOLFOX6 versus placebo plus mFOLFOX6 in patients with CLDN18.2-positive, The Fifth Medical Center of the HER2-negative, locally advanced unresectable or metastatic gastric or gastro-oesophageal junction adenocarcinoma. PLA General Hospital, Zolbetuximab plus mFOLFOX6 might represent a new first-line treatment in these patients. Beijing, China (Prof J Xu MD); City of Hope Comprehensive www.thelancet.com Vol 401 May 20, 2023 1655  Articles Cancer Center, Duarte, CA, USA Funding Astellas Pharma, Inc. (J Chao MD); Hospital Universitario Miguel Servet, Zaragoza, Aragón, Spain Copyright © 2023 Elsevier Ltd. All rights reserved. (R Pazo-Cid MD); Department of Oncology, Asan Medical Introduction improve survival, although there are few validated Center, University of Ulsan, Gastric and gastro-oesophageal junction adenocar­cinomas molecular targets in this disease.9 In the approximately 15% Seoul, South Korea (Prof Y-K Kang MD); Astellas are among the malignancies with the highest unmet of patients with human epidermal growth factor receptor 2 Pharma Global Development, medical needs, with the highest incidence rates in Asia, (HER2)-positive disease, trastuzumab is approved in Inc, Northbrook, IL, USA Latin America, and eastern Europe.1–3 The standard combination with chemotherapy.2,4–8 Checkpoint inhibition (J Yang PhD, D Moran PhD, first-line chemotherapy for patients with locally with nivolumab is approved as a first-line treatment in P Bhattacharya DrPH, A Arozullah MD, J W Park PhD, advanced unresectable or metastatic gastric or gastro- combination with chemotherapy in some countries; M Oh PharmD); The University oesophageal junction adenocarcinoma has been platinum– however, the efficacy of this treatment is mainly in patients of Texas, MD Anderson Cancer fluoropyrimidine chemotherapy, including mFOLFOX6 with a programmed death-ligand 1 (PD-L1) combined Center, Houston, TX, USA (modified folinic acid, fluorouracil, and oxaliplatin positive score (CPS) of 5 or more.2,4,7,8,10,11 There is an unmet (Prof J A Ajani MD) regimen), resulting in a median overall survival duration of need for additional targeted therapies to prolong survival Correspondence to: Prof Jaffer A Ajani, The University about 1 year.2–8 in the large population of patients with HER2-negative of Texas, MD Anderson Cancer Combining targeted therapies with chemotherapy can disease, many of whom have a PD-L1 CPS less than 5.6,9,12,13 Center, Houston, TX 77030, USA [email protected] Research in context Evidence before this study improved progression-free survival and overall survival versus At the time of publication of this study, to our knowledge, there placebo plus chemotherapy in patients with CLDN18.2- were no non-immunotherapy, biomarker-based therapies positive, HER2-negative, locally advanced unresectable or approved for the first-line treatment of patients with human metastatic gastric or gastro-oesophageal junction epidermal growth factor receptor 2 (HER2)-negative, locally adenocarcinoma. To our knowledge, SPOTLIGHT is the first advanced unresectable or metastatic gastric or gastro- global, phase 3 study to examine the efficacy and safety of the oesophageal junction adenocarcinoma. We used Ovid to search addition of CLDN18.2-targeted therapy (ie, zolbetuximab) Embase on Feb 3, 2023, using the terms (“claudin 18” to standard chemotherapy (ie, mFOLFOX6 [modified folinic OR “CLDN18”) AND (“stomach cancer” OR “gastroesophageal acid, fluorouracil, and oxaliplatin regimen]) in gastric or gastro- junction”) AND (“phase 2 clinical trial” OR “phase 3 clinical trial”), oesophageal junction adenocarcinoma and to show prolonged with no language or time restrictions. We filtered the results by survival with a CLDN18.2-targeted therapy. Nearly 40% of articles with no time restrictions, and conference abstracts screened patients had tumours that met the definition of published within the past 3 years. We removed results related to CLDN18.2 positivity used in this study. Although nausea and the current study, conference abstracts reporting the same study vomiting leading to