Supplementary Table S1. RATIONALE-213 ethics approval. Supplementary Table S2. Representativeness of study participants. Supplementary Table S3. Summary of treatment exposure to tislelizumab. Supplementary Table S4. Summary of treatment exposure to chemotherapy. Supplementary Table S5. Summary of treatment exposure to radiotherapy. Supplementary Table S6. Baseline characteristics of surgical and nonsurgical patients by PET response status. Supplementary Table S7. Summary of surgery (safety analysis set). Supplementary Table S8. Subgroup analyses of pathological response by baseline clinical characteristics. Supplementary Table S9. Analysis of pathological response assessed by local pathologist by baseline PD-L1 status (efficacy-evaluable analysis set). Supplementary Table S10. Analysis of disease response by investigator per RECIST v1.1 (safety analysis set with measurable disease at baseline). Supplementary Table S11. Disease recurrence by PET response status. Supplementary Table S12. TRAEs with ≥10% incidence (safety analysis set). Supplementary Table S13. Summary of surgery-relevant adverse events (efficacy-evaluable analysis set). Supplementary Table S14. Summary of all deaths (safety analysis set). Supplementary Table S15. Baseline demographic and clinical characteristics (efficacy- and GEP-evaluable populations).

BACKGROUND:The optimal extent of lymphadenectomy following neoadjuvant chemoimmunotherapy (nCIT) for esophageal squamous cell carcinoma (ESCC) remains unclear. Current recommendations are largely derived from neoadjuvant chemoradiotherapy cohorts, and their applicability in the era of immunotherapy is unclear. This study evaluated the association between lymph node dissection (LND) yield and disease-free survival (DFS) in patients with ESCC treated with nCIT. METHODS:This retrospective multicenter cohort study included 465 patients with ESCC who underwent nCIT followed by radical esophagectomy at six hospitals in China between January 2019 and December 2023. The median follow-up was 40.7 months. The total number of dissected lymph nodes was analyzed in relation to DFS. The restricted mean survival time at 5 years was estimated by using random survival Forest-based models, with subgroup analyses by posttherapy pathological stage (ypT/ypN). RESULTS:A higher lymph node yield was associated with improved DFS, although the relationship was nonlinear and varied by pathological subgroup. In patients with ypT0-2N0 disease, DFS improved with increasing LND up to approximately 20-30 lymph nodes, after which the benefit plateaued. In patients with residual nodal disease (ypN1-3), higher lymph node yields were associated with longer DFS, with greater yields observed in more advanced disease stages. Across all subgroups, lower lymph node yields were consistently associated with inferior DFS. CONCLUSIONS:In this multicenter cohort of patients with ESCC treated with nCIT, lymph node yield was associated with DFS in a stage-dependent manner, suggesting its potential role as a postoperative quality indicator for surgical lymphadenectomy after nCIT.

Background:Neoadjuvant chemoimmunotherapy improves pathological outcomes and survival in resectable non-small cell lung cancer (NSCLC), but its effect on quality of life and the association between pathological response and patient-reported outcomes (PROs) are not well established. This analysis from phase II of the TD-FOREKNOW trial explores the preliminary relationship between pathologic complete response (pCR) and PROs. Methods:In this randomized phase 2 trial, patients with stage IIIA-IIIB (T3N2) NSCLC received camrelizumab plus chemotherapy or chemotherapy alone followed by surgery. PROs [global health status/quality of life (GHS/QoL), cough, dyspnea, hemoptysis] were compared between pCR and non-pCR groups using change-from-baseline analyses. Results:A total of 94 patients were randomized. Among 40 patients who received camrelizumab plus chemotherapy followed by surgery and completed ≥1 PRO assessment, 14 achieved pCR and 26 did not. GHS/QoL scores declined during neoadjuvant therapy but recovered postoperatively in both groups. During neoadjuvant treatment, pCR patients showed significantly greater improvement in cough [least square (LS) mean difference: -18.2; 95% confidence interval (CI): -33.0 to -3.5; P=0.02] and dyspnea (-21.3; 95% CI -38.3 to -4.3; P=0.02) compared to non-pCR patients. Conclusions:These findings suggest that pCR is associated with superior PROs, highlighting its potential as a marker for patient-centered benefits beyond survival.

