Browse Articles
Discover research articles across all indexed journals
All that glitters is not gold: The importance of ruling out other causes of leptomeningeal enhancement on MRI that may resemble leptomeningeal disease (LMD).
e14018 Background: Leptomeningeal metastatic disease (LMD) is increasingly common entity which at times, in the proper clinical context, can be diagnosed based on MRI imaging findings without CSF confirmation of positive CSF cytology, CSF genomics or CSF Circulating Tumor Cells (CTCs). CSF studies remain very insensitive, LMD treatments are very invasive (e.g. cranial spinal RT, intra-thecal therapy, Ommaya reservoir placement etc.) and therefore it is critical to carefully consider non-LMD causes of meningeal enhancement. The imaging findings may include linear or nodular enhancement of the sub arachnoid space, hydrocephalus, or cranial nerve enhancement. As these findings are non-specific and appear with other disease processes, we studied these in a single institutional case series at the Moffitt Cancer Center. Methods: In this single institution retrospective analysis, Neuro-oncology Tumor Board records from 2021-2025 at Moffitt Cancer Center were reviewed for cases in which LMD was initially considered by imaging, but whose imaging findings were subsequently shown to be from an alternative and often reversible diagnosis. Results: We identified a total of 13 patients whose leptomeningeal enhancement were due to causes other than LMD. Three (23%) had acute ischemia and with accompanying luxury perfusion, three (23%) had acute venous occlusion or compression, two (15%) were due to posterior reversible encephalopathy syndrome (PRES), two (15%) were caused by status epilepticus (including one with non-convulsive status epilepticus), and one each (7%) due cryptococcal meningitis, laminar necrosis from remote ischemia and intracranial hypotension. In 12 (92%) patients had improvement/resolution of the radiographic abnormalities on subsequent examinations and other relevant studies (e.g. CSF microbiological studies showing cryptococcus with institution of anti-biotics). The one (8%) patient that didn’t improve radiographically had chronic infarction with laminar necrosis which was unchanged. Conclusions: Leptomeningeal enhancement in cancer patients has other often treatable causes other than LMD. In addition to CSF cytology & genomics we routinely order infectious and cytokine panels to rule out infectious or inflammatory causes and EEGs. Other etiologies besides LMD should be considered when investigating possible LMD on MRI. Follow up imaging and Neurology consultations are essential. In our patient population and the small sample size we did not encounter other inflammatory diseases such as MS, other autoimmune diseases or other CNS infections such as HIV, etc.
Long-term survival, recurrence patterns, and postrecurrence outcomes after neoadjuvant chemoradiotherapy versus surgery plus adjuvant therapy in locally advanced esophageal squamous cell carcinoma: Final results of a randomized phase III trial.
e16092 Background: The optimal treatment strategy for locally advanced esophageal squamous cell carcinoma (ESCC) remains uncertain. Although neoadjuvant chemoradiotherapy (NCRT) improves pathological response, its long-term survival advantage over surgery plus adjuvant therapy (AT) is unclear. The impact of these strategies on recurrence patterns and postrecurrence survival (PRS) also requires clarification. Methods: In this prospective, open-label, randomized phase III trial, 245 patients with cT1N+M0 or cT2–4aNxM0 thoracic ESCC (2018–2020) were assigned to NCRT followed by surgery or surgery plus AT; 118 and 112 patients were included in the final analysis. NCRT consisted of IMRT (40 Gy/20 fractions) with weekly paclitaxel and carboplatin. AT was individualized by a multidisciplinary team. The primary endpoint was overall survival (OS); secondary endpoints included disease-free survival (DFS), pathological outcomes, recurrence patterns, recurrence timing, and PRS. Results: After a median follow-up of 59.1 months, OS and DFS did not differ significantly between groups (5-year OS: 59.2% vs 59.6%, HR 1.01, P = 0.97; 5-year DFS: 53.1% vs 56.5%, HR 1.13, P = 0.53). The NCRT group achieved a 28.8% pathological complete response (pCR) rate, and pCR was associated with markedly improved OS (76.5% vs 52.1%; HR 0.39, P = 0.01). Among 230 evaluable patients, the recurrence rate was 33.5% (34.7% vs 32.1%, P = 0.696), with distant metastasis as the predominant pattern. Early recurrence ( < 12 months) occurred in 14.3% of patients and predicted significantly poorer OS (6.1% vs 27.1%; HR 4.22, P < 0.001) and shorter PRS (6.93 vs 10.28 months; HR 1.75, P = 0.026). It was the only independent predictor of PRS. Conclusions: NCRT did not improve long-term survival over surgery plus AT in an unselected ESCC population. Nevertheless, patients achieving pCR experienced substantial benefit. Recurrence remained common, dominated by distant metastasis, and early recurrence was strongly associated with poor PRS. Clinical trial information: NCT06775652 .
Closing the information gap in oncology: A retrieval-augmented generation multi-modal AI platform for personalized patient education and physician efficiency.
11075 Background: Effective patient education is critical in oncology but hindered by the complexity of treatment guidelines and the time constraints of clinicians. Traditional static materials often fail to address the personalized needs of diverse demographics, particularly the elderly. We developed an Intelligent Medical Science Popularization Platform integrating Multimodal Large Language Models (MLLM) with Retrieval-Augmented Generation (RAG) to automate the production of reliable, guideline-based oncology educational content. Methods: The platform utilizes a three-tier architecture deployed at a tertiary cancer center. The core engine combines a self-developed MLLM with RAG, anchoring content generation to authoritative sources (e.g., CSCO/NCCN guidelines and peer-reviewed journals) to ensure clinical accuracy. The system features: (1) Physician Portal: Automates the conversion of clinical protocols into patient-friendly text, layouts, and animations; (2) Multimodal Synthesis: Generates virtual avatar videos explaining diagnoses and treatments; (3) Patient Interface: Delivers personalized, accessible content via mobile apps, supporting voice interaction for elderly adherence. Results: Implementation data demonstrated significant efficiency and engagement gains. The MLLM+RAG model achieved > 98% accuracy in content generation as verified by expert oncologists. The production time for educational videos was reduced from 72 hours to ~10 minutes (> 99% reduction), enabling rapid updates aligned with new trial data. Animation production costs decreased by 10-fold. Daily content output exceeded 15 units. Notably, in a cohort of elderly cancer patients, the platform increased user retention rates by 40% and extended average session duration by 50 seconds, indicating improved engagement with health information. Conclusions: This policy-compliant AI platform validates the utility of RAG-based MLLMs in oncology. By shifting from passive information supply to active, personalized intelligent services, it significantly reduces the workload for oncologists while bridging the digital divide for vulnerable patient populations. This model offers a scalable solution for enhancing health literacy and treatment adherence in cancer care.
