TPS616 Background: Biliary tract cancers (BTCs) are aggressive malignancies with limited treatment options and poor prognosis. Since the ABC-02 trial established gemcitabine plus cisplatin (Gem/Cis) as the standard first-line regimen, overall survival has remained unsatisfactory. Recent phase 3 trials, such as TOPAZ-1 and KEYNOTE-966, demonstrated that the addition of immune checkpoint inhibitors (ICIs) to Gem/Cis improved survival, yet the incremental benefit was modest. Nab-paclitaxel has been shown to disrupt desmoplastic stroma and enhance drug delivery in pancreatic cancer and BTCs, although its efficacy in phase 3 trials was offset by toxicity. Tislelizumab, a novel PD-1 inhibitor, has demonstrated robust activity across multiple tumor types. We therefore designed the GENTIS trial (NCT06893380) to evaluate whether strategic low-dose nab-paclitaxel combined with Gem/Cis and tislelizumab could optimize efficacy while maintaining tolerability in treatment-naïve patients with advanced BTC. Methods: GENTIS (NCT06893380) is a prospective, multicenter, open-label, phase 1b/2 trial. In phase 1b, a 3+3 dose-escalation design determines the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D) of nab-paclitaxel combined with fixed-dose Gem/Cis and tislelizumab. In phase 2, a single-stage design with 49 patients is being conducted to assess efficacy at the RP2D. Eligible patients are ≥19 years with histologically confirmed, locally advanced, or metastatic BTC, ECOG performance status 0–1, and measurable disease by RECIST v1.1. Treatment consists of tislelizumab 200 mg on day 1 every 3 weeks, gemcitabine 1000 mg/m² and cisplatin 25 mg/m² on days 1 and 8, and nab-paclitaxel at 50–100 mg/m² on days 1 and 8 of a 21-day cycle. The primary endpoints are MTD and RP2D in phase 1b, and objective response rate in phase 2. Secondary endpoints include progression-free survival, overall survival, disease control rate, duration of response, safety, and quality of life. Exploratory endpoints include biomarker analyses from tumor tissue and peripheral blood. Results: This trial is ongoing, with enrollment planned for 6–12 patients in phase 1b and 49 patients in phase 2 across seven Korean institutions. The first patient was enrolled in May 2025, and study completion is expected by April 2027. Conclusions: GENTIS (NCT06893380) is the first investigator-initiated trial to evaluate the combination of Gem/Cis, low-dose nab-paclitaxel, and tislelizumab in the first-line treatment of advanced BTC. This study is designed to determine whether stromal modulation and PD-1 inhibition added to chemotherapy can improve clinical outcomes while minimizing toxicity. The results will provide important evidence for developing novel chemo–immunotherapy platforms in BTC. Clinical trial information: NCT06893380 .

Background & Aims:The membrane proteoglycan glypican-3 (GPC3) is expressed in ∼70%-80% of hepatocellular carcinomas (HCCs), and its soluble fragment is released into the circulation by protease cleavage. We investigated correlations between tumor and plasma GPC3 levels and their associations with clinical outcomes in advanced HCC. Methods:We analyzed 186 patients with advanced HCC treated between 2017 and 2023. Tumor GPC3 expression was assessed using immunohistochemistry and RNA sequencing, and plasma GPC3 by ELISA. Clinical outcomes were investigated in 106 patients receiving first-line atezolizumab plus bevacizumab. Results:Plasma GPC3 was detectable in 59.3% of patients (median 11.4 pg/ml), and correlated with tumor expression (H-score: R = 0.40, p <0.001). Plasma GPC3 positivity (>0 pg/ml) increased across H-score strata (p <0.001) but remained undetectable in 25.9% with H-scores >200. High plasma GPC3 (>10 pg/ml) was associated with shorter progression-free survival (3.4 vs. 9.4 months, p = 0.003), overall survival (12.3 vs. 30.6 months, p <0.001) and lower objective response rates (17.9% vs. 47.9%, p = 0.001). Compared to tumor values, plasma GPC3 better predicted 6- and 12- month outcomes, and was independently associated with worse progression-free survival (hazard ratio 1.70, 95% CI 1.07-2.69, p = 0.02) and overall survival (1.96, 1.12-3.40, p = 0.02) on multivariable analyses. Conclusions:While plasma GPC3 levels showed moderate concordance with tumor expression, meaningful discrepancies were also noted in a significant fraction of patients. High plasma GPC3 independently predicted worse survival outcomes in patients treated with atezolizumab plus bevacizumab, showing superior prognostic performance over tumor-based measures. Impact and implications:Glypican-3 (GPC3) is a promising biomarker and therapeutic target in hepatocellular carcinoma (HCC), yet the relationship between plasma and tumor GPC3 levels remains unclear, and their clinical relevance has not been well defined, particularly in patients receiving atezolizumab plus bevacizumab. Our study provides clinical evidence supporting plasma GPC3 as a non-invasive biomarker in HCC, showing that circulating GPC3 levels are associated with tumor expression but more closely linked to clinical outcomes, serving as an independent predictor of survival and treatment response in patients treated with atezolizumab plus bevacizumab. These findings suggest that plasma-based assessment may complement tissue-based evaluation and support real-time risk stratification in advanced HCC, although further validation in diverse, prospective cohorts is warranted.