The efficacy and safety of tislelizumab combined with chemotherapy as neoadjuvant therapy have been presented and evaluated in previously published TD-NICE study. However, the long-term impacts of patients who received tislelizumab plus chemotherapy followed by surgery were not observed in the study. From September 2020 to March 2021, a total of 45 patients were enrolled in the TD-NICE trial. The 4 -year follow-up data of the TD-NICE ITT population were collected by investigators. The primary endpoint was OS, and the secondary endpoint was EFS. Kaplan-Meier method and cox proportional hazards model were used to proceed statistical analysis. The median OS and EFS were not reached in all patients. The OS was 82.2

8064 Background: For stage II-III driver-negative non–small cell lung cancer (NSCLC), neoadjuvant chemoimmunotherapy followed by surgery is standard. However, patients without pathological complete response (non-pCR) have suboptimal outcomes and heterogeneous benefit from adjuvant immunotherapy. Baseline driver-negative status may be confounded by false-negative PCR- or DNA-only testing and detecting missed drivers postoperatively could alter adjuvant strategies. This study characterized post-treatment driver alterations and its prognostic value in non-pCR NSCLC after neoadjuvant immunotherapy. Methods: The retrospective multicenter study enrolled 247 stage II-III lung adenocarcinoma (LUAD) patients initially tested as EGFR L858R/19del- and ALK-negative and treated with neoadjuvant immunotherapy (2018-2025). 76 lung squamous cell carcinoma (LUSC) cases were included for exploratory analysis. Postoperative samples underwent 35-gene synchronous DNA/RNA next-generation sequencing (DR-NGS) to identify SNV/indel and fusion events and assess prognostic associations. Results: Of 247 LUAD cases, 179 passed NGS quality control (QC). Driver alterations were identified in 104 (58.1%) patients, including 26 fusions: RETn = 9), MET ex14 skipping (n = 6), ALK(n = 4), ROS1(n = 3), NRG1(n = 2), MET fusion(n = 1), NTRK(n = 1) and 80 SNVs/indels :EGFR (n = 30), KRAS G12C/D (n = 16), HER2/3 (n = 16), BRAF (n = 2), KRAS non-G12C/D (n = 16). Notably, 10 classic EGFR mutations (19del/L858R) 、9 rare EGFR mutations and 4 ALK fusions were newly identified, indicating baseline omissions. Driver-positive tumors had significantly higher residual tumor burden(median: 51.8% vs. 35.8%, p = 0.003) and a higher proportion of females (43.3% vs 20.0%, p = 0.001). Median recurrence-free survival (mRFS) differed significantly among fusion-positive, mutation-positive, and driver-negative groups (20.9 vs 41.5 vs 60.3 months; p = 0.024). In patients receiving adjuvant immunotherapy, mRFS was 34.6 months in driver-positive and not reached in driver-negative patients. Driver detection was rare ( < 3%) in QC-failed samples (n = 68), which exhibited a higher major pathological response than QC-passed samples (58.1% vs 14.5%), indicating tumor cellularity as critical for detection. Exploratory analysis in 55 QC-passed LUSC samples showed a low driver detection rate (7.3%), suggesting limited utility in this cohort. Conclusions: DR-NGS reveals a high prevalence of drivers in non-pCR LUAD after neoadjuvant immunotherapy, with positivity associated with higher residual tumor burden. Fusion-positive patients have the poorest prognosis, identifying a subgroup that may benefit from tailored adjuvant strategies. These findings support routine postoperative molecular re-profiling in non-pCR patients to guide individualized treatment.

Lung adenocarcinoma (LUAD) is the most common lung cancer histological subtype. Although the unfolded protein response (UPR) has been linked to various human diseases, its role in LUAD remains unclear. To identify UPR-related genes, we applied various methods, including weighted gene co-expression network analysis, differential expression analysis, and multivariate Cox regression. Ten machine learning algorithms were used to construct a UPR-related signature (UPRRS), which was validated using multiple public LUAD datasets. The UPRRS was integrated into a nomogram used in clinical practice for prognosis prediction. We also evaluated predicted drug sensitivity patterns across different risk subgroups. We identified 33 UPR-associated hub genes. A UPRRS was developed through systematic evaluation of 101 machine-learning combinations, exhibiting stable prognostic performance across multiple cohorts. Integration of the UPRRS into a nomogram facilitated the construction of a quantitative prognostic model. Significant differences in biological processes and tumor microenvironment immune cell infiltration were observed between the high- and low-risk UPRRS groups. All five UPRRS genes (ALDH2, FKBP4, KLF4, LAIR1, SIDT2) were validated at the protein level in LUAD cell lines, and FKBP4 was further confirmed by IHC in clinical tissues. Functional experiments showed that FKBP4 knockdown inhibited proliferation, migration, and invasion of A549 and H1975 cells, supporting a potential role for FKBP4 in LUAD progression. Our UPRRS provides a promising tool for prognostic stratification and may offer additional insights into tumor immune microenvironment characterization and therapeutic response prediction in LUAD.