discontinuation were more frequent with or later published as articles, reviews, editorials, a phase 1 study, zolbetuximab plus chemotherapy versus placebo plus and a study in patients with locally advanced resectable gastric chemotherapy, the safety profile was manageable in the cancer. Our search yielded four studies with efficacy and safety context of the significant benefits for both progression-free data. Three studies evaluated zolbetuximab: the phase 2a MONO and overall survival. study of zolbetuximab monotherapy showed restricted Implications of all the available evidence antitumour efficacy, with a manageable safety profile in patients The SPOTLIGHT trial addresses a high unmet need for with claudin-18 isoform 2 (CLDN18.2)-positive tumours; the additional biomarker-targeted therapies to improve survival phase 2b FAST study of zolbetuximab plus chemotherapy outcomes in patients with previously untreated, showed improved progression-free survival and overall survival HER2-negative, locally advanced unresectable or metastatic versus chemotherapy, with a manageable safety profile in gastric or gastro-oesophageal junction adenocarcinoma for patients with CLDN18.2-positive tumours; and the whom, other than immunotherapy, no advances have been phase 2 ILUSTRO study (cohort 2) of zolbetuximab plus reported in over a decade. The results of SPOTLIGHT suggest chemotherapy showed a median progression-free survival of that CLDN18.2 tumour positivity defines a large population of 13·7 months and an objective response in 63% of patients, with patients whose survival is prolonged by targeted therapy with a manageable safety profile in patients with CLDN18.2-positive, zolbetuximab plus chemotherapy, and that zolbetuximab HER2-negative tumours. A phase 1b–2 study evaluated a could represent a new potential first-line therapy in CLDN18.2-specific chimeric antigen receptor T-cell therapy combination with chemotherapy in patients with CLDN18.2- and showed promising anti-tumour efficacy with a manageable positive, HER2-negative, locally advanced unresectable or safety profile in patients with CLDN18.2-positive tumours. metastatic gastric or gastro-oesophageal junction Added value of this study adenocarcinoma. With over 550 patients randomly assigned in the SPOTLIGHT trial, first-line zolbetuximab plus chemotherapy showed 1656 www.thelancet.com Vol 401 May 20, 2023  Articles Claudin-18 isoform 2 (CLDN18.2) is a tight junction performance status score of 0 or 1; and adequate organ protein that is normally expressed exclusively in gastric function. Additional eligibility criteria are listed in the mucosa cells.14,15 CLDN18.2 expression is retained in most appendix (pp 165–68). In SPOTLIGHT, the patient gastric and gastro-oesophageal junction adenocarcinoma selection assay was switched from a manual cells and is the dominant CLDN18 isoform expressed in immunohistochemistry test used in the FAST study both normal and malignant gastric cells.12,15–20 During (Claudetect 18.2; Ganymed Pharmaceuticals; Mainz, malignant transformation, cell polarity is lost and Germany) to the VENTANA CLDN18 (43-14A) RxDx CLDN18.2 might become exposed on the surface of Assay (both assays use the CLDN18 [43-14A] antibody). gastric and gastro-oesophageal junction adenocarcinoma The cutoff for CLDN18.2 positivity was adapted (from cells, which might render CLDN18.2 more accessible to ≥70% in FAST to ≥75% in SPOTLIGHT) to best identify antibodies.15–21 Therefore, CLDN18.2 is a promising a similar patient population using the new test. emerging therapeutic target.15–21 Sex was reported by study site staff through an Zolbetuximab is a first-in-class chimeric immuno­ interactive response technology system with the options globulin G1 monoclonal antibody that targets and binds to male or female. CLDN18.2.15,20 This binding mediates cell death of As an ad-hoc analysis, PD-L1 immunohistochemistry CLDN18.2-positive gastric and gastro-oesophageal junction was investigated using the Dako PD-L1 IHC 28-8 pharmDx adenocarcinoma cells via antibody-dependent cellular assay on the available remaining