Time is a very important element here. When we give patients more of their time and we prevent them from having to come to the hospital every day, Monday through Friday, for 5 and a half weeks, there could potentially be benefits associated with that.
The other element that was quite unique about our study, and different from other studies that have applied SBRT, is that we did adaptive MRI-guided SBRT in most of the patients who had SBRT. Of the just over 50 patients who had SBRT, the majority received adaptive MRI guidance. We used an MRI linear accelerator. We recontoured their tumors each day, and we recontoured their normal structures. While we did not test this, that precision, daily adaptation, and daily accuracy may have had something to do with the preservation of their quality of life and function.
How did surgical outcomes, including resection rates and complications, compare between the arms?
R0 resection rates were very high for both approaches, well over 80%, which is impressive. The other thing to note is that the majority of patients went to surgery. We enrolled 102 patients, and 4 patients withdrew, so we had 98 who were available for the analysis. Of those 98, 83 went to surgery. That’s a high rate of surgical resection, and it speaks volumes to the expertise of the surgeons in our program. We have amazing surgical oncologists. They’re wonderful to work with, very multidisciplinary, and very supportive of and open-minded to radiation and new ways of treating patients with pancreatic cancer. We work very closely together. We meet every week and review all the patients closely, and that contributed to the high rates of surgical resection. These rates were much higher than in other trials that have tested SBRT in the neoadjuvant setting.
It speaks to a couple of important elements. Number one, patients who have pancreatic cancer need to go to a high-volume center. They need to take the time to find a qualified team that has a deep breadth of experience treating pancreatic cancer. That’s something we do daily and weekly at the Medical College of Wisconsin and the LaBahn Pancreatic Cancer Program. We have an incredible amount of expertise in treating pancreatic cancer, and it’s clear that makes a big difference. When patients come in with pancreatic cancer and we review them as a multidisciplinary team, most of the time, if they are surgical candidates, they end up going to surgery, which is a wonderful thing.
The other thing to note is that there were no differences in surgical complications. When we started the study, we didn’t know if there would be more complications associated with giving SBRT preoperatively. That had not been studied robustly in a prospective randomized trial compared with more traditional, conventionally fractionated radiation schedules, so we just didn’t know what the toxicity was going to look like. We saw no differences in surgical complications or toxicity. We also collected surgical perspectives on intraoperative complexity, fibrotic findings, and the difficulty of the surgical procedure, and we didn’t see any differences there either. That gives us a lot of confidence in building on SBRT as a neoadjuvant platform.
The other thing to note is that we did SBRT using slightly different volumes, and a lot of that detail will be in the presentation as well as in the manuscript. We used a low-dose clinical target volume to cover regional nodes. For example, we treated the celiac axis, the superior mesenteric artery, and the superior mesenteric vein. There’s a common region called the triangle region, which has been published by a group at Johns Hopkins looking at high-risk areas of recurrence. We treated all those areas with 25 Gy in 5 fractions as a preventative dose on the SBRT arm, and then we simultaneously boosted the tumor to the mid-30s, so 33 Gy to 35 Gy. The surgical outcomes were essentially identical in both cohorts. Margin negativity rates were extremely similar.
Does a noninferior but more convenient result change how you sequence radiation therapy with surgery and systemic therapy?
We have to be careful labeling them as noninferior. It was a superiority trial, not a noninferiority trial. We would typically need a much larger trial for that. We didn’t see any differences between the arms, but one thing I want to emphasize is that radiation in the neoadjuvant setting needs more prospective evaluation.
We’ve seen many trials published in the past 3 to 4 years, many of which have used radiation and shown it’s very effective. For example, PREOPANC-2 used radiation with a more moderately hypofractionated schedule, not an SBRT-type schedule, and showed very similar efficacy to FOLFIRINOX, which is quite impressive.2 Toxicity rates were low. We’ve seen multiple other trials published over the past several years that have used radiation, either single-arm, single-institution trials or smaller randomized trials, but it’s been done in a very homogeneous way, and there haven’t been many trials that have compared different radiation schedules head-to-head. That’s valuable.
What we showed is that if an institution is committed to enrolling patients into trials, a single institution can enroll very meaningful numbers of patients into prospective randomized trials. As a relatively modest-sized single institution, we were able to enroll 102 patients into this prospective randomized trial, which speaks to the value of commitment, collaboration, and perseverance in enrolling these patients into clinical trials. It also speaks volumes about the patients who are willing to participate and engage in this.