BackgroundAs the leading histological form of lung cancer, lung adenocarcinoma (LUAD) displays considerable intratumoral heterogeneity, frequent therapeutic resistance, and an unfavorable clinical outcome. Although rewiring of mitochondrial energy metabolism is known to drive tumor progression and treatment failure, a comprehensive understanding of its dual role in LUAD drug resistance and prognosis has yet to be established. Here, we built a robust predictive signature that integrates mitochondrial metabolism with drug resistance through multi-omics integration and machine learning frameworks.MethodsWe explored single-cell RNA sequencing profiles together with TCGA-LUAD transcriptomic data. Weighted gene co-expression network analysis (WGCNA) was applied to extract gene modules linked to mitochondrial-related genes (MRGs) and drug resistance-related genes (DRGs). From these, a five-gene (KLF4, KLF10, CAT, ALDOA, HLA-DRA) prognostic classifier, designated MDrisk, was formulated using LASSO-Cox regression and 101 combinations of 10 machine learning algorithms.ResultsThe MDrisk model demonstrated reliable and precise prognostic capacity across training, internal test, and external GEO cohorts, serving as an independent risk factor. Elevated MDrisk scores correlated with an immunosuppressive microenvironment, higher tumor mutational burden, distinct copy-number alteration profiles, and decreased drug sensitivity in computational predictions. In vitro experiments further validated that silencing ALDOA—a central component of the signature—suppressed the proliferation, migration, and invasive capacity of LUAD cells.ConclusionThe MDrisk signature derived from mitochondrial energy metabolism and drug resistance may be useful for distinguishing prognosis, immune contexture, and computationally inferred drug susceptibility in LUAD. It may offer a tool for further exploration of individualized therapy and sheds light on the interplay between metabolic dysregulation and antitumor immunity.

Supplementary Figure S5. Comparison of immune-related and tumor-intrinsic signatures at baseline between PET responders and PET nonresponders (A) and indicated signatures between patients with and without pathological response in (B, C) PET responders and (D, E) PET nonresponders.

Supplementary Figure S2. Distribution of residual viable tumor percentages in non-pCR patients by PET response status

BACKGROUND:Patients with nonpathological complete response constitute the majority of esophageal squamous cell carcinoma (ESCC) cases after neoadjuvant therapy, yet the prognostic implications of different pathological outcomes-particularly major pathological response (MPR) combined with lymph node (LN) status-remain poorly defined. This study aimed to elucidate their prognostic value and to refine postoperative evaluation systems. METHODS:This multicenter retrospective cohort study included patients with ESCC from 8 high-volume esophagectomy centers in China who underwent neoadjuvant chemoradiotherapy (nCRT) or neoadjuvant immunochemotherapy (nICT) followed by R0 resection between 2016 and 2023. The primary comparison was between the nCRT and nICT cohorts, with additional planned analyses of pathological response subgroups and pathological stage cohorts (early vs advanced). Propensity score matching was used to balance baseline characteristics, and survival outcomes were compared using Kaplan-Meier and Cox regression analyses. RESULTS:After matching, nICT demonstrated similar pathological complete response (pCR) rates to nCRT but achieved more extensive nodal dissection and significantly reduced distant recurrence (P < .001). Among pathological subtypes, MPR(+)LN(-) yielded the most favorable long-term outcomes, surpassing or equaling those of pCR, whereas MPR(-)LN(+) was associated with the worst prognosis (P < .001). Notably, patients with MPR(+)LN(-) had comparable survival between early- and advanced-stage cohorts, suggesting a reduced prognostic impact of pathological stage and underscoring its strong predictive value. CONCLUSIONS:Patients with MPR and LN negativity had long-term survival comparable to pCR, suggesting that multidimensional pathological evaluation may improve prognostic accuracy. Compared with nCRT, nICT demonstrated improved nodal clearance and systemic disease control across different pathological outcomes.

