KM Research Hub
← 返回论文库
已完成 入排:排除 图源:可得 入排置信度 0.96

Nivolumab plus chemotherapy or ipilimumab in gastroesophageal cancer: exploratory biomarker analyses of a randomized phase 3 trial.

Kohei Shitara, Yelena Y Janjigian, Jaffer Ajani, Markus Moehler, Jin Yao, Xuya Wang, Aparna Chhibber, Dimple Pandya, Lin Shen, Marcelo Garrido, Carlos Gallardo, Lucjan Wyrwicz, Kensei Yamaguchi, Tomasz Skoczylas, Arinilda Bragagnoli, Tianshu Liu, Michael Schenker, Patricio Yañez, Ruben Kowalyszyn, Michalis Karamouzis, Thomas Zander, Kynan Feeney, Elena Elimova, Parul Doshi, Mingshun Li, Ming Lei

Nat Med 31(5) 1519-1530

PMID40055521
PMCIDPMC12092258
DOI10.1038/s41591-025-03575-0
发表日期2025-03-07
入排依据:属于事后亚组、二次、生活质量或卫生经济学分析,不作为原始随机试验结果纳入。
摘要

摘要

First-line nivolumab-plus-chemotherapy demonstrated superior overall survival (OS) and progression-free survival versus chemotherapy for advanced gastroesophageal adenocarcinoma with programmed death ligand 1 combined positive score ≥ 5, meeting both primary end points of the randomized phase 3 CheckMate 649 trial. Nivolumab-plus-ipilimumab provided durable responses and higher survival rates versus chemotherapy; however, the prespecified OS significance boundary was not met. To identify biomarkers predictive of differential efficacy outcomes, post hoc exploratory analyses were performed using whole-exome sequencing and RNA sequencing. Nivolumab-based therapies demonstrated improved efficacy versus chemotherapy in hypermutated and, to a lesser degree, Epstein-Barr virus-positive tumors compared with chromosomally unstable and genomically stable tumors. Within the KRAS-altered subgroup, only patients treated with nivolumab-plus-chemotherapy demonstrated improved OS benefit versus chemotherapy. Low stroma gene expression signature scores were associated with OS benefit with nivolumab-based regimens; high regulatory T cell signatures were associated with OS benefit only with nivolumab-plus-ipilimumab. Our analyses suggest that distinct and overlapping pathways contribute to the efficacy of nivolumab-based regimens in gastroesophageal adenocarcinoma.

KM 提取结果

3 个 Figure · 14 个 KM 面板

进入结果中心
KM Figure2 个面板

Fig. 3

Efficacy by KRAS alterations in patients treated with nivolumab-plus-chemotherapy versus chemotherapy. a , Kaplan–Meier estimates of OS in all WES-evaluable patients with altered (left) or unaltered (right) KRAS mutations + amplifications. b , OS by baseline KRAS pathway alteration status derived from the model, including interaction of KRAS alteration status with treatment, MSI and TMB status. Data are presented as unstratified HRs and 95% CI. c , KRAS mutation map. a MSI status not evaluable/available in two patients. KRAS , Kirsten rat sarcoma viral oncogene.

Altered_KRAS_mutations_amplifications

OS

Fig. 3 · Panel Altered_KRAS_mutations_amplifications

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Altered_KRAS_mutations_amplifications 是同一 Fig. 3 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
Unaltered_KRAS_mutations_amplifications

OS

Fig. 3 · Panel Unaltered_KRAS_mutations_amplifications

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Unaltered_KRAS_mutations_amplifications 是同一 Fig. 3 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
KM Figure6 个面板

Extended Data Fig. 5

OS by GES scores and PD-L1 CPS status in patients treated with nivolumab-plus-chemotherapy versus chemotherapy. a , Forest plot showing the correlation between OS and selected signatures used to stratify patients into tertiles (high, medium, or low). Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was less than 5. b and c , Kaplan–Meier estimates of OS in all RNA-seq-evaluable patients with low, medium or high angiogenesis GES scores (5-gene angiogenesis) and PD-L1 CPS ≥ 5 ( b ) or PD-L1 CPS < 5 ( c ) at baseline. Chemo, chemotherapy; CI, confidence interval; CPS, combined positive score; GES, gene expression signature; HR, hazard ratio; NIVO, nivolumab; OS, overall survival; PD-L1, programmed death ligand 1.

