Reframing the immunotherapy landscape in borderline resectable pancreatic ductal adenocarcinoma: progress or plateau?

Pancreatic ductal adenocarcinoma (PDAC) is still a highly lethal gastrointestinal cancer, with a 5-year survival rate of less than 10% (1,2). Despite advances in systemic therapies for other solid tumors, PDAC continues to evade meaningful clinical progress, in large part due to its immunosuppressive tumor microenvironment (TME), high stromal content, and aggressive biology (3). In recent years, the neoadjuvant setting for borderline resectable PDAC (BR-PDAC) has emerged as a therapeutic window, offering an opportunity to administer multimodal treatment aimed not only at cytoreduction but also at systemic disease control (4). In this context, novel immunotherapy strategies are being explored to enhance tumor antigenicity and promote an effective anti-tumor immune response (5).

In their recent phase II clinical trial, Agarwal et al. evaluated the feasibility, safety, and immunologic effects of a neoadjuvant combination regimen including GVAX [a granulocyte-macrophage colony-stimulating factor (GM-CSF)-secreting whole-cell pancreatic cancer vaccine], low-dose cyclophosphamide, anti-programmed death-1 (anti-PD-1) checkpoint inhibitor nivolumab, and stereotactic body radiotherapy (SBRT), administered after standard neoadjuvant chemotherapy (6). Their study, conducted in patients with BR-PDAC, aimed to overcome the inherent immune resistance of PDAC and interrogated the immunologic changes induced by this approach using robust translational endpoints.

This commentary seeks to situate the findings of Agarwal et al. within the broader landscape of neoadjuvant immunotherapy for PDAC, exploring the strengths and limitations of the study, its translational implications, and the critical unanswered questions that it raises for future research (6).

We read with great interest the paper by Agarwal et al., which focuses on evaluating the safety and immune effects of GVAX/cyclophosphamide/nivolumab and SBRT on the PDAC TME (6). A total of 31 patients were enrolled, of whom 18 received at least one dose of combination immunotherapy. The trial was well designed to reflect real-world treatment sequencing: all patients first received standard-of-care neoadjuvant chemotherapy [mFOLFIRINOX (5-fluorouracil, leucovorin, irinotecan, and oxaliplatin) or gemcitabine/nab-paclitaxel if intolerant to mFOLFIRINOX], followed by the study intervention comprising cyclophosphamide, GVAX, nivolumab, and SBRT. Importantly, the study was carefully structured to deliver immunotherapy during the window between chemotherapy completion and surgery, minimizing interference with surgical planning. Ultimately, 14 patients underwent definitive resection, with an R0 resection rate of 93%. Nonetheless, the authors did not correlate pathological response categories (responder vs. non-responder) with survival outcomes, which represents a missed opportunity. Such correlation would provide important mechanistic insight into whether pathological response serves as a meaningful surrogate endpoint in the context of combined immunotherapy and SBRT.

From a safety standpoint, the combination was well tolerated, with no treatment-related adverse events (TRAEs). The most common adverse events included vaccine site reactions (VSRs), fatigue, and anorexia. Only two patients experienced grade 3 TRAEs; one patient experienced grade 3 alanine aminotransferase (ALT) increase and aspartate aminotransferase (AST) increase, and the other patient experienced grade 3 neutrophil count decrease, white blood cell decrease, and grade 3 anorexia. These data support the feasibility of integrating multi-agent immunotherapy into the neoadjuvant timeline for BR-PDAC and locally advanced PDAC (LA-PDAC).

The study’s primary endpoint was CD8⁺ T-cell density in resected tumors compared to historical controls treated with chemotherapy and SBRT alone. Disappointingly, no statistically significant difference was observed. Mean CD8⁺ density was numerically higher in the experimental group (266.7 vs. 199.9 cells/mm²), but the difference did not reach statistical significance (P=0.246). These finding echoes similar results from other pancreatic cancer immunotherapy studies, underscoring the challenge of inducing durable T-cell infiltration in such a hostile TME (7).

However, this result should not be viewed as a definitive failure. The study was limited by small sample size (n=13 for CD8⁺ analysis) and by the use of historical controls, which, while pragmatic, introduces potential biases and limits the strength of any comparative conclusions. A randomized control arm would have provided more definitive insights into treatment efficacy. Additionally, the reliance on a single immunologic endpoint, such as CD8⁺ density, while practical, may overlook broader immune remodeling within the tumor.