Sensory neuropathy and neuropathic pain during taxane chemotherapy in CIPN-free patients.
e24186 Background: Chemotherapy-induced peripheral neuropathy (CIPN) is a common toxicity of taxane chemotherapy and includes both sensory neuropathy and neuropathic pain. The longitudinal course of these symptoms among patients without neuropathy at treatment initiation is not well characterized. Describing symptom trajectories during taxane therapy may inform monitoring and prevention strategies. Methods: PACT, a multinational coordinated study, pooled data from three parallel randomized trials in patients with early-stage breast cancer receiving taxane chemotherapy. Participants screened negative for neuropathy at baseline using PRO-CTCAE items. Trials compared acupuncture with an active relaxation–nature-video control. Sensory neuropathy was assessed using the EORTC CIPN-20 sensory subscale (0–100), and neuropathic pain using a 0–10 worst pain Likert scale at baseline, week 12 (primary timepoint), and week 24 (follow-up). Longitudinal mixed-effects models estimated symptom trajectories, adjusting for site, chemotherapy schedule, and number of neurotoxic agents. Pairwise time comparisons use model-adjusted estimates with Tukey–Kramer adjustment. Given no treatment differences, values are presented as averages across arms. Results: Among 127 evaluable participants, model-adjusted mean CIPN-20 sensory scores increased 7.7 points (95% adj CI: 4.8 to 10.6, adj p<0.0001) from baseline to week 12 and remained stable at 24 weeks. Worst neuropathic pain increased 0.64 points (95% adj CI: 0.18–1.09; adj p=0.004) at 12 weeks with no further significant change at 24 weeks. Median worst pain scores remained 0 at all timepoints, while sensory scores rose from 0 at baseline to 3.7 at week 12. (Table 1) Conclusions: Sensory neuropathy and worst pain scores increased during taxane chemotherapy. Sensory changes were clinically significant, while changes in pain were subclinical and the majority of patients remained free of pain at all time points. These findings highlight the heterogeneous manifestation of CIPN during standard chemotherapy in an integrative medicine primary CIPN prevention trial and underscore the importance of carefully assessing both sensory and pain-related outcomes in prevention strategies. Clinical trial information: NCT05528263, ChiCTR2200066714, KCT0008470 . Model-adjusted symptom trajectories. Mean Change (95% adj CI; adj p-value) Outcome Baseline Week 12 Week 24 Week 12 - Baseline Week 24 – Week 12 Week 24 - Baseline CIPN-20 sensory(0–100) 0.6 8.3 8.5 7.7 (4.8 to 10.6; <0.0001) 0.2 (-2.8 to 3.1; 0.99) 7.8 (4.7 to 10.9; <0.0001) Worst neuropathic pain(0–10) 0.33 0.97 1.19 0.64 (0.18 to 1.09; 0.004) 0.22 (-0.16 to 0.60; 0.35) 0.86 (0.35 to 1.36; 0.0003) Predicted means and pairwise changes averaged across arms; CI and p-values adjusted for multiple comparisons (Tukey–Kramer).
Partial breast irradiation followed by neoadjuvant immunochemotherapy in early TNBC.
e12649 Background: Neoadjuvant immunochemotherapy (NAICT) incorporating immune checkpoint blockade has become standard for triple-negative breast cancer (TNBC). However, a substantial proportion of patients, particularly those with immune-cold tumor microenvironment (TME), fail to achieve pathological complete response (pCR) despite pembrolizumab treatment, highlighting the need for strategies to enhance antitumor immunity. Emerging evidence suggests that radiotherapy can favorably modulate TME and augment immune checkpoint blockade. We retrospectively evaluated outcome of early TNBC patients treated with partial breast irradiation (PBI) prior to NAICT with pembrolizumab. Methods: Patients with TNBC who received PBI to the primary tumor before NAICT between July 2022 and June 2025 were retrospectively analyzed. Eligible criteria included age ≥ 20 years, ER and PR expression < 1%, HER2-negative invasive carcinoma, and clinical stage II-III disease (cT2-4N0-3 or cT1N1-3). All patients received single-fraction PBI (8 Gy), followed by neoadjuvant NAICT with four cycles of carboplatin every 3 weeks plus paclitaxel weekly, followed by four cycles of anthracycline-based chemotherapy every 3 weeks. Pembrolizumab was administered concurrently every 3 weeks for a total of eight cycles. For comparison, 186 patients who did not receive PBI were included in the analysis. Results: 21 patients completed PBI and NAICT followed by curative surgery. Median age was 43 years (range, 29-71). Overall, 95.2% had cT2-4 disease, 47.6% had cN2-3 disease, and 66.7% had grade III tumors. The overall pCR rate was 71.4% (15/21). Among the patients with immune-cold TME, defined as stromal tumor-infiltrating lymphocytes (sTIL) < 30% and PD-L1 combined positive score (CPS) < 10, the pCR rate was 64.7% (11/17). Among patients with pCR, 63.6% (7/11) had grade III tumors, 81.8% (9/11) had cT2-4 disease, and 54.5% (6/11) had cN2-3 disease. With short-term follow-up, no patient with pCR experienced disease recurrence. In the comparison group of 186 patients without PBI, the pCR rate was 66.1% (123/186). No new or notable adverse effects were identified in patients who received PBI. Conclusions: A single dose of PBI 8Gy to primary tumor before NAICT yielded a promising pCR rate, supporting the hypothesis that radiotherapy modulates the TME and enhances local antitumor immunity. This approach was particularly beneficial for patients with immune-cold TME. A prospective phase II study (NCT07011823) exclusively enrolling patients with immune-cold TME is underway to validate these findings and define the clinical role of PBI prior to NAICT in early TNBC.