I do think it reflects that a more convenient schedule is very meaningful to patients. It’s a wonderful thing to be able to sit with patients who are [experiencing] a life-threatening diagnosis and tell them that you feel comfortable that they can come for only 5 total treatments, as opposed to having to come for 5 and a half weeks. When we designed the study, we were hypothesizing that 5 and a half weeks may have been better, and we certainly did not see that. It’s a nice thing to be able to build on the 5-treatment platform. We’re opening a new study right now that builds on this SBRT platform in a slightly different way, and that is something I look forward to continuing to study in the future.
What stood out to you most in the results, including the safety findings?
The toxicities were not different between the cohorts. There were no major perioperative differences. What stood out to me the most is the fact that there wasn’t much of a difference between the arms. I was very surprised. The only thing that was slightly different was the quality of life, which is remarkable. That, to me, was surprising. I thought that 5 and a half weeks of daily radiation with concurrent chemotherapy would result in more pathologic downstaging, less nodal positivity, and perhaps higher rates of margin-negative resections, but none of that panned out.
This is why we do randomized trials. You can have your biases and preconceived notions as an oncologist, but this is why we lean hard into randomized trials: to say, no, we don’t know which one is better; we need to prospectively study this, and we’re going to randomly assign these patients to different interventions. What stood out to me is the fact that there was very little to no difference between the cohorts.
What would a phase 3 trial building on this need to test?
That is a great question. Over the past several years, there’s been a little bit of the wind taken out of the sails in the use of neoadjuvant radiation therapy for pancreatic cancer, and I would confidently submit that has been done very erroneously. It’s very unfortunate that the tide has temporarily turned away from focusing on methods to give neoadjuvant radiation therapy and is instead only focusing on systemic therapy.
In a malignancy like pancreatic adenocarcinoma, systemic therapy is the absolute backbone. These patients, unfortunately, have very high rates of metastatic disease, and they need robust systemic therapy. Ignoring locoregional modalities such as radiation, which are quick and convenient and significantly reduce rates of locoregional recurrence and improve rates of margin-negative resection, is a mistake for our patients. Novel ways of giving radiation can be very quickly, very conveniently, and very safely incorporated into a neoadjuvant paradigm. There is no reason why, in a malignancy as deadly as pancreatic cancer, we should not be robustly considering all forms of neoadjuvant therapy, inclusive of chemotherapy, radiation therapy, and, ideally, margin-negative surgical resection. That is how we cure patients with pancreatic cancer, and we are seeing meaningful rates of cure in the cohorts we’re publishing.
As we consider phase 3 randomized clinical trials, we cannot ignore significant advances in the delivery and use of radiation. A lot of people talk about radiation as if it’s one modality or one option. Occasionally, though not from very experienced pancreatic cancer groups, you’ll hear phrases like, “Radiation doesn’t have a role,” or “It doesn’t work,” or “It’s not effective for patients with pancreatic cancer.” My next question is: What type of radiation? How are you giving the radiation? What’s the total dose? What’s the dose rate? What volumes did you treat? There are so many nuances to how radiation can be given. It’s a real mistake to dismiss it as an ineffective modality because it is extremely effective at preventing locoregional recurrence.
My ideal phase 3 trial in this setting would focus on entirely new ways of giving radiation therapy. For example, how are we giving daily adaptive MRI-guided radiation? How are we using novel dose rates, giving radiation at either faster or slower dose rates? How are we combining different forms of radiation with novel systemic therapies such as daraxonrasib and the pan-RAS inhibitors? These are important questions, and our patients deserve opportunities to participate in trials that are testing them. What we need in the neoadjuvant space for pancreatic adenocarcinoma is, number one, large phase 3 clinical trials that are testing robust and innovative ways of delivering radiation neoadjuvantly. That’s something that, over the next several years, we’re very committed to pushing forward and promoting the importance and value of because it is an incredibly important space that we need to keep exploring.
References
- Hall WA, Tsai S, Banerjee A, et al. A randomized, phase II clinical trial of stereotactic body radiation therapy or conventionally fractionated concurrent chemotherapy and radiation therapy preoperatively for pancreatic adenocarcinoma, the SOFT Preop study. Presented at: 2026 ASTRO Annual Meeting; September 26–30, 2026; Boston, MA. Abstract LBA 08.
- Janssen QP, van Dam JL, van Bekkum ML, et al. Neoadjuvant FOLFIRINOX versus neoadjuvant gemcitabine-based chemoradiotherapy in resectable and borderline resectable pancreatic cancer (PREOPANC-2): a multicentre, open-label, phase 3 randomised trial. Lancet Oncol. 2025;26(10):1346-1356. doi:10.1016/S1470-2045(25)00363-8
- FDA approves daraxonrasib for metastatic pancreatic adenocarcinoma. News release. FDA. August 26, 2026. Accessed September 29, 2026. https://tinyurl.com/2kp6d3jm
