4023 Background: Sone-Ve (AZD0901), an antibody-drug conjugate comprising an anti-CLDN18.2 antibody and monomethyl auristatin E, showed promising antitumor activity and a manageable safety profile in pts with advanced or metastatic gastric/GEJ cancers in the first-in-human study conducted in China. Methods: CLARITY-PanTumor01 is an ongoing, global, phase 2 study (NCT06219941) evaluating Sone-Ve in gastric/GEJ cancers, PDAC, and BTC outside of China. Here we report data for pts with CLDN18.2+ advanced or metastatic gastric/GEJ cancers and ≤2 lines of prior systemic therapy (substudy 1) receiving Sone-Ve 2.2 mg/kg IV Q3W (pts randomized to 1.8 mg/kg are not reported). Primary endpoints were safety and objective response rate (ORR; 30 pts were selected to estimate ORR; lack of efficacy is concluded if ≤6 confirmed responses). Secondary endpoints included duration of response (DoR) and progression-free survival (PFS). Molecular responses (MRs) based on circulating tumor DNA (ctDNA) were evaluated. Results: As of Oct 31, 2025, 67 pts received Sone-Ve 2.2 mg/kg, including 30, 31, and 6 pts from the randomized (RC), paired biopsy (PBC), and Japanese safety (JSC) cohorts, respectively. All pts had adverse events (AEs; grade ≥3 in 34.3% overall, 36.7% of the RC, and 35.5% of the PBC). The most common grade ≥3 AEs were neutrophil count decreased (7.5%), anemia (7.5%), vomiting (6.0%), and nausea (6.0%). Dose reductions were most commonly due to gastrointestinal (GI) AEs and occurred in 40.3% of pts overall, 56.7% of the RC, and 19.4% of the PBC (adoption of a 4-drug antiemetic regimen improved GI tolerability in the PBC). The ORR was 28.4% (95% CI: 18.0–40.7). Median DoR has not yet been reached. Additional data are in the Table. Among evaluable pts, 58.6% (17/29) had sufficient baseline ctDNA levels to enable MR analysis; of these, 10 (58.8%) had at least a partial MR (>50% ctDNA reduction) within 2 cycles of Sone-Ve initiation. Conclusions: Sone-Ve 2.2 mg/kg demonstrated clinically meaningful efficacy and a manageable safety profile consistent with the first-in-human study. ctDNA MRs were observed in over half the evaluable pts. Clinical trial information: NCT06219941 . RC(n=30) PBC (n=31) Total*(n=67) ORR, % (95% CI) † 26.7 (12.3–45.9) 29.0 (14.2–48.0) 28.4 (18.0–40.7) Best overall response, n (%) † Complete response Partial response Stable disease Progressive disease Not evaluable 2 (6.7)6 (20.0)14 (46.7)7 (23.3)1 (3.3) 1 (3.2)8 (25.8)18 (58.1)4 (12.9)0 3 (4.5)16 (23.9)35 (52.2)12 (17.9)1 (1.5) 6-month response rate, % (95% CI) 87.5 (38.7–98.1) NC (NC–NC) 71.9 (39.2–89.1) PFS † Events, n/N (%) Median, months (95% CI) 6-month, % (95% CI) 22/31 (71.0)4.2 (2.9–8.6)44.8 (26.5–61.6) 19/31 (61.3)3.1 (2.8–5.5)30.3 (13.3–49.3) 47/68 (69.1)4.2 (3.0–7.0)39.9 (27.7–51.9) NC, not calculable. *Includes the JSC. † Investigator-assessed per RECIST v1.1.

