Abstract
Background
Chemotherapy (chemo) combined with an immune checkpoint inhibitor (ICI) or dual ICI therapy with nivolumab and ipilimumab (nivo + ipi) is the standard first-line treatment for patients with advanced esophageal squamous cell carcinoma (ESCC). In this study, we evaluated real-world clinical outcomes for first-line ICI-based therapy and explored its prognostic factors.
Patients and methods
This single-center retrospective study included patients with ESCC who received ICI-based therapy between January 2021 and July 2024.
Results
In total, 92 patients received either ICI + chemo (n = 60) or nivo + ipi (n = 32). The median progression-free survival and overall survival (OS) were 5.0 and 16.0 months for ICI + chemo and 3.5 and 16.9 months for nivo + ipi, respectively. Of the 70 patients with measurable lesions, early tumor shrinkage (ETS) was achieved in 37% for ICI + chemo and 33% for nivo + ipi. ETS was significantly associated with a lower performance status and neutrophil-to-lymphocyte ratios, but not with the treatment regimen or programmed death-ligand 1 (PD-L1) status. Patients who achieved ETS showed significant tumor reduction and a durable response. ETS was an independent predictor of favorable OS (hazard ratio 0.34, 95% confidence interval 0.11-0.88, P = 0.04), whereas neither the treatment regimen nor the PD-L1 status influenced OS. Immune-related adverse events of grade ≥3 occurred in 12% of patients.
Conclusions
First-line immunotherapy is effective and safe for the treatment of patients with ESCC. Rapid and deep tumor shrinkage may serve as an early predictive biomarker for longer survival.
Key words: esophageal cancer, esophageal squamous cell carcinoma, chemotherapy, immunotherapy, early tumor shrinkage, depth of response
Highlights
-
•
First-line ICI-based therapy is effective and safe for real-world ESCC cases.
-
•
ETS is associated with deeper tumor reduction, a durable response, and greater survival benefits.
-
•
Patients with ETS are more likely to have a lower ECOG PS and NLR.
-
•
PD-L1 TPS status may guide nivo + ipi treatment decisions.
Introduction
Esophageal cancer (EC) remains a public health concern with a high mortality and disease burden worldwide.1 EC can be histologically classified into two main types: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma, each with distinct epidemiological, molecular, and clinical characteristics.2 ESCC is the predominant histological subtype of EC, accounting for ∼85% of all EC cases.3 Geographically, ESCC is more prevalent in Asia, where it constitutes >80% of all global cases.3 By contrast, esophageal adenocarcinoma is more commonly observed in North America, Western Europe, and Northern Europe. Because of the absence of early clinical signs, ESCC is often diagnosed at an advanced stage, and thus not amenable to curative treatment.4 Systemic chemotherapy is the main therapeutic option for patients with ESCC, which is aimed at prolonging survival time and improving symptoms and quality of life.5, 6, 7
The cancer immunity cycle is often disrupted by inhibitory immune checkpoint molecules such as programmed cell death protein 1 (PD-1) and its ligand programmed death-ligand 1 (PD-L1), which facilitate antitumor immunity evasion.8 ESCC is characterized by high PD-L1 expression,9,10 potentially indicating its susceptibility to PD-1/PD-L1 blockade. Notably, immune checkpoint inhibitors (ICIs) that target PD-1 have substantially transformed therapeutic paradigms, providing durable clinical responses in several landmark trials.9,11 Phase III trials, including KEYNOTE-59012 and CheckMate 648,13 have demonstrated the survival benefits of ICIs, including anti-PD-1 antibodies pembrolizumab and nivolumab (nivo), combined with first-line chemotherapy (ICI + chemo) for patients with ESCC.14, 15, 16, 17, 18 As a result, this regimen has been established as a new standard first-line therapy. Cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) functions as a negative regulator during the initial priming of T cells in the early stages of the immune response, whereas PD-1/PD-L1 suppresses antitumor T-cell responses in later stages.19 Therefore, the dual blockade of PD-1/PD-L1 and CTLA-4 synergistically enhances antitumor immunity by inhibiting complementary immunosuppressive pathways. In the CheckMate 648 trial, nivo combined with the anti-CTLA-4 monoclonal antibody ipilimumab (nivo + ipi) was compared with chemotherapy alone.13 The nivo + ipi regimen resulted in longer overall survival (OS) and prolonged duration of response (DOR) in patients with ESCC, establishing dual ICI therapy as another new standard first-line treatment. However, treatment-related serious adverse events (AEs) were more frequent with nivo + ipi than with chemotherapy alone, despite the treatment being chemotherapy free. The European Medicines Agency has authorized the use of pembrolizumab + chemotherapy for patients whose tumors express PD-L1 with a combined positive score (CPS) of ≥10 only. Both nivo + chemotherapy and nivo + ipi have been approved for patients with a tumor proportion score (TPS) of ≥1%. The European Society for Medical Oncology clinical practice guidelines assign a lower grade of recommendation to nivo + ipi compared with nivo + chemotherapy due to a lower radiological response rate and an increased risk of early progression and death in patients treated without chemotherapy.20 By contrast, in the USA and Japan, all ICI-based therapies have been approved as first-line treatments for ESCC, regardless of the PD-L1 TPS or CPS status. The National Comprehensive Cancer Network recommends nivo + ipi as one of the preferred first-line regimens for all patients with ESCC, with a category 2A recommendation.5 The Japanese EC practice guidelines recommend both ICI + chemo and nivo + ipi as equally preferred options for the treatment of all patients with ESCC.7
Early tumor shrinkage (ETS) and depth of response (DpR) are emerging as key metrics for evaluating treatment efficacy. ETS reflects a reduction in tumor size within the initial weeks of treatment, indicating timely treatment sensitivity. By contrast, DpR measures the maximum tumor reduction achieved during therapy and serves as a marker of cumulative response. ETS is a practical and reliable marker because of its early availability during treatment and is strongly associated with improved survival outcomes across various cancer types, including ESCC.21, 22, 23, 24, 25, 26, 27 In patients with ESCC receiving pembrolizumab combined with chemotherapy in the KEYNOTE-590 trial, an ETS of ≥20% was associated with improved OS compared with those without ETS.28 However, data on the clinical utility of ETS in patients receiving ICI + chemo in clinical practice remain limited.29 Furthermore, the association between ETS and PD-L1 expression and its relevance in patients treated with nivo + ipi remains unclarified.
Real-world data may help bridge this gap by providing valuable insights into treatment effectiveness and safety in diverse patient populations in clinical practice, thereby supporting more informed and patient-centered decision making. Notably, real-world data on the use of first-line ICI-based therapies for ESCC cases are limited,30 particularly those reporting on the nivo + ipi regimen, treatment efficacy stratified by PD-L1 status, and prognostic markers. To address these gaps, we conducted a retrospective observational study to assess the clinical efficacy and safety of first-line ICI-based therapies and investigate the utility of ETS as a predictor of long-term efficacy and survival in patients with ESCC in clinical practice.
Patients and methods
Study population
This retrospective single-institution study was conducted at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research. The key inclusion criteria were patients with (i) nonresectable, recurrent/metastatic EC, (ii) histologically confirmed squamous cell carcinoma, (iii) ICI + chemo or nivo + ipi as first-line treatment therapies between January 2021 and July 2024, and (iv) no other advanced cancers.