Many real-world patients with advanced biliary tract cancer (BTC) are excluded by TOPAZ-1 criteria; this study assessed cisplatin and gemcitabine plus durvalumab (CGD) efficacy and safety in these patients. Data from 1358 patients with advanced BTC treated with first-line CGD across 55 international centers were retrospectively analyzed. Patients were classified as TOPAZ-1-in (meeting all inclusion criteria) or TOPAZ-1-out (meeting ≥ 1 exclusion criterion). Primary endpoints were overall survival (OS) and progression-free survival (PFS); secondary endpoints included objective response rate (ORR), disease control rate (DCR), and safety. Furthermore, OS and PFS of the TOPAZ-1-out cohort were compared with reconstructed survival data from the phase III trial. 912 (67.1%) patients were classified as TOPAZ-1-in, and 446 (32.9%) as TOPAZ-1-out. After a median follow-up of 14.5 months (95% CI: 13.7-36.5), OS was 16.1 months in TOPAZ-1-in and 12.5 months in TOPAZ-1-out (HR 0.69, 95% CI: 14.6-16.5, p = 0.0001); PFS was 8.2 versus 6.5 months (HR 0.73, 95% CI: 7.1-8.1, p < 0.0001). Median OS in the TOPAZ-1-out cohort (12.5 months) was nearly identical to the phase III trial experimental arm (12.9 months; HR 1.15, p = 0.13); for PFS the HR was 1.12 (95% CI 0.93-1.42; p = 0.09). Among TOPAZ-1-out patients, active infection, elevated bilirubin, and ECOG PS > 1 were linked to poorer OS, while no detrimental impact emerged for ALT/AST abnormalities, corticosteroid use, renal or hematologic parameters, or prior surgery within 6 months, suggesting treatment effectiveness was broadly maintained across clinical subgroups. Furthermore, TOPAZ-1-out patients experienced no significant increase in adverse events compared with the TOPAZ-1-in group. Real-world data suggest CGD may be effective in TOPAZ-1-out patients, with no evident increase in toxicity, supporting the potential expanded use of the regimen in clinical practice.

TPS4247 Background: Recurrent or metastatic gastric cancer (GC) progressing after standard first- and second-line therapies has limited treatment options, leading to poor survival outcomes. Nesuparib is an orally available, next-generation dual inhibitor of tankyrase and poly(ADP-ribose) polymerase (PARP). It disrupts DNA damage repair and induces a “BRCAness” phenotype by modulating Wnt/β-catenin and Hippo signaling pathways. In a previous phase I study, nesuparib monotherapy demonstrated promising clinical activity with an overall response rate (ORR) of 28.2% and a disease control rate (DCR) of 64.1%. Preclinical models have shown synergistic anti-tumor efficacy when nesuparib is combined with irinotecan. This study aims to evaluate the safety, tolerability, and preliminary efficacy of nesuparib plus irinotecan in patients with advanced GC. Methods: This multicenter, open-label, single-arm, dose-finding and expansion phase 1b/2 study enrolls adults with histologically confirmed recurrent or metastatic GC or gastroesophageal junction adenocarcinoma who have progressed after ≥2 prior systemic therapies and have ECOG PS 0–1, measurable disease per RECIST v1.1, and UGT1A1 wild-type genotype. Phase 1b follows a standard 3+3 design with multiple predefined dose levels. Each dose level consists of a fixed combination of (1) a specific nesuparib dose (25–100 mg) administered orally once daily on either a 5-days-on/2-days-off or 3-days-on/4-days-off schedule and (2) irinotecan administered every 2 weeks at 100, 120, or 150 mg/m². The dose-limiting toxicity evaluation window is 28 days. Phase 2 enrolls patients at RP2D to further evaluate anti-tumor activity. The primary endpoint is objective response rate. Secondary endpoints include safety, disease control rate, duration of response, progression-free survival, time to response, and overall survival. Exploratory objectives include population pharmacokinetics, pharmacodynamic biomarkers, homologous recombination deficiency profiling, and circulating tumor DNA analyses. Approximately 43–49 patients are planned (18–24 in phase 1b; 25 in phase 2). Treatment continues until disease progression, unacceptable toxicity, or study withdrawal. Clinical trial information: NCT07364422 .

