ASCO 2025 highlights: emerging therapies and biomarkers in pancreatic ductal adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies, ranking as the fourth leading cause of cancer-related mortality in Europe and showing an upward trend in death rates across both sexes [1]. Most patients are diagnosed at an unresectable or metastatic stage, where standard palliative chemotherapy has not achieved a median overall survival beyond 12 months in phase III clinical trials. Notably, at this year’s ASCO annual meeting, several pivotal studies were presented, including positive phase III trials in both early and advanced PDAC, complemented by encouraging early-phase data.

The randomized phase III CASSANDRA trial investigated the efficacy of the quadruple chemotherapy regimen PAXG (nab-paclitaxel, capecitabine, cisplatin, gemcitabine) versus a current standard regimen modified (m)FOLFIRINOX (5-fluorouracil + leucovorin, irinotecan, oxaliplatin) in stage I–III resectable and borderline resectable PDAC in the neoadjuvant setting [2]. A total of 260 patients were randomized to receive either PAXG or mFOLFIRINOX every 14 days for 4 months. Patients without disease progression or unacceptable toxicity underwent a second randomization to either immediate surgery followed by 2 months of the same chemotherapy or completion of 6 months of chemotherapy prior to surgical resection. The results of the first randomization with the primary endpoint event-free survival (EFS)—defined as the absence of disease progression, recurrence, intraoperative metastases or unresectability, a significant rise in CA19-9 levels, or death—were reported at ASCO 2025.

The PAXG regimen demonstrated a significant EFS improvement, with a median EFS of 16.0 months compared to 10.2 months with mFOLFIRINOX (hazard ratio [HR] 0.64, p = 0.003). The 1‑ and 3‑year EFS rates were 61% vs 45% and 31% vs 13%, respectively. Furthermore, PAXG also achieved superior disease control (DCR) (98% vs 91%; p = 0.01), higher CA19‑9 response rates (≥ 50% reduction: 88% vs 64%; p < 0.001) and fewer intra- or postoperative metastases (5% vs 12%; p = 0.03).

In contrast, resection rates (75% vs 67%; p = 0.16) and R0 resection rate (51% vs 52%; p = 0.63) were similar between the two regimens. At the time of data presentation, overall survival (OS) data were still immature but showed a trend toward improved outcomes with PAXG (median OS: 37.3 vs 26.0 months; p = 0.07). Although PAXG represents a quadruple-agent intensification compared to the triple-agent mFOLFIRINOX regimen, their safety profiles were largely comparable. The PAXG regimen was associated with higher rates of hematologic adverse events and hand–foot syndrome, whereas diarrhea more frequently occurred in the mFOLFIRINOX group. The majority of patients completed the assigned preoperative chemotherapy (PAXG: 76%, mFOLFIRINOX: 67%).

In conclusion, CASSANDRA was a positive phase III trial demonstrating improved EFS with PAXG compared to mFOLFIRINOX. However, resection rates were comparable and the final OS data have not been published yet. Currently, adjuvant mFOLFIRINOX is regarded as the standard of care for resectable PDAC based on the PRODIGE trial [3]. Given the high relapse rate, shifting chemotherapy into the neoadjuvant setting might represent a promising approach and may serve as a “biological probe” to identify patients with more favorable tumor biology. In this context, PAXG could represent an attractive option for patients with a good performance status in the future.

By gene expression profiling, PDAC can be classified into two major subtypes: “classical” and “basal-like” [4]. The classical subtype, characterized by a GATA6-driven gene signature, is associated with better survival outcomes than the basal-like counterpart [4]. Furthermore, the basal-like subtype is characterized by resistance to platinum-based regimens such as mFOLFIRINOX in the palliative setting, whereas patients appear to derive comparatively greater benefit from nab-paclitaxel plus gemcitabine [5].

NeoPancONE is a multicenter phase II study evaluating perioperative mFOLFIRINOX in resectable PDAC patients stratified by GATA6 expression [6]. Following an initial biopsy for GATA6 assessment, patients received six rounds of mFOLFIRINOX, underwent surgery, and completed an additional six rounds of chemotherapy. The primary endpoint was 1‑year EFS, with secondary endpoints including overall response rate (ORR), R0 resection rate, OS, and biomarker analyses. A numerical trends toward improved 1‑year EFS (73% vs. 50%; p = 0.12) and median OS (34.2 vs 18.7 months; p = 0.12) were observed in the GATA6-high cohort, although differences were not statistically significant. Notably, no patients with low GATA6 expression responded to neoadjuvant chemotherapy, whereas the ORR reached 22% in the GATA6-high group. Similar results were obtained when patients were reclassified into classical and basal-like subtypes by RNA sequencing according to the PurIST (purity independent subtyping of tumors) algorithm, confirming the concordance between GATA6 in situ hybridization (ISH) and transcriptomic subtyping.