In the ICI + chemo group, the patients received pembrolizumab intravenously at a dose of either 200 mg every 3 weeks or 400 mg every 6 weeks. Alternatively, nivo at a dose of 240 mg every 2 weeks or 480 mg every 4 weeks combined with fluorouracil and cisplatin chemotherapy was administered. Chemotherapy comprised intravenous cisplatin at a dose of 80 mg per square meter every 3-4 weeks and continuous intravenous infusion of fluorouracil at a dose of 800 mg/m2/day on days 1-5 of each 3-4-week cycle.12,13 In the nivo + ipi group, patients received nivo at a dose of 240 mg every 2 weeks or 360 mg every 3 weeks, combined with ipi at 1 mg/kg administered intravenously every 6 weeks.13 ICI-based therapy was regarded as the first-line treatment for residual disease or recurrence after chemoradiotherapy.
This study was approved by the Ethics Review Board of our hospital (IRB number 2021-GB-095) and conducted in accordance with the principles of the Declaration of Helsinki. The study protocol was publicly available on the hospital’s website, and participants were provided with the opportunity to opt out of the study. No additional informed consent was required.
Assessments
A standardized immunohistochemical PD-L1 antibody assay (Dako 28-8 pharmDx), which was approved as a complementary diagnostic tool, was used to evaluate PD-L1 expression. The TPS was calculated as the percentage of viable tumor cells with partial or complete PD-L1 membrane staining relative to the total number of viable tumor cells in the sample. Data on patient characteristics, AEs, and clinical outcomes were obtained from electronic medical records, and performance status (PS) was assessed using the Eastern Cooperative Oncology Group (ECOG) PS. The symptoms of dysphagia were assessed using a standardized dysphagia score.31 Clinicopathological data included age, sex, metastatic sites, number of metastatic organs, primary resection, and history of systemic chemotherapy or radiotherapy. Baseline diseases before ICI-based therapy were classified into the following categories: (i) unresectable advanced; (ii) synchronous metastatic; (iii) recurrent, locoregional; and (iv) recurrent, distant. Unresectable advanced diseases included cases with residual tumors after esophagectomy or chemoradiotherapy. Recurrent diseases included cases of recurrence following a clinical complete response (CR) after chemoradiotherapy for locally advanced disease. The neutrophil-to-lymphocyte ratio (NLR) was used to assess the systemic inflammatory status. The NLR was calculated using the following formula: NLR = absolute neutrophil count/lymphocyte count. Hematological and nonhematological toxicities were assessed during ICI-based therapy, following the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0.
Treatment assessment was based on RECIST version 1.1, conducted among patients with at least one measurable lesion. The overall response rate (ORR) was defined as the percentage of patients who achieved the best overall response of either CR or partial response (PR). Disease control rate (DCR) was defined as the percentage of patients with the best overall response to CR, PR, or stable disease. ETS was defined as a ≥20% reduction in the sum of the longest diameters of RECIST target lesions from baseline, assessed within the first 12 weeks of initiating treatment.21 In addition, the DpR was defined as the percentage change from baseline in the sum of the longest diameters of the RECIST target lesions at the nadir without new lesions or progression of nontarget lesions.21 Investigators who were not blinded to the study evaluated not only the objective tumor response according to the RECIST version 1.1 criteria but also the ETS and DpR based on treatment assessment reports reviewed by experienced, independent radiologists.
Statistical analysis
To evaluate patient characteristics, summary statistics were constructed using frequencies and proportions of categorical variables, and medians and ranges of continuous variables. A two-sample t-test or Mann–Whitney U-test was used to compare continuous variables between groups, whereas Fisher’s exact test was used for categorical data. Progression-free survival (PFS) was defined as the time from the first dose of ICI-based therapy to the onset of either documented progressive disease (PD) or death from any cause. The DOR was defined as the time from the first documented objective response (CR or PR) to disease progression or death from any cause. OS was defined as the time from the first dose of ICI-based therapy to death from any cause. For the prognostic analysis, PFS, DOR, and OS were censored at the last known contact date, when the patient was confirmed alive, for those with no events reported before the August 2024 data cut-off. Survival curves were estimated and compared using the Kaplan–Meier method and log-rank test. Prognostic variables that were significantly associated with PFS or OS in the univariate analyses were further evaluated using multivariable Cox proportional hazards model analyses with adjusted hazard ratios (HRs) and 95% confidence intervals (CIs). Statistical significance was set at P < 0.05.
All statistical analyses were carried out using EZR statistical software (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). EZR is a modified version of the R commander designed to add specific statistical functions frequently used in biostatistics.32
Ethics approval and consent to participate
This study involves human participants and was approved by the Certified Review Board at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research (IRB number: 2021-GB-095). The protocol was described on the hospital website, and patients were provided the opportunity to opt out; therefore, no additional consent was required from patients.
Consent for publication
Consent to publish was received from all individual participants included in the study.
Results
Cohort characteristics
Ninety-two patients underwent ICI-based treatment at our institute during the study period. Overall, the median age was 66 years (range 42-88 years); 73% were male; and ECOG PS was 0, 1, and 2 in 59%, 35%, and 7% of patients, respectively (Table 1). Seventy patients (76%) had measurable lesions, and 47 (51%) had two or more metastatic sites. Prior treatments included esophagectomy and chemoradiotherapy in 26 (28%) and 38 (41%) patients, respectively. PD-L1 TPS was evaluated in 54 patients, with TPS <1% in 34 patients (63%) and ≥1% in 20 patients (37%). Sixty patients received ICI + chemo, whereas 32 received nivo + ipi.
Table 1.