B

Overall survival

Extended Data Fig. 5 · Panel B

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel B 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线交叉次数异常,需检查跨图跳线或误跟踪曲线缺失:预期 2 条,实际恢复 1 条
查看结果
B_2

Overall survival

Extended Data Fig. 5 · Panel B_2

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel B_2 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
B_3

Overall survival

Extended Data Fig. 5 · Panel B_3

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel B_3 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
C

Overall survival

Extended Data Fig. 5 · Panel C

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel C 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
C_5

Overall survival

Extended Data Fig. 5 · Panel C_5

complete 待 EXE 对齐 预期 2 条 已提取 2 条

Panel C_5 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

查看结果
C_6

Overall survival

Extended Data Fig. 5 · Panel C_6

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel C_6 是同一 Extended Data Fig. 5 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
KM Figure6 个面板

Extended Data Fig. 7

OS by PD-L1 CPS and Treg status in patients treated with nivolumab-plus-ipilimumab versus chemotherapy. a and b , Kaplan–Meier estimates of OS in all RNA-seq-evaluable patients with high, medium or low regulatory T cell GES scores (2-gene regulatory T cell) and PD-L1 CPS ≥ 5 ( a ) or PD-L1 CPS < 5 ( b ) at baseline. Chemo, chemotherapy; CI, confidence interval; CPS, combined positive score; HR, hazard ratio; IPI, ipilimumab; NIVO, nivolumab; OS, overall survival; PD-L1, programmed death ligand 1.

Treg-high_a_

OS

Extended Data Fig. 7 · Panel Treg-high_a_

complete 待 EXE 对齐 预期 2 条 已提取 2 条

Panel Treg-high_a_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示曲线 1:extreme jump曲线 1:off origin start
查看结果
Treg-high_b_

OS

Extended Data Fig. 7 · Panel Treg-high_b_

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Treg-high_b_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
Treg-medium_a_

OS

Extended Data Fig. 7 · Panel Treg-medium_a_

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Treg-medium_a_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
Treg-medium_b_

OS

Extended Data Fig. 7 · Panel Treg-medium_b_

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Treg-medium_b_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
Treg-low_a_

OS

Extended Data Fig. 7 · Panel Treg-low_a_

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Treg-low_a_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
Treg-low_b_

OS

Extended Data Fig. 7 · Panel Treg-low_b_

quality_warning 待 EXE 对齐 预期 2 条 已提取 1 条

Panel Treg-low_b_ 是同一 Extended Data Fig. 7 中的独立 KM 子图;人群、终点或分析条件以本 Panel 图注语义为准。

质量提示疑似重复轨迹:同一条 KM 曲线可能被重复识别疑似图例或人工水平线:轨迹出现不合理的共同平台曲线缺失:预期 2 条,实际恢复 1 条疑似重复轨迹:同一条 KM 曲线可能被重复识别
查看结果
其他 Figure(12,默认折叠)