Moreover, the immunologic heterogeneity observed within the experimental cohort hints at potential benefit in a subset of patients. Specifically, patients with minimal residual disease or moderate pathological responses showed higher CD8⁺ densities on average compared to poor responders, suggesting that immune activation may be achievable in select cases.

Further support for immunologic activity comes from the presence of intratumoral lymphoid aggregates in resection specimens. In this trial, the patient achieving pathological complete response (pCR) exhibited low CD8⁺ density, a result more likely related to the minimal residual tissue available for analysis than to a lack of immune activation. While these findings are anecdotal and require validation, they raise intriguing possibilities about localized immune education within the TME.

Notably, these findings are aligned with prior evidence from a randomized trial by Zheng et al., in which patients with resectable PDAC received neoadjuvant GVAX with or without low-dose cyclophosphamide. In this study, the density of tertiary lymphoid aggregate (TLA) was significantly associated with longer overall survival (OS), with patients showing OS >24 months exhibiting higher TLA compared to those with OS <15 months. These data not only reinforce the biological relevance of vaccine-induced TLA but also support their potential as predictive biomarkers in future neoadjuvant immunotherapy trials for PDAC (8).

The recent study by Wang et al. adds mechanistic depth to these findings. Using multiplex immunohistochemistry (IHC) and spatial analysis in a similar therapeutic context (GVAX, ani-PD-1, and SBRT), Wang et al. demonstrated that not all CD8⁺ T cells are equal (5). Specifically, infiltration of granzyme B-expressing CD8⁺ (GZMB⁺CD8⁺) effector T cells, rather than CD8⁺ density per se, correlated with long-term survival. Moreover, these high-quality T cells were enriched in TLAs, reinforcing the hypothesis that these aggregates may serve as immune activation niches. These data help explain the lack of correlation between bulk CD8⁺ density and outcome in Agarwal et al.’s study and argue for refined immune endpoints in future trials (5,6).

One of the study’s most valuable contributions lies in its exploratory RNA expression analysis. Using the Nanostring PanCancer IO 360 panel, the authors evaluated immune-related gene expression in resected tumors, comparing responders to non-responders. Among the 38 differentially expressed genes, responders showed upregulation of multiple transcripts associated with immune cell activation and trafficking, including KLRK1 (NKG2D), CD48, PTPRC (CD45), GZMK, and CCL4.

Interestingly, several upregulated genes in responders are also implicated in immunosuppressive or tumor-promoting pathways, such as CXCL8 and IL7R. This paradox reflects a broader theme in PDAC immunotherapy: immune activation is often accompanied by compensatory mechanisms that blunt therapeutic efficacy. The presence of such escape pathways suggests that future strategies may need to incorporate targeted interventions against stromal and myeloid components of the TME to fully unleash the potential of T-cell-based therapies.

Wang et al. provide further evidence in this direction, showing that the density of M2-like tumor-associated macrophages (TAMs) increased in patients with poor outcomes, while a higher M1/M2 TAM ratio was associated with longer survival (5). Their study emphasizes the dual role of radiotherapy: while SBRT may facilitate T-cell priming and positioning, it also enhances recruitment of suppressive myeloid populations, potentially limiting efficacy. These findings argue for the inclusion of macrophage-targeting agents (e.g., CSF1R inhibitors) in future combination strategies (5).

Moreover, the transcriptomic shifts observed underscore the limitations of single immunologic metrics such as CD8⁺ density. While useful, CD8⁺ infiltration alone may not capture the complex interplay of effector and suppressive signals in the TME. Without parallel quantification of total T cells (CD3⁺), it remains unclear whether the observed infiltration reflects selective enrichment of cytotoxic T lymphocytes or a more generalized increase in lymphocytes. Calculating the ratio of CD8⁺ cells to total T cells (CD3⁺ cells) would have contextualized the composition of the T-cell infiltrate and strengthened the interpretation of the immunologic results. Moreover, CD8⁺ density alone does not distinguish functional cytotoxic cells from exhausted or dysfunctional lymphocyte populations. As the study did not include phenotypic markers, it is not possible to determine whether the infiltrating CD8⁺ cells were capable of mounting an effective antitumor response—a critical distinction in PDAC, where T-cell exhaustion is frequently observed. Multimodal profiling, including spatial transcriptomics and single-cell analyses, may be better suited to deconvolve these interactions and guide patient selection.