Background:Limited-stage small-cell lung cancer (LS-SCLC) is an aggressive malignancy, and patients with this disease have a poor long-term survival when treated with current standard chemoradiotherapy. Although neoadjuvant or conversion immunotherapy combined with chemotherapy has shown efficacy in patients with non-SCLC (NSCLC), its effect in those with LS-SCLC remains underexplored, particularly regarding long-term survival outcomes. This study aims to evaluate the efficacy and long-term survival of neoadjuvant or conversion immunotherapy plus chemotherapy in patients with LS-SCLC. Methods:This retrospective study enrolled 34 patients with stage IIB-IIIB SCLC who received neoadjuvant or conversion anti-programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) immunotherapy plus platinum-etoposide chemotherapy, followed by surgical resection. The primary end points were overall survival (OS) and event-free survival (EFS). Pathological complete response (pCR) and major pathological response (MPR) were assessed, and subgroup analyses were performed based on pathological response, demographic factors, and tumor stage. Results:With a median follow-up of 34.2 months, the 2-year EFS and OS rates were 72.1% [95% confidence interval (CI): 53.1-84.5%] and 82.4% (95% CI: 64.9-91.7%), respectively. Among the 34 patients, a pCR rate of 47.1% and an MPR rate of 70.6% were achieved. Patients achieving pCR or MPR had significantly better survival outcomes than did the nonresponders. In the subgroup analysis, a lower body mass index (BMI) (<25.0 kg/m2) and earlier disease stage II were associated with improved survival, although this was not statistically significant in all comparisons. Conclusions:Neoadjuvant or conversion chemoimmunotherapy followed by surgery demonstrates promising efficacy and survival benefits in patients with stage IIB-IIIB SCLC, with high pCR/MPR rates strongly predicting superior outcomes. These findings support the integration of immunotherapy into multimodal treatment strategies for patients with stage II-IIIB SCLC and warrant validation in larger randomized controlled trials.

8512 Background: HLX43 (a PD-L1 ADC) has demonstrated encouraging efficacy with a manageable safety in a phase 1 study. Here, we present the pooled findings from patients with NSCLC in a phase 1 study (HLX43-FIH101) as well as a global phase 2 study investigating HLX43 in NSCLC (HLX43-NSCLC201). Methods: The phase 1 study included a dose-escalation phase in patients with advanced solid tumors (0.5–4.0 mg/kg Q3W), followed by a dose-expansion phase in patients with advanced or metastatic NSCLC at doses of 2.0, 2.5, and 3.0 mg/kg Q3W. The phase 2 study was conducted in NSCLC and comprised two parts: Part A was dose-exploration for patients who had failed prior first-line therapy and had no actionable genomic alterations, to receive HLX43 at 2.0 or 2.5 mg/kg Q3W; Part B was dose-expansion in which patients received HLX43 at the recommended dose determined from Part A. This analysis integrated data from heavily pretreated NSCLC patients enrolled across both studies. Efficacy and safety outcomes were evaluated in the pooled population. Results: As of December 31, 2025, 205 patients were enrolled and received HLX43 at 1 mg/kg (n = 3), 2 mg/kg (n = 89), 2.5 mg/kg (n = 85), 3 mg/kg (n = 23), and 4 mg/kg (n = 5). Patients received a median of 2 lines of prior antitumor therapy (range, 1–9). Among the 161 response-evaluable patients (2, 69, 64, 21, and 5 in the 1, 2, 2.5, 3, and 4 mg/kg groups, respectively), the investigator-assessed ORR was 31.1%. In the 2.0 mg/kg group, investigator-assessed ORR was 36.4% for squamous NSCLC (n = 33); among these patients, ORR was 40.0% for those who previously failed docetaxel (n = 15). ORR was 47.4%, and 50.0% for patients with EGFR-wildtype (n = 19), and EGFR-mutant (n = 16) nonsquamous NSCLC receiving HLX43 at 2.5 mg/kg. Biomarker exploratory analyses showed that efficacy was not associated with PD-L1 expression, with ORRs of 30.1% and 32.1% in patients with PD-L1-positive (n = 83) and PD-L1-negative tumors (n = 78), respectively. Overall, 199 (97.1%) patients experienced treatment-related adverse events (TRAEs), of whom 88 (42.9%) had grade ≥3 in severity. Most common Grade ≥3 TRAEs (incidence ≥10%) included lymphocyte count decreased (n = 47, 22.9%), white blood cell count decreased (n = 27, 13.2%), anemia (n = 25, 12.2%), and neutrophil count decreased (n = 23, 11.2%). TRAEs led to treatment discontinuation in 17 (8.3%) patients. Conclusions: HLX43 exhibited promising efficacy in patients with heavily pretreated advanced NSCLC, regardless of histology subtypes, and PD-L1 expression, along with manageable tolerability. Further investigation is warranted. Clinical trial information: NCT06115642 (HLX43-FIH101-phase 1 study), NCT06907615 (HLX43-NSCLC201- phase 2 study).