这些图片未产生 KM 曲线结果,不在主页面重复展示。

Fig. 1Genomic characterization of baseline tumors from CheckMate 649. a , Flowchart depicting method of genomic classification of tumors (left). Scatter-plot of TMB versus genomic instability in EBV-negative tumors (right). b , Tumor subtyping by anatomical location. c , d , Forest plot of OS by genomic subsets in patients treated with nivolumab-plus-chemotherapy ( c ) or nivolumab-plus-ipilimumab ( d ) versus chemotherapy. Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was <5. Open circles represent MSI-H, filled circles represent MSS, and squares represent not available (NA). IPI, ipilimumab; MSI-H, MSI-high; NIVO, nivolumab.
非 KM Figure
Fig. 2OS by TMB in patients treated with nivolumab-plus-chemotherapy or nivolumab-plus-ipilimumab versus chemotherapy. a , Forest plot showing OS by baseline TMB status in patients who received nivolumab-plus-chemotherapy versus chemotherapy. Data are presented as unstratified HRs and 95% CI. b , Kaplan–Meier curves of OS by microsatellite stability status in patients treated with nivolumab-plus-chemotherapy versus chemotherapy. c , Forest plot showing OS by baseline TMB status in patients who received nivolumab-plus-ipilimumab versus chemotherapy. d , Kaplan–Meier curves of OS by microsatellite stability status in patients treated with nivolumab-plus-ipilimumab versus chemotherapy. TMB-high denotes ≥199 mutations per exome, whereas TMB-low denotes <199 mutations per exome. Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was <5. a TMB not evaluable (NE)/available in 896 patients (NIVO + chemo: n = 431; Chemo: n = 465). b TMB not evaluable/available in 727 patients (NIVO + chemo: n = 355; Chemo: n = 372). c TMB not evaluable/available in 447 patients (NIVO + IPI: n = 226; Chemo: n = 221). d TMB not evaluable/available in 351 patients (NIVO + IPI: n = 181; Chemo: n = 170).
非 KM Figure
Fig. 4Relationship between key biomarkers in all patients treated with nivolumab-plus-chemotherapy, nivolumab-plus-ipilimumab or chemotherapy. a , Correlation of gene signatures among all RNA-seq evaluable patients. Numbers indicate the Spearman’s correlation. The strength of association is indicated by the color scale shown on the right. b , Heatmap of baseline tumors clustered by the gene signatures of interest in patients treated with nivolumab-plus-chemotherapy, nivolumab-plus-ipilimumab or chemotherapy. For visualization purposes, each gene signature was first normalized by z -score method where samples with high z -score (red) indicates relative high gene signature score and low z -score (blue) indicates relative low gene signature score. Patients (column) were ordered based on similarity of gene signatures of their tumor samples via hierarchical clustering method. Dendrogram (top) was added to show the hierarchical relationship between samples. Similarly, gene signatures (rows) were ordered and dendrogram (left) was added to show the hierarchical relationship between gene signatures. a Mutations per exome.
非 KM Figure
Fig. 5OS by GES scores in patients treated with nivolumab-plus-chemotherapy versus chemotherapy. a , Forest plot showing the correlation between OS and selected signatures used to stratify patients into tertiles (high, medium or low). Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was <5. b , Kaplan–Meier estimates of OS in all RNA-seq-evaluable patients with low, medium or high angiogenesis GES scores (five-gene angiogenesis) at baseline.
非 KM Figure
Fig. 6OS by GES scores in patients treated with nivolumab-plus-ipilimumab versus chemotherapy. a , Forest plot showing the correlation between OS and selected signatures used to stratify patients into tertiles (high, medium or low). Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was <5. b , Kaplan–Meier estimates of OS in all RNA-seq-evaluable patients with high, medium or low T reg cell GES scores (two-gene T reg cell) at baseline.
非 KM Figure
Extended Data Fig. 1Genomic characterization of baseline tumors from CheckMate 649. a and b , Tumor characterization subtyping by genomic subtypes and histology ( a ), or PD-L1 CPS ( b ). P values are based on two-sided Chi-squared test. Exact P value for ( a ) was 6.673e-09. c and d , Forest plot of OS by genomic subsets in all WES-evaluable patients treated with nivolumab-plus-chemotherapy ( c ) or nivolumab-plus-ipilimumab ( d ) versus chemotherapy. Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was less than 5. CIN, chromosomal instability; CPS, combined positive score; EBV, Epstein–Barr virus; GS, genomically stable; IPI, ipilimumab; MSI-H, microsatellite instability high; MSS, microsatellite stable; N/A, not available; NIVO, nivolumab; PD-L1, programmed death ligand 1; TMB, tumor mutational burden; WES, whole-exome sequencing.
非 KM Figure