596 Background: Glypican‐3 (GPC3) is a membrane proteoglycan expressed in ~80% of hepatocellular carcinomas (HCC). In addition, GPC3 is cleaved by proteases, shedding its N-terminal fragment into the circulation. While GPC3-targeted therapies are under clinical development, the relationship between tumor GPC3 and plasma GPC3 (pGPC3) remains unclear. Here, we investigated the correlation between tumor GPC3 and pGPC3 and their associations with clinical outcomes in HCC patients. Methods: This study included 186 patients with advanced HCC treated at CHA Bundang Medical Center (2017–2023). Tissue GPC3 levels were assessed by immunohistochemistry (IHC) and RNA-seq, while pGPC3 was assessed using ELISA in 166 patients. Clinical outcomes were investigated in 106 patients treated with first-line atezolizumab plus bevacizumab (atezo/bev). H-score groups were defined as: 0 (negative), 1–100 (low), 101–200 (intermediate), and 201–300 (high). Results: The median IHC extent of GPC3 was 80%, reflecting widespread expression in tumor tissue. The median pGPC3 level was 11.4 pg/mL, with 40.7% of patients undetectable. pGPC3 significantly correlated with H-score (R=0.40, p<0.001) and RNA-seq (R=0.28, p=0.006). pGPC3 positivity (>0 pg/mL) increased proportionally across H-score groups (p<0.001); however, even in the high group (201–300), 25.9% of patients had undetectable pGPC3. pGPC3 was higher in patients with BCLC stage C, PVTT, AFP ≥400 ng/mL and no prior local therapy. High pGPC3 (>10 pg/mL) was associated with shorter progression-free survival (PFS) (3.4 vs 9.4 months, p=0.003), overall survival (OS) (12.3 vs 30.6 months, p<0.001), and lower objective response rates (17.9% vs 47.9%, p=0.001). Time-dependent ROC showed that pGPC3 had a higher predictive accuracy than tumor GPC3 for PFS and OS at both 6- and 12-month time points. In multivariable analysis, high pGPC3 was independently associated with worse PFS (Hazard Ratio [HR]=1.68, 95% Confidence Interval [CI]: 1.06–2.66, p=0.03) and OS (HR=1.93, 95% CI: 1.11–3.36, p=0.02). Conclusions: pGPC3 showed positive correlation with tumor tissue GPC3 and high pGPC3 was associated with poor survival outcomes in patients with advanced HCC treated with atezo/bev.

e15018 Background: SHR-A1904, a novel CLDN18.2 targeted ADC, showed promising antitumor activity in pretreated CLDN18.2-positive gastric/gastroesophageal junction cancer (GC/GEJC) in a China-only phase 1 trial (Nat Med. 2025; NCT04877717). Here, we report a 2-part multicenter global study assessing SHR-A1904 in patients (pts) with CLDN18.2-expressing advanced solid tumors (NCT05277168). Methods: During dose escalation (DE), pts with CLDN18.2-expressing (H score ≥1 by central lab IHC) advanced relapsed or refractory (R/R) solid tumors were enrolled to receive SHR-A1904 at 0.6–6.0 mg/kg (Q3W IV) in an i3+3 design. During dose optimization (DO), pts with CLDN18.2-positive (≥50% of cells with 2+ or 3+ staining) advanced R/R GC/GEJC were enrolled to receive SHR-A1904 at 6.0 and 8.0 mg/kg. The primary endpoints were DLT and safety in DE, and efficacy and safety in DO. Exploratory subgroup analyses in Asians vs non-Asians were done at 3.6 mg/kg (the minimal effective dose) or higher. Results: As of Dec 1, 2025, 51 pts were enrolled from Australia, South Korea, the United States, and Moldova, including 41 with GC/GEJC, 9 with pancreatic cancer, and 1 with lung adenocarcinoma. 42 pts received SHR-A1904 at 3.6 mg/kg or higher, 24 of whom were Asian. All pts had prior therapy (≥2 lines, 72.5%). The median follow-up was 6.7 mo (range, 0.2–27.9). During DE, 1 DLT (grade 3 vomiting) occurred at 6.0 mg/kg. Among all pts, TRAEs were reported in 48 (94.1%) pts; the most common were nausea (66.7%), vomiting (52.9%), and fatigue (23.5%). Grade ≥3 TRAEs and serious TRAEs occurred in 21 (41.2%) and 9 (17.6%) pts. No TRAEs led to death. In Asians, TRAEs and Grade ≥3 TRAEs occurred in 22 (91.7%) pts and 11 (45.8%) pts; in non-Asians, TRAEs and Grade ≥3 TRAEs occurred in 17 (94.4%) pts and 9 (50.0%) pts. Overall, ORR, DCR, and CBR (CR + PR + SD ≥24 weeks) were 25.5%, 58.8%, and 33.3%, respectively. The median PFS, OS, and DoR were 2.8 mo (95% CI, 1.7–5.4), 9.8 mo (95% CI, 6.7–15.2), and 5.7 mo (95% CI, 2.8–NR), respectively. Exploratory subgroup analyses by race are shown in Table. After a single dose, C max and AUC of SHR-A1904 increased with the dose except the 4.8 mg/kg group. The mean t 1/2 of SHR-A1904 is around 4.3–7.2 days. Conclusions: SHR-A1904 showed tolerable safety and promising antitumor activity in pretreated CLDN18.2-expressing advanced solid tumors. Furthermore, exploratory subgroup analyses by race support global development of SHR-A1904 in both Asians and non-Asians. Clinical trial information: NCT05277168 . Efficacy summary. Overall (N=51) Asians (N=24) # Non-Asians (N=18) # ORR 25.5 (13; 14.3–39.6) 20.8 (5; 7.1–42.2) 44.4 (8; 21.5–69.2) DCR 58.8 (30; 44.2–72.4) 66.7 (16; 44.7–84.4) 66.7 (12; 41.0–86.7) CBR 33.3 (17; 20.8–47.9) 20.8 (5; 7.1–42.2) 66.7 (12; 41.0–86.7) DoR, mo 5.7 (2.8–NR) 8.5 (2.9–NR) 5.7 (2.8–NR) PFS, mo 2.8 (1.7–5.4) 2.8 (1.5–5.4) 9.7 (1.9–NR) OS, mo 9.8 (6.7–15.2) 9.8 (5.2–NR) 15.2 (9.0–NR) Data are % (n; 95% CI) or median (95% CI). # at 3.6 mg/kg or higher.