tumour samples. cytotoxicity and complement-dependent cyto­ toxicity.20,22,23 The phase 2b FAST study suggested that zolbetuximab Randomisation and masking improves progression-free survival and overall survival Enrolled patients were randomly assigned (1:1) to when combined with first-line chemotherapy versus zolbetuximab plus mFOLFOX6 or placebo plus chemotherapy alone in patients with advanced gastric or mFOLFOX6 by masked site staff via interactive response gastro-oesophageal junction adenocarcinoma; the survival technology by block randomisation (block sizes of two) benefit was enhanced in patients whose tumours had and stratified according to region (Asia vs non-Asia), higher CLDN18.2 expression compared with those with number of organs with metastases (0–2 vs ≥3), and lower CLDN18.2 expression.16 previous gastrectomy (yes vs no). The randomisation list Here, we report the results of the SPOTLIGHT trial, and study drug masking were maintained by the which investigated the efficacy and safety of first-line interactive response technology system. The funder, zolbetuximab plus mFOLFOX6 versus placebo plus investigators, clinical staff, and patients remained mFOLFOX6 in patients with CLDN18.2-positive, HER2- masked to treatment throughout the study. Zolbetuximab negative, locally advanced unresectable or metastatic and placebo were identical in appearance and form and gastric or gastro-oesophageal junction adenocarcinoma. were provided to the investigator or designee by an unmasked pharmacist and administered in identical Methods volumes, routes, and schedules to maintain masking. Study design and patients SPOTLIGHT is a global, randomised, placebo-controlled, Procedures double-blind, phase 3 trial that enrolled patients from Patients received either an intravenous infusion of 215 centres in 20 countries (appendix p 4). The trial was zolbetuximab 800 mg/m² (cycle 1, day 1) followed by done in accordance with the standards of Good Clinical 600 mg/m² (cycle 1, day 22, and days 1 and 22 of See Online for appendix Practice and the Declaration of Helsinki. The protocol subsequent cycles) plus intravenous infusion of and all amendments were approved by the appropriate mFOLFOX6 (folinic acid 400 mg/m², or optionally in institutional review board or ethics committee at each Japan, levofolinate 200 mg/m²; fluorouracil 400 mg/m² participating institution. All patients provided written bolus followed by 2400 mg/m² in a 46-h to 48-h infusion; informed consent before entering the trial. and oxaliplatin 85 mg/m²; on days 1, 15, and 29) or Eligible patients were aged 18 years or older with placebo plus mFOLFOX6, for four 42-day cycles. Patients CLDN18.2-positive (defined as ≥75% of tumour cells without disease progression continued beyond four cycles showing moderate-to-strong membranous CLDN18 with zolbetuximab or placebo plus, at the investigator’s staining, determined by central immunohistochemistry discretion, folinic acid (or optionally in Japan, levofolinate) using the investigational VENTANA CLDN18 [43-14A] and fluorouracil. Treatment continued until disease RxDx Assay [Roche Diagnostic Solutions; Tucson, AZ, progression, development of toxic effects, start of another USA]), HER2-negative (based on local or central anticancer treatment, or other discontinuation criteria evaluation), previously untreated, locally advanced were met, as specified in the protocol. unresectable or metastatic gastric or gastro-oesophageal The radiological tumour response was assessed by junction adenocarcinoma, with radiologically evaluable imaging at screening, then every 9 weeks in the first disease (measurable or non-measurable) according to 54 weeks, and every 12 weeks thereafter until disease Response Evaluation Criteria in Solid Tumors (RECIST) progression or the start of another anticancer treatment. version 1.1; an Eastern Cooperative Oncology Group During follow-up, survival was assessed at least every www.thelancet.com Vol 401 May 20, 2023 1657  Articles 12 weeks. Patients completed health-related quality-of-life were objective response and duration