Pancreatic ductal adenocarcinoma (PDAC) remains highly resistant to treatment, with mortality rates largely unchanged despite advances in cancer therapies. For ∼80% of borderline, non-resectable, or metastatic cases, chemotherapy is predominantly palliative, underscoring the need for improved drug delivery approaches. This study presents the development, characterization and in vivo evaluation of a novel polymeric implant loaded with 5-fluorouracil, irinotecan, and oxaliplatin (FIRINOX). Scanning electron microscopy of FIRINOX implants showed internal microstructure was preserved upon drug loading, while micro-CT and X-ray imaging revealed valuable insights into the morphology and degradation of implants retrieved from in vivo experiments. In murine PDAC models, dose-escalation identified 4 × FIRINOX implants as the maximum tolerated dose, while 2 × implants achieved significant therapeutic efficacy at lower doses than IV administration, without compromising animal safety. In healthy pigs, 20 × FIRINOX implants were well-tolerated, as confirmed by histopathology and blood analysis. Finally, laser ablation-inductively coupled plasma-mass spectrometry imaging and microbiome analysis confirmed localized drug perfusion within tissues, and minimal off-target effects, including preservation of gut microbiota diversity. These findings support the potential of this implantable platform to improve outcomes in borderline or non-resectable PDAC and enhance tolerability of cytotoxic chemotherapy through localized, controlled delivery, addressing a key gap where current treatment options are limited.

Glioblastoma remains one of the most lethal solid tumours, with recurrence driven by highly migratory residual cells that escape resection and resist adjuvant radiochemotherapy and arise from underlying genetic heterogeneity that promotes both a migratory phenotype and therapy resistance. These invasive populations exploit endogenous chemotactic cues, including hypoxia induced signalling, postoperative wound-healing cytokines, perivascular niche factors, and interstitial or cerebrospinal-fluid-driven transport, to disperse through the brain and re-establish tumour mass. While efforts to therapeutically inhibit chemokine pathways have been extensive, clinical trials targeting CXCR4, CCR2, CCR5, and CXCR1/2 have demonstrated limited survival benefit, in part due to ligand-receptor redundancy, network compensation, and the inherently spatial nature of chemotactic signalling.Emerging strategies instead seek to exploit chemotaxis by engineering chemoattractant gradients capable of directing glioblastoma cells toward a defined location, such as an implantable scaffold or hydrogel placed in the resection cavity. This review synthesises the biological principles underlying chemokine- driven migration, the spatial organisation of tumour- associated gradients, and the engineering considerations required to generate stable, localised chemotactic fields. We give an overview of biomaterial platforms designed for chemokine delivery and summarise the preclinical evidence supporting tumour cell trapping or redirection in vivo.Finally, we outline the major biological, engineering, and regulatory barriers to translation, including gradient stability, unintended immune recruitment, fluid-flow-driven distortion, chemokine instability, and sterilisation challenges. Together, these insights establish a conceptual and practical framework for using engineered chemotactic guidance as a therapeutic strategy and highlight opportunities for next-generation “migration-sink” devices to improve local control of glioblastoma.

Systemic chemotherapy and immunotherapy can disrupt gut microbial homeostasis, contributing to inflammation, treatment-related toxicity, and diminished anti-tumour immunity in pancreatic ductal adenocarcinoma (PDAC). Here, we evaluated whether localised delivery of chemo-immunotherapy via biodegradable implants could mitigate these adverse effects and preserve gut microbiota integrity. Using a syngeneic KPC mouse model of PDAC, we compared systemic versus implant-based delivery of gemcitabine/nab-paclitaxel and anti-CD40/anti-PD1 antibodies. 16S rRNA sequencing of faecal samples revealed that systemic chemo-immunotherapy significantly reduced alpha diversity, depleted immunoregulatory species (e.g. Akkermansia muciniphila, Bifidobacterium longum), and enriched pathobionts (Escherichia coli, Clostridium septicum), accompanied by elevated intestinal pro-inflammatory cytokines. In contrast, localised delivery preserved microbial diversity, maintained beneficial taxa and suppressed inflammatory cytokine levels. Further, high-dose localised chemotherapy promoted M1 macrophage polarisation while preserving microbiota more effectively than even low-dose systemic regimens. This is the first study to demonstrate that spatial control of drug exposure via localised delivery can protect the gut microbiome and modulate systemic immunity in PDAC. These findings subsequently provide proof-of-concept that implant-based approaches can enhance tolerability and efficacy of chemo-immunotherapy by minimising microbiome disruption.

