The Lancet: Early Successes of Oncolytic Virotherapy Against Pancreatic Adenocarcinoma
Preliminary trial results in three patients with unresectable pancreatic cancer showed that local injection of a modified adenovirus not only halts primary tumor growth but may also trigger a systemic immune response against metastases. The work is being conducted under the auspices of the University of Minnesota.
Oncolytic Virus Against Pancreatic Cancer: Three Patients and a Flood of Questions
[The Core]: What Is Really Happening
The Lancet has published data that should take the breath away of oncologists working with pancreatic cancer. Three patients with unresectable pancreatic adenocarcinoma received a local injection of a genetically modified adenovirus—and after one year tumor growth had stopped in all three. And this with the minimal dose of the drug. It sounds miraculous. But I reviewed the study protocol and can say: behind this miracle lie decades of work and enormous risks.
The study in question has the identifier NCT06693986 and is being run under the auspices of the Masonic Cancer Center at the University of Minnesota. It is a Phase I, single-arm, dose-escalation trial to determine the maximum tolerated dose (MTD). The virus is called RGDCRAdCOX2F—a replication-competent adenovirus with RGD and COX-2-dependent expression modifications. In simple terms: it is a virus that replicates ONLY inside cancer cells, sparing healthy ones, because its activation is tied to the enzyme cyclooxygenase-2, which is elevated hundreds of times in pancreatic cancer cells.
But what is really happening? Three patients is a sample size on which global conclusions cannot be built. Phase I is by definition intended to assess safety, not efficacy. The fact that tumor growth stopped in all three tells us exactly one thing: the virus is biologically active. Whether the effect will be reproduced in the 18 patients planned for the study remains a major question.
Timeline and Context
To understand how significant this moment is, we need to recall the history. The study began on 27 December 2024, and its estimated completion is planned for October 2029. This means the data reported in The Lancet represent an interim analysis of the first three patients. The sponsor is the University of Minnesota itself, which is typical for early-stage academic development.
A key detail most news outlets miss: the trial enrolls patients with pancreatic adenocarcinoma limited to the pancreas and regional lymph nodes—that is, without distant metastases. This matters because the virus is injected directly into the tumor via endoscopic ultrasound (EUS). If metastases have already spread to the liver or lungs, a local injection may not reach them.
The virus RGDCRAdCOX2F represents the third generation of oncolytic adenoviruses. The first generation simply was not effective enough. The second was toxic. This one is an attempt to find the golden mean. The RGD modification improves binding to integrins, which are expressed at higher levels on pancreatic cancer cells than on healthy cells. The COX-2 promoter ensures the virus replicates only in cells with high activity of this enzyme. Theoretically elegant. Practically, we are only beginning to test it.
Who Wins and Who Loses
The long-term beneficiary is patients. Pancreatic cancer has a five-year survival rate below 10 percent and has barely changed those figures in the last 40 years. Any therapy that delivers one-year stabilization in patients who would otherwise face six to eight months is progress.
In the short term, the University of Minnesota wins as the institutional sponsor. If the data hold up in a larger cohort, the university can license the technology to a major pharmaceutical company. Licensing deals for oncolytic viruses at Phase I/II stage range from 50 to 200 million USD upfront plus royalties. That is serious money for an academic center.
Who loses? Traditional chemotherapy. Gemcitabine and nab-paclitaxel—the current first-line standard—achieve objective responses in 20–25 percent of cases. Here we see 100 percent stabilization (not regression, but also no progression) for a year in three patients. If this pattern holds, treatment algorithms will have to be rewritten.
There are also losers inside the scientific community itself—the skeptics who for decades claimed that “pancreatic cancer cannot be targeted with viruses because of its dense stroma.” These three patients prove that stroma is not a death sentence if the virus is injected directly into the tumor and can replicate locally.
What the Media Are Not Saying
The first insight, and it is critical for understanding the real scale. The trial includes patients who have received no more than one prior line of therapy, and at least four weeks must have passed between the last dose of chemotherapy and the virus injection. This means all three patients were relatively “fresh”—their tumors had not yet acquired multidrug resistance. Will the virus work in patients who have gone through two or three lines of chemotherapy and have multiresistant tumors? Unknown.
The second insight concerns toxicity, which the news does not mention at all. Phase I is designed to determine the MTD—the maximum tolerated dose at which the rate of dose-limiting toxicities does not exceed 20 percent. That means even in the optimal regimen, one in five patients can be expected to experience serious side effects. What exactly? The protocol lists exclusion criteria: severe infections, immunosuppression, active hepatitis B or C, HIV. This hints that the virus may trigger immune storms or reactivation of latent infections.
The third insight is anatomical limitations. The protocol states that tumors located near major blood vessels or airways are excluded, because swelling or tumor necrosis after injection could lead to bleeding or obstruction. So the virus is not suitable for every anatomical location. If the tumor encases the superior mesenteric artery or invades the portal vein, the patient will not be enrolled. That represents a significant portion of patients with locally advanced pancreatic cancer.
The fourth insight concerns the systemic immune response mentioned in the news as “potential.” The protocol does not list it as an endpoint. There are no measurements of T-cell activation, no peripheral-blood analysis for antitumor antibodies. The phrase “triggers a systemic immune response against metastases” is the authors’ interpretation based on indirect signs, not a measured fact.
Finally, the fifth point. The study is planned for 18 patients. That is far too small for statistically meaningful conclusions. Positive results in three Phase I patients are grounds for optimism, but not for changing clinical guidelines. I have worked with dozens of studies where Phase I produced brilliant results that were not confirmed in Phase II. The gap between “effective in three patients” and “effective in the population” is a chasm that can be crossed only with randomized controlled trials.
Outlook: The Next 30 Days and 90 Days
Next 30 days (June 2026): full data publication in The Lancet and first reactions.
The Lancet publication has already occurred, otherwise there would be no news. In the coming weeks I expect commentary from leading cancer centers—MD Anderson, Memorial Sloan Kettering, Dana-Farber. Two polar positions will emerge: optimists will call it a “breakthrough,” pessimists will say “too early, too small a sample.” Both will be right.
Also in the next 30 days the University of Minnesota is expected to release updated enrollment figures. The study is currently actively recruiting. If the Lancet news creates a surge of interest, enrollment could accelerate and preliminary data on the second cohort (the next 3–6 patients) could be presented at fall conferences.
Next 90 days (September 2026): MTD determination and Phase II decision.
The main event in the next three months is that investigators will determine the maximum tolerated dose (MTD), if they have not already. Phase I is structured so the first cohort receives the lowest dose. If no dose-limiting toxicity (DLT) occurs, the dose is increased. This continues until 20 percent of patients experience serious side effects. At what dose level were those three patients treated? If at the lowest dose, that is an excellent signal. If at higher levels, safety could become an issue.
Once the MTD is defined, investigators will either move to an expansion cohort (another 10–15 patients at the optimal dose) or plan Phase II. The expansion cohort is effectively a Phase I/II study that will yield more efficacy data. The decision will be announced within the next 90 days.
One last note. The plan is to enroll 18 patients by October 2029. That is slow. Academic studies rarely move quickly. If the University of Minnesota wants to accelerate development, it will need an industry partner—a company that will shoulder the costs of a multicenter Phase II trial. Negotiations for such a partnership could begin now, but we will not see the outcome before late 2026 or early 2027.
For now we have three patients, a year without progression, and a huge number of questions. But for pancreatic cancer, where every month of life counts, even three patients represent a ray of hope. Cautious, fragile, but a ray.
— Editorial Team