of response per assessments, including the European Organization for RECIST version 1.1 as determined by an independent Research and Treatment of Cancer QLQ-C30, QLQ-OG25, review committee, safety and tolerability of and Global Pain and the EuroQOL Five-Dimensions zolbetuximab, additional patient-reported outcomes, and Questionnaire, at screening, then every 3 weeks while on pharmacokinetics and immunogenicity of zolbetuximab; treatment, at treatment discon­tinuation, and at 30 days additional patient-reported outcome data will also be and 90 days after treatment discontinuation. disclosed in a subsequent publication. As an ad-hoc analysis, objective response per RECIST version 1.1 was Outcomes assessed in patients with measurable lesions. Adverse The primary endpoint was progression-free survival per events, graded according to National Cancer Institute RECIST version 1.1, as determined by an independent Common Terminology Criteria for Adverse Events review committee. Key secondary endpoints were overall version 4.03, were monitored throughout the trial and for survival and time to confirmed deterioration; statistical 90 days after treatment discontinuation. Adverse event hypothesis testing for time to confirmed deterioration in preferred terms were defined according to the Medical key patient-reported outcomes is pending the clinically Dictionary for Regulatory Activities terminology meaningful threshold obtained from the ongoing exit version 25.0. survey study per protocol and will be disclosed in a subsequent publication. Additional secondary endpoints Statistical analysis The Kaplan-Meier method was used to estimate the 2735 patients assessed for eligibility distribution of progression-free survival, overall survival, and duration of response, and a stratified log-rank test was used to assess between-group differences. A Cox 2170 ineligible 332 CLDN18.2 not available proportional hazard model, stratified according to region 1481 CLDN18.2-negative (Asia vs non-Asia), number of organs with metastases 357 other reasons* (0–2 vs ≥3), and previous gastrectomy (yes vs no), was used to estimate hazard ratios (HRs) and corresponding 565 enrolled and randomly assigned 95% CIs. The Cochran-Mantel-Haenszel test was used to assess between-group differences in objective response rate. Prespecified multiplicity adjustment methods were used to control the overall one-sided type I error rate 283 assigned to zolbetuximab plus 282 assigned to placebo plus mFOLFOX6 at 0·025. Efficacy boundaries were calculated for the mFOLFOX6 interim overall survival analysis based on the information fraction at the time of analysis. The reported 95% CIs 4 untreated 4 untreated describe the precision of the point estimates and might not correspond to the significance of the test. 279 received assigned treatment 278 received assigned treatment We planned to include 550 patients in the study. The sample size analysis was done by a statistician at the protocol development stage. The final analysis of 218 discontinued treatment† 232 discontinued treatment† progression-free survival was planned for when 133 disease progression 177 disease progression 25 adverse events 12 adverse events 300 patients had disease progression or died, to provide 29 withdrawal by subject 16 withdrawal by subject 93% power to detect a between-group difference with an 13 death 16 death 1 protocol deviation 17 other reasons assumed median progression-free survival of 9 months 1 lost to follow-up versus 6 months (HR 0·67) with zolbetuximab plus 20 other reasons mFOLFOX6 versus placebo plus mFOLFOX6, at an overall one-sided significance level of 0·025. A 10% 61 treatment ongoing 46 treatment ongoing dropout rate was considered in the sample size calculation for the progression-free survival analysis. Based on the planned event number, the statistically significant 283 included in efficacy analysis 282 included in efficacy analysis 279 included in safety analysis 278 included in safety analysis boundary for the progression-free survival analysis was an HR of 0·80. To strictly control the one-sided