Triple negative breast cancer (TNBC) is associated with the poorest prognosis among breast cancer subtypes and while immunotherapy has demonstrated some promise, its effectiveness as a monotherapy remains limited, offering only modest improvements in clinical outcomes when combined with chemotherapy. Immune checkpoint therapies, such as PD-1 inhibitors, elicit responses in a subset of patients with metastatic TNBC, and few patients experience durable effects. One reason for this limited efficacy may be due to defects in antigen presentation, as the deficiency of dendritic cells, essential for effective antigen presentation, correlates with inadequate anti-tumour immunity. To address this challenge, we report the development of a novel implantable drug delivery device that enables the localized administration of a CD40 agonist in combination with anti-PD1, doxorubicin and nanoparticle albumin-bound paclitaxel (nab-paclitaxel). CD40 agonists are a unique class of agents that activate antigen-presenting cells, including dendritic cells and B cells, and reprogram macrophages to support anti-tumour immunity. By enabling targeted delivery to the tumour site, we aimed to enhance immune priming while mitigating systemic toxicities often observed with combinations of intravenous chemo-immunotherapy. In a 4T1 murine model of TNBC, repeated systemic administration of the therapeutic combination led to fatal xenogeneic reactions, which were not observed with the localized delivery approach. Localized delivery also slowed tumour growth compared to systemic administration of the therapeutic compounds. Immune profiling further revealed that the addition of anti-CD40 agonist antibody promoted the activation of PD-1+ CD8+ T lymphocytes within the lymph nodes in both locally and systemically treated animals. However, the localized approach achieved equivalent or enhanced immune activation without inducing the fatal immune reactions observed with systemic dosing, suggesting that localized treatment can offer a significant therapeutic advantage. This study demonstrates that our innovative localized delivery approach has the potential to significantly improve patient outcomes by maximizing efficacy and minimizing adverse effects for this aggressive subtype of breast cancer.

183 Background: MNK inhibition has been shown to downregulate phosphorylation at serine 209 of eIF4E, a potent regulatory point of CAP-mediated translation of 5’ polyadenylated mRNA associated with growth factor and pro-oncogenic signals in cells. Elevated levels of eIF4E phosphorylation are observed in a broad range of tumors. MNK1/2 knock out mice are healthy and viable, as cellular house-keeping mRNA translation occurs readily using the IRES mechanism, whereas mRNA with more complex and longer 5’ untranslated ends such as those involved in growth factor and pro-oncogenic signaling, are MNK activated. Cells from MNK knockout animals become relatively resistant to subsequent oncogenic transformation. In a program of SAD and MAD Phase I studies conducted in normal healthy volunteers, Tinodasertib was found to be safe and well tolerated with no dose-limiting toxicity (DLT). The objectives of this ongoing Phase 2 study are to evaluate safety, and preliminary efficacy of Tinodasertib as monotherapy and in combination with either pembrolizumab or irinotecan in patients with advanced colorectal cancer (CRC). Methods: Patients had to have advanced CRC and previously received ≥2 lines of therapy. The dose escalation phase of the study was open to all patients with CRC. As of 23 July 2024, 22 patients were dosed in a modified 3X3 dose escalation design with Tinodasertib from 20 to 80 mg on alternate days. Out of 22 patients, 12 received monotherapy, 4 were treated with Tinodasertib and Irinotecan and 6 were treated with Tinodasertib and pembrolizumab. Majority (19) had MSS CRC and 3 had unknown status. Nine patients had KRAS-mut CRC. Results: No DLTs were observed and MTD was not reached. Grade 3 treatment-related adverse events (TRAEs) were observed in 2 (9%) patients and were related to irinotecan while no Grade 3 AEs were attributed to Tinodasertib by either the investigator or the sponsor. Most common TRAEs were related to gastrointestinal system organ class. There were no Grade 4-5 TRAEs. Of the 22 patients, 18 were evaluable for RECIST 1.1 response. No patient had an objective response to treatment, 12 had stable disease with disease control rate of 67% and a progression free survival of 2.99 months. Patients remained on therapy for up to 28 weeks. Overall survival (OS) at 52 weeks was 52% (CI 29 to 93). Conclusions: Tinodasertib either as monotherapy or combined with irinotecan or pembrolizumab, was well tolerated with no DLTs at the dose levels evaluated. Prolonged median time to progression and OS compared to historical controls were observed even during the dose escalation phase for the study population as a whole and for each of the monotherapy and combination arms. Enrolment in the dose escalation phase continues. Clinical trial information: NCT05462236 .