The formation of a hardened fibrous membrane through in situ degradation of polycaprolactone (PCL) in soft tissues has emerged as a promising alternative to conventional rigid bone implants for chest wall reconstruction. However, strategies to enhance the mechanical integrity and biological performance of such fibrous constructs remain limited. While nano-hydroxyapatite (nHA) is known to promote osteoblast proliferation and mineralization, its role in regulating fibroblast behavior remains unclear, particularly within a dynamically strained environment mimicking respiratory motion. In this study, we developed PCL scaffolds incorporating various concentrations of nHA using fused deposition modeling (FDM). The PCL/10 wt% nHA scaffold exhibited optimal mechanical properties and significantly enhanced fibroblast proliferation, adhesion, and extracellular matrix deposition in vitro. Notably, higher nHA content led to excessive Piezo1 activation, resulting in Ca2+ overload and increased fibroblast apoptosis. Under dynamic mechanical stimulation, the PCL/10 wt% nHA scaffold markedly promoted fibroblast functionality and tissue fibrosis, facilitating robust soft tissue hardening in vivo. Mechanistic investigations revealed that the Piezo1/TGF-β1 signaling axis plays a central role in mediating fibroblast responses to cyclic shear stress. These findings demonstrate that the PCL/10 wt% nHA scaffold effectively supports tissue-engineered structural reinforcement through fibroblast-driven fibrosis, presenting a biodegradable and mechanically adaptive approach with potential for future chest wall repair applications.

Inactivating vascular endothelial growth factor receptor (VEGFR) may improve the efficacy of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in EGFR-mutant non-small cell lung cancer (NSCLC). The ATTENTION study (phase II, open-label, randomized, multicenter trial (Registration number: ChiCTR2100047453), evaluated the efficacy and safety of aumolertinib plus apatinib vs. aumolertinib alone in untreated, EGFR-mutant, advanced NSCLC. The primary endpoint was the 18-month PFS rate. Across 18 centers in China, 104 patients were enrolled to receive aumolertinib alone (n = 51) or with apatinib (n = 53). At a median follow-up duration of 19.4 months, aumolertinib plus apatinib outperformed aumolertinib alone in terms of the 18-month progression-free survival (PFS) rate (74% vs. 50%, P = 0.036), median PFS (not reached [NR] vs. 20.1 months, hazard ratio [HR] = 0.41, P = 0.017), and objective response rate (79% vs. 59%, P = 0.024). No grade 4/5 treatment-related adverse effects (TRAEs) were observed, whereas grade 3 TRAEs occurred in 38% vs. 27% of patients, with hypertension (11%) and platelet count decrease (9%) being most common in the combination arm. Exploratory analysis revealed that PFS benefits from aumolertinib plus apatinib predominantly in those with TP53 mutations. As an infusion-free option, aumolertinib plus apatinib demonstrated PFS benefits with manageable safety in patients with untreated, EGFR-mutant, advanced NSCLC.

Polyether Ether Ketone (PEEK) is a commonly used implant for bone defect reconstruction surgery. However, postoperative infection remains a critical complication because PEEK lacks antibacterial properties, and conventional detection methods lack sensitivity, speed, and real-time capability, risking delayed diagnosis and even the need for a second surgery. Herein, we report an electrochemical aptamer-based PEEK biosensing system for early detection of wound infections. Due to the non-conductive nature of PEEK and its inability to sense the surrounding environment, the PEEK surface was rendered conductive by laser ablation and electrodeposited with gold nanoparticles, followed by the immobilization of thiolated aptamers specific to Escherichia coli and Staphylococcus aureus through Au-S bonds, thereby forming a three-electrode system. This biosensor enables rapid, real-time detection of single or mixed bacterial infections with high sensitivity (LOD: 5.776 CFU mL-1 for E. coli and 11.254 CFU mL-1 for S. aureus), a wide linear range (10-10⁸ CFU mL-1), and excellent selectivity. Biocompatibility tests verified that the functionalized PEEK induces no additional inflammatory response. In murine wound models, the system accurately monitored bacterial infections throughout the healing process. By converting non-conductive PEEK into an intelligent sensing platform, this work affords a promising strategy for real-time postoperative infection monitoring, enabling early diagnosis and personalized wound management. This approach is expected to provide new insights into reducing postoperative wound infections.