Effects of microplastics on the colorectal tumor immune microenvironment and immunotherapy resistance via JAK–STAT–microbiota perturbation.
e15731 Background: Microplastics (MPs) are pervasive environmental pollutants with rising human exposure. Although potential health effects have been recognized, their link to cancer remains poorly defined. The colorectum represents a primary absorption site for MPs, prompting investigation of their impact on colorectal cancer (CRC) progression and immunotherapy response. Methods: MPs were isolated and characterized from CRC tumor tissues and matched blood samples. Functional studies were conducted in murine CRC models to evaluate tumor progression, immune cell infiltration, gut microbiota alterations, and response to anti–PD-L1 therapy. Transcriptomic profiling and immune analyses were performed to elucidate underlying mechanisms. Results: MP exposure aggravated CRC progression and significantly reduced sensitivity to anti–PD-L1 therapy. Mechanistically, tumor-infiltrating MPs suppressed JAK-STAT signaling, downregulated IFN-γ, CXCL9, CXCL11, B2m, and H2-K1 expression, and reduced CD8⁺ and CD4⁺ T-cell infiltration, resulting in impaired antigen presentation and T-cell recruitment. MPs also induced gut microbiota dysbiosis that further contributed to immunotherapy resistance. Conclusions: MPs promote CRC immunotherapy resistance by altering the tumor immune microenvironment through disruption of the JAK-STAT–microbiota axis. These data highlight MPs as an unrecognized determinant of CRC immunotherapy outcomes and a potential modifiable environmental target amid rising global MP contamination.
Zanubrutinib plus rituximab as front-line treatment for mucosa-associated lymphoid tissue (MALT) lymphoma: A single-arm, phase II study (ZAMA).
7080 Background: Mucosa-associated lymphoid tissue (MALT) lymphoma is the most common subtype of marginal zone lymphoma and follows an indolent clinical course. For patients with advanced-stage disease with an indication for treatment, rituximab-based immunochemotherapy is recommended as frontline therapy; however, toxicity and tolerability remain concerns. Zanubrutinib, a next-generation Bruton tyrosine kinase inhibitor, has shown promising antitumor activity with a manageable safety profile in marginal zone lymphoma. This study aimed to evaluate zanubrutinib plus rituximab as frontline treatment for MALT lymphoma. Methods: The ZAMA study (NCT06647732), a multicenter, single-arm, phase II trial, enrolled patients aged ≥18 years with histologically confirmed CD20-positive MALT lymphoma. Eligible patients had newly diagnosed advanced-stage disease or relapsed disease after prior local therapy, had not received prior systemic treatment, and had at least one measurable lesion. Patients received rituximab (375 mg/m 2 ) plus zanubrutinib (160 mg twice daily) in 28-day cycles. Treatment was administered for six cycles, after which patients achieving complete response (CR) entered observation, while those with partial response or stable disease received two additional cycles. The primary endpoint was the CR rate as best response. Results: From October 30, 2024 to December 12, 2025, 39 patients were enrolled. At the data cutoff of January 10, 2026, 23 patients had completed protocol-specified treatment, including 14 patients who received six cycles and 9 patients who received eight cycles. The median age was 58 years (range 24–75); 7 patients were males (30.4%). Most patients (n = 21, 91.3%) had an ECOG performance status of 0–1, and 21 patients (91.3%) had Ann Arbor stage III–IV disease; two patients had relapsed disease after prior local therapy. Eighteen patients (78.2%) achieved CR, and the remaining five patients achieved PR. Treatment-emergent adverse events (TEAEs) occurred in 20 patients (87.0%); hematologic toxicity were most common: neutropenia (n = 2, 8.7%), and thrombocytopenia (n = 2, 8.7%). Grade ≥ 3 TEAEs occurred in 4 patients (17.4%), most frequently neutropenia (n = 2, 8.7%), and thrombocytopenia (n = 2, 8.7%). Conclusions: Preliminary results from the ZAMA study show that frontline zanubrutinib plus rituximab provides encouraging efficacy with manageable safety in MALT lymphoma. The study is ongoing, and patient enrollment is continuing. Clinical trial information: NCT06647732 . Baseline characteristics. Characteristics Patients (n=23) Age, years (median [range]) 58 (24-75) Sex MaleFemale 7 (30.4%)16 (69.6%) ECOG PS 0-12 21 (91.3%)2 (8.7%) Ann Arbor stage III-IV 21 (91.3%) Prior local therapy 2 (8.7%) MALT-IPI score 0 1 (4.3%) 1 21 (91.3%) 2 1 (4.3%) Bone marrow involvement 4 (17.4%)
KEYNOTE-B59: An open-label, multicenter, phase 1/2 study of efdelikofusp alfa (GI-101A; CD80-IgG4 Fc-IL2v2) in advanced solid tumors (part E & F of GII-101-P101).
2519 Background: Efdelikofusp alfa is a novel immunocytokine consisting of CD80 fused to an engineered IL-2 variant (IL-2v2) optimized to reduce IL-2Rα affinity. It blocks CTLA-4–mediated immune suppression while preferentially activating and expanding immune effector cells over regulatory T cells. Here, we report results from the dose-escalation phases of Part E (monotherapy) and Part F (combination with pembrolizumab) of the ongoing KEYNOTE-B59 study. Methods: KEYNOTE-B59 is an open-label, phase 1/2 study evaluating efdelikofusp alfa alone or in combination with pembrolizumab in patients (pts) with advanced solid tumors. In Part E, escalating doses of efdelikofusp alfa (0.05–0.3 mg/kg) were administered intravenously every 3 weeks (Q3W). In Part F, pts received efdelikofusp alfa (0.05–0.3 mg/kg) plus pembrolizumab (200 mg) Q3W. Primary objectives were to assess safety, tolerability, and to determine maximum tolerated dose (MTD) and/or recommended phase 2 dose. Results: As of 29 December 2025, a total of 84 pts who had progressed on available therapies were enrolled, including 36 pts in Part E and 48 pts in Part F. In Part E, one dose-limiting toxicity (DLT) occurred at 0.1 mg/kg (grade 3 hypertension). However, MTD was not reached up to 0.3 mg/kg, and efdelikofusp alfa was generally well tolerated. At the biologically effective dose range (0.2–0.3 mg/kg), monotherapy demonstrated clinical activity in immunotherapy (IO)-experienced bladder cancer (confirmed CR; DoR 5.8 months; PFS 13.9 months) and mesothelioma (confirmed PR; DoR 5.6 months; PFS 6.7 months). In Part F, 48 pts were treated with efdelikofusp alfa in combination with pembrolizumab. Median age was 63 years and 45.8% of pts had received prior IO. The overall safety profile of the combination was comparable to monotherapy, with common treatment-related adverse events including pyrexia (89.6%), liver enzyme elevation (47.9%), chills (37.5%), and decreased platelet count (33.3%). Objective responses were observed across multiple tumor types, including ccRCC, urothelial cancer, squamous NSCLC, cutaneous squamous cell carcinoma, pancreatic adenocarcinoma, and cervical cancer, irrespective of prior IO exposure. In pts with ccRCC (n = 10), the most frequently enrolled indication (70.0% prior IO-treated), the objective response rate was 40.0% and the disease control rate was 70.0%, with durable responses observed from 5.5+ to 16.7+ months; median DoR was not reached. Efdelikofusp alfa induced robust peripheral lymphocyte expansion, mainly CD8 + T and NK cells, which correlated with longer median PFS. Conclusions: Efdelikofusp alfa was well tolerated as a monotherapy and in combination with pembrolizumab. Antitumor activity was observed with a favorable benefit–risk profile, supporting continued development in selected tumors, including ccRCC. Clinical trial information: NCT04977453 .
