SUMMARY: The American Cancer Society estimates that in 2025, about 67,440 people will be diagnosed with pancreatic cancer and 51,980 people will die of the disease. Pancreatic Ductal AdenoCarcinoma (PDAC) remains one of the most lethal malignancies, with most cases diagnosed at advanced stages and few modifiable risk factors identified to date.
Pancreatic ductal adenocarcinoma is characterized by marked biological heterogeneity and limited therapeutic durability. While combination chemotherapy regimens have modestly extended survival in the metastatic setting, outcomes remain poor for the majority of patients, underscoring the urgent need for better treatment selection strategies. Molecular stratification has emerged as a promising approach in PDAC, supported by well-established predictive biomarkers such as germline BRCA2 and PALB2 alterations, which identify a subset of patients more likely to benefit from platinum-based chemotherapy and PARP inhibition. Beyond DNA damage repair defects, transcriptomic profiling has further refined the molecular landscape of PDAC, consistently identifying two dominant expression subtypes-Classical and Basal-like, with important prognostic and potentially predictive implications.
The Classical subtype is generally associated with a more differentiated epithelial phenotype and improved survival, whereas Basal-like tumors exhibit stem-like features, relative chemoresistance, and inferior outcomes. Prior nonrandomized and prospective studies have suggested differential chemotherapy sensitivity between these subtypes, raising the question of whether transcriptional classification could inform first-line regimen selection. The Pancreatic Adenocarcinoma Signature Stratification for Treatment-01 (PASS-01) trial was designed to address this gap by prospectively comparing two commonly used first-line chemotherapy regimens, modified FOLFIRINOX (mFOLFIRINOX) and Gemcitabine plus nab-Paclitaxel (GnP), while embedding comprehensive molecular and translational analyses.
PASS-01 Trial Design and Study Objectives
PASS-01 was a randomized, open-label, multinational Phase II study conducted across centers in Canada and the United States. The trial enrolled patients with de novo metastatic PDAC who were chemotherapy-naïve between October 2020 and January 2024, and excluded individuals harboring germline pathogenic variants in BRCA1, BRCA2, or PALB2, thereby removing a population with known platinum sensitivity. Patients were randomized 1:1 to receive either mFOLFIRINOX (N=80) or GnP (N=80. The Primary endpoint was Progression-Free Survival (PFS) in the intention-to-treat population, using a relaxed significance threshold appropriate for a signal-seeking Phase II design. Key Secondary objectives included Overall Survival (OS), Safety, Objective Response Rates (ORR), and exploratory analyses evaluating outcomes according to RNA expression subtype, GATA6 expression, Patient-Derived Organoid (PDO) data, and other molecular correlatives.
Importantly, PASS-01 incorporated mandatory pretreatment tumor biopsies whenever feasible. These samples underwent whole-genome and transcriptome sequencing, RNA-based subtype classification, and PDO generation, with results reviewed in a molecular tumor board to inform later-line treatment decisions. This design allowed for a real-world assessment of the feasibility and clinical relevance of upfront molecular profiling in metastatic PDAC.
First-Line Efficacy Outcomes in the Overall Study Population
With a median follow-up of 8.3 months, PFS was numerically longer with GnP compared with mFOLFIRINOX, although the difference did not reach conventional statistical significance. Median PFS in the intention-to-treat population was 5.3 months with GnP versus 4.0 months with mFOLFIRINOX. Similar trends were observed in the per-protocol analysis.
Overall Survival outcomes favored GnP more clearly. Median OS approached 10 months with GnP and was under 9 months with mFOLFIRINOX, translating into a statistically significant hazard ratio favoring the Gemcitabine-based regimen. Notably, these differences persisted after adjustment for key clinical covariates, including performance status, liver metastases, and KRAS mutation status. While absolute survival gains were modest, these findings are clinically relevant given the lack of head-to-head randomized data comparing these regimens in Western populations. Objective Response Rates were comparable between treatment arms. However, Disease Control Rate and Durability of Response favored GnP. Patients treated with GnP experienced a higher Disease Control Rate and a longer Duration of Response, suggesting more sustained benefit in a subset of patients.
Safety Profile and Treatment Tolerability
Treatment-related toxicity differed meaningfully between regimens. Hospitalizations due to adverse events were more frequent in the mFOLFIRINOX arm, driven primarily by gastrointestinal complications, febrile neutropenia, and serious infections. In contrast, severe toxicities with GnP were less common and more limited in scope. These safety differences are particularly relevant in a population with aggressive disease biology and limited physiologic reserve, where treatment tolerability may influence both quality of life and the ability to receive subsequent therapy.