The emergence of messenger RNA (mRNA)-based cancer vaccines introduces a further promising avenue. A recent review by Laila et al. discusses the rapid evolution of mRNA vaccine platforms, highlighting their adaptability, personalized antigen targeting, and synergistic potential with other immunotherapies (9). In contrast to fixed-antigen platforms like GVAX, mRNA vaccines offer the possibility of encoding patient-specific neoantigens, improving immune specificity, and potentially overcoming tumor heterogeneity. Their integration with lipid nanoparticle delivery systems also enhances antigen presentation and intracellular stability, critical for durable responses in immune-cold tumors such as PDAC (9).

Clinical outcomes in the study were comparable to other BR-PDAC neoadjuvant trials. Among patients who received study therapy, the R0 resection rate was high (93%), and median OS for the entire intention-to-treat cohort (n=30) was 20.4 months. Among resected patients, median OS reached 30 months. These figures align with those reported in the Alliance A021501 and PREOPANC trials, indicating that the addition of immunotherapy did not compromise oncologic efficacy (10,11). Notably, the resection rate in the intention-to-treat population (47%) was lower than what is typically observed in dedicated BR-PDAC neoadjuvant series (≈60–68%) (10,11). This discrepancy largely reflects the substantial attrition that occurred during induction chemotherapy: seven patients progressed radiographically or biochemically on mFOLFIRINOX before reaching the immunotherapy/SBRT phase, and two additional patients were deemed medically inoperable prior to treatment start. When considering only patients who actually received the full GVAX/cyclophosphamide/nivolumab/SBRT regimen (n=18), the resection rate rises to 78%, aligning more closely with contemporary neoadjuvant datasets. The small median tumor size in the resected specimens (1.6 cm) is also consistent with profound treatment-induced fibrosis following prolonged chemotherapy and SBRT, which often leads radiographic and gross measurements to underestimate residual tumor burden.

However, the absence of a contemporaneous control arm and the limited sample size restrict definitive conclusions on additive benefit. Cross-trial comparisons remain inherently flawed due to heterogeneity in staging, patient selection, and treatment intensity. Larger, randomized studies are essential to assess whether multi-agent immunotherapy improves long-term survival beyond what standard chemotherapy and radiotherapy can achieve alone. Furthermore, the lack of carbohydrate antigen 19-9 (CA19-9) levels—both at baseline and following neoadjuvant therapy—limits the ability to interpret the biological heterogeneity of the cohort. CA19-9 kinetics are often useful for distinguishing chemo-sensitive from chemo-resistant tumors and could have contributed meaningfully to the assessment of tumor biology and treatment response.

GVAX has a long history in PDAC immunotherapy research, having been studied in both the adjuvant and neoadjuvant settings. It has demonstrated an ability to induce immune cell trafficking, tertiary lymphoid structure formation, and programmed death-ligand 1 (PD-L1) expression. Recent work by Montagne et al. further clarifies these mechanisms, showing that GVAX combined with anti-PD-1 expands clonal CD8⁺ T cell populations (particularly CD3⁺CD8⁺CD137⁺ subsets) and alters extracellular matrix interactions (12). However, despite promising biological effects, clinical benefit has remained elusive (12).

The study by Zheng et al. highlights a promising direction for GVAX-based strategies: the vaccine’s ability to induce intratumoral lymphoid aggregates in the TME appears to correlate strongly with improved survival, particularly when administered in the neoadjuvant setting (8). These findings suggest that beyond general immune activation, GVAX may act as an immune-organizing agent capable of initiating lymphoid structure formation within the tumor, a feature associated with better outcomes and potentially synergistic with other immunomodulators (8).

The trial by Agarwal et al. (6) raises several key questions for future investigation:

• Patient selection: are there molecular or immunologic signatures that can identify patients more likely to benefit from immunotherapy-based neoadjuvant regimens? Baseline TME characteristics, circulating biomarkers, or circulating tumor DNA (ctDNA) kinetics may serve as potential predictors.
• Combination strategies: should future regimens incorporate dual checkpoint blockade (e.g., PD-1 plus LAG3/TIGIT), TME-modulating agents (e.g., CXCR4 or CSF1R inhibitors), or novel vaccine platforms?
• Timing and sequencing: is there an optimal window during or after chemotherapy in which the TME is most permissive to immune modulation? Can radiotherapy be better timed to act as an immune primer?
• Endpoints and trial design: should surrogate endpoints such as CD8⁺ TLA density remain the standard for early-phase trials, or should broader immune profiling be used? Are adaptive trial designs needed to accelerate progress?
• Multi-omic integration: integration of transcriptomic, proteomic, and spatial data may reveal actionable nodes of immune resistance and enable rational combination therapies.

