Searching for Signals: PDAC Atlas Charts Path to New Immunotherapy Combos

In recent years, immunotherapy has rapidly redefined the standard of care across multiple solid tumor types. However, in pancreatic cancer, progress with this modality has been far more elusive. Outside of rare biomarker-defined subsets, the cancer remains largely refractory to checkpoint inhibition, and trials evaluating combination immunotherapy have frequently fallen short of meaningful clinical benefit.

Oncology researchers at Johns Hopkins Medicine have revisited those efforts with a different objective. Led by Won Jin Ho, MD, associate professor of oncology at Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins Medicine in Baltimore, Maryland, and Dimitrios Sidiropoulos, PhD, assistant professor in the Department of Oncology at Johns Hopkins School of Medicine, the team has created an online, publicly accessible cytometric atlas compiling immune response data from clinical trials of patients with metastatic pancreatic cancer treated with vaccines or checkpoint inhibitors.^1,2

Their efforts, recently reported in Cancer Immunology Research,¹ document biologic signals that may shape the next generation of immunotherapy combinations, particularly as new modalities such as RAS-targeted agents and next-generation vaccine platforms enter clinical development.

Behind the Atlas: Facing an Immunotherapy-Resistant Disease

The atlas is driven by decades of work in pancreatic cancer immunotherapy at Johns Hopkins. As a physician-scientist, Ho is no stranger to the lethal, difficult-to-treat nature of pancreatic cancer, which has a 5-year survival rate in the single digits for metastatic disease.³ The dismal prognosis in finding a cure reflects not only late detection and complex anatomy, but also profound treatment resistance.

Pancreatic cancer is widely notorious for its therapeutic resistance, with dense desmoplastic stroma, poor vascularity, tumor heterogeneity, and a highly immunosuppressive tumor microenvironment (TME) collectively limiting drug delivery and blunting antitumor immune responses.⁴ The disease is characterized by abundant myeloid-derived suppressor cells, tumor-associated macrophages, and regulatory T cells, along with sparse effector T-cell infiltration—features that help explain why it has historically been refractory to immunotherapy.^5,6

Sustained efforts have evaluated immunotherapy combinations designed to overcome immune exclusion and reprogram the TME.⁷ Vaccine platforms, in particular, have been explored as a means of priming T-cell responses and converting an immunologically “cold” tumor into one more susceptible to checkpoint inhibition.⁸

Although many studies did not meet their primary end points, they have nonetheless generated critical biologic insights, clarifying mechanisms of resistance, identifying immune correlates of response, and refining hypotheses for subsequent trials. In that sense, each trial has contributed another layer of understanding, Ho said in an interview with Targeted Oncology.

“We [have been] making incremental advances…because you don’t want to make dramatic changes and leaps in a given trial [design],” Ho reflected. “We want to start from what we know and then try to understand what the next best iteration might be. And so, as we’re doing this over a decade plus of clinical trials, we are also accumulating a lot of opportunity.”

That opportunity came in the form of a substantial repository of biospecimens amassed over years of clinical investigation. With advances in high-throughput technologies such as mass cytometry and single-cell profiling, Ho and colleagues recognized they could revisit these samples with greater depth, characterizing immune cell subsets to better understand combination-therapy responses and refine which combinations warrant further development.

“We have these valuable biospecimens that we can learn from, and [so we asked]: Can we integrate all of that information? Then [we] asked the very important question: What combination of immunotherapies out of the things that we have tried thus far would be a critical combination—a backbone, if you will—to then build future therapies on?”

Why Revisit “Negative” Trials?

For Sidiropoulos, who has expertise in computational biology and immunology, the atlas serves as an effort to move beyond single-trial conclusions and toward a more cumulative understanding of immune responses in pancreatic cancer.