Charge‐Engineered COFs for Biointegrated Memristor Nerves
ABSTRACT Restoring motor function after neurological injury requires artificial neural interfaces that emulate biological rate coding with low power and stability. Here, we present a molecular‐level strategy to engineer covalent organic frameworks (COFs) for biointegrated memristors as artificial efferent nerves. Leveraging intrinsic porosity and chemical tunability, we modulate ionic transport and memristive dynamics via charged group functionalization. We synthesize positively and negatively charged COF nanosheets and reveal polarity‐dependent memristive behaviors. In a conductive‐filament memristor architecture, negatively charged COFs enhance electrostatic interactions with mobile metal ions, more effectively regulating filament nucleation and rupture. Consequently, negatively charged devices reduce the switching voltage to 0.5 V, deliver an ON/OFF ratio > 10 5 , and lower power consumption to 0.04 nW, with suppressed leakage of ∼5 pA and stable operation over 5000 bending cycles. In vivo, the COF memristor translates neuronal spike trains into smooth, graded muscle contractions in a mouse leg, emulating physiological motor control. This work establishes charge‐engineered COFs as a platform for neuromorphic and bioelectronic technologies.
Efficacy of adjuvant S-1 vs capecitabine plus oxaliplatin (XELOX) chemotherapy among gastric cancer patients after D2 gastrectomy: A systematic review and meta-analysis.
e16115 Background: Adjuvant chemotherapy improves outcomes after D2 gastrectomy for gastric cancer. S-1 and XELOX are commonly used regimens, but their relative efficacy remains uncertain. We conducted a systematic review and meta-analysis comparing S-1 therapy with XELOX in gastric cancer patients undergoing D2 gastrectomy. Methods: We conducted a comprehensive literature search across five databases: PubMed, Cochrane, Scopus, Embase, and ClinicalTrials.gov for articles comparing efficacy and safety of S-1 vs XELOX adjuvant chemotherapies in gastric cancer patients after D2 gastrectomy. Nine studies fulfilling the eligibility criteria were included in the review and analysis. The analysis was conducted on RStudio v4.5.0 using the 'meta' package. The generic inverse variance method was used for continuous data, and the Mantel–Haenszel method was used for binary data. The pooled analysis is presented as hazard ratios (HRs) and odds ratios (ORs) with 95% confidence intervals (CIs). Results: Nine studies evaluating S-1 vs XELOX were included. There was no significant difference in DFS (HR: 0.94, 95% CI: 0.72 – 1.22; p-value = 0.634; I 2 = 88.1%) or OS (HR: 0.85, 95% CI: 0.59 – 1.21; p-value = 0.363; I 2 = 79%) between the two regimens. One-year DFS (OR: 1.26, 95% CI: 0.71 – 2.23; p = 0.421; I 2 = 76.9%), 3-year DFS (OR: 1.18, 95% CI: 0.75 – 1.85; p = 0.483; I 2 = 85.3%), and 5-year DFS (OR: 1.34, 95% CI: 0.95 – 1.89; p = 0.101; I 2 = 74.4%) were similar between S-1 and XELOX. Likewise, 1-year OS, 3-year OS, and 5-year OS showed no significant differences between the two regimens: (OR: 0.97, 95% CI: 0.33 – 2.90; p = 0.961; I 2 = 86.6%), (OR: 1.20, 95% CI: 0.69 – 2.08; p = 0.517; I 2 = 85.3%), and (OR: 1.55, 95% CI: 0.99 – 2.42; p = 0.054; I 2 = 87.7%), respectively. Heterogeneity was moderate to high across outcomes, but results were robust on sensitivity analyses. Conclusions: Among gastric cancer patients following D2 gastrectomy, adjuvant S-1 demonstrates efficacy comparable to XELOX, with no significant differences observed in disease-free or overall survival across short- and long-term follow-up.
Clinical validation of a metabolomics-based multi-cancer early detection test.
e22512 Background: Multi-cancer detection (MCD) tests can improve the efficacy of existing screening tests by detecting a greater number of cancers. For exerting a clinical impact, however, it is important that the MCD is able to accurately detect early-stage cancers, a feature that is limited in currently available tests. We had previous developed an alternate approach that combined serum metabolomics with machine learning-powered data analytics. Using this approach we had developed an MCD test that could concurrently detect 30 cancers with high accuracy. The goal of the present study was to clinically validate the performance of this test. Methods: A prospective, multicenter, observational study was conducted in which de-identified blood samples were collected from 10,074 participants with ( n = 7246, 72%) and without ( n = 2828, 28%) cancer. A blinded arm was also included to validate test performance. Sensitivity and specificity of cancer detection, and the accuracy of tissue of origin (TOO) identification was measured. Results: The overall sensitivity obtained for cancer detection was 98.55% while the specificity was >99%. Importantly, detection of early-stage cancers (i.e. Stage I and II) was also achieved with high sensitivity, which ranged from 95% to 100% for the different cancers. The overall sensitivity obtained for TOO was 95.62%, with accuracies ranging from 90% to 100% for the early-stage cancers. Conclusions: This study validates that our serum-based MCD test is uniquely capable of detecting early-stages of diverse cancers with high sensitivity and specificity, while also ascribing TOO with high fidelity. This test can, therefore, potentially complement existing single-cancer screening tests and contribute significantly to the detection of early-stage cancers. Clinical trial information: CT/MD/2024/000007.
Applying a large language model to detect immune-related adverse events in electronic health records: Effect of integrating prescription and laboratory data on accuracy.