Patient characteristics
| Characteristics | Total (N = 92) | Immune checkpoint inhibitor + chemotherapy (n = 60) | Nivolumab plus ipilimumab (n = 32) | P value |
|---|---|---|---|---|
| Age, years | <0.01 | |||
| Median (range) | 65.5 (42-88) | 64.5 (42-79) | 71 (48-88) | |
| Sex, n (%) | 0.81 | |||
| Male | 67 (72.8) | 43 (71.7) | 24 (75.0) | |
| Female | 25 (27.2) | 17 (28.3) | 8 (25.0) | |
| Eastern Cooperative Oncology Group performance status, n (%) | 0.06 | |||
| 0 | 54 (58.7) | 33 (55.0) | 21 (65.6) | |
| 1 | 32 (34.8) | 25 (41.7) | 7 (21.9) | |
| 2 | 6 (6.5) | 2 (3.3) | 4 (12.5) | |
| Dysphagia score, n (%) | <0.01 | |||
| 0 | 46 (50.0) | 28 (46.7) | 18 (56.3) | |
| 1 | 15 (16.3) | 8 (13.3) | 7 (21.9) | |
| 2 | 10 (10.9) | 10 (16.7) | 0 (0) | |
| 3 | 8 (8.7) | 6 (10.0) | 2 (6.3) | |
| 4 | 9 (9.8) | 8 (13.3) | 1 (3.1) | |
| Smoking history (any tobacco product, including cigarettes, cigars, or pipes), n (%) | 0.08 | |||
| Never smoker | 30 (32.6) | 21 (35.0) | 9 (28.1) | |
| Ex-smoker | 24 (26.1) | 19 (31.7) | 5 (15.6) | |
| Current smoker | 38 (41.3) | 20 (33.3) | 18 (56.3) | |
| Tumor location, n (%) | 0.42 | |||
| Cervical esophagus | 13 (14.1) | 8 (13.3) | 5 (15.6) | |
| Upper thoracic esophagus | 19 (20.7) | 13 (21.7) | 7 (21.9) | |
| Middle thoracic esophagus | 34 (37.0) | 19 (31.7) | 15 (46.9) | |
| Lower thoracic esophagus | 23 (25.0) | 18 (30.0) | 5 (15.6) | |
| Esophagogastric junction | 2 (2.2) | 2 (3.3) | 0 (0) | |
| Measurable lesions, n (%) | 0.12 | |||
| Yes | 70 (76.1) | 49 (81.7) | 21 (65.6) | |
| No | 22 (23.9) | 11 (34.4) | 11 (34.4) | |
| Baseline sum of target lesions | 0.36 | |||
| Median (range) | 23.7 (10.1-142.3) | 26.0 (10.1-142.3) | 19.5 (10.1-98.7) | |
| Disease, n (%) | 0.02 | |||
| Unresectable advanced | 17 (18.5) | 11 (18.3) | 6 (18.8) | |
| Synchronous metastatic | 47 (51.1) | 35 (58.3) | 12 (37.5) | |
| Recurrent, locoregional | 4 (4.4) | 2 (3.3) | 2 (6.3) | |
| Recurrent, distant | 24 (26.1) | 12 (20.0) | 12 (37.5) | |
| Metastatic sites, n (%) | ||||
| Liver | 18 (19.6) | 12 (20.0) | 6 (18.8) | 0.54 |
| Lymph node | 79 (85.9) | 52 (86.7) | 27 (84.4) | 0.55 |
| Lung | 23 (25.0) | 15 (25.0) | 8 (25.0) | 0.95 |
| Peritoneum | 8 (8.7) | 6 (10.0) | 2 (6.3) | 0.49 |
| Bone | 14 (15.2) | 11 (18.3) | 3 (9.4) | 0.24 |
| Number of metastases, n (%) | 0.26 | |||
| 0 | 3 (3.3) | 2 (3.3) | 1 (3.1) | |
| 1 | 42 (45.7) | 24 (40.0) | 18 (56.3) | |
| ≥2 | 47 (51.1) | 36 (60.0) | 13 (40.6) | |
| Programmed death-ligand 1 tumor proportion score, n (%) | <0.01 | |||
| <1% | 34 (37.0) | 16 (26.7) | 18 (56.3) | |
| ≥1% | 20 (21.7) | 12 (20.0) | 8 (25.0) | |
| Unknown | 38 (41.3) | 32 (53.3) | 6 (18.8) | |
| Prior treatment, n (%) | ||||
| Neoadjuvant chemotherapy | 20 (21.7) | 9 (15.0) | 11 (34.4) | 0.03 |
| Chemoradiotherapy | 38 (41.3) | 24 (40.0) | 14 (43.8) | 0.73 |
| Esophagectomy | 26 (28.3) | 10 (16.7) | 18 (56.3) | <0.01 |
| Immune checkpoint inhibitor agents, n (%) | <0.01 | |||
| Pembrolizumab | 27 (29.3) | 27 (45.0) | 0 (0) | |
| Nivolumab | 65 (70.7) | 33 (55.0) | 32 (100) | |
| Neutrophil-to-lymphocyte ratio | 0.60 | |||
| Median (range) | 4.4 (0-88.0) | 4.4 (1.0-88.0) | 4.2 (0-46.8) |
Efficacy
In the ICI + chemo group, the median PFS (mPFS) and OS were 5.0 months (95% CI 3.6-7.9 months) and 16.0 months (95% CI 8.9-21.2 months), respectively, with a median follow-up period of 22.5 months (Figure 1A and B). In the nivo + ipi group, the mPFS was 3.5 months (95% CI 1.9-4.7 months), and the median OS (mOS) was 16.9 months [95% CI 6.4-not reached (NR) months], with a median follow-up of 6.9 months (Figure 1C and D). Among the 70 patients with measurable lesions, including 49 in the ICI + chemo group and 21 in the nivo + ipi group, the ORR and DCR were 46.9% and 77.5%, respectively, in the ICI + chemo group, and 23.8% and 61.9%, respectively, in the nivo + ipi group (Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171). The median DOR was 12.9 months (95% CI 8.1-23.6 months) in the ICI + chemo group; however, this duration was NR in the nivo + ipi group (95% CI 3.6-NR months; Supplementary Table S1 and Figure S1A and B, available at https://doi.org/10.1016/j.esmogo.2025.100171). In both groups, patients who achieved CR or PR showed the longest mOS, followed by those with stable disease and PD (Supplementary Figure S1C and D, available at https://doi.org/10.1016/j.esmogo.2025.100171). In addition, pembrolizumab and nivo showed comparable clinical efficacy in the ICI + chemo group (Supplementary Figure S2, available at https://doi.org/10.1016/j.esmogo.2025.100171). Chemotherapy was administered for a median of five cycles in both the pembrolizumab and nivo groups. The number of cycles did not differ between patients with TPS >1% and those with TPS <1% in either group. Thirteen patients who received radiotherapy with target lesions located within the radiation field before ICI-based therapy were included in the study. However, similar results were obtained after excluding these patients (Supplementary Figure S3, available at https://doi.org/10.1016/j.esmogo.2025.100171).
Figure 1.
Kaplan–Meier estimates for 92 patients with advanced esophageal squamous cell carcinoma treated with an immune checkpoint inhibitor (ICI)-based therapy. (A) Progression-free survival (PFS) and (B) overall survival (OS) in ICI + chemotherapy (chemo). (C) PFS and (D) OS in nivolumab + ipilimumab (nivo + ipi). CI, confidence interval; NR; not reached.
The median baseline sum of target lesion diameters was 26.0 mm (range 10.1-142.3 mm) and 19.5 mm (range 10.1-98.7 mm) in the ICI + chemo group and nivo + ipi groups, respectively. The median maximum percentage change from baseline was −34.7% (range −100% to +96.5%) in the ICI + chemo group and −5.1% (range −49.0% to +219.2%) in the nivo + ipi group, respectively, indicating a median DpR of 34.7% and 5.1% (Figure 2). To assess prognostic outcomes, tumor reduction was categorized into four groups: DpR ≥ 50%, 30% ≤ DpR < 50%, −20% ≤ DpR < 30%, and DpR < −20%. Of the 49 patients with target lesions treated with ICI + chemo, 30.6%, 16.3%, 30.6%, and 22.4% achieved DpR ≥ 50%, 30% ≤ DpR < 50%, −20% ≤ DpR < 30%, and DpR < −20%, respectively (Supplementary Figure S4A, available at https://doi.org/10.1016/j.esmogo.2025.100171). Patients who achieved DpR ≥ 50% had the most favorable mOS, which was NR (95% CI 11.4-NR months). This was followed by an mOS of 20.7 months (95% CI 11.2-NR months) for those with DpR of 30% to <50%, 16.8 months (95% CI 6.1-NR months) for those with DpR between −20% and <30%, and 7.6 months (95% CI 1.6-18.5 months) for those with DpR < −20%. To minimize potential immortal time bias when analyzing prognostic outcomes across varying intervals for maximum tumor reduction assessment, an 18-week landmark analysis was conducted (Supplementary Figure S4B, available at https://doi.org/10.1016/j.esmogo.2025.100171). Patients with a higher DpR had a more favorable OS. Among the 21 patients with target lesions treated with nivo + ipi, 4.8%, 19.0%, 33.3%, and 42.9% achieved DpR ≥ 50%, 30% ≤ DpR < 50%, −20% ≤ DpR < 30%, and DpR < −20%, respectively (Supplementary Figure S4C and D, available at https://doi.org/10.1016/j.esmogo.2025.100171). The nivo + ipi group showed a similar trend.