“At the end of these [clinical trials], we typically create the static…report of what happened,” Sidiropoulos observed. “Immunologically, there could be a multitude of things, but we really want to move away from that and start building a more reusable resource that we can keep going back to and keep building [so] that [it] can essentially provide long-lasting utility as we continue to develop novel therapies,” Sidiropoulos said in an interview with Targeted Oncology.

He emphasized that a lack of clinical benefit in a trial does not necessarily mean a lack of biological activity. In fact, partial biologic responses—which may be dismissed in isolation—may become meaningful when viewed across aggregated cohorts.

“There's also this notion of, if a given immunological agent wasn't successful in providing clinical benefit, it's a negative finding,” Sidiropoulos said. “But I think what we're seeing is even in ‘negative’ trials, we might see positive signals, and by aggregating multiple cohorts together, we can amplify their smallest immunological signals and learn a lot from them and use that information in future clinical trials.”

Inside the Atlas

The resulting atlas contains more than 200 immune profiles derived from blood samples collected from 64 patients with metastatic pancreatic ductal adenocarcinoma (PDAC) who participated in 3 clinical trials. Patients received combinations of a granulocyte-macrophage colony-stimulating factor gene-transfected tumor cell vaccine (GVAX), CRS-207, anti–CTLA-4, and/or anti–PD-1 therapies.^1,2

Crucially, samples were collected at 2 time points—before and after treatment—allowing the investigators to assess temporal changes in immune cell populations.

“We get to then take the specific time point data, which we have from a given patient…so we can understand what the effect of treatment [is] on the immune profile, and what happens when we add another therapy on top,” Ho explained.

An analysis of the atlas, also included in the Cancer Immunology Research publication, began to tease apart how individual components of these regimens shape immune dynamics. Results showed that GVAX activated T cells and increased checkpoint expression without changing overall T-cell numbers; pairing it with anti–PD-1 amplified T-cell activation, whereas anti–CTLA-4 preferentially boosted memory formation.¹ Although both vaccines had comparable immune effects, GVAX more strongly expanded plasmacytoid dendritic cells than CRS-207.

In parallel, the team has also invested substantial effort in immunophenotyping and profiling multiple lineages of the immune system, enabling examination of both expected responses, such as T-cell activation following vaccination, and unanticipated shifts in other compartments that may drive resistance or synergy.

“We have a focus on T cells, but we also have myeloid cells, so it's really trying to, in parallel and in tandem, query immune responses in the entire immune system as a whole, in peripheral blood and patient tumors,” Sidiropoulos said. “From a hypothesis point of view, we can immediately look at how the immune cell types that we know are responsive to the therapeutics are responding in the patients, but also any other effects that might be happening in other lineages and other components of the immune system, which can potentially drive us in new directions.”

In collaboration with the Johns Hopkins Institute for Data-Intensive Engineering and Science, the investigators built an interactive, web-based interface that allows users to explore immune profiles alongside annotated clinical outcomes. Built-in analytical tools enable users to generate and test hypotheses directly within the platform.

Ho and Sidiropoulos intend the atlas to be a living public resource, with plans to add to the atlas over time as new trials mature and more information becomes available.

“This is our first iteration of this atlas,” Ho noted. “We envision that it's going to be revised and added and expanded over time in the future as well.”

A Blueprint for Immune Profiling in Practice

Although the atlas is primarily a research tool, Ho believes that anyone involved in the pancreatic cancer space, including those caring for patients and those with general interests in the immune system, would find the insights interesting.

He also likens the atlas’ long-term potential to that of large-scale genomic databases such as The Cancer Genome Atlas, which helped catalyze the commercialization of tumor sequencing platforms and normalized mutation profiling in routine oncology practice.

“The Cancer Genome Atlas [was] established quite some time ago, where we're looking at the genetic makeup of many different cancer types. These atlas efforts [are] what really inspired a lot of the commercially available genetic tests out there for tumors. We have these large-scale databases that allow us to understand what mutations are in cancers. And now, as the field progresses forward, and we have more of these mutation-targeted approaches, everybody wants to know the genetic makeup of the cancer,” Ho explained. “To make the point that there is potential for immune profiles for patients, while this is one example of an atlas down the road, we want to get to a point where this can be useful and widely applied clinically as well. We're not quite there yet, but I think this is a step forward.”