12133 Background: Accurate identification of immune-related adverse events (irAEs) is essential; however, manual review of electronic health records poses a significant burden on clinicians. Furthermore, conventional text-only approaches may fail to capture clinically important events. In this study, we evaluated whether a Large Language Model (LLM) with a multimodal approach could improve patient-level identification of irAEs. Methods: Among patients who received immune checkpoint inhibitors (ICIs) monotherapy at National Cancer Center Hospital, Tokyo, Japan between October 2014 and July 2023, a subset of patients was randomly sampled for analysis. A zero-shot prompt was designed to identify irAEs. Using a standalone local LLM (gpt-oss-20b), single-modal analysis utilized clinical notes alone, whereas multimodal analysis additionally incorporated systemic steroid prescriptions and laboratory data. Patient-level predictions were aggregated, using the threshold that maximizes the micro-averaged F1-score. The irAEs were clinically adjudicated by board-certified oncologists. Analyses were limited to major irAEs, and performance was evaluated using precision, recall, and the F1-score (α = 0.05). Results: The median number of cases reviewed per day by physicians was 18, whereas the LLM processed 25. During the study period, 296 patients (207 with irAEs and 89 without irAEs) were randomly sampled. Among patients with irAEs, irAE types included rash, pneumonitis, endocrine disorders, colitis, and hepatitis. For the identification of irAE occurrence, the precision of the multimodal approach was significantly higher than that of the single-modal approach (88.7% to 83.7%, p = 0.01). In organ-specific analyses, precision for colitis and rash and recall for endocrine were significantly improved. In addition, the multimodal approach reduced false-positive classifications from 33% to 22% and false-negative classifications from 37% to 35%. IrAE onset timing also differed significantly between the two approaches, with earlier identification by the multimodal LLM (−21.9 vs −17.7 days, p = 0.04); both approaches detected onset earlier than clinical adjudication. Conclusions: A multimodal LLM framework improves irAE detection accuracy and may reduce clinician workload. IrAEs with significant differences between single-modal and multimodal analyses. Precision Recall F1-score % 95%CI P % 95%CI P % Any irAE Multi 88.7 84.2-93.1 0.01 83.1 78.0-88.2 0.62 85.8 Single 83.7 78.7-88.8 82.1 76.9-87.3 82.9 Rash Multi 64.5 53.7-75.2 0.03 60.5 49.8-71.1 0.10 62.4 Single 58.9 48.7-69.1 65.4 55.1-75.8 62.0 Colitis Multi 80.4 69.5-91.3 <0.01 77.4 66.1-88.6 0.32 78.8 Single 61.8 50.2-73.3 79.2 68.3-90.2 69.4 Endocrine Multi 71.2 59.6-82.7 0.37 75.0 63.7-86.3 0.03 73.0 Single 76.6 64.5-88.7 64.3 51.7-76.8 69.9
Zelenectide pevedotin (BT8009) monotherapy in previously treated metastatic urothelial carcinoma (mUC): Update on Duravelo-1.
4563 Background: Zelenectide pevedotin (zele, formerly BT8009) is a highly selective Bicycle Drug Conjugate (BDC) targeting Nectin-4, a protein that is overexpressed in mUC and other cancers. Zele has a lower molecular weight (4.2 kDa) and a shorter plasma half-life (<1 hour) than antibody drug conjugates (ADCs), with potential to rapidly penetrate solid tumors and minimize healthy tissue exposure. Here, we present updated safety and efficacy results of zele monotherapy in enfortumab vedotin (EV)–naïve patients with mUC from the Phase 1/2 Duravelo-1 study (NCT04561362). Methods: Eligible patients had previously treated unresectable mUC, creatinine clearance ≥50 mL/min, prior anti-PD-1/PD-L1 exposure, had progressed on or were ineligible for platinum-based chemotherapy, and had not received EV. All patients treated with the recommended Phase 2 dose of zele monotherapy 5 mg/m 2 once weekly across the dose-escalation and dose expansion phases were included in the analysis. Results: As of July 1, 2025, 53 patients with a median age of 63 years (range 25–84) and a median of 2 (range 1–7) prior lines of therapy were included. Median time on treatment was 4.4 months and maximum duration was 24.9 months, and median follow-up time was 9.9 months (range 0.5–43.9). Median progression-free survival was 5.4 months (95% CI 2.3–7.1). Confirmed objective response rate (cORR) in all patients was 32% (n=17/53), and disease control rate (DCR) was 66% (n=35/53), including 2 complete responses and 15 partial responses (PR); 3 additional patients had unconfirmed PR. Of the 13 patients who had received prior taxane therapy, cORR was 31% and DCR was 69%, including 4 PR and 5 with stable disease. Median duration of response was 10.0 months (95% CI 5.3–16.6) among patients with confirmed responses (n=17). Nausea (36%), asthenia (28%), and diarrhea (28%) were the most common treatment-related adverse events (TRAEs). TRAEs of Grade ≥3 occurred in 25% of patients and serious TRAEs occurred in 9%. TRAEs led to treatment withdrawal in 4% of patients. Grade 3 TRAEs of clinical interest (AECIs) were single reports of skin reaction and hyperglycemia (n=1 each, 2%). Grade 1–2 treatment-related AECIs of peripheral neuropathy were reported in 47% of patients and were all sensory-based, with 1 patient requiring treatment withdrawal. Grade 1–2 treatment-related skin reaction and eye disorder AECIs were reported in 15% and 13%, respectively. Conclusions: Zele monotherapy at 5 mg/m 2 once weekly continues to demonstrate an encouraging efficacy profile and a promising safety profile with the potential to differentiate from ADCs in EV-naïve patients with mUC. Antitumor activity was preserved in patients with exposure to taxanes. Treatment withdrawal resulting from zele-related AECIs was uncommon. A Phase 2/3 study of zele in patients with locally advanced or mUC (NCT06225596; Duravelo-2) is currently active. Clinical trial information: NCT04561362 .