Figure 2.
Depth of response for 70 patients with measurable advanced esophageal squamous cell carcinoma, with clinical features. The waterfall plots show the percentage reduction in tumor burden from baseline, with red bars representing nivolumab (nivo) + chemotherapy (chemo), blue bars indicating pembrolizumab + chemo, and green bars representing nivo + ipilimumab (nivo + ipi) treatment. DpR, depth of response; ETS, early tumor shrinkage; OS, overall survival; pem, pembrolizumab; PFS, progression-free survival; RT, prior radiotherapy; TPS, tumor proportion score.
The association between ETS and clinical outcomes was analyzed to assess the impact of early tumor reduction on the prognosis of patients treated with ICI-based regimens; 18 of the 49 (36.7%) patients with target lesions in the ICI + chemo group and 7 of the 21 (33.3%) patients in the nivo + ipi group achieved ETS (Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171). Patients who achieved ETS were significantly associated with a lower ECOG PS and reduced NLR but not with PD-L1 TPS status, metastatic sites, or treatment regimen (Supplementary Table S2, available at https://doi.org/10.1016/j.esmogo.2025.100171). The maximum tumor reduction from baseline was significantly greater in patients with ETS (median −47.8%; range −100% to 0%) than those without (median −4.7%; range −100% to +219.2%), with a strong correlation observed between ETS and the DpR (Supplementary Figure S5, available at https://doi.org/10.1016/j.esmogo.2025.100171). Most patients with DpR ≥ 50% were included in the ETS group, and 13 out of 16 (81.3%) patients had DpR ≥ 50%. In the ICI + chemo group, the ORR and DCR were 88.9% and 100%, respectively, in the ETS group and 20.0% and 63.3%, respectively, in the non-ETS group (Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171). The mPFS was 14.3 months in the ETS group (95% CI 5.1-23.6 months), which was significantly longer than the 3.6 months (95% CI 2.3-4.9 months) in the non-ETS group (P = 0.01; Figure 3A). The mOS was also prolonged in the ETS group, which was NR (95% CI 11.7-NR months), compared with the 11.2 months (95% CI 7.6-18.5 months) in the non-ETS group (P = 0.01; Figure 3B). Sensitivity analysis was conducted to evaluate the role of ETS in patients who experienced clinical benefits (Supplementary Figure S6A and B, available at https://doi.org/10.1016/j.esmogo.2025.100171). Even when patients were categorized based on the presence of ETS and the best overall response (ETS, non-ETS, and non-PD, and non-ETS and PD), the survival benefit of ETS remained evident. In the nivo + ipi group, the ORR and DCR were 71.4% and 100%, respectively, in the ETS group and 0% and 42.9%, respectively, in the non-ETS group (Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171). Patients who achieved ETS showed a significantly greater survival benefit than those who did not (Figure 3C and D and Supplementary Figure S6C and D, available at https://doi.org/10.1016/j.esmogo.2025.100171). In the multivariable Cox proportional hazards analysis, ETS was significantly associated with prolonged PFS (HR 0.34, 95% CI 0.15-0.78, P = 0.01; Supplementary Table S3, available at https://doi.org/10.1016/j.esmogo.2025.100171) and OS (HR 0.34, 95% CI 0.11-0.88, P = 0.04; Table 2). In addition, ECOG PS, dysphagia score, and liver metastasis were independent prognostic factors for OS but not PFS. Neither the treatment regimen nor the PD-L1 TPS status had a significant impact on PFS or OS.
Figure 3.
Kaplan–Meier estimates for immune checkpoint inhibitor (ICI)-based therapy according to the early tumor shrinkage (ETS) status. (A) Progression-free survival (PFS) and (B) overall survival (OS) in ICI + chemotherapy (chemo). (C) PFS and (D) OS in nivolumab + ipilimumab (nivo + ipi). CI, confidence interval; NR; not reached.
Table 2.
Cox proportional hazard model analysis for overall survival
| Variables | Univariate |
Multivariate |
||||
|---|---|---|---|---|---|---|
| Hazards ratio | 95% Confidence interval | P value | Hazards ratio | 95% Confidence interval | P value | |
| Sex (male versus female) | 0.79 | 0.41-1.51 | 0.48 | |||
| Age (≥65 versus <65 years) | 1.33 | 0.71-2.49 | 0.38 | |||
| Eastern Cooperative Oncology Group performance status (≥1 versus 0) | 2.94 | 1.57-5.52 | <0.01 | 2.33 | 1.08-5.00 | 0.03 |
| Dysphagia score | ||||||
| 0 | Reference | — | — | Reference | — | — |
| 1-2 | 2.33 | 1.12-4.86 | 0.02 | 1.88 | 0.81-4.36 | 0.14 |
| 3-4 | 4.29 | 1.99-9.27 | <0.01 | 2.97 | 1.17-7.53 | 0.02 |
| Smoking history (any tobacco product, including cigarettes, cigars, or pipes) | ||||||
| Never smoker | Reference | — | — | |||
| Ex-smoker | 1.02 | 047-2.22 | 0.97 | |||
| Current smoker | 1.06 | 052-2.18 | 0.87 | |||
| Tumor location | ||||||
| Cervical esophagus/upper thoracic esophagus | Reference | — | — | |||
| Middle thoracic esophagus | 1.29 | 0.63-2.66 | 0.48 | |||
| Lower thoracic esophagus/esophagogastric junction | 1.09 | 0.50-2.37 | 0.83 | |||
| Metastasis (metachronous versus synchronous) | 0.74 | 0.35-1.56 | 0.43 | |||
| Recurrence (distant versus locoregional) | 5.43 × 108 | 0-not reached | 1 | |||
| Prior esophagectomy (yes versus no) | 0.81 | 0.39-1.65 | 0.56 | |||
| Prior radiotherapy (yes versus no) | 0.74 | 0.40-1.38 | 0.35 | |||
| Programmed death-ligand 1 tumor proportion score (≥1% versus <1%) | 0.95 | 0.36-2.47 | 0.92 | |||
| Liver metastasis (yes versus no) | 3.76 | 1.83-7.72 | <0.01 | 2.90 | 1.18-7.10 | 0.02 |
| Lung metastasis (yes versus no) | 1.05 | 0.54-2.04 | 0.88 | |||
| Peritoneal metastasis (yes versus no) | 2.69 | 0.93-7.75 | 0.07 | |||
| Lymph node metastasis (yes versus no) | 0.93 | 0.36-2.38 | 0.88 | |||
| Number of metastasis | ||||||
| 0 | Reference | — | — | |||
| 1 | 0.30 | 0.07-1.38 | 0.12 | |||
| ≥2 | 0.70 | 0.16-2.95 | 0.63 | |||
| Treatment regimen (immune checkpoint inhibitor + chemotherapy versus nivolumab plus ipilimumab) | 1.17 | 0.58-2.38 | 0.66 | |||
| Immune checkpoint inhibitor agent (pembrolizumab versus nivolumab) | 1.09 | 0.55-2.15 | 0.80 | |||
| Early tumor shrinkage (yes versus no) | 0.21 | 0.07-0.60 | <0.01 | 0.34 | 0.11-0.88 | 0.04 |
| Neutrophil-to-lymphocyte ratio (≥ median 4.4 versus < median 4.4) | 1.70 | 0.91-3.16 | 0.10 | — | — | — |
Among the 53 patients in the ICI + chemo group and 26 patients in the nivo + ipi group who discontinued their respective treatments, 40% and 54%, respectively, received subsequent chemotherapy (Supplementary Table S4, available at https://doi.org/10.1016/j.esmogo.2025.100171). The PFS for paclitaxel (Taxol; Nippon Kayaku, Tokyo, Japan) treatment was 3.3 months (95% CI 1.7-5.5 months) in the ICI + chemo group and 3.5 months (95% CI 1.2-6.6 months) in the nivo + ipi group.