Purpose EP4, a key receptor in the PGE 2 axis, mediates tumor immunosuppression; the EP4 antagonist ONO-4578 plus nivolumab showed manageable safety, immune activation, and preliminary antitumor activity in previously treated gastric/gastroesophageal junction cancer (G/GEJC). This study explored whether ONO-4578 enhances the efficacy of nivolumab plus chemotherapy in unresectable advanced or recurrent G/GEJC. Patients and Methods This multicenter, double-blind, randomized phase 2 study enrolled chemotherapy-naïve patients with HER2-negative unresectable advanced or recurrent G/GEJC. Patients were randomized (2:1) to receive oral ONO-4578 or matching placebo, each in combination with nivolumab and oxaliplatin-based chemotherapy. The primary endpoint was investigator-assessed progression-free survival (PFS). Secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. Results At the data cutoff, 226 patients were randomized to the ONO-4578 group (n = 150) and the placebo group (n = 76). Adding ONO-4578 significantly improved PFS (hazard ratio of 0.67; 90% confidence interval, 0.48–0.92; p-value, 0.040 [prespecified two-sided α = 0.10]), with favorable OS at a prespecified analysis with limited follow-up (hazard ratio of 0.60; 95% confidence interval, 0.37–0.96) and ORR (62.0% vs 48.7%). Exploratory subgroup analyses suggested that ONO-4578 regimen provided greater benefit in PD-L1 CPS ≥1, whereas no clear benefit in CPS <1/indeterminate patients. In an extended follow-up exploratory OS analysis with a minimum follow-up of 16.1 months, OS numerically favored ONO-4578 regimen. Common treatment-emergent adverse events in the ONO-4578 group were diarrhoea (55.7% vs 45.3%), and anaemia (55.0% vs 34.7%). Conclusion This study demonstrated promising efficacy and acceptable safety of an ONO-4578 regimen as first-line treatment for HER2-negative unresectable advanced or recurrent G/GEJC. These findings warrant confirmation in a phase 3 trial.