This phase II study confirms that neoadjuvant combination immunotherapy using GVAX, cyclophosphamide, nivolumab, and SBRT is safe and feasible in patients with BR-PDAC. While the primary immunologic endpoint was not met, translational analyses suggest immune modulation in a subset of patients, laying the groundwork for future hypothesis-driven trials.

As PDAC immunotherapy continues to evolve, the focus must shift from broad application to intelligent personalization. Studies such as this represent a critical step in building the biological rationale and translational toolkit necessary to bring immunotherapy into meaningful clinical relevance for pancreatic cancer. In the face of a historically immunotherapy-resistant malignancy, progress will likely be incremental, but with persistence and a commitment to methodological rigor, increasingly precise.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, AME Clinical Trials Review. The article has undergone external peer review.

Peer Review File: Available at https://actr.amegroups.com/article/view/10.21037/actr-25-101/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://actr.amegroups.com/article/view/10.21037/actr-25-101/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Hu JX, Zhao CF, Chen WB, et al. Pancreatic cancer: A review of epidemiology, trend, and risk factors. World J Gastroenterol 2021;27:4298-321. [Crossref] [PubMed]
2. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
3. Ho WJ, Jaffee EM, Zheng L. The tumour microenvironment in pancreatic cancer - clinical challenges and opportunities. Nat Rev Clin Oncol 2020;17:527-40. [Crossref] [PubMed]
4. Wu ST, Zhu L, Feng XL, et al. Advances in neoadjuvant therapy for pancreatic cancer: Current trends and future directions. World J Clin Oncol 2025;16:105849. [Crossref] [PubMed]
5. Wang J, Gai J, Zhang T, et al. Neoadjuvant radioimmunotherapy in pancreatic cancer enhances effector T cell infiltration and shortens their distances to tumor cells. Sci Adv 2024;10:eadk1827.
6. Agarwal P, Guo M, Munjal K, et al. A phase II study of neoadjuvant GVAX and cyclophosphamide combined with nivolumab and SBRT followed by surgery in borderline resectable pancreatic adenocarcinoma. Clin Cancer Res 2025;31:3205-14. [Crossref] [PubMed]
7. Lutz ER, Wu AA, Bigelow E, et al. Immunotherapy converts nonimmunogenic pancreatic tumors into immunogenic foci of immune regulation. Cancer Immunol Res 2014;2:616-31. [Crossref] [PubMed]
8. Zheng L, Ding D, Edil BH, et al. Vaccine-Induced Intratumoral Lymphoid Aggregates Correlate with Survival Following Treatment with a Neoadjuvant and Adjuvant Vaccine in Patients with Resectable Pancreatic Adenocarcinoma. Clin Cancer Res 2021;27:1278-86. [Crossref] [PubMed]
9. Laila UE, An W, Xu ZX. Emerging prospects of mRNA cancer vaccines: mechanisms, formulations, and challenges in cancer immunotherapy. Front Immunol 2024;15:1448489. [Crossref] [PubMed]
10. Katz MHG, Shi Q, Meyers J, et al. Efficacy of Preoperative mFOLFIRINOX vs mFOLFIRINOX Plus Hypofractionated Radiotherapy for Borderline Resectable Adenocarcinoma of the Pancreas: The A021501 Phase 2 Randomized Clinical Trial. JAMA Oncol 2022;8:1263-70. [Crossref] [PubMed]
11. Versteijne E, van Dam JL, Suker M, et al. Neoadjuvant Chemoradiotherapy Versus Upfront Surgery for Resectable and Borderline Resectable Pancreatic Cancer: Long-Term Results of the Dutch Randomized PREOPANC Trial. J Clin Oncol 2022;40:1220-30. [Crossref] [PubMed]
12. Montagne JM, Mitchell JT, Tandurella JA, et al. CD137 agonism enhances anti-PD1 induced activation of expanded CD8+ T cell clones in a neoadjuvant pancreatic cancer clinical trial. iScience 2024;28:111569. [Crossref] [PubMed]