Programmed death-ligand 1 tumor proportion score analysis
The treatment efficacy was analyzed according to PD-L1 TPS status. In the ICI + chemo group, PFS did not differ significantly between patients with PD-L1 TPS ≥1% and those with TPS <1% [mPFS, 11.0 months (95% CI 1.3-16.0 months) versus 5.1 months (95% CI 2.1-9.5 months); P = 0.44; Figure 4A]. Similarly, OS did not differ significantly between patients with TPS ≥1% and those with TPS <1% [mOS, 19.8 months (95% CI 7.7-NR months) versus NR (95% CI 7.0-NR months); P = 0.29]. Notably, patients with TPS >1% had a 1-year OS rate of 68.7%, and those with TPS <1% had a rate of 72.7% (Figure 4B). Comparable outcomes were observed for the TPS status (TPS ≥1% versus TPS <1%) in terms of ORR (60.0% versus 53.3%), DCR (70.0% versus 80.0%), and ETS (50.0% versus 46.7%; Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171). In the nivo + ipi group, PFS was significantly longer for patients with TPS ≥1% compared with those with TPS <1% [mPFS, 8.5 months (95% CI 2.3-NR months) versus 2.6 months (95% CI 1.6-4.4 months); P = 0.01; Figure 4C]. The OS did not differ significantly between patients with TPS ≥1% and <1% [mOS, NR months (95% CI 6.4-NR months) versus 16.9 months (95% CI 3.6-NR months); P = 0.20]; however, the 1-year OS rates were 80.0% for those with TPS ≥1% and 51.7% for those with TPS <1% (Figure 4D). The ORR (66.7% versus 20.0%), DCR (100% versus 60.0%), and ETS (66.7% versus 33.3%) were favorable in patients with TPS ≥1% compared with those with TPS <1% (Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2025.100171).
Figure 4.
Kaplan–Meier estimates for immune checkpoint inhibitor (ICI)-based therapy according to the programmed death-ligand 1 (PD-L1) tumor proportion score (TPS) status. (A) Progression-free survival (PFS) and (B) overall survival (OS) in ICI + chemotherapy (chemo). (C) PFS and (D) OS in nivolumab + ipilimumab (nivo + ipi). CI, confidence interval; ETS, early tumor shrinkage; NR, not reached.
Safety
Treatment-related AEs with potential immunological etiologies in the study population are summarized in Supplementary Table S5, available at https://doi.org/10.1016/j.esmogo.2025.100171. Overall, 36 (39.1%) and 11 (12.0%) patients had AEs of any grade and grade ≥3, respectively. The most common AEs were hypothyroidism, skin toxicity, adrenal insufficiency, colitis, pneumonia, and hepatitis, which occurred in 11 (12.0%), 7 (7.6%), 5 (5.4%), 4 (4.3%), 4 (4.3%), and 4 (4.3%) patients, respectively. The most common grade ≥3 AEs were colitis, skin toxicity, pneumonia, hemolytic anemia, and myositis, which occurred in 3 (3.3%), 2 (2.2%), 2 (2.2%), 2 (2.2%), and 2 (2.2%) patients, respectively. The incidence of AEs was considerably higher in the nivo + ipi group than in the ICI + chemo group. In total, 59.4% of the patients in the nivo + ipi group versus 28.3% in the ICI + chemo group experienced any-grade AEs. In addition, grade ≥3 AEs were observed in 21.9% of patients in the nivo + ipi group compared with 6.7% in the ICI + chemo group. In the nivo + ipi group, several rare AEs were observed, including myositis in two patients and hemolytic anemia in one patient. Four patients discontinued the ICI + chemo treatment because of toxicities, including hepatitis, adrenal insufficiency, and pneumonia. Ten patients discontinued nivo + ipi treatment because of immune-related AEs, including hepatitis, pneumonia, myositis, colitis, and cystitis. No treatment-related AEs or deaths were observed and the frequency of immune-related AEs was not influenced by the PD-L1 TPS status.
Discussion
This study evaluated the efficacy and safety of ICI-based therapies for patients with ESCC in real-world clinical practice. The key findings were as follows: (i) ICI-based therapies were practically efficacious and manageably safe; (ii) PD-L1 TPS status was a moderate predictor of nivo + ipi treatment efficacy but not for ICI + chemo; and (iii) ETS was associated with high tumor reduction, durable response, and improved OS, regardless of the treatment regimen. Patients with ETS were more likely to have a lower ECOG PS and NLR. These findings provide insights that will facilitate informed and patient-centered decision making.
Patients who would not have been enrolled in clinical trials due to comorbidities or frailty might still receive these treatment regimens in real-world settings. Therefore, real-world data are required to guide therapeutic decisions and optimize treatment. The median age in this study was similar to that in the phase III CheckMate 648 and KEYNOTE-590 trials12,13; however, 6.5% of patients had an ECOG PS of 2, despite the exclusion of patients with a PS of ≥2 in these phase III trials. Furthermore, 18.5% of the patients in this study experienced difficulties with oral intake, as indicated by a dysphagia score ≥3, which contributed to poor nutritional status and was identified as an independent poor prognostic factor for OS in the multivariate analysis. Nevertheless, the clinical outcomes in this study were consistent with those reported in phase III trials,12,13 and the safety profiles of the ICI-based therapies observed in this study were generally consistent with the established profile. The differences between randomized trials and real-world data must be recognized. Notably, this study provides data that support clinically meaningful efficacy and a manageable safety profile in clinical practice, providing compelling evidence for ICI-based therapy as a valuable first-line treatment regimen.