EMB-02 is a symmetric bispecific antibody targeting programmed cell death protein-1 and lymphocyte-activation gene 3 simultaneously. Here, we present the first-in-human study results of EMB-02 in patients with advanced solid tumors. Patients were treated with intravenous infusions of EMB-02 at doses of 6–900 mg. The primary objective was to evaluate the safety and tolerability and to determine the maximum tolerated dose and/or recommended phase II dose(s). Secondary objectives included characterizing the pharmacokinetic (PK) profile, assessing preliminary antitumor activity and the immunogenicity. A total of 47 patients were enrolled. All grade and grade 3/4 treatment-emergent and treatment related adverse events occurred in 97.9%, 48.9%, 68.1% and 12.8% patients, respectively. The objective response rate (ORR) was 6.4% and clinical benefit rate at 24 weeks (CBR-24) was 25.5% in overall population. The CBR-24 was 33.3% in checkpoint inhibitor (CPI)-naïve patients, and 15% in CPI-treated. No clear relationship was observed between the efficacy and PD-L1, LAG-3, or MHC II expression level. Doses 360 mg or higher resulted in sustained saturation of PD-1 receptors on circulating CD3 + T cells. EMB-02 demonstrated a favorable safety profile and early efficacy signals in multiple solid tumors, warranting further development. (NCT04618393).

Pancreatic ductal adenocarcinoma (PDAC) remains largely unresponsive to immune checkpoint blockade due to its immunosuppressive tumor microenvironment and the systemic toxicity associated with combination immunotherapies. Here, this work reports the development of a biodegradable implant, designed for the localized co‐delivery of CD40 agonist and anti–PD‐1 antibodies, and evaluate its therapeutic effect in an aggressive KPC mouse model of PDAC. This localized approach significantly reduces tumor burden compared to controls and markedly attenuated systemic toxicities relative to conventional systemic administration. Immune profiling via transcriptomic analysis and flow cytometry revealed tumor microenvironment reprogramming, characterized by elevated dendritic cell activation (CD86⁺CD11c⁺), increased CD8⁺ T cell infiltration, and reduced expression of genes associated with myeloid‐derived suppressor cells. Histopathology confirms improved tolerability. These findings demonstrate that localized co‐delivery of CD40 and PD‐1 immunotherapy via a biodegradable implant promotes anti‐tumor immunity while reducing systemic toxicity and immunotherapy‐associated complications. This strategy offers a promising alternative to systemic immunotherapy for PDAC and may help overcome the barriers of toxicity and immune resistance in this disease.

12 Background: Adjuvant chemotherapy (CT) following neoadjuvant chemoradiation and surgery for locally advanced rectal cancer (LARC) is widely adopted, despite uncertain survival benefit. Circulating tumor DNA (ctDNA) detection after surgery has been shown to be a strong prognostic marker in localized colorectal cancer and potentially could inform adjuvant treatment decision making. Methods: AGITG DYNAMIC-Rectal is a multi-centre randomized controlled phase II trial. Eligible patients (pts) had LARC (cT3-4 and/or cN+) treated with neoadjuvant chemoradiation, total mesorectal excision, and were fit for adjuvant CT. Pts were randomly assigned 2:1 to ctDNA-guided management or standard management (clinician decision). A tumor-informed personalized ctDNA assay was used. For the ctDNA-guided group, a positive result at 4 and/or 7 weeks after surgery prompted 4 months of oxaliplatin-based or fluoropyrimidine CT; for ctDNA-negative patients, no chemotherapy if ypN0 or clinician’s choice if ypN+. The primary endpoint was adjuvant CT use. A key secondary endpoint was non-inferiority in RFS rate at 3 years. The target sample size of 408 would provide 80% power with 95% confidence to demonstrate non-inferiority between the two arms with a margin of at most 10%. Results: The study ceased recruitment prematurely due to COVID-19 and increasing adoption of total neoadjuvant therapy. 230 eligible pts were enrolled from Jul 2018 to Nov 2021, median follow-up was 37 months. In the 155 ctDNA-guided patients, ctDNA analysis was successful in 150 (97%) pts and 42 (28%) were ctDNA-positive. Fewer pts in the ctDNA-guided arm received adjuvant CT (71/155, 46%) compared to 58/75 (77%) pts receiving standard management (p < 0.001). Overall, an oxaliplatin-based doublet was administered in 43 (28%) and 19 (25%) of ctDNA-guided compared to standard management pts (P = 0.82). 3-year recurrence-free survival for ctDNA-guided and standard management was 76% and 82% respectively (difference 6%, 95% CI: -6% to 17%). Cumulative probability of distant and locoregional recurrence at 3 years for ctDNA-positive pts treated with CT were 36% and 11% respectively; for ctDNA-negative pts without adjuvant CT were 12% and 1%. Of the 15 ctDNA-negative pts who recurred, 12 (80%) were lung only, 1 (7%) peritoneal and lung, and 2 (13%) nodal only relapse. Conclusions: A ctDNA-guided approach to adjuvant therapy for LARC post neoadjuvant chemoradiation and surgery was associated with a reduced rate of CT administration. The small sample size precludes any conclusions to be drawn about the non-inferiority of ctDNA-guided vs standard management. The recurrence rate in treated ctDNA positive pts appears low when compared to untreated historical controls. Our data confirms a lower risk of recurrence in pts with undetectable post-op ctDNA, with a notable proportion of these being in the lung. Clinical trial information: ACTRN12617001560381 .