type I error rate at 0·025, overall survival was tested only if the null Figure 1: Trial profile CLDN18.2=claudin-18 isoform 2. mFOLFOX6=modified folinic acid, fluorouracil, and oxaliplatin regimen. hypothesis for progression-free survival was rejected. An *Represents patients whose tumours were CLDN18.2-positive but who failed screening for other reasons, interim analysis of overall survival was planned at the final including withdrawal by patient, laboratory findings, human epidermal growth factor receptor 2 expression status, progression-free survival analysis, and a final analysis of and Eastern Cooperative Oncology Group performance status score. †If a patient discontinued from both overall survival was planned after 396 deaths to provide zolbetuximab or placebo and mFOLFOX6 on the same day, all reasons for discontinuation are summarised; therefore, the sum of the values for individual reasons for discontinuation is more than 218 for the zolbetuximab 81% power to detect a between-group difference with an group and more than 232 for the placebo group. assumed median overall survival of 14·7 months versus 1658 www.thelancet.com Vol 401 May 20, 2023  Articles 11 months (HR 0·75) with zolbetuximab plus mFOLFOX6 Zolbetuximab plus Placebo plus mFOLFOX6 versus placebo plus mFOLFOX6, at an overall one-sided mFOLFOX6 group (n=283) group (n=282) significance level of 0·025. A 5% dropout rate was Age, years 62·0 (51·0–69·0) 60·0 (50·0–69·0) considered in the sample size calculation for the interim Sex overall survival analysis. Based on the planned event Male 176 (62%) 175 (62%) number, the statistically significant boundary for the Female 107 (38%) 107 (38%) interim overall survival analysis was an HR of 0·78. At the Region time of the interim overall survival analysis, a one-sided Asia 88 (31%) 89 (32%) level of significance of 0·0135 was used with an 82·3% Non-Asia 195 (69%) 193 (68%) information fraction. The study enrolled around Ethnicity 14 patients per month. The study accrual window was from Oct 29, 2018 (first patient randomly assigned), to Hispanic or Latino 36 (13%) 37 (13%) April 1, 2022 (last patient randomly assigned). The Not Hispanic or Latino 225 (80%) 213 (76%) assumed study duration was from June 21, 2018, to Missing 22 (8%) 32 (11%) Sept 9, 2022 (approximately 51 months). Organs with metastases Progression-free survival, overall survival, objective 0–2 219 (77%) 219 (78%) response, and duration of response were assessed in the ≥3 64 (23%) 63 (22%) full analysis set, which comprised all randomised Location of metastases* patients. Safety was assessed in the safety analysis set, Lymph node 101 (36%) 109 (39%) which comprised all randomised patients who received Peritoneum 94 (33%) 76 (27%) at least one dose of any study drug (ie, zolbetuximab, Liver 62 (22%) 75 (27%) placebo, or mFOLFOX6). Lung 36 (13%) 33 (12%) Full details of analyses are provided in the statistical Bone 28 (10%) 23 (8%) analysis plan in the appendix (pp 617–62). Abdominal cavity 19 (7%) 17 (6%) Sample size calculations were done with East Ovary 16 (6%) 19 (7%) version 6.4. Statistical data analyses were done with SAS Previous gastrectomy version 9.3 or later. An independent data monitoring Yes 84 (30%) 82 (29%) committee reviewed safety and efficacy data. The study is No 199 (70%) 200 (71%) registered with ClinicalTrials.gov, NCT03504397. Primary site Stomach 219 (77%) 210 (74%) Role of the funding source Gastro-oesophageal junction 64 (23%) 72 (26%) The funder of the study was involved in the study design, Lauren classification data collection, data analysis, data interpretation, and Diffuse 82 (29%) 117 (41%) writing of the report. Intestinal 70 (25%) 66 (23%) Mixed 31 (11%) 13 (5%) Results Unknown 49 (17%) 40 (14%) Between June 21, 2018, and April 1, 2022, 2735 patients Other 50 (18%) 42 (15%) were screened (figure 1). Of the 2403 patients assessable Missing 1 (75 8/16 8·34 13/22 10·41 0·96 (0·39–2·34) Sex Male 95/176 10·61 106/175 8·94 0·78 (0·59–1·02) Female 51/107 11·04 61/107 8·38 0·71 (0·49–1·03) Region Asia 45/88 12·55 47/89 8·21 0·56 (0·37–0·85) Non-Asia 101/195 9·69 120/193 9·36 0·85 (0·65–1·11) Number of metastatic sites 0–2 107/219 12·35 123/219 9·20 0·73 (0·56–0·94) ≥3 39/64 8·11 44/63 8·18 0·84 (0·55–1·30) Previous gastrectomy No 106/199 9·30 117/200 8·57 0·81 (0·62–1·05) Yes 40/84 12·42 50/82 9·26 0·62 (0·41–0·94) Primary site Stomach 109/219 12·22 126/210 8·38 0·69 (0·53–0·89) Gastro-oesophageal junction 37/64 8·77 41/72 8·94 1·02 (0·65–1·59) Lauren classification Diffuse 40/82 12·48 64/117 10·28 0·76 (0·51–1·13) Intestinal 41/70 10·28 46/66 6·57 0·58 (0·38–0·89) Mixed or other 49/81 9·79 35/55 8·67 0·93 (0·60–1·43) Country Japan 12/32 18·07 17/33 8·28 0·48 (0·23–1·01) Non-Japan 134/251 10·41 150/249 8·74 0·79 (0·63–1·00) China 12/19 8·54 10/17 6·24 0·50 (0·20–1·26) Non-China 134/264 11·04 157/265 9·07 0·75 (0·60–0·95) Race White 77/140 8·94 82/134 10·15 0·93 (0·68–1·27) Asian 47/96 13·96 51/97 8·21 0·53 (0·35–0·79) Tobacco history Never 75/142 10·35 82/137 8·25 0·74 (0·54–1·01) Current 15/26 10·28 14/25 8·67 1·00 (0·48–2·09) Former 56/113 12·32 70/118 9·76 0·71 (0·50–1·02) 0·25 0·50 1·0 2·0 4·0 Favours zolbetuximab + mFOLFOX6 Favours placebo + mFOLFOX6 Figure 2: Progression-free survival in the full analysis set (A) Kaplan-Meier plot by treatment group. (B) Subgroup analyses by treatment group. HR=hazard ratio. mFOLFOX6=modified folinic acid, fluorouracil, and oxaliplatin regimen. 283 patients in the zolbetuximab group versus set, an objective response, as evaluated by the investigator, 134 (48%, 42–54) of 282 patients in the placebo group. The was observed in 150 (53%, 95% CI 47–59) of median duration of response was 9·00 months (95% CI 283 patients in the zolbetuximab group versus 6·87–10·25) in the zolbetuximab group versus 124 (44%, 38–50) of 282 patients in the placebo group. The 8·05 months (6·47–10·81) in the placebo group (appendix median duration of response was 9·00 months (95% CI p 6). As an ad-hoc analysis, in patients with measurable 7·49–10·25) in the zolbetuximab group versus 6·80 months disease, an objective response was observed in 128 (61%, (6·21–8·31) in the placebo group (appendix p 8). 95% CI 54–67) of 211 patients in the zolbetuximab group Subsequent anticancer therapies were received by versus 131 (62%, 55–69) of 211 patients in the placebo 135 (48%) of 283 patients in the zolbetuximab group versus group. The median duration of response was 8·51 months 148 (53%) of 282 patients in the placebo group. Types of (95% CI 6·80–10·25) in the zolbetuximab group versus therapies were similar between the two arms (appendix p 9). 8·11 months (6·47–11·37) in the placebo group The median duration of exposure for each treatment (appendix p 7). As a sensitivity analysis in the full analysis component is reported in the appendix (p 10); the median www.thelancet.com Vol 401 May 20, 2023 1661  Articles duration of exposure was 6·2 months (IQR 2·5–12·0) for appendix (pp 15–22). Nausea led to discontinuation of any the zolbetuximab group and 6·4 months (3·7–10·3) for the study drug in 18 (6%) of 279 patients in the zolbetuximab placebo group. Treatment-emergent grade 3 or worse group versus three (1%) of 278 patients in the placebo adverse events occurred in 242 (87%) of 279 patients in the group, and vomiting led to discontinuation of any study zolbetuximab group versus 216 (78%) of 278 patients in the drug in 20 (7%) of 279 patients in the zolbetuximab group placebo group (table 2). The most common all-grade versus one (65 60/102 16·16 65/101 15·28 0·76 (0·53–1·09) ≤75 138/267 18·86 165/260 15·54 0·71 (0·57–0·90) >75 11/16 9·76 12/22 15·64 1·32 (0·58–3·00) Sex Male 98/176 17·38 113/175 15·57 0·76 (0·58–1·00) Female 51/107 23·75 64/107 15·28 0·73 (0·50–1·05) Region Asia 47/88 21·49 59/89 17·74 0·64 (0·44–0·95) Non-Asia 102/195 16·99 118/193 13·73 0·80 (0·61–1·04) Number of metastatic sites 0–2 110/219 19·68 129/219 15·84 0·77 (0·59–0·99) ≥3 39/64 16·43 48/63 12·02 0·67 (0·44–1·03) Previous gastrectomy No 109/199 16·95 125/200 14·32 0·84 (0·65–1·09) Yes 40/84 24·80 52/82 15·74 0·58 (0·38–0·87) Primary site Stomach 111/219 20·24 135/210 13·83 0·67 (0·52–0·86) Gastro-oesophageal junction 38/64 15·80 42/72 16·39 1·07 (0·69–1·67) Lauren classification Diffuse 46/82 20·63 75/117 15·84 0·77 (0·53–1·11) Intestinal 38/70 22·90 48/66 13·83 0·55 (0·36–0·85) Mixed or other 48/81 16·13 34/55 14·69 0·99 (0·64–1·54) Country Japan 23/32 23·10 27/33 17·71 0·71 (0·41–1·25) Non-Japan 126/251 17·87 150/249 14·69 0·76 (0·60–0·96) China 9/19 16·16 9/17 12·25 0·91 (0·36–2·32) Non-China 140/264 18·86 168/265 15·57 0·74 (0·59–0·92) Race White 81/140 16·13 81/134 15·28 0·95 (0·70–1·29) Asian 49/96 23·33 65/97 16·53 0·57 (0·39–0·83) Tobacco history Never 70/142 18·96 87/137 13·47 0·68 (0·49–0·93) Current 15/26 19·35 15/25 15·57 0·82 (0·40–1·69) Former 64/113 17·81 74/118 15·84 0·81 (0·58–1·13) 0·25 0·50 1·0 2·0 4·0 Favours zolbetuximab + mFOLFOX6 Favours placebo + mFOLFOX6 Figure 3: Overall survival in the full analysis set (A) Kaplan-Meier plot by treatment group. (B) Subgroup analyses by treatment group. HR=hazard ratio. mFOLFOX6=modified folinic acid, fluorouracil, and oxaliplatin regimen. the placebo group might have been partly due to the patients from various countries compared with distribution of patients from various countries—there SPOTLIGHT, and might provide additional insights into were more patients from Japan and Korea in SPOTLIGHT the survival outcomes in the control groups in this patient compared with other global studies (eg, CheckMate 649).10 population. Although the median difference in Similarly, in the ATTRACTION-4 study in patients in progression-free survival was less than the 3 months Japan, South Korea, and Taiwan, the median overall expected per statistical assumptions, this might have been survival of the control group was longer than in the global affected by the longer than expected median progression- CheckMate 649 study.10,24 Another phase 3 study of free survival in the placebo group. The longer than zolbetuximab plus chemotherapy in patients with anticipated median progression-free survival and overall CLDN18.2-positive, HER2-negative, locally advanced survival in patients in the placebo group might also have unresectable or metastatic gastric or gastro-oesophageal contributed to the delayed separation of the survival curves junction adenocarcinoma (GLOW; NCT03653507) has of patients between the two treatment groups. Other completed enrolment, but trial patients continue to be possible explanations include early discontinuation of followed up. This study has different distributions of patients in the zolbetuximab group because of increased www.thelancet.com Vol 401 May 20, 2023 1663  Articles Zolbetuximab plus mFOLFOX6 group (n=279) Placebo plus mFOLFOX6 group (n=278) All grade Grade ≥3 All grade Grade ≥3 All treatment-emergent events 278 (>99%) 242 (87%) 277 (>99%) 216 (78%) Treatment-emergent serious events 125 (45%) ·· 121 (44%) ·· Treatment-emergent events leading to discontinuation of 120 (43%) ·· 106 (38%) ·· any study drug Treatment-related events leading to discontinuation of 106 (38%) ·· 82 (29%) ·· any study drug Treatment-emergent events leading to discontinuation of 55 (20%) ·· 30 (11%) ·· zolbetuximab or placebo Treatment-related events leading to discontinuation of 38 (14%) ·· 6 (2%) ·· zolbetuximab or placebo Treatment-emergent events leading to death ·· 22 (8%) ·· 24 (9%) Treatment-related events leading to death ·· 5 (2%) ·· 4 (1%) Treatment-emergent events* by preferred terms Nausea 230 (82%) 45 (16%) 169 (61%) 18 (6%) Vomiting 188 (67%) 45 (16%) 99 (36%) 16 (6%) Decreased appetite 131 (47%) 16 (6%) 93 (33%) 9 (3%) Diarrhoea 110 (39%) 12 (4%) 122 (44%) 9 (3%) Peripheral sensory neuropathy 106 (38%) 11 (4%) 118 (42%) 15 (5%) Neutropenia 102 (37%) 79 (28%) 94 (34%) 65 (23%) Anaemia 100 (36%) 24 (9%) 104 (37%) 26 (9%) Constipation 99 (35%) 3 (1%) 112 (40%) 2 (1%) Neutrophil count decreased 95 (34%) 69 (25%) 91 (33%) 69 (25%) Fatigue 78 (28%) 17 (6%) 91 (33%) 14 (5%) Asthenia 74 (27%) 20 (7%) 64 (23%) 7 (3%) Abdominal pain 67 (24%) 12 (4%) 82 (29%) 6 (2%) Stomatitis 58 (21%) 7 (3%) 57 (21%) 3 (1%) Weight decreased 55 (20%) 5 (2%) 54 (19%) 2 (1%) Pyrexia 54 (19%) 1 (