The clinical benefits of ICI are for a limited number of patients with ESCC,2 highlighting the need for robust predictive biomarkers. PD-L1 expression status is among the candidate biomarkers for ESCC, with higher expression levels being associated with survival benefits.9,12,13 In a meta-analysis of individual patient data from first-line phase III trials that evaluated PD-L1 expression based on TPS (CheckMate-648 and ESCORT-1st), there was no survival benefit for ICI-based regimens in the first-line setting compared with chemotherapy alone in the TPS <1% subgroup.33 However, its role as a reliable biomarker is limited because of its heterogeneity34 and the complexities introduced by both cancer cells and the tumor immune microenvironment.34, 35, 36 Notably, the clinical benefits of ICIs have been observed even in populations with a PD-L1 TPS <1%.11,37, 38, 39, 40 The addition of nivo to chemotherapy resulted in a higher ORR than chemotherapy alone in patients with PD-L1 TPS <1%.13 Although a meta-analysis of five phase III trials for ESCC also demonstrated that PD-L1 TPS was the strongest predictor of OS benefits associated with ICI, the treatment with ICI proved OS benefits even in the low TPS subgroup, defined as TPS <1%, with an HR of 0.84 (95% CI 0.75-0.95).37 In addition, in a meta-analysis of five first-line phase III trials (KEYNOTE-590, CheckMate 648, ESCORT-1st, JUPITER-06, and ORIENT-15), the addition of anti-PD-1 antibody to chemotherapy showed significant OS benefits in both the TPS <1% (HR 0.74, 95% CI 0.56-0.97) and CPS <10 (HR 0.77, 95% CI 0.66-0.89) subgroups, with improved PFS and ORR,38 consistent with findings of other meta-analyses.39,40 Thus, patients with PD-L1 TPS ≥1% experienced a greater clinical benefit with ICI + chemo compared with chemotherapy alone. However, whether chemotherapy alone should be considered for patients with a TPS <1% instead of ICI-based therapies as a first-line therapy remains debatable. Furthermore, data on the clinical practice utility of ICI-based therapies based on PD-L1 TPS status, especially in patients with TPS <1%, are lacking. Based on the results of the phase III CheckMate 648 trial, which demonstrated that patients with TPS <1% had similar OS compared with those receiving chemotherapy alone,13 PD-L1 TPS testing was approved in Japan as a complementary diagnostic tool for nivo-based treatment. The Japanese guidelines recommend nivo-based regimens as first-line therapy, emphasizing the importance of the PD-L1 TPS status, although testing is not mandatory.7 Thus, there is an opportunity to introduce ICI-based therapies for patients with a TPS <1% in Japan. In this study, patients with a PD-L1 TPS ≥1% showed a trend toward improved PFS compared with those with a TPS <1% in the ICI + chemo group. However, OS, ORR, and DCR were similar irrespective of PD-L1 TPS status. Overall, our findings may challenge the decision to use ICI-based therapies over chemotherapy alone based only on PD-L1 TPS status. However, it remains unclear whether the additive effect of ICIs on chemotherapy is consistent across different PD-L1 TPS statuses, as PD-L1 expression may negatively affect the efficacy of first-line chemotherapy alone.41 A combined analysis of pembrolizumab and nivo based on PD-L1 TPS status may not be appropriate. Therefore, we carried out separate subanalyses for pembrolizumab and nivo. When combined with chemotherapy, both pembrolizumab and nivo showed similar therapeutic effects (Figure 2 and Supplementary Figure S2, available at https://doi.org/10.1016/j.esmogo.2025.100171). Although the small sample size limited the statistical power of the study, pembrolizumab and nivo showed a similar trend in the relationship between TPS status and treatment efficacy (data not shown). The relationship between CPS status and treatment efficacy was not evaluated in this study. Notably, in the nivo + ipi group, patients with PD-L1 TPS ≥1% demonstrated significantly prolonged PFS (median, 8.5 months versus 2.6 months) and a trend toward improved OS (median NR versus 16.9 months), compared with those with a TPS <1%. Furthermore, the ORR, DCR, and incidence of non-PD at the initial evaluation were also improved in cases with a PD-L1 TPS ≥1%. These findings suggest that PD-L1 TPS status may play a modest role as a predictive biomarker when considering treatment with nivo + ipi; however, it may still be inadequate as a reliable biomarker for ICI + chemo in ESCC cases.
ETS has emerged as a promising early on-treatment predictor of treatment efficacy for various cancer types.21, 22, 23, 24, 25, 26, 27 However, real-world data on the clinical utility of ETS in patients receiving ICI + chemo remain limited.29 Furthermore, there are currently no data available on ETS in patients treated with nivo + ipi as a chemotherapy-free regimen. In this study, ETS independently and effectively stratified prognostic outcomes among real-world patients treated with ICI-based therapies, irrespective of the PD-L1 TPS status and treatment regimen. Notably, ETS was achieved at a similar rate despite the different ORRs between the ICI + chemo and nivo + ipi groups. Furthermore, most patients who achieved ETS experienced substantial tumor reduction. In the ICI + chemo group, patients who achieved ETS had an mPFS of 14.3 months, compared with 3.6 months in those who did not. In the nivo + ipi group, the mPFS was NR for patients with ETS, whereas it was 1.4 months for those without ETS. Therefore, those who achieve ETS are likely to experience a durable response. By contrast, for patients without ETS, close monitoring and timely imaging are critical to facilitate the early initiation of subsequent therapies. Additionally, patients who achieved ETS were significantly associated with a lower ECOG PS and NLR. Tumor-associated inflammation plays a crucial role in cancer development by influencing the host’s immune status.42 The NLR represents the balance between a protumor inflammatory state and an antitumor immune response and has been reported as a predictive biomarker for ICI in patients with ESCC.43 Patients who achieved ETS were more likely to have both a favorable general condition and an antitumor immune response, as evidenced by the fact that ETS was observed in 56% of patients with both PS0 and low NLR, compared with only 5% in those with PS ≥1 and high NLR. Collectively, the ETS provides early insights into treatment efficacy, informs clinical decision making, and optimizes patient outcomes. Further studies are required to establish standardized criteria, although the incorporation of ETS metrics into clinical practice could promote a more patient-centered approach.
The nivo + ipi regimen showed a low ORR, a high proportion of PD at the initial evaluation, and shorter PFS than the ICI + chemo regimen. However, the initiation rates of subsequent chemotherapies were 54% and 40% in the nivo + ipi group and ICI + chemo group, respectively. The mOS was comparable to that of the ICI + chemo group, highlighting the importance of an appropriate transition to subsequent therapies before clinical deterioration occurs. This approach may explain the lack of a significant difference in OS between patients treated with ICI + chemo and those treated with nivo + ipi. Post hoc exploratory multivariate analyses from the CheckMate 648 trial revealed that several factors, including a large baseline tumor burden and liver metastasis, were associated with the delayed benefit of the nivo + ipi treatment.44 In this study, the nivo + ipi regimen was more commonly used than the ICI + chemo regimen in patients who had previously undergone esophagectomy or had a smaller tumor burden at baseline. In addition, liver metastasis was an independent prognostic factor for OS. Thus, the nivo + ipi regimen may be the optimal choice for patients who are expected to benefit from treatment, such as those with a PD-L1 TPS ≥1% and no liver metastasis, and who are likely to transition to subsequent therapies, such as those with smaller tumor burdens or asymptomatic disease. Notably, the nivo + ipi treatment group experienced immune-related AEs at more than twice the frequency observed in the ICI + chemo group. This suggests the need for careful management of these AEs, especially with the nivo + ipi treatment.
This study had some limitations. Firstly, this was conducted as a single-center retrospective analysis at a cancer-specialized hospital in Japan, potentially introducing selection bias and limiting the generalizability of our findings. The retrospective and exploratory nature of this study restricts the interpretability of the results. Because of this design, AEs may not have been fully captured in the medical records. Secondly, the relatively small sample size limited the statistical power of the study. The follow-up period in the nivo + ipi group was relatively short compared with that in the ICI + chemo group, which may have limited the number of OS events. This difference could hinder the interpretation of the results based on the treatment regimen, emphasizing the need for extended observations in this subgroup. Thirdly, this study lacked data on PD-L1 CPS status and had a substantial amount of missing data on PD-L1 TPS status. In the phase III KEYNOTE-590 trial, pembrolizumab combined with chemotherapy significantly improved PFS and OS compared with chemotherapy alone.12 While patients with ESCC and a high PD-L1 CPS experienced the greatest OS benefit, all patients with ESCC also showed modest improvements (HR 0.72, 95% CI 0.60-0.88). As a result, this regimen was approved in Japan as a first-line treatment for all patients with ESCC in November 2021, regardless of PD-L1 CPS status. In the Japanese guidelines for EC, CPS testing is not required, even as a complementary diagnostic tool, for the first-line pembrolizumab + chemotherapy regimen.7 Therefore, we could not assess the CPS status for pembrolizumab + chemotherapy in clinical practice. The results of the phase III CheckMate 648 trial13 subsequently led to the approval of nivo + chemotherapy as well as nivo + ipi, as first-line treatments for patients with ESCC in Japan, along with a complementary diagnostic tool for nivo-based therapies, in May 2022. Owing to the retrospective design of this study, we were unable to assess the PD-L1 expression status in the first-line setting before the approval of PD-L1 TPS testing, particularly in patients who received pembrolizumab + chemotherapy before its approval. These limitations highlight the need for further validation and large-scale studies; however, our findings offer valuable insights into the potential benefits of ICI-based first-line treatments.