Photodynamic Therapy Combined with Biliary Drainage for Extrahepatic Cholangiocarcinoma: A propensity score–matched, single-center retrospective study.
e16279 Background: Inoperable extrahepatic cholangiocarcinoma (eCCA) carries a poor prognosis. This study evaluates the efficacy and safety of photodynamic therapy (PDT) combined with biliary drainage (BD) versus BD alone. Methods: We retrospectively analyzed 86 eCCA patients from June 2021 to June 2025. Patients received PDT+BD (n = 39) or BD alone (n = 47). A 1:1 propensity score matching (PSM) was performed to balance baseline characteristics, including age, gender, tumor location, and TNM stage. Primary endpoint was overall survival (OS); secondary endpoints were complications and stent removal rate. Results: After PSM (n = 22/group), all baseline variables were well-balanced ( p > 0.05). The PDT+BD group showed significantly superior median OS compared to the BD group (17.8 vs. 6.4 months; p = 0.011). The 1-year OS rate was 60.5% (95% CI: 48.5–72.5) for PDT+BD versus 16.4% (95% CI: 8.1–24.7) for BD ( p < 0.001). 10.3% (4/39) in the PDT group achieved drainage independence due to lumen relief, versus 0% in the BD group ( p = 0.039). Subgroup analysis confirmed that the survival benefit remained significant even for patients receiving only a single PDT session (mOS 17.8 vs. 6.4 months; p = 0.022). The overall complication rate post-PSM was identical between the PDT+BD and BD groups (18.2% vs. 18.2%; p = 1.000). Conclusions: PDT combined BD significantly prolongs OS in eCCA patients without increasing procedural risks. The higher rate of drainage independence highlights PDT's efficacy in local tumor control. Characteristics and clinical outcomes of patients before and after propensity score matching. Variable Pre-PSM PDT (n=39) Pre-PSM BD (n=47) P -value¹ Post-PSM PDT (n=22) Post-PSM BD (n=22) P -value² Tumor Location, n (%) <0.001 0.750 - Hilar Cholangiocarcinoma 16 (41.0%) 36 (76.6%) 15 (68.2%) 14 (63.6%) - Distal Cholangiocarcinoma 23 (59.0%) 11 (23.4%) 7 (31.8%) 8 (36.4%) TNM Stage, n (%) 0.188 0.728 - Stage III 30 (76.9%) 30 (63.8%) 17 (77.3%) 16 (72.7%) - Stage IV 9 (23.1%) 17 (36.2%) 5 (22.7%) 6 (27.3%) Median Overall Survival 17.8 mo 5.7 mo <0.01 17.8 mo 6.4 mo 0.011 1-Year Survival Rate, % 59.5% 22.0% <0.001 60.5% 16.4% <0.001 Stent Removal Rate, % (n/N) 10.3% (4/39) 0% (0/47) 0.039 — — — 1 P -value for comparison between PDT and Stent groups before PSM. 2 P -value for comparison between PDT and Stent groups after PSM.
Incidence and risk of musculoskeletal and rheumatologic (MSK-R) immune-related adverse events (irAEs) in a large real-world cohort.
e23409 Background: Immunotherapy (IT) improves cancer outcomes but is associated with irAEs, including MSK-R toxicities, which occur in 5–10% of patients, predominantly as arthralgias & myalgias, & may impair quality of life or lead to treatment interruption. Real-world comparative data on incidence & risk across tumor types remain limited. We evaluated the incidence & risk of MSK-R irAEs in patients receiving IT versus chemotherapy alone. Methods: We conducted a retrospective cohort study of adult cancer patients (n = 5,556) treated between 2010–2022 at Cleveland Clinic Ohio centers. Patients (≥18 years) with cancers eligible for first-line IT were included. Patients receiving IT (± chemotherapy ± radiation) were compared with those not receiving IT. Follow-up began at systemic therapy initiation & continued until irAE occurrence, last follow-up, or death. MSK-R irAEs included inflammatory arthritis, myositis, polymyalgia rheumatica, keratitis, salivary gland dysfunction, & vasculitis; arthralgia & myalgia were excluded. Data on attribution, CTCAE grade, management, IT interruption or rechallenge, & outcomes were collected. Incidence rates per 100 person-years were calculated using Poisson methods. Cumulative incidence (CIR) was estimated with Kaplan-Meier methods. Adjusted hazard ratios (HRs) were estimated using time-dependent Cox models treating IT exposure as a time-varying covariate. Results: Among the cohort, 1,291 (23%) & 948 (17%) received first- & second-line IT. Median age was 66 years; 38% were female & 87% were White. Anti-PD-1/PD-L1 &/or anti–CTLA-4 agents accounted for 99.9% of IT exposure. The cohort was dominated by advanced lung cancer, particularly stage IV non-small cell lung cancer (29%), followed by esophageal/gastric (stages II–III: 14%; stage IV: 11%) and other advanced solid tumors. During median follow-up of 14 months (CI: 6-34), 46 MSK-R irAEs were observed, including 23 events among patients receiving first-line IT. Incidence rates were higher in IT than non-IT group (0.09 vs 0.02 per 100 person-years; P < 0.05). Inflammatory arthritis accounted for most events (0.06 vs 0.01 per 100 person-years). The 24-month CIR of MSK-R irAEs was 2.3% with IT vs 0.6% in non-IT (log-rank P < 0.01; HR 7.1; P < 0.01). The 24-month CIR of inflammatory arthritis was 1.7% versus 0.2% (P < 0.01). Among inflammatory arthritis cases, 72% were CTCAE grade II–III; immunomodulatory therapy was used in 65%, mainly corticosteroids, with an 89% response rate. Conclusions: In this large real-world cohort, IT was associated with a substantially increased risk of MSK-R irAEs, driven primarily by inflammatory arthritis, although events were uncommon, consistent with prior reports. These toxicities often required immunosuppressive therapy and were associated with IT interruption or discontinuation, underscoring the need for multidisciplinary management.
National trends in mortality from gastric cancer with metastatic involvement of the digestive and respiratory systems in the United States.