107 Background: Recurrence rates following upfront resection of pancreatic adenocarcinoma are high, with some benefit from adjuvant chemotherapy (AC). A biomarker that improves risk stratification and/or provides real time indication of AC benefit could improve routine clinical management and accelerate trial progress. Previous studies in pancreas cancer suggest that patients with detectable ctDNA post surgery are at an elevated risk of recurrence. Detectable ctDNA at the completion of AC may also be associated with an elevated recurrence risk. Methods: Patients with early stage pancreatic adenocarcinoma were enrolled following upfront resection at 26 Australian centres. Patients were ECOG 0-1 and fit for AC. A tumour-informed ctDNA assay was used to identify somatic mutations for tracking in circulating cell-free DNA. ctDNA+ patients received 6 months of AC (FOLFIRINOX or gemcitabine/capecitabine selected at the clinician's discretion), while ctDNA- patients could de-escalate to 3 months at the clinician’s discretion. ctDNA was assessed again at the end of AC. The primary study endpoint was the feasibility of ctDNA-guided therapy. Secondary endpoints included the association of ctDNA with clinicopathologic risk factors and survival outcomes. Results: A total of 102 patients were enrolled from March 2019 to Nov 2023. Median age was 68 years (range 41 – 86), with 50% male and 95% ECOG 0-1. Tumours were located in the head of pancreas in 72%. Histology revealed T1 (18%), T2 (50%), and T3 (32%) tumors, with nodal involvement in 71%, and an R0 resection in 77%. Post-operative Ca19-9 was elevated in 29%. Forty patients (40%) were ctDNA+ve post resection, 54 (53%) were ctDNA-ve, and no result was obtained in 4 (4%) due to the absence of tumor mutation, and as a result, they were considered ineligible. The presence or absence of ctDNA was not associated with known clinicopathologic risk factors. Median time to ctDNA collection was 5 weeks (range 3 – 9) and to commencing AC was 6 weeks (range 4-12). Of 54 ctDNA-ve patients, 24 (44%) were de-escalated to receive a planned 3 months of AC. With a median follow-up of 36 months (range 2-56) the median recurrence free survival (RFS) in ctDNA+ve patients was 13 months compared to 22 months for ctDNA-ve patients (HR 0.52, p = 0.003). Conclusions: A tumor informed ctDNA approach to AC selection is feasible for patients undergoing upfront resection of pancreatic adenocarcinoma, with the first blood draw to be scheduled at week 5, allowing time for ctDNA analysis to determine AC selection. A high proportion of patients had detectable ctDNA, which appears independent of known prognostic markers. ctDNA detection was associated with earlier recurrence. Analyses of the impact of changes in ctDNA over time on survival are ongoing. Clinical trial information: ACTN12618000335291.