Impact of Transcriptional Subtypes on Clinical Outcomes
One of the most informative aspects of PASS-01 was its prospective evaluation of RNA expression subtypes. Among patients with adequate tissue for analysis, approximately 75% were classified as Classical and 25% as Basal-like, consistent with prior reports. Across the entire cohort, Basal-like tumors were associated with numerically shorter PFS and OS compared with Classical tumors, reinforcing their adverse prognostic significance.
When outcomes were examined by treatment arm within each subtype, important patterns emerged. In patients with Classical PDAC, PFS was similar between regimens, but OS was notably longer with GnP compared with mFOLFIRINOX. Conversely, in Basal-like disease, outcomes were uniformly poor regardless of regimen, though trends consistently favored GnP across PFS, Response Rate, and Duration of Response. These findings suggest that Basal-like tumors may derive limited benefit from intensified multi-agent chemotherapy and may be particularly resistant to Fluorouracil and Irinotecan-based approaches.
GATA6 Expression as a Pragmatic Surrogate Biomarker
Given prior evidence linking GATA6 expression with the Classical subtype, PASS-01 also evaluated GATA6 RNA in situ hybridization as a pragmatic surrogate biomarker. High GATA6 expression correlated strongly with Classical transcriptional identity. While patients with high GATA6 expression demonstrated a trend toward longer PFS, GATA6 status alone did not reliably predict differential benefit from mFOLFIRINOX versus GnP. These findings suggest that while GATA6 may serve as a useful prognostic marker, its role as a standalone predictive tool for chemotherapy selection remains limited and may require integration into broader multiplex or composite biomarker platforms.
Early CA 19-9 Dynamics as a Biomarker of Treatment Response
PASS-01 also provided important insights into the utility of early CA 19-9 changes as a biomarker of treatment response. Among patients with evaluable markers, a decline in CA 19-9 within four weeks of therapy initiation was associated with significantly prolonged PFS, whereas early increases were linked to inferior outcomes. However, a subset of patients with early CA 19-9 rises subsequently achieved radiographic disease control, underscoring that CA 19-9 kinetics should not be used in isolation to prompt premature treatment discontinuation. These findings support the potential role of early biomarker dynamics, particularly when combined with emerging tools such as circulating tumor DNA, in adaptive treatment strategies.
Translational Findings and the Challenge of Second-Line Therapy
Despite the extensive molecular profiling and use of correlate-guided recommendations, outcomes in the second-line setting were uniformly poor. Only about half of patients were able to receive subsequent therapy, and survival following progression was measured in months. Correlate-guided treatment selection did not meaningfully improve outcomes compared with standard approaches, highlighting the clinical reality that opportunities for precision intervention in PDAC are often lost once patients progress beyond first-line therapy.
Clinical Implications for First-Line Treatment Selection
PASS-01 confirms that outcomes with standard first-line combination chemotherapy for metastatic PDAC remain disappointing, even in carefully selected clinical trial populations. Within this context, the modest but consistent efficacy and safety advantages observed with GnP over mFOLFIRINOX are practice-informing, particularly for patients without known DNA repair defects. More importantly, the trial reinforces the prognostic importance of transcriptional subtypes and supports the concept that molecular features should be assessed early, when they are most likely to influence meaningful treatment decisions.
As novel therapeutic strategies, including KRAS-targeted agents and rational combination approaches, move into earlier lines of therapy, transcriptional subtype may prove critical in guiding regimen selection and trial design. PASS-01 demonstrates that comprehensive upfront molecular profiling is feasible in multicenter settings and provides a framework for future biomarker-driven trials aimed at improving first-line outcomes in this highly lethal disease.
Key Takeaways and Conclusions
In the Phase II PASS-01 trial, Progression-Free Survival was similar between mFOLFIRINOX and Gemcitabine plus nab-Paclitaxel. However, Overall Survival, treatment durability, and safety trends favored the Gemcitabine-based regimen. Molecular analyses confirmed the adverse prognosis associated with Basal-like PDAC and suggested limited benefit from intensified chemotherapy in this subgroup. Collectively, these findings emphasize the critical importance of optimizing first-line treatment strategies and integrating molecular stratification early in the disease course, as opportunities for effective intervention rapidly diminish after progression.
PASS-01: Randomized Phase II Trial of Modified FOLFIRINOX Versus Gemcitabine/Nab-Paclitaxel and Molecular Correlatives for Previously Untreated Metastatic Pancreatic Cancer. Knox JJ, O’Kane G, King D, et al. J Clin Oncol. 2025;43:3355-3368.