4007 Background: Anti-PD-1 antibodies combined with chemo are the standard 1L treatment for HER2-negative G/GEJ cancer. Prostaglandin E 2 (PGE 2 )-EP4 signaling is known to induce immunosuppressive tumor microenvironment, by inducing the differentiation of MDSCs and M2 macrophages, potentially contributing to resistance to immunotherapy. ONO-4578, an EP4 antagonist, is expected to modulate tumor immunosuppression through inhibiting the PGE 2 -EP4 signaling and to enhance the efficacy of anti-PD-1 antibody. The addition of ONO-4578 appeared to augment the efficacy of NIVO in patients with G/GEJ cancer in the previous phase 1 trial. This study aimed to evaluate the efficacy and safety of ONO-4578 combined with NIVO and chemo compared with placebo combined with NIVO and chemo as 1L treatment for patients with unresectable adv/rec G/GEJ cancer. Methods: ONO-4578-08 study is a randomized, double-blind, phase 2 trial conducted in Japan, Korea, and Taiwan. Chemo-naïve patients with HER2-negative adv/rec G/GEJ cancer were randomized in a 2:1 ratio (ONO-4578 group:placebo group), stratified by PD-L1 expression level (CPS < 5 vs CPS≥5), ECOG performance status (0 vs 1), presence of peritoneal metastasis (yes vs no). Patients in each group received ONO-4578 40 mg or placebo once daily, and all received NIVO 360 mg every 3 weeks and chemo (SOX [S-1 + oxaliplatin]/CAPOX [capecitabine + oxaliplatin]). The primary endpoint was investigator-assessed progression-free survival (PFS), powered (two-sided α = 0.10) to detect a HR of 0.65 with 117 events in a planned enrollment of 210 patients; secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. Results: Between December 2023 and September 2024, 226 patients were randomized to each group (150 vs 76 patients). With a median follow-up of 8.5 months, PFS was significantly longer in the ONO-4578 group than the placebo group with a hazard ratio (HR) of 0.67 (90% confidence interval [CI]: 0.48–0.92; P = 0.040; median PFS: 9.0 vs 6.9 months). At a minimum follow-up of 7.4 months, OS favored the ONO-4578 group (median OS: not reached vs 12.7 months; HR: 0.60 [95% CI: 0.37–0.96]). ORR was also higher in the ONO-4578 group (62.0% vs 48.7%). The most common treatment-emergent adverse events (TEAEs) in the ONO-4578 group were diarrhea (55.7% vs 45.3%), anemia (55.0% vs 34.7%), and peripheral sensory neuropathy (50.3% vs 46.7%). Serious TEAEs occurred in 53.7% of patients in the ONO-4578 group and 42.7% in the placebo group. Conclusions: ONO-4578 combined with NIVO and chemo significantly improved PFS compared with placebo combined with NIVO and chemo in treatment-naive patients with HER2-negative unresectable adv/rec G/GEJ cancer with no new safety concerns. Clinical trial information: NCT06256328 .

PURPOSE:EP4, a key receptor in the prostaglandin E2 axis, mediates tumor immunosuppression; the EP4 antagonist ONO-4578 plus nivolumab showed manageable safety, immune activation, and preliminary antitumor activity in previously treated gastric/gastroesophageal junction cancer (G/GEJC). This study explored whether ONO-4578 enhances the efficacy of nivolumab plus chemotherapy in unresectable advanced or recurrent G/GEJC. METHODS:This multicenter, double-blind, randomized phase II study enrolled chemotherapy-naïve patients with human epidermal growth factor receptor 2 (HER2)-negative unresectable advanced or recurrent G/GEJC. Patients were randomly assigned (2:1) to receive oral ONO-4578 or matching placebo, each in combination with nivolumab and oxaliplatin-based chemotherapy. The primary end point was investigator-assessed progression-free survival (PFS). Secondary end points included overall survival (OS), objective response rate (ORR), and safety. RESULTS:At the data cutoff, 226 patients were randomly assigned to the ONO-4578 group (n = 150) and the placebo group (n = 76). Adding ONO-4578 significantly improved PFS (hazard ratio [HR] of 0.67 [90% CI, 0.48 to 0.92]; P = .040 [prespecified two-sided α = .10]), with favorable OS at a prespecified analysis with limited follow-up (HR of 0.60 [95% CI, 0.37 to 0.96]) and ORR (62.0% v 48.7%). Exploratory subgroup analyses suggested that the ONO-4578 regimen provided greater benefit in PD-L1 combined positive score (CPS) ≥1, whereas no clear benefit was observed in those with CPS <1/indeterminate patients. In an extended follow-up exploratory OS analysis with a minimum follow-up of 16.1 months, OS numerically favored the ONO-4578 regimen. Common treatment-emergent adverse events in the ONO-4578 group were diarrhea (55.7% v 45.3%) and anemia (55.0% v 34.7%). CONCLUSION:This study demonstrated promising efficacy and acceptable safety of an ONO-4578 regimen as first-line treatment for HER2-negative unresectable advanced or recurrent G/GEJC. These findings warrant confirmation in a phase III trial.