BackgroundThis study investigated the safety and efficacy of an anti-CTLA-4 monoclonal antibody (CS1002) as monotherapy and in combination with an anti-PD-1 monoclonal antibody (CS1003) in patients with advanced/metastatic solid tumors.MethodsThe phase 1 study involved phase 1a monotherapy dose-escalation (part 1) and phase 1b combination therapy dose escalation (part 2) and expansion (part 3). Various dosing schedules of CS1002 (0.3, 1, or 3 mg/kg every 3 weeks, or 3 mg/kg every 9 weeks) were evaluated with 200 mg CS1003 every 3 weeks in part 3.ResultsParts 1, 2, and 3 included a total of 13, 18, and 61 patients, respectively. No dose-limiting toxicities or maximum tolerated doses were observed. Treatment-related adverse events (TRAEs) were reported in 30.8%, 83.3%, and 75.0% of patients in parts 1, 2, and 3, respectively. Grade >= 3 TRAEs were experienced by 15.4%, 50.0%, and 18.3% of patients in each part. Of 61 patients evaluable for efficacy, 23 (37.7%) achieved objective responses in multiple tumor types. Higher objective response rates were observed with conventional and high-dose CS1002 regimens (1 mg/kg every 3 weeks or 3 mg/kg every 9 weeks) compared to low-dose CS1002 (0.3 mg/kg every 3 weeks) in microsatellite instability-high/mismatch repair-deficient tumors, melanoma, and hepatocellular carcinoma (50.0% vs. 58.8%, 14.3% vs. 42.9%, and 0% vs. 16.7%).ConclusionCS1002, as monotherapy, and in combination with CS1003, had a manageable safety profile across a broad dosing range. Promising antitumor activities were observed in patients with immune oncology (IO)-naive and IO-refractory tumors across CS1002 dose levels when combined with CS1003, supporting further evaluation of this treatment combination for solid tumors.Plain Language Summary CS1002 is a human immunoglobulin (Ig) G1 monoclonal antibody that blocks the interaction of CTLA-4 with its ligands and increases T-cell activation/proliferation. CS1003, now named nofazinlimab, is a humanized, recombinant IgG4 monoclonal antibody that blocks the interaction between human PD-1 and its ligands. In this original article, we determined the safety profile of CS1002 as monotherapy and in combination with CS1003. Furthermore, we explored the antitumor activity of the combination in anti-programmed cell death protein (ligand)-1 (PD-[L]1)-naive microsatellite instability-high/mismatch repair-deficient (MSI-H/dMMR) pan tumors, and anti-PD-(L)1-refractory melanoma and hepatocellular carcinoma (HCC). CS1002 in combination with CS1003 had manageable safety profile across a broad dosing range and showed promising antitumor activities across CS1002 dose levels when combined with CS1003. This supports further assessment of CS1002 in combination with CS1003 for the treatment of solid tumors. This was a phase 1a/1b, dose-escalation and dose-expansion study evaluating the safety and efficacy of an anti-CTLA-4 monoclonal antibody (CS1002) as monotherapy and in combination with an anti-PD-1 monoclonal antibody (CS1003/nofazinlimab), in patients with advanced/metastatic solid tumors. Promising antitumor activities were observed in heavily pretreated patients with immune oncology (IO)-naive and IO-refractory tumors across CS1002 dose levels when combined with CS1003, which supports further assessment of CS1002 in combination with CS1003 for the treatment of solid tumors.