Background: Lung cancer remains a major global health burden. RNA-binding proteins (RBPs) play crucial roles in post-transcriptional gene regulation, and their dysregulation is frequently implicated in tumorigenesis. The present study aimed to elucidate the molecular network governed by the highly expressed RBP TIMELESS in lung adenocarcinoma (LUAD) and determine its mechanistic role in LUAD progression. Methods: The Cancer Genome Atlas-LUAD, Gene Expression Omnibus, and single-cell RNA sequencing datasets were analyzed to identify aberrantly expressed RBP genes. The RBP gene TIMELESS exhibited the most significant effect on LUAD cell death and was selected for further study. Photoactivatable ribonucleoside-enhanced crosslinking and immunoprecipitation sequencing and RNA sequencing were employed to identify ferroptosis-related targets directly bound by TIMELESS. Molecular mechanisms underlying the TIMELESS-mediated regulation of ferroptosis in LUAD were investigated via immunoprecipitation–mass spectrometry, glutathione S-transferase pull-down, immunofluorescence–fluorescence in situ hybridization, RNA immunoprecipitation, poly(A)-tail, and RNA stability assays. In an orthotopic lung cancer mouse model treated with erastin (a ferroptosis inducer) and programmed cell death protein 1 (PD-1) blockade, the role of TIMELESS in therapeutic response was assessed via flow cytometry and multiplex immunofluorescence (mIF). Infiltrating immune cells in LUAD were analyzed by tissue microarrays (TMAs) via mIF. Results: TIMELESS significantly affected LUAD cell proliferation and death, and TIMELESS knockdown significantly enriched RNA-binding and ferroptosis pathways. Transferrin (TF) was identified as a direct TIMELESS target governing ferroptosis. TIMELESS was revealed to bind Ccr4-Not transcription complex subunit 3 (CNOT3) to promote TF mRNA degradation. TIMELESS depletion combined with erastin and PD-1 blockade enhances efficacy, prolongs survival, increases T cell and M1 macrophage infiltration, and reduces M2 macrophage infiltration. Further, high TIMELESS expression was inversely correlated with ferroptosis marker 4-hydroxynonenal but positively correlated with programmed cell death ligand 1 (PD-L1), reduced T cell and M1 macrophage infiltration, and increased M2 macrophage infiltration. Conclusions: TIMELESS recruits CNOT3 to accelerate TF mRNA degradation, thereby suppressing ferroptosis and promoting LUAD growth. These findings suggest that the TIMELESS/TF regulatory axis may be a promising therapeutic target for LUAD.

AbstractPurpose: The aim of this study is to assess positron emission tomography/computed tomography (PET/CT)–guided neoadjuvant treatment with tislelizumab plus chemotherapy/chemoradiotherapy in patients with resectable esophageal squamous cell carcinoma (ESCC). Patients and Methods: RATIONALE-213, a multicenter, open-label, two-cohort phase II study, enrolled 70 patients with resectable ESCC. After one induction chemotherapy cycle, patients were categorized as PET responders (PET maximum standardized uptake value decrease ≥35%) or PET nonresponders (<35%). Subsequently, all patients received three cycles of tislelizumab 200 mg intravenously. In addition, PET responders received chemotherapy and PET nonresponders received chemoradiotherapy as neoadjuvant treatment followed by surgery. The primary endpoint was the pathologic complete response (pCR) rate. Results: As of October 25, 2024 (median follow-up 25.5 months), pCR rates were 30% [95% confidence interval (CI), 11.9–54.3] in PET responders and 34.4% (95% CI, 18.6–53.2) in PET nonresponders. Higher pCR rates were associated with higher baseline programmed cell death ligand 1 tumor area positivity scores in both cohorts. One-year disease-free survival rates were 79% and 74.2%, and event-free survival rates were 87.1% and 67.8%, respectively. Any-grade and grade ≥3 treatment-related adverse events (TRAE) occurred in 93.3% and 50% of PET responders and 97.5% and 82.5% of PET nonresponders, with no new safety signals. The most common grade ≥3 TRAE was decreased neutrophil count (36.7% and 70%, respectively). Conclusions: Tislelizumab with chemotherapy/chemoradiotherapy as neoadjuvant treatment for resectable ESCC yielded promising pCR rates and favorable survival outcomes with a tolerable safety profile in both PET/CT-guided responders and PET nonresponders.