Sidiropoulos agreed, emphasizing the growing importance of defining a patient’s baseline immune state before initiating immunotherapy. “To echo Won’s point, I think…it is really important now, but it'll become increasingly more important to know…the baseline immune state of a patient as they're entering these immunotherapy treatments,” he said.

Looking Toward the Next Wave of Combinations

Despite noting the historically disappointing immunotherapy results in PDAC, both Ho’s and Sidiropoulos’ outlooks on the future of pancreatic cancer treatment are laced with optimism.

“I think we're in such an exciting time,” Ho said. “We're super hopeful like never before. The landscape is changing quite rapidly.”

Beyond checkpoint blockade, RAS inhibitors are gaining traction in pancreatic cancer, and vaccine platforms are advancing with more potent designs. Single-cell technologies are illuminating the TME with unprecedented resolution. With these advances, the central question is no longer whether to combine modalities, but how to do so smartly.

“When we think about immunotherapy, it's not just dealing with immunotherapy [alone],” Ho said. “Can we combine chemotherapy? Can we combine targeted therapies, [such as] RAS inhibitors? Can we combine other modalities in very intelligent ways? And because of single-cell technology, we are also [becoming] able to understand the [TME] like never before. For all those reasons, we're very optimistic that treating pancreatic cancer much more effectively is going to be a reality soon.”

For Sidiropoulos, the atlas provides a scaffold for answering those questions. “These trials, these deeply annotated clinical-grade data sets, can get us there. It can get us to effective pancreatic cancer treatment strategies.”

The atlas can be accessed on the SciServer website here.

REFERENCES

1. Sidiropoulos DN, Zhang Z, Durham JN, et al. Cytometric atlas of combination immunotherapy in pancreatic cancer: blood-based signatures reveal vaccine and checkpoint inhibitor responses. Cancer Immunol Res. 2026;14(3):399-410. doi:10.1158/2326-6066.CIR-25-1126

2. Johns Hopkins investigators offer free atlas of immunotherapy responses in pancreatic cancers, report new clinical trial findings. News release. Johns Hopkins Medicine. January 12, 2026. Accessed February 19, 2026. https://tinyurl.com/2jp3a6hp

3. Sohal DPS, Kennedy EB, Cinar P, et al. Metastatic pancreatic cancer: ASCO guideline update. J Clin Oncol. 2020;38(27):3217-3230. doi:10.1200/jco.20.01364

4. Espona-Fiedler M, Patthey C, Lindblad S, Sarró I, Öhlund D. Overcoming therapy resistance in pancreatic cancer: new insights and future directions. Biochemical Pharmacol. 2024;229:116492-116492. doi:10.1016/j.bcp.2024.116492

5. Torphy RJ, Zhu Y, Schulick RD. Immunotherapy for pancreatic cancer: Barriers and breakthroughs. Ann Gastroenterol Surg. 2018;2(4):274-281. doi:10.1002/ags3.12176

6. Ju Y, Xu D, Liao M, et al. Barriers and opportunities in pancreatic cancer immunotherapy. NPJ Precis Oncol. 2024;8(1):199. doi:10.1038/s41698-024-00681-z

7. Giurini EF, Ralph O, Pappas SG, Gupta KH. Looking beyond checkpoint inhibitor monotherapy: uncovering new frontiers for pancreatic cancer immunotherapy. Mol Cancer Ther. 2024;24(1):18-32. doi:10.1158/1535-7163.mct-24-0311

8. Wang Y, Huang P, Li C, Tu S, Yang H. Therapeutic cancer vaccines in pancreatic cancer. Front Immunol. 2025;16:1674743. doi:10.3389/fimmu.2025.1674743