3037 Background: Prognosis for advanced pancreatic ductal adenocarcinoma (PDAC) and biliary tract cancer (BTC) remains poor, with limited treatment options available. Expression of claudin18.2 (CLDN18.2) has emerged as a potential target for anti-cancer treatment. Herein, we report preliminary safety and efficacy results of IBI343, an antibody-drug conjugate (ADC) consisting of anti-CLDN18.2 monoclonal antibody conjugated to exatecan (topoisomerase I inhibitor), in patients (pts) with advanced PDAC or BTC in a phase 1 study. Methods: Eligible pts who failed or were intolerant to standard treatment were enrolled. IBI343 were intravenously administered at 6 mg/kg or 8 mg/kg Q3W. In dose escalation, pts were enrolled regardless of CLDN18.2 expression. In dose expansion, pts were required to have CLDN18.2 expression ≥40% (1+/2+/3+ staining intensity by immunohistochemistry). Primary endpoint was safety. Secondary endpoints included objective response rate (ORR), disease control rate (DCR), duration of response (DoR) and progression free survival (PFS) assessed by investigator per RECIST v1.1. Results: As of December 19, 2023, 35 pts (1 pt in dose escalation and 34 pts in dose expansion) were enrolled from China and Australia (males: 57.1%, median age: 58.0 years, ECOG PS 1: 71.4%, stage IV: 91.4%, median lines of prior treatment: 2) including 28 PDAC pts and 7 BTC pts. Pts received IBI343 at 6 mg/kg (n=17) or 8 mg/kg (n=18). Median duration of treatment was 7.0 weeks (range: 3.0-23.6) with 23 (65.7%) pts still on treatment. In all pts, treatment-related adverse events (TRAEs) occurred in 28 (80.0%) pts including grade ≥3 TRAEs in 9 (25.7%) pts. Common TRAEs (≥20%) were anemia (42.9%), neutrophil count decreased (28.6%), nausea (25.7%), vomiting (25.7%) and white blood cell count decreased (22.9%). Serious TRAEs occurred in 4 (11.4%) pts. TRAEs leading to dose interruption and treatment discontinuation occurred in 7 (20.0%) pts and 1 (2.9%) pt respectively. No TRAE led to death. Safety profiles of IBI343 in PDAC and BTC were comparable with the whole study cohort and no new safety signal was observed. As of January 15, 2024, 25 pts at 6 mg/kg and 8 mg/kg were efficacy evaluable. Partial response (PR) was observed in 7 pts (5 PDAC and 2 BTC). The ORR was 28.0% (95%CI: 12.1-49.4) and DCR was 80.0% (95%CI: 59.3-93.2). In evaluable pts at 6 mg/kg with CLDN18.2 expression ≥60% (1+/2+/3+, n=13), 5 pts had PR with ORR of 38.5% (95%CI: 13.9-68.4) and DCR of 84.6% (95%CI: 54.6-98.1). Among 10 PDAC pts in this subgroup, ORR was 40% (95%CI: 12.2-73.8). DoR and PFS data were immature. More updated data on safety and efficacy will be presented at the meeting. Conclusions: IBI343 was well tolerated with favorable safety profiles and encouraging efficacy in CLDN18.2-positive PDAC and BTC. Clinical trial information: NCT05458219 .

Pancreatic cancer (PC) remains the predominant type of upper gastrointestinal tract cancer, associated with heightened morbidity and a survival rate below 12%. While immunotherapy has brought about transformative changes in the standards of care for most solid tumors, its application in PC is hindered by the ''cold tumor'' microenvironment, marked by the presence of immunosuppressive cells. Modest response rates in PC are attributed, in part to, the fibrotic stroma that obstructs the delivery of systemic immunotherapy. Furthermore, the occurrence of immune-related adverse events (iRAEs) often necessitates the use of sub-therapeutic doses or treatment discontinuation. In the pursuit of innovative approaches to enhance the effectiveness of immunotherapy for PC, implantable drug delivery devices and scaffolds emerge as promising strategies. These technologies offer the potential for sustained drug delivery directly to the tumor site, overcoming stromal barriers, immunosuppression, T cell exclusion, immunotherapy resistance, optimizing drug dosage, and mitigating systemic toxicity. This review offers a comprehensive exploration of pancreatic ductal adenocarcinoma (PDAC), the most common and aggressive form of PC, accompanied by a critical analysis of the challenges the microenvironment presents to the development of successful combinational immunotherapy approaches. Despite efforts, these approaches have thus far fallen short in enhancing treatment outcomes for PDAC. The review will subsequently delve into the imperative need for refining delivery strategies, providing an examination of past and ongoing studies in the field of localized immunotherapy for PDAC. Addressing these issues will lay the groundwork for the development of effective new therapies, thereby enhancing treatment response, patient survival, and overall quality of life for individuals diagnosed with PDAC.