e15694 Background: Gastric cancer remains a major contributor to cancer-related mortality in the United States (US), particularly in advanced stages. However, long-term national trends and sociodemographic disparities in mortality associated with metastatic involvement of the respiratory (RS) and digestive systems (DS) remain incompletely characterized. Methods: We conducted a national population-based analysis using the CDC WONDER Multiple Cause of Death database, including individuals aged ≥45 years from 1999–2023. Deaths with GC as the underlying cause (ICD-10 C16) and secondary metastases of the RS and DS as contributing causes (ICD-10 C78) were identified. Age-adjusted mortality rates (AAMRs) per 100,000 population were calculated and stratified by age, sex, race, ethnicity, geographic region, and urbanization. Temporal trends were evaluated using Joinpoint regression to identify inflection points and estimate annual percent change (APC) and average annual percent change (AAPC). Results: Our analysis showed that a total of 18,885 deaths were attributed to GC with metastatic involvement of the RS and DS, corresponding to an AAMR of 1.5 per 100,000. National mortality demonstrated a U-shaped temporal pattern, with a significant decline from 1999 to 2008 (APC -8.1%; 95% CI [-9.2; -6.9]), followed by a sustained increase through 2023 (APC +3.4%; 95% CI [2.6; 4.2]). Among individuals aged ≥ 65 years, mortality declined significantly during the early study period (APC -8.3%; 95% CI [-11.6; -6.2]), before reversing after 2008 (APC 3.9%; 95% CI [2.7; 5.7]). Both men and women exhibited similar trend reversals, with significant post-2008 increases (men: APC 5.7%; 95% CI [4.3; 10.4], and women: APC 5.9%; 95% CI [4.6; 8.2]. Racial and ethnic disparities persisted throughout the study period. Non-Hispanic Black and Asian/Pacific Islander populations demonstrated higher AAMRs and steeper post-inflection increases compared with non-Hispanic White populations, with a post-2008 APC range approximately 4.0% to 6.5%. Urbanization analyses revealed significantly increasing mortality in large central metropolitan areas (APC 4.7%, 95% CI [3.6; 6.2]), whereas nonmetropolitan regions exhibited declining mortality (AAPC -1.05%; 95% CI [-1.9; -0.2]). Conclusions: After a decade of declining mortality, deaths associated with metastatic GC involving the RS and DS have risen steadily in the US since the late 2000s. This reversal disproportionately affects older adults, ethnic minority populations, and residents of large metropolitan areas, highlighting widening inequities in advanced disease outcomes. These findings suggest that prior gains in GC mortality have not been sustained and underscore the need for targeted strategies focused on earlier detection, equitable access to specialized care, and improved management of advanced disease in high-risk populations.
Clinical efficiency of remote symptom monitoring using electronic patient-reported outcome (ePRO) in lung cancer: Results of the prospective, multicenter REAL-MOOV-LUNG trial.
11107 Background: Remote symptom monitoring using electronic patient-reported outcome (ePRO) has been shown to provide clinical benefits and is being implemented for routine care in cancer patients (pts). It may represent a significant operational burden for healthcare providers, that has to be balanced with its actual efficiency in generating clinically meaningful alerts leading to modify pts management. Methods: REAL-MOOV-LUNG is a prospective, multicenter, phase IV trial evaluating the MOOVCARE system in patients (pts) ≥18yo with non-progressive lung cancer, any histology, any stage, who had completed anticancer treatment for < 8weeks (cohort 1), or who were receiving adjuvant, consolidation or maintenance anticancer treatment for ≥8weeks (cohort 2). Primary endpoint was the proportion of patients who had a clinically meaningful modification of the management, predefined as unplanned consultation at the hospital or CT-scan imaging following a MOOVCARE alert. Secondary endpoints included management of alerts, patients and healthcare providers satisfaction, observance, quality-of-life, and survival. Results: From October 2021 to October 2024, 188 pts were enrolled, including 107 men, 81 women, 155 pts with non-small cell lung cancer, 36 with other histologies; 36 pts had oncogene-addicted tumors; 70 pts had metastatic disease, 58 locally-advanced disease, and 60 had early-stage disease. 92% of pts filled at least 1 ePRO questionnaire, median number of questionnaires was 56/pts; 81% of pts had at least 1 alert, median number of alerts was 11, and total number of alerts was 2294. After a median follow-up of 23.9 mo, 39 (21%) pts had a modification of management between 2 scheduled visits, 29 (74%) of which related to preplanned visit, not to an alert; among the remaining 10 (26%) pts, only 5 (13%) pts had clinically meaningful modifications as per the predefined definition. 18-month overall survival was 79.6% [IC95% 73.6%-86.1%]. Conclusions: While adoption of remote symptom monitoring on ePRO was high, with frequent alerts generated from surveys, the actual impact on clinical management of patients with lung cancer was limited in the context of frequent preplanned visits at the hospital as per clinical guidelines. Clinical trial information: NCT04934865 .
A retrospective analysis of outcomes in metastatic solid tumors harboring alterations in the PI3K pathway as stratified by the presence or absence of diseases of insulin resistance.
e22668 Background: Epidemiologic links are established between diseases of insulin resistance/hyperinsulinemia [pre-diabetes mellitus (P-DM) and type 2 diabetes mellitus (T2DM)] and increased cancer incidence and worsened oncologic outcomes. The phosphoinositide 3-kinase (PI3K) pathway is central to insulin signaling. Hyperinsulinemia may cooperate with somatic alterations in this pathway to impact clinical outcomes. We hypothesize that patients with either gain of function (GOF) alterations or loss of function (LOF) of major regulators/suppressors (R/S) in the PI3K pathway may have worse clinical outcomes in the setting of P-DM and T2DM. Methods: We performed a retrospective analysis of patients treated at MaineHealth with incurable solid tumors that harbored GOF alterations or LOF of R/S in the PI3K pathway from 1/1/2017-1/1/2024 (sequencing by Tempus (Chicago, IL) and Jackson Laboratory (Bar Harbor, ME)). Patients must have received at least one line of systemic therapy to be included. Overall survival (OS) was measured from the start of systemic therapy to death or last known follow up. Log-rank (Mantel-Cox) test was used for survival analysis. Results: 109 patients were included, 41 with P-DM or T2DM (Insulin Resistant (IR) group) at the time of diagnosis and 68 without (non-IR Group). Median age was 66 years with 59% women in the IR group and 65 years with 68% women in the non-IR group. Mean BMI was 31.0kg/m2 in the IR group and 27.7 kg/m2 in the non-IR group (p=.016). Mean glycosylated hemoglobin (HbA1c) was 7.0% in the IR group and 5.3% in the non-IR group. The most common malignancies in the IR group were NSCLC (27%), colorectal (20%), breast (12%), endometrial (10%), and ovarian (3%) cancer, in contrast to NSCLC (18%), breast (18%), colorectal (12%), endometrial (12%), HNSCC (10%), and ovarian (9%) in the non-IR group. The most common genetic alterations in the IR group were point mutations in PIK3CA (39%), PTEN loss (37%) or PIK3R1 loss (17%), in contrast to PIK3CA (59%), PTEN loss (37%) and PIK3R1 loss (3%) in the non-IR group. For patients with recorded values, ECOG was 0-1 in 76% of the IR group and 88% in the non-IR group. Median survival was 774 days (d) in the IR and 1187d in the non-IR group (Hazards ratio (HR)= 1.65, 95% Confidence interval [CI](1.02 to 2.69). In multivariate Cox regression, ECOG was independently associated with survival (ECOG 2–3 vs 0–1: HR 2.50, p=0.003), while female sex was associated with lower hazard (HR 0.52, p=0.016). IR (HR 1.42, p=0.174), age (HR 1.03, p=0.35), and BMI (HR 0.97, p=0.24) were not statistically significant predictors of survival. Conclusions: Patients with metastatic solid tumors harboring GOF or LOF alterations in the PI3K pathway who had P-DM or T2DM showed a trend toward worse OS, although this did not reach statistical significance in multivariate analysis.