BACKGROUND/AIMS:Immune checkpoint inhibitors (ICIs) have transformed advanced HCC treatment. The benefit of sequential immunotherapy after prior ICI failure remains unclear. Given the expanded use of atezolizumab plus bevacizumab (Ate/Bev) over the past 5 years, we explored the real-world outcomes of nivolumab plus ipilimumab (Nivo/Ipi) in patients with advanced HCC, with more focus on those previously exposed to Ate/Bev. METHODS:Patients treated with Nivo/Ipi for advanced HCC from six referral hospitals in Korea, Hong Kong, Taiwan and Singapore were included. Patients with prior non-Ate/Bev ICI or Child-Pugh B-C were excluded. Outcomes were compared between the ICI-naïve and Ate/Bev-experienced groups. RESULTS:Among 116 patients with advanced HCC treated with Nivo/Ipi, 57 were ICI-naïve and 59 had prior Ate/Bev exposure. Overall objective response rate was 31.2%, higher in the ICI-naïve group (42.6% vs. 20.0%, p = 0.01). However, the median duration of response was comparable between groups (24.8 vs. 23.7 months; p = 0.71), suggesting durable benefits regardless of prior Ate/Bev therapy. Median progression-free survival (PFS) and overall survival (OS) were 2.5 and 11.3 months, respectively, with longer PFS (5.3 vs. 1.6 months; p < 0.01) and OS (16.2 vs. 7.8 months; p = 0.06) in the ICI-naïve. Immune-related adverse events (irAEs), especially thyroid dysfunction, were associated with longer PFS and OS. Notably, most Nivo/Ipi responders post-Ate/Bev (8/11) had irAEs during Nivo/Ipi treatment, whereas no irAEs occurred during prior Ate/Bev. Nivo/Ipi responders post-Ate/Bev revealed a high tumour mutational burden (5.71-12.75 mutations/Mb). CONCLUSION:Nivo/Ipi demonstrated meaningful clinical activity in patients with advanced HCC, even after Ate/Bev failure.

Purpose:Unplanned readmissions of patients with cancer increase healthcare costs and disrupt care. Although well-studied in surgical oncology, data on patients receiving active treatment for advanced cancer remain limited. This study examined the causes, clinical characteristics, and outcomes of unplanned readmissions. Materials and Methods:This retrospective, multicenter study included patients with advanced solid tumors from six South Korean university hospitals who had unplanned readmissions within 1 month of prior hospitalization in 2019. Patients with terminal cancer who did not receive active treatment were excluded. Readmissions were categorized as Cancer Progression (e.g., worsening symptoms), treatment-related (e.g., therapy complications), or other (e.g., non-cancer conditions). Additional unplanned hospital use within 1 month post-discharge was analyzed in survivors with 6-month follow-up data. Results:Among the 542 patients, readmissions were classified as cancer progression (42.6%), treatment-related (37.3%), or other (20.1%). The cancer progression group had the longest hospital stay (median, 12 days) and the highest mortality (23.4%). The Treatment-Related group had shorter stays (8 days) and lower mortality (8.4%). Among the 445 survivors, 24.9% had unplanned hospital visits within 1 month post-discharge. Home discharge increased the likelihood of these events (adjusted odds ratio: 4.82 for readmissions, 2.65 for emergency department visits). Conclusion:Cancer progression was the leading cause of readmission and was associated with prolonged hospital stays and high mortality rates. Home discharge is a key predictor of early additional unplanned hospital visits, indicating the need for careful post-discharge monitoring in this population.

PURPOSE:Targeting aberrant metabolism in tumors with alterations in genes encoding Krebs cycle enzymes, a central component of glucose metabolism, is a promising therapeutic strategy. These tumors rely on aerobic glycolysis and promote VEGF-dependent angiogenesis; furthermore, EGFR signaling enhances aerobic glycolysis. This phase II trial evaluated bevacizumab and erlotinib in patients with solid tumors harboring Krebs cycle gene mutations. PATIENTS AND METHODS:Eligible patients had solid tumors bearing pathogenic mutations in fumarate hydratase (FH), isocitrate dehydrogenase 1/2, succinate dehydrogenase, or maleate dehydrogenase 2 and measurable disease per RECIST version 1.1 or Response Assessment in Neuro-Oncology 2.0 criteria. Participants received bevacizumab (10 mg/kg IV on day 1) and erlotinib (150 mg orally once daily) every 14 days until progression or unacceptable toxicity. The primary endpoint was the objective response rate (ORR); secondary endpoints were progression-free survival (PFS), overall survival, and safety. RESULTS:From February to November 2023, 35 participants were enrolled: 19 with biliary tract cancer (54.3%), seven with brain tumors (20.0%), and five with FH-deficient renal cell carcinomas (FH-deficient RCC, 17.1%). ORR was 37.1% (one complete and 12 partial responses). The disease control rate was 85.7%. Subgroup ORR were 80.0% in FH-deficient RCC, 36.8% in biliary tract cancer, and 28.6% in brain tumors. At a median follow-up of 11.7 months, the median PFS was 8.3 months; median overall survival was not reached. No new safety signals were observed. Exploratory transcriptomic analyses revealed VEGF and immune pathway enrichment in patients with favorable PFS, whereas amino acid, fatty acid metabolism, and oxidative phosphorylation-related pathways were enriched in those with poor PFS. CONCLUSIONS:Bevacizumab plus erlotinib demonstrated promising efficacy in tumors with Krebs cycle gene mutations, warranting further investigation beyond FH-deficient RCC.