Extended Data Fig. 2OS by PD-L1 CPS and TMB status in patients treated with nivolumab-plus-chemotherapy or nivolumab-plus-ipilimumab versus chemotherapy. a , Forest plot showing OS by baseline PD-L1 CPS and TMB status in patients who received nivolumab-plus-chemotherapy versus chemotherapy. b , Forest plot showing OS by baseline PD-L1 CPS and TMB status in patients who received nivolumab-plus-ipilimumab versus chemotherapy. Data are presented as unstratified HRs and 95% CI. a TMB not evaluable/available in 504 patients (NIVO+chemo: n = 238; chemo: n = 266). b TMB not evaluable/available in 376 patients (NIVO+chemo: n = 185; chemo: n = 191). c TMB not evaluable/available in 243 patients (NIVO+IPI: n = 121; chemo: n = 122). d TMB not evaluable/available in 191 patients (NIVO+IPI: n = 98; chemo: n = 93). Chemo, chemotherapy; CI, confidence interval; CPS, combined positive score; HR, hazard ratio; NA, not available; NIVO, nivolumab; OS, overall survival; PD-L1, programmed death ligand 1; TMB, tumor mutational burden; WES, whole-exome sequencing.
非 KM Figure
Extended Data Fig. 3OS by gene alterations. a , Oncoprints of gene alterations among all WES-evaluable patients. Frequency and overlap of select gene alterations. b and c , Forest plots showing the association between OS and selected gene alterations in patients who received nivolumab-plus-chemotherapy versus chemotherapy ( b ) or nivolumab-plus-ipilimumab versus chemotherapy ( c ). Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was less than 10. HR were derived from model including interaction of gene alteration status with treatment, MSI and TMB status. a MSI status not evaluable/available in 2 patients. Chemo, chemotherapy; CI, confidence interval; CPS, combined positive score; HR, hazard ratio; IPI, ipilimumab; MSI-H, microsatellite instability high; N/A, not available; NIVO, nivolumab; OS, overall survival; WES, whole-exome sequencing.
非 KM Figure
Extended Data Fig. 4Association test of differential OS treatment effect by GES scores in patients treated with nivolumab-plus-chemotherapy or with nivolumab-plus-ipilimumab versus chemotherapy. a , P value of interaction between GES score and treatment arm among all RNA-seq-evaluable patients and by PD-L1 CPS status who received nivolumab-plus-chemotherapy versus chemotherapy. Selected gene signatures with P values < 0.1 from Likelihood Ratio Test are colored in the table; color scale is shown at the bottom. b , P value of interaction between GES score and treatment arm among all RNA-seq-evaluable patients and by PD-L1 CPS status who received nivolumab-plus-ipilimumab versus chemotherapy. Selected gene signatures with P values < 0.1 from Likelihood Ratio Test are colored in the table; color scale is shown at the bottom. Selected gene signatures with P values < 0.1 from Likelihood Ratio Test are colored in the table; color scale is shown at the bottom.
非 KM Figure
Extended Data Fig. 6OS by GES scores and PD-L1 CPS status in patients treated with nivolumab-plus-ipilimumab versus chemotherapy. Forest plot showing the correlation between OS and selected signatures used to stratify patients into tertiles (high, medium versus low). Data are presented as unstratified HRs and 95% CI. HR was not calculated if the number of patients in each arm was less than 5. CI, confidence interval; CPS, combined positive score; GES, gene expression signature; HR, hazard ratio; PD-L1, programmed death ligand 1.
非 KM Figure
Extended Data Fig. 8Correlations between biomarkers in patients treated with nivolumab-plus-chemotherapy, nivolumab-plus-ipilimumab or chemotherapy. Summary of factors potentially associated with efficacy with nivolumab-plus-chemotherapy or nivolumab-plus-ipilimumab. a Indicates significant association with treatment when treated as a continuous variable ( P < 0.1). These exploratory P values are not meant to show statistical significance and are intended to describe the relative performance of the different signatures for association with response. Chemo, chemotherapy; CPS, combined positive score; GES, gene expression signatures; IPI, ipilimumab; NIVO, nivolumab; PD-L1, programmed death ligand 1; TMB, tumor mutational burden.
非 KM Figure
Extended Data Fig. 9Gene set enrichment analysis in patients treated with nivolumab-plus-chemotherapy or nivolumab-plus-ipilimumab versus chemotherapy. Enrichment of genes associated with OS in nivolumab-plus-chemotherapy, nivolumab-plus-ipilimumab, or chemotherapy as well as the genes with differential OS association by treatment arm (interaction) using chemotherapy as the reference. The data are shown as NES (adjusted P value) and the cells are colored if adjusted P value is < 0.01. The strength of association is shown as a color scale at the bottom. Chemo, chemotherapy; IPI, ipilimumab; NES, normalized enrichment score; NIVO, nivolumab; OS, overall survival.
非 KM Figure