Quality of life analysis from a multicenter, randomized, phase 2, investigator-initiated ETCTN trial of olaparib + radium-223 versus radium-223 in metastatic castration-resistant prostate cancer with bone metastases (COMRADE).
5040 Background: Radium-223 is an α-emitting radioisotope that improves overall survival in men with metastatic castration resistant prostate cancer (mCRPC) and bone metastases (BM). In the ALSYMPCA trial, radium-223 was also associated with improved quality of life (QOL), delayed time to first opioid use, and pain palliation. We previously reported superior radiographic progression free survival (rPFS) with the addition of olaparib to radium-223 in the randomized phase 2 COMRADE study (NCT03317391). We now report patient-reported QOL outcomes from the trial. Methods: Patients were randomized 1:1 to olaparib 200 mg twice daily plus radium-223 (Arm A) or radium-223 alone (Arm B). QOL was assessed using the Functional Assessment of Cancer Therapy-Prostate (FACT-P) and Brief Pain Inventory (BPI) at baseline and every 12 weeks. Mixed-effects ANCOVA models assessed treatment effects on QOL changes from baseline, adjusting for baseline score, timepoint, age, and Eastern Cooperative Oncology Group performance status. Least-squares (LS) mean changes were reported for each arm with between-arm differences and 95% confidence intervals. Analysis was limited to patients with baseline and ≥1 post-baseline assessment within 24 weeks. Results: Of 114 treated patients, 74 (65%) were evaluable for QOL analysis (Arm A, n = 40; Arm B, n = 34), including 71 (62%) for FACT-P and 72 (63%) for BPI pain severity and pain interference. Baseline characteristics were well balanced: median age was 70 and 72 years, ECOG PS 1 was present in 60% and 62%, prior docetaxel was received by 55% and 50%, > 20 bone metastases were present in 45% and 47%, and pain medication use at baseline was reported in 60% and 72% of patients in Arms A and B, respectively. There were no significant differences in LS mean change from baseline in FACT-P total score between arms at week 12 (Arm A: -5.09 vs Arm B: -0.23; difference -4.87, 95% CI -13.4 to 3.62) or week 24 (Arm A: -1.41 vs Arm B: -3.20; difference 1.79, 95% CI -8.28 to 11.9). Similarly, no significant between-arm differences were observed in LS mean changes in BPI pain severity at week 12 (difference 0.44, 95% CI -0.44 to 1.33) or week 24 (difference -0.20, 95% CI -1.32 to 0.93). Arm B showed greater improvement in BPI pain interference at week 12 (LS mean change: Arm A +0.21 vs Arm B -0.88; difference 1.09, 95% CI 0.14 to 2.05), but this difference was not sustained at week 24 (difference -0.27, 95% CI -1.37 to 0.82). Conclusions: In this phase 2 trial, the addition of olaparib to radium-223 achieved superior rPFS without significant detriment to patient-reported QOL or pain compared to radium-223 alone, supporting the tolerability of this combination in mCRPC patients with BM. Clinical trial information: NCT03317391 .
Real-world study of adebrelimab for immune resumption after immune-related adverse events (irAEs).
e23334 Background: "Management of Immunotherapy-Related Toxicities. NCCN Guidelines V2.2024" suggests considering restarting immunotherapy after resolution of some grade 2-3 irAEs to ≤ grade 1. Adebrelimab, a humanized PD-L1 inhibitor, demonstrates a favorable safety profile due to its unique structural optimization. In the real world, whether patients can continue to derive benefit from restarting immunotherapy after experiencing irAEs remains worthy of further exploration. Methods: This was a prospective, open-label, real-world study conducted across 83 hospitals in Jiangsu Province, China. Adult patients (≥18 years) with advanced solid tumors who developed irAEs during ICI treatment without disease progression were enrolled when irAEs resolved to ≤ grade 1. They received treatment based on adebrelimab. The primary endpoint was the 2-year survival rate. Secondary endpoints included progression-free survival (PFS), overall survival (OS), duration of response (DoR), objective response rate (ORR), and safety. Results: From April 16, 2025, to July 28, 2025, 141 patients were enrolled. Among them, 106 (75.18%) were male; the median age was 69 years (range, 40-93); 133 (94.33%) had an ECOG PS of 1; diagnoses included SCLC (30 patients, 21.28%), ESCC (26, 18.44%), NSCLC (22, 15.60%), and HCC (15, 10.64%); 118 patients (83.69%) were in second-line therapy; 131 patients (92.91%) had received anti-PD-1 therapy. At the data cutoff, the median duration of follow-up was 4.9 months (95% CI, 4.77-5.1). Treatment regimens included adebrelimab monotherapy (41 patients, 29.08%), combination with chemotherapy (68, 48.23%), combination with antiangiogenic agents (24, 17.02%), combination with antiangiogenic agents and chemotherapy (4, 2.84%), and combination with other therapies (4, 2.84%). The overall response rate(ORR) was 17.73%, with confirmed 1 (0.71%) complete response (CR) and 24 (17.02%) partial responses (PR). Stable disease (SD) was observed in 88 patients (62.41%), resulting in a disease control rate (DCR) of 80.14%. The median PFS was 5.23 months (95% CI, 5.2-NA). The median OS was not reached yet, the 6-month OS rate was 83% (95% CI, 75.9%-90.8%). After immune Resumption, the incidences of all-grade and grade ≥3 treatment-related adverse events (TRAEs) were 99.29% and 22.70%, respectively. Regarding irAEs post-resumption: 17.02% of patients experienced the same irAE, 17.73% experienced a different irAE, 2.84% experienced both the same and a different irAE, and 62.41% did not experience an irAE. The irAEs more likely to recur were rash, fatigue, and pneumonitis. Newly occurring irAEs included hepatotoxicity, thyroid dysfunction, fever, pneumonitis, and rash. Conclusions: Adebrelimab is safe and effective for immune resumption after irAEs. Clinical trial information: ChiCTR2500101774.