In conclusion, we demonstrated the real-world clinical efficacy and manageable safety profiles of first-line ICI-based therapies in patients with ESCC. Rapid and deep tumor shrinkage may serve as an early predictive biomarker for a patient’s longer survival outcomes. In addition, PD-L1 TPS status may provide moderate guidance for treatment decisions for nivo + ipi. However, its clinical utility appears limited for ICI + chemo.
Acknowledgements
We are deeply indebted to the patients who participated in this study and to their families. The authors thank Ms. Yuki Horiike, Ms. Hitomi Hannan, and Ms. Yukie Naito for providing data management.
Funding
None declared.
Disclosure
The authors have declared no conflicts of interest.
Data sharing
The data used for this study, although not available in a public repository, will be made available to other researchers upon reasonable request.
Supplementary data
References
- 1.Sung H., Ferlay J., Siegel R.L., et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209–249. doi: 10.3322/caac.21660. [DOI] [PubMed] [Google Scholar]
- 2.Ooki A., Osumi H., Chin K., Watanabe M., Yamaguchi K. Potent molecular-targeted therapies for advanced esophageal squamous cell carcinoma. Ther Adv Med Oncol. 2023;15 doi: 10.1177/17588359221138377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Arnold M., Ferlay J., van Berge Henegouwen M.I., Soerjomataram I. Global burden of oesophageal and gastric cancer by histology and subsite in 2018. Gut. 2020;69:1564–1571. doi: 10.1136/gutjnl-2020-321600. [DOI] [PubMed] [Google Scholar]
- 4.Smyth E.C., Lagergren J., Fitzgerald R.C., et al. Oesophageal cancer. Nat Rev Dis Primers. 2017;3 doi: 10.1038/nrdp.2017.48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.The National Comprehensive Cancer Network NCCN Clinical Practice Guidelines in Oncology, Esophageal and Esophagogastric Junction Cancer. https://www.nccn.org/professionals/physician_gls/pdf/esophageal.pdf Available at. version 1.
- 6.Muro K., Lordick F., Tsushima T., et al. Pan-Asian adapted ESMO Clinical Practice Guidelines for the management of patients with metastatic oesophageal cancer: a JSMO-ESMO initiative endorsed by CSCO, KSMO, MOS, SSO and TOS. Ann Oncol. 2019;30:34–43. doi: 10.1093/annonc/mdy498. [DOI] [PubMed] [Google Scholar]
- 7.Kitagawa Y., Ishihara R., Ishikawa H., et al. Esophageal cancer practice guidelines 2022 edited by the Japan esophageal society: part 1. Esophagus. 2023;20:343–372. doi: 10.1007/s10388-023-00993-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ooki A., Shinozaki E., Yamaguchi K. Immunotherapy in colorectal cancer: current and future strategies. J Anus Rectum Colon. 2021;5:11–24. doi: 10.23922/jarc.2020-064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kojima T., Shah M.A., Muro K., et al. Randomized phase III KEYNOTE-181 study of pembrolizumab versus chemotherapy in advanced esophageal cancer. J Clin Oncol. 2020;38:4138–4148. doi: 10.1200/JCO.20.01888. [DOI] [PubMed] [Google Scholar]
- 10.Salem M.E., Puccini A., Xiu J., et al. Comparative molecular analyses of esophageal squamous cell carcinoma, esophageal adenocarcinoma, and gastric adenocarcinoma. Oncologist. 2018;23:1319–1327. doi: 10.1634/theoncologist.2018-0143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kato K., Cho B.C., Takahashi M., et al. Nivolumab versus chemotherapy in patients with advanced oesophageal squamous cell carcinoma refractory or intolerant to previous chemotherapy (ATTRACTION-3): a multicentre, randomised, open-label, phase 3 trial. Lancet Oncol. 2019;20:1506–1517. doi: 10.1016/S1470-2045(19)30626-6. [DOI] [PubMed] [Google Scholar]
- 12.Sun J.M., Shen L., Shah M.A., et al. Pembrolizumab plus chemotherapy versus chemotherapy alone for first-line treatment of advanced oesophageal cancer (KEYNOTE-590): a randomised, placebo-controlled, phase 3 study. Lancet. 2021;398:759–771. doi: 10.1016/S0140-6736(21)01234-4. [DOI] [PubMed] [Google Scholar]
- 13.Doki Y., Ajani J.A., Kato K., et al. Nivolumab combination therapy in advanced esophageal squamous-cell carcinoma. N Engl J Med. 2022;386:449–462. doi: 10.1056/NEJMoa2111380. [DOI] [PubMed] [Google Scholar]
- 14.Luo H., Lu J., Bai Y., et al. Effect of camrelizumab vs placebo added to chemotherapy on survival and progression-free survival in patients with advanced or metastatic esophageal squamous cell carcinoma: the ESCORT-1st randomized clinical trial. JAMA. 2021;326:916–925. doi: 10.1001/jama.2021.12836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xu J., Kato K., Raymond E., et al. Tislelizumab plus chemotherapy versus placebo plus chemotherapy as first-line treatment for advanced or metastatic oesophageal squamous cell carcinoma (RATIONALE-306): a global, randomised, placebo-controlled, phase 3 study. Lancet Oncol. 2023;24:483–495. doi: 10.1016/S1470-2045(23)00108-0. [DOI] [PubMed] [Google Scholar]
- 16.Wang Z.X., Cui C., Yao J., et al. Toripalimab plus chemotherapy in treatment-naïve, advanced esophageal squamous cell carcinoma (JUPITER-06): a multi-center phase 3 trial. Cancer Cell. 2022;40:277–288.e273. doi: 10.1016/j.ccell.2022.02.007. [DOI] [PubMed] [Google Scholar]
- 17.Lu Z., Wang J., Shu Y., et al. Sintilimab versus placebo in combination with chemotherapy as first line treatment for locally advanced or metastatic oesophageal squamous cell carcinoma (ORIENT-15): multicentre, randomised, double blind, phase 3 trial. BMJ. 2022;377 doi: 10.1136/bmj-2021-068714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Song Y., Zhang B., Xin D., et al. First-line serplulimab or placebo plus chemotherapy in PD-L1-positive esophageal squamous cell carcinoma: a randomized, double-blind phase 3 trial. Nat Med. 2023;29:473–482. doi: 10.1038/s41591-022-02179-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zappasodi R., Merghoub T., Wolchok J.D. Emerging concepts for immune checkpoint blockade-based combination therapies. Cancer Cell. 2018;33:581–598. doi: 10.1016/j.ccell.2018.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Obermannová R.L., Leong T. ESMO Clinical Practice Guideline interim update on the treatment of locally advanced oesophageal and oesophagogastric junction adenocarcinoma and metastatic squamous-cell carcinoma. ESMO Open. 2025;10 doi: 10.1016/j.esmoop.2025.104134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Heinemann V., Stintzing S., Modest D.P., Giessen-Jung C., Michl M., Mansmann U.R. Early tumour shrinkage (ETS) and depth of response (DpR) in the treatment of patients with metastatic colorectal cancer (mCRC) Eur J Cancer. 2015;51:1927–1936. doi: 10.1016/j.ejca.2015.06.116. [DOI] [PubMed] [Google Scholar]