Background and Aims:Metastatic pancreatic ductal adenocarcinoma (mPDAC) has a 5-year survival rate of 3%. In nonmetastatic, locally advanced pancreatic cancer, the addition to chemotherapy of EUS-guided intratumoral phosphorus-32 (32P) microparticle implantation has achieved good local disease control. The aim of this study was to report the clinical outcomes of this treatment in patients with mPDAC. Methods:Patients with mPDAC treated with chemotherapy and intratumoral 32P-microparticles from 5 centers in Australia and the United Kingdom were analyzed retrospectively. Results:Fourteen patients were treated (7 female subjects; median age, 64.5 years; Eastern Cooperative Oncology Group performance status scores 0/1/2, 21.4%/57.1%/21.4%). The median baseline primary tumor longest diameter was 40.5 mm. Patients had a median of 3 metastases (interquartile range, 2.25 to 5) and received either 5-fluorouracil, leucovorin, irinotecan, and oxaliplatin (n = 4) or gemcitabine/nab-paclitaxel (n = 10). 32P microparticles were implanted at a median 3.1 months from chemotherapy commencement. Local disease control rate at 3 months' postimplantation was 100%. Primary tumor longest diameter decreased by 25% (interquartile range, -31.4% to -16.7%; P = .008), and serum cancer antigen 19-9 levels declined from 134 U/mL to 66 U/mL (P = .018). Local progression-free survival was 12.2 months (95% CI, 9.0-15.4 months) from chemotherapy commencement and 8.3 months (95% CI, 2.6-16.0 months) from 32P microparticle implantation. Therapy was associated with improved quality of life, including global health status at 12 weeks' postimplantation (P = .037). Median overall survival was 13.8 months (95% CI, 10.5-17.1 months) from chemotherapy commencement and 11 months (95% CI, 5.6-17.4 months) from 32P microparticle implantation. No grade 4/5 acute toxicities were observed. Conclusions:This first multicenter analysis of combined chemotherapy and EUS-guided 32P microparticle implantation in mPDAC shows the potential clinical benefits of local tumor control in a cohort for whom outcomes are historically poor.

140 Background: 5-fluorouracil (5-FU) and its biomodulator leucovorin (LV) are an efficacious treatment option for various solid tumours, and are considered the backbone of therapy for colorectal cancer. Standard administration schedules require sequential administration of the two agents due to their chemical incompatibility; given the importance of LV in modulating 5-FU activity, their simultaneous administration is expected to enhance anti-tumour efficacy. Deflexifol is an all-in-one injectable reformulation of 5-FU and LV at physiological pH. An initial phase I trial demonstrated the safety and efficacy of Deflexifol when administered as a bolus or infusion, with a maximum tolerated dose (MTD) of ≥25% more 5-FU than established for standard regimens. This subsequent phase I study sought to investigate Deflexifol administered as a combined bolus and infusion schedule. Methods: Patients with advanced malignancy who had failed standard treatment were administered Deflexifol as a bolus followed by a continuous 46-hr infusion. The infusional dose was escalated across four levels (2400 mg/m 2 to 3800 mg/m 2 ) in the absence of Dose Limiting Toxicity using a traditional 3+3 design; the bolus dose was fixed at the previously declared MTD of 525 mg/m 2 . The stated Deflexifol dose represents the dose of standard 5-FU delivered. Primary trial objectives were to determine the safety and tolerability of a bolus plus infusion regimen of Deflexifol, with a secondary objective to determine the pharmacokinetics and MTD. Efficacy was an exploratory objective. Results: Nineteen patients with mainly colorectal (n=13) and breast (n=4) cancers were enrolled and treated with Deflexifol. Approximately two-thirds of patients had previously received fluoropyrimidine treatment. Deflexifol was generally well tolerated, with the MTD declared as a 525 mg/m 2 bolus plus 3400 mg/m 2 infusion. Treatment-related toxicities were consistent with standard 5-FU/LV therapy; the most frequent Grade 1/2 adverse events were fatigue, mucositis, and diarrhea. Grade ≥3 adverse events were reported in 6/19 patients, including mucositis (2/19) and haematological toxicity (5/19). One treatment-related death occurred at the highest infusion dose level of 3800 mg/m 2 due to haematological toxicity. Quantifiable 5-FU and LV concentrations were exhibited at all time-points. Disease control, including one partial response, was achieved in 9/13 evaluable patients. Conclusions: Deflexifol is safe and well tolerated at doses of 5-FU up to 40% higher than typically administered by the standard modified de Gramont regimen. 5-FU and LV co-exposure extended throughout the entire dosing period, substantially longer than reported for standard treatment. Disease control was achieved in 69% of patients. Clinical trial information: ACTRN12619001533189 .