Toripalimab Plus Modified FOLFIRINOX Extends Event-Free Survival in Pancreatic Cancer
Borderline resectable and locally advanced pancreatic cancer remain among the most difficult settings in gastrointestinal oncology because local control, systemic spread, and surgical feasibility all compete in the same treatment window. In this open-label phase II randomized trial, adding the PD-1 inhibitor toripalimab to modified FOLFIRINOX produced a statistically significant improvement in event-free survival versus chemotherapy alone, without a parallel rise in treatment-related toxicity or postoperative morbidity.
The result matters because the therapeutic ceiling in pancreatic cancer has long been set by biology as much as by regimen choice. Modified FOLFIRINOX already delivers an aggressive cytotoxic backbone, but resistance emerges early, microscopic metastases often declare themselves quickly, and even patients who appear technically resectable can move through a narrow path to surgery. Immune checkpoint blockade has been slow to gain traction in pancreatic cancer, in part because the tumor microenvironment is typically dense, suppressive, and poorly permissive to immune-cell entry. In this trial, combining PD-1 blockade with cytotoxic therapy appeared to push that boundary enough to improve the time patients stayed free of progression, surgery failure, or death.
Trial Design and the Primary Endpoint
The study enrolled patients with histologically confirmed borderline resectable pancreatic cancer or locally advanced pancreatic cancer and randomly assigned them in a 1:1 ratio to modified FOLFIRINOX plus toripalimab or modified FOLFIRINOX alone. Treatment response was assessed every 2 months, and surgical eligibility was decided by a multidisciplinary team. The primary endpoint was event-free survival in the intention-to-treat population.
Between April 2019 and March 2025, 392 patients entered the trial, with 196 assigned to each arm. After a median follow-up of 39.6 months, median EFS reached 11.1 months in the immunochemotherapy arm versus 9.6 months with chemotherapy alone. The hazard ratio of 0.74, with a 95% confidence interval of 0.60 to 0.92, and a P value of .007, places the finding squarely in statistically significant territory.
The separation in EFS was modest in absolute terms but clinically relevant in a disease where each additional month can determine whether a patient remains eligible for operative exploration or slips into unresectability. In pancreatic cancer, small gains can carry disproportionate value if they preserve the chance of surgery.
Survival, Response, and the Surgical Pathway
Overall survival trended in the same direction, although the confidence interval crossed 1. Median OS was 20.6 months with toripalimab-based immunochemotherapy and 17.1 months with chemotherapy alone, with an HR of 0.81 and a 95% CI of 0.63 to 1.03. The data support a survival signal, but not yet a definitive OS separation.
Objective response also favored the combination. The response rate reached 30.1% in the toripalimab arm versus 18.9% in the control arm. That difference is clinically meaningful in a presurgical setting because tumor shrinkage and biologic stabilization can improve candidacy for resection, even when radiographic downstaging does not fully capture the underlying change in tumor behavior.
Despite the improved response rate, the overall tumor resection rates were nearly identical: 42.3% with immunochemotherapy and 43.9% with chemotherapy alone. R0 resection rates were 84.1% among resected patients in the toripalimab arm and 74.7% in the control arm. The data suggest that the combination may have improved the quality of surgery among those who reached the operating room, even if it did not increase the overall proportion of patients who underwent resection.
Why the Biology Still Matters
The biologic rationale is straightforward enough to be credible, even in a disease that has been historically resistant to immunotherapy. Modified FOLFIRINOX can generate tumor antigen release and immune priming, while PD-1 blockade can relieve the inhibitory checkpoint pressure that keeps exhausted T cells from functioning. In a tumor type known for immune exclusion, this sequencing may help convert a cytotoxic response into a more durable antitumor state.
BRPC and LAPC are not interchangeable clinical entities, though both live on the same difficult continuum. The prespecified subgroup analyses suggested the benefit concentrated in borderline resectable pancreatic cancer, with improved EFS and OS there but not in locally advanced pancreatic cancer. That split is biologically plausible: borderline resectable tumors likely sit closer to a threshold where modest tumor control can change the surgical endpoint, while locally advanced disease may already be too anatomically and biologically entrenched for checkpoint addition to shift outcomes meaningfully.
A one-size-fits-all strategy did not emerge from these data.
Toxicity, Operability, and Postoperative Safety
Safety signals did not worsen with the addition of PD-1 blockade. Grade 3 or higher adverse events occurred in 47.9% of patients receiving immunochemotherapy and 49.2% of those receiving chemotherapy alone. The near overlap argues against a major additive toxicity burden from toripalimab in this regimen.
Postoperative morbidity also remained comparable among resected patients. Grade III or higher postoperative complications occurred in 30.1% of those in the immunochemotherapy group and 25.6% in the chemotherapy group. For surgeons and medical oncologists working together on pancreatic cancer, that finding is important: intensifying induction treatment did not obviously compromise surgical safety in this cohort.
The absence of an excess toxicity penalty gives the combination practical appeal, especially when the goal is to preserve a chance at curative-intent surgery. In pancreatic cancer, a regimen that improves EFS without increasing perioperative harm occupies a narrow but valuable therapeutic space.
What This Means for Current Practice
The trial adds a randomized data set to a question that has been clinically unsettled for years: whether checkpoint inhibition has a place before surgery in nonmetastatic pancreatic cancer. Here, the answer is not universal, but it is clearly more favorable than previous skepticism would have predicted. The benefit was real for EFS, the response rate improved, and the safety profile remained manageable.
Still, the resection rate did not rise, and overall survival has not yet crossed the statistical threshold needed for a definitive claim. The subgroup pattern also warns against assuming the same effect in every locally advanced case. Patients with BRPC may be the group most likely to gain from this approach, while LAPC may require additional biologic or local strategies before immunotherapy can change the trajectory.
The study leaves a few concrete questions in place: whether the EFS advantage will mature into a firmer OS benefit, whether biomarker selection can identify the patients most likely to benefit, and whether the same pattern would hold in broader practice outside a randomized trial framework. For now, the data support toripalimab plus modified FOLFIRINOX as a credible preoperative option in selected patients with borderline resectable disease and as a regimen that deserves cautionary, not dismissive, attention in locally advanced pancreatic cancer.