Abstract Background: Current prognostic models for hepatocellular carcinoma (HCC) treated with atezolizumab plus bevacizumab (AB) rely on limited variables and lack prospective validation. We aimed to develop and externally validate a machine-learning model integrating multiple clinical variables to predict clinical benefit to first-line AB in advanced HCC. Methods and Results: This multicenter study included 637 patients with unresectable HCC from three hospitals and one phase III prospective trial (IMbrave150), grouped into four AB cohorts. The training set comprised patients from CHA Bundang Medical Center (Korea, n=301) and the Medical University of Vienna (Austria, n=53), while external validation used IMbrave150 (n=99) and Severance Hospital (Korea, n=184) cohort. Clinical benefit (CB) was defined as CR, PR, or SD with PFS ≥6 months by RECIST v1.1; all other cases were classified as non-clinical benefit (NCB). Among 14 candidate variables, nine were identified by univariable Cox regression for OS and PFS. A recursive elimination procedure maximizing five-fold cross-validated AUC for NCB classification identified six optimal predictors—CRP, AFP, platelet, total bilirubin, lymphocyte, and neutrophil—which were used to train an XGBoost classifier, termed CAPTYN. In the training set, CAPTYN achieved an AUC of 0.93. SHAP-based interpretation showed that elevated CRP, AFP, and total bilirubin and reduced lymphocyte counts contributed to NCB, whereas platelet and neutrophil counts exhibited U-shaped associations. In external validation, CAPTYN achieved AUCs of 0.70 (95% CI, 0.59-0.81) in IMbrave150 cohort and 0.67 (0.59-0.75) in Severance cohort, outperforming CRAFITY, ALBI, and CRAPT-M (DeLong's test p<0.05). Calibration was acceptable (Brier score=0.22 and 0.24, respectively), and CAPTYN significantly stratified OS and PFS (IMbrave150 cohort: both p<0.001; Severance cohort: p=0.012 for OS, p=0.009 for PFS), whereas comparator models failed to discriminate PFS. Subgroup analyses across demographics and disease features in IMbrave150 consistently showed higher hazard ratios (>1.5) for OS and PFS in CAPTYN-predicted NCB patients. Conclusion: CAPTYN, a six-variable machine-learning model predicting CB to AB, was externally validated using a prospective trial and a real-world cohort, providing calibrated, interpretable probabilities that may inform individualized treatment decisions. Citation Format: Gae Hoon Jo, Sohyun Hwang, Bernhard Scheiner, Won Suk Lee, Beodeul Kang, Jung Sun Kim, Ho Yeong Lim, Chansik An, Dong Yun Kim, Inyoung Kim, Dong-hyuk Heo, Matthias Pinter, Beom Kyung Kim, Chan Kim, Hong Jae Chon. CAPTYN, a six-variable machine-learning model predicting clinical benefit of atezolizumab-bevacizumab in hepatocellular carcinoma: Development and external validation in IMbrave150 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4221.

Abstract Background: Combination immunotherapy of nivolumab plus ipilimumab (Nivo/Ipi) has shown durable responses in advanced hepatocellular carcinoma (HCC). However, reliable biomarkers predicting clinical benefit remain limited. CD16+ (FcγRIIIa) NK cells mediate antibody-dependent cellular cytotoxicity (ADCC) and may influence the efficacy of Ipi, a human IgG1 monoclonal antibody. We investigated whether baseline peripheral CD16+ NK cell levels correlate with clinical outcomes in advanced HCC patients receiving Nivo/Ipi. Method: Patients with advanced HCC were prospectively enrolled from 4 tertiary cancer centers in Korea from Mar 2020 to Aug 2023. Patients received Nivo 1 mg/kg plus Ipi 3 mg/kg every 3 weeks (4 cycles) followed by Nivo maintenance (240 mg every 2 weeks). Peripheral blood mononuclear cell samples were collected at baseline and on day 22 (Cycle 2 Day 1 [C2D1]) and analyzed by flow cytometry. Patients were classified into high and low CD16+ NK groups based on a cut-point determined by maximally selected rank statistics. Single-cell RNA sequencing from 10 responders and 20 non-responders was performed to identify transcriptional correlates of treatment response. Results: 55 patients were enrolled and analyzed. Most patients were male (82%) with Hepatitis B virus infection (81.8%), and preserved liver function (Child-Pugh A 78.2%, B 21.8%). 74.5% had received ≥2 prior systemic therapies, and 23.6% were immunotherapy naïve. The median follow-up duration was 38.7 months (range, 18.3-50.1). Median progression-free survival (PFS) was 1.3 months (95% CI, 1.2-1.7), and median overall survival (OS) was 4.7 months (95% CI, 3.8-9.1). Although there were no differences in total NK cell numbers between responders and non-responders (p = 0.139), baseline CD16+ NK cells were significantly higher in responders. The high CD16+ NK group showed superior objective response rate (60% vs 17.8%, p = 0.012), disease control rate (70% vs 26.7%, p = 0.023), and longer PFS (14.4 months, 95% CI 1.4-not reached [NR], p = 0.017) and OS (14.7 months, 95% CI 1.4-NR, p = 0.079). High baseline CD16+ NK cell proportion were associated with the decline of regulatory T cells at C2D1, suggesting enhanced ADCC by Ipi. Consistently, single cell analysis also revealed higher baseline CD16+ expression in NK cells at baseline in responders than non-responders. Independent validation of these findings is ongoing. Conclusion: Baseline peripheral CD16+ NK cell levels were significantly associated with improved clinical outcomes in advanced HCC patients treated with Nivo/Ipi. High CD16+ NK cell frequency correlated with higher response rates and prolonged survival, potentially through enhanced ADCC and suppression of regulatory T cells. These findings suggest that baseline CD16+ NK cells may serve as a predictive biomarker for Nivo/Ipi in HCC, warranting further validation cohorts. Citation Format: Hongjae Chon, Narim Lee, Hannah Yang, Junho Kang, Won Suk Lee, So Jung Kong, Beodeul Kang, Jung Sun Kim, Ho Yeong Lim, Jung Kyoon Choi, Chan Kim. Baseline CD16+ NK cells are associated with clinical outcome of nivolumab plus ipilimumab immunotherapy in advanced hepatocellular carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7737.