These findings suggest that biomarkers such as GATA6 and RNA subtypes may help identify patients most likely to benefit from (neoadjuvant) mFOLFIRINOX, moving toward a precision-medicine approach in pancreatic cancer and away from the current “one-size-fits-all” strategy. The phase II PANCREAS trial (NCT04683315) is currently recruiting, assigning patients to either neoadjuvant nab-paclitaxel/gemcitabine (basal-like) or mFOLFIRINOX (classical subtype) based on the PurIST classification of pretreatment endoscopic ultrasound fine-needle aspiration (EUS/FNA) samples.

In the PANOVA‑3 phase III trial, patients with locally advanced PDAC were randomized to receive palliative chemotherapy with gemcitabine and nab-paclitaxel, either alone or in combination with tumor treating fields (TTFields) [7].

TTFields is a locoregional antitumoral therapy that delivers alternating electrical fields via a portable device and is already approved for the treatment of glioblastoma [8]. Its cytotoxic activity is mediated through multiple mechanisms, including interference with mitotic processes and cell motility, enhancement of antitumor immune responses and increased cell membrane permeability [8].

In the PANOVA-3 trial, the primary endpoint OS was significantly improved with the addition of TTFields (median OS 16.2 vs 14.2 months; HR = 0.82, p = 0.039). Another notable endpoint—pain-free survival (time to a ≥ 20-point increase from baseline on a patient-reported visual analogue scale)—was also significantly prolonged with TTFields, showing a 6-month improvement and sustained separation of the curves (median 15.2 vs 9.1 months; HR = 0.74; p = 0.027). Overall, the local treatment was well tolerated, with dermatitis, rash and pruritus being the most common side effects. Of note, there was no significant difference in median local progression-free survival (PFS; 12.5 vs 10.4 months; HR = 0.84, p = 0.151). Overall response rates were comparable between the groups (36.1% vs 30.0%; p = 0.094) and no improvement in resectability was observed.

In summary, PANOVA‑3 represents a positive phase III trial in advanced PDAC, demonstrating a moderate improvement in OS and prolonged pain-free survival with TTFields. However, no benefit was observed in local PFS or resectability. Future studies will need to address remaining open questions, including comparisons with mFOLFIRINOX and potential combinations with other chemotherapy regimens. Of note, stereotactic radiotherapy, including celiac plexus radiosurgery, may serve as an effective local alternative for pain relief as Lawrence et al. reported. In a cohort composed predominantly of PDAC patients, pain improvement was achieved in 53% of cases, accompanied by a significant reduction in opioid requirements [9].

Claudin 18.2 (CLDN18.2) is an emerging, potentially tumor-agnostic biomarker, with membrane-bound expression observed in approximately 60% of PDAC cases [10]. At the 2025 ASCO annual meeting, a phase I dose-expansion trial was presented evaluating the antibody–drug conjugate (ADC) IBI343 in locally advanced unresectable or metastatic PDAC [11]. IBI343 consists of a CLDN18.2-targeting antibody linked to the topoisomerase‑1 inhibitor exatecan. In terms of the primary endpoint of safety, hematologic toxicities were most frequently reported, followed by gastrointestinal adverse events. Exploratory analyses showed preliminary signs of efficacy, with an ORR of 22.7%, a DCR of 81.8%, and a median DOR of 6.7 months in patients with CLDN18.2 expression (1+, 2+, or 3+ in ≥ 60% of tumor cells).

Although this was an early-phase trial and clinical application remains distant, the study demonstrated promising clinical activity of CLDN18.2-targeted therapy in advanced pancreatic cancer, with a manageable safety profile. Currently, the anti-CLDN18.2 antibody zolbetuximab combined with gemcitabine/nab-paclitaxel is evaluated in a randomized phase II trial in patients with advanced PDAC (NCT03816163). With further ongoing trials—including ADCs and chimeric antigen receptor (CAR) T cell therapies—CLDN18.2 has the potential to become a novel therapeutic target for a substantial proportion of PDAC patients in the future.

Although PDAC remains one of the most lethal tumor entities, several encouraging studies presented at this year’s ASCO annual meeting offer hope for improved patient outcomes. Novel biomarkers such as GATA6 and Claudin 18.2 hold promise for bringing precision medicine into the treatment of pancreatic cancer. In addition, the PAXG regimen was introduced as a novel quadruple chemotherapy protocol for patients with resectable and borderline resectable PDAC, while tumor treating fields (TTFields) demonstrated clinically meaningful efficacy as a localized therapeutic approach for locally advanced disease, particularly in terms of pain management.