- 22.Osumi H., Takahari D., Shinozaki E., et al. Associations between early tumor shrinkage and depth of response and clinical outcomes in patients treated with 1st-line chemotherapy for advanced gastric cancer. Gastric Cancer. 2018;21:267–275. doi: 10.1007/s10120-017-0729-2. [DOI] [PubMed] [Google Scholar]
- 23.Piessevaux H., Buyse M., Schlichting M., et al. Use of early tumor shrinkage to predict long-term outcome in metastatic colorectal cancer treated with cetuximab. J Clin Oncol. 2013;31:3764–3775. doi: 10.1200/JCO.2012.42.8532. [DOI] [PubMed] [Google Scholar]
- 24.Stintzing S., Modest D.P., Rossius L., et al. FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab for metastatic colorectal cancer (FIRE-3): a post-hoc analysis of tumour dynamics in the final RAS wild-type subgroup of this randomised open-label phase 3 trial. Lancet Oncol. 2016;17:1426–1434. doi: 10.1016/S1470-2045(16)30269-8. [DOI] [PubMed] [Google Scholar]
- 25.Ooki A., Morita S., Tsuji A., et al. Impact of early tumor shrinkage on quality of life in patients treated with first-line cetuximab plus chemotherapy for unresectable metastatic colorectal cancer: results of phase II QUACK trial. BMC Cancer. 2022;22:711. doi: 10.1186/s12885-022-09811-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Shimozaki K., Fukuda K., Ooki A., et al. Safety and efficacy of first-line nivolumab plus chemotherapy for HER2-negative advanced gastric or gastroesophageal junction adenocarcinoma: real-world data analysis. ESMO Gastrointest Oncol. 2024;5 doi: 10.1016/j.esmogo.2024.100072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ura T., Hironaka S., Tsubosa Y., et al. Early tumor shrinkage and depth of response in patients with metastatic esophageal cancer treated with 2-weekly docetaxel combined with cisplatin plus fluorouracil: an exploratory analysis of the JCOG0807. Esophagus. 2023;20:272–280. doi: 10.1007/s10388-022-00968-9. [DOI] [PubMed] [Google Scholar]
- 28.Kato K., Kojima T., Hara H., et al. First-line pembrolizumab plus chemotherapy for advanced/metastatic esophageal cancer: 1-year extended follow-up in the Japanese subgroup of the phase 3 KEYNOTE-590 study. Esophagus. 2024;21:306–318. doi: 10.1007/s10388-024-01053-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sugase T., Kanemura T., Takeoka T., et al. Clinical impact of early tumour shrinkage in metastatic or unresectable oesophageal cancer treated with pembrolizumab plus chemotherapy. Oncology. 2024;102:484–493. doi: 10.1159/000535186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gao L., Tang L., Peng J., Hu Z., Yang J., Liu B. PD-1 inhibitor combined with chemotherapy for first-line treatment of esophageal squamous cell carcinoma patients with distant metastasis: a real-world retrospective study. Front Immunol. 2024;15 doi: 10.3389/fimmu.2024.1353445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Mellow M.H., Pinkas H. Endoscopic laser therapy for malignancies affecting the esophagus and gastroesophageal junction. Analysis of technical and functional efficacy. Arch Intern Med. 1985;145:1443–1446. [PubMed] [Google Scholar]
- 32.Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant. 2013;48:452–458. doi: 10.1038/bmt.2012.244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Yap D.W.T., Leone A.G., Wong N.Z.H., et al. Effectiveness of immune checkpoint inhibitors in patients with advanced esophageal squamous cell carcinoma: a meta-analysis including low PD-L1 subgroups. JAMA Oncol. 2023;9:215–224. doi: 10.1001/jamaoncol.2022.5816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Yan T., Cui H., Zhou Y., et al. Multi-region sequencing unveils novel actionable targets and spatial heterogeneity in esophageal squamous cell carcinoma. Nat Commun. 2019;10:1670. doi: 10.1038/s41467-019-09255-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Dinh H.Q., Pan F., Wang G., et al. Integrated single-cell transcriptome analysis reveals heterogeneity of esophageal squamous cell carcinoma microenvironment. Nat Commun. 2021;12:7335. doi: 10.1038/s41467-021-27599-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Zheng Y., Chen Z., Han Y., et al. Immune suppressive landscape in the human esophageal squamous cell carcinoma microenvironment. Nat Commun. 2020;11:6268. doi: 10.1038/s41467-020-20019-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yoon H.H., Jin Z., Kour O., et al. Association of PD-L1 expression and other variables with benefit from immune checkpoint inhibition in advanced gastroesophageal cancer: systematic review and meta-analysis of 17 phase 3 randomized clinical trials. JAMA Oncol. 2022;8:1456–1465. doi: 10.1001/jamaoncol.2022.3707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wu H.X., Pan Y.Q., He Y., et al. Clinical benefit of first-line programmed death-1 antibody plus chemotherapy in low programmed cell death ligand 1-expressing esophageal squamous cell carcinoma: a post hoc analysis of JUPITER-06 and meta-analysis. J Clin Oncol. 2023;41:1735–1746. doi: 10.1200/JCO.22.01490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Gao T.T., Shan J.H., Yang Y.X., et al. Comparative efficacy and safety of immunotherapy for patients with advanced or metastatic esophageal squamous cell carcinoma: a systematic review and network meta-analysis. BMC Cancer. 2022;22:992. doi: 10.1186/s12885-022-10086-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Chen W., Cao K., Zhang L., et al. Efficacy and safety evaluation of frontline immunotherapy combinations in advanced esophageal squamous cell carcinoma: a network meta-analysis highlighting the value of PD-L1 expression positivity scores. Front Immunol. 2024;15 doi: 10.3389/fimmu.2024.1414753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sachdeva M., Zhao J.J., Zhu K.X., et al. PD-L1 expression as a negative predictive biomarker in advanced esophageal squamous-cell cancer treated with chemotherapy alone. ESMO Gastrointest Oncol. 2025;7 doi: 10.1016/j.esmogo.2024.100109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Grivennikov S.I., Greten F.R., Karin M. Immunity, inflammation, and cancer. Cell. 2010;140:883–899. doi: 10.1016/j.cell.2010.01.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Xu J., Li Y., Fan Q., et al. Clinical and biomarker analyses of sintilimab versus chemotherapy as second-line therapy for advanced or metastatic esophageal squamous cell carcinoma: a randomized, open-label phase 2 study (ORIENT-2) Nat Commun. 2022;13:857. doi: 10.1038/s41467-022-28408-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Chau I., Ajani J., Doki Y., et al. O-3 Nivolumab (NIVO) plus chemotherapy (chemo) or ipilimumab (IPI) vs chemo as first-line treatment for advanced esophageal squamous cell carcinoma (ESCC): expanded efficacy and safety analyses from CheckMate 648. Ann Oncol. 2022;33:S379–S380. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.