Surgery after response to first-line chemoimmunotherapy may offer a potential curative option for patients with locally advanced biliary tract cancer (BTC) initially considered unresectable. However, robust real-world evidence in this setting remains limited. We retrospectively analyzed patients with locally advanced BTC treated with first-line cisplatin, gemcitabine, and durvalumab (CGD) across 55 centers in 12 countries. Endpoints included overall survival (OS), progression-free survival (PFS), disease-free survival (DFS) among resected patients, objective response rate (ORR), and the prognostic impact of surgery. A multivariable logistic regression model was developed to predict surgical conversion using baseline clinical and laboratory variables. Among 1358 screened patients, 219 had locally advanced disease and were included in the analysis. Median follow-up was 14.5 months. ORR was 34.6%, with median PFS and OS of 10.0 and 20.7 months, respectively. Twenty-four patients (10.9%) underwent surgical resection after systemic therapy. Median OS was 22.5 months in resected patients versus 19.9 months in those who did not undergo surgery. Median DFS after resection was 15.8 months. Pathological lymph node involvement was independently associated with shorter DFS. Larger baseline tumor size correlated with higher odds of radiological response but not with OS. An exploratory elastic-net model retained four baseline variables and showed moderate discriminative ability for surgical conversion, with an apparent AUC of 0.72 and an optimism-corrected AUC of 0.65. First-line CGD enabled secondary resection in approximately 11% of patients with locally advanced BTC. A simple baseline model may support patient selection for conversion surgery, although prospective validation is needed.

Introduction: The phase 1/2 CheckMate-040 and the phase 3 CheckMate-9DW trials consistently demonstrated meaningful activity of nivolumab plus ipilimumab (Nivo/Ipi) in previously treated and treatment-naïve patients with advanced HCC. However, patients with ≥50% liver involvement, Vp4 portal vein tumor thrombosis (PVTT), or bile duct invasion were excluded from these trials. We aimed to assess the real-world efficacy and safety of Nivo/Ipi in this high-risk population. Methods: This international, multicenter, retrospective study was conducted in patients with advanced HCC who received Nivo/Ipi at six hospitals in Korea, Hong Kong, Singapore, and Taiwan between 2016 and 2024. Patients with Child-Pugh B/C or ECOG >2 were excluded. Results: Of the 206 patients assessed for eligibility, 145 were included in the final analysis: 44 classified as high risk (Vp4 PVTT, n = 10; ≥50% liver involvement, n = 34; bile duct invasion, n = 7) and 101 as non-high risk. Baseline characteristics were comparable, aside from differences in liver function and high-risk features. Nivo/Ipi was administered as third-line or later-line therapy in 76.6% of patients. The high-risk group exhibited a significantly lower objective response rate (ORR) (12.8% vs. 34.3%) and shorter median progression-free survival (PFS) (1.2 vs. 2.9 months) and overall survival (OS) (3.6 vs. 16.5 months). Among the high-risk features, ≥50% liver involvement was strongly associated with inferior ORR, PFS, and OS, whereas Vp4 PVTT and bile duct invasion alone were not significantly associated with outcomes. A higher intrahepatic tumor burden was also inversely associated with treatment response and survival. Adverse events were more frequent in high-risk patients, particularly hyperbilirubinemia (36.4% vs. 13.9%; grade ≥3: 11.4% vs. 1.0%). On multivariable analysis, ≥50% liver involvement and prior immune-checkpoint inhibitor exposure independently predicted shorter PFS and OS. Conclusion: Extensive liver involvement emerged as the key predictor of poor prognosis, whereas Vp4 PVTT and bile duct invasion were not associated with significantly worse outcomes.