Neoadjuvant mFOLFIRINOX Exceeds Historical OS Rates in Pancreatic Cancer

Matthew H. G. Katz, MD, CMQ, FACS, FASCO

mFOLFIRINOX prior to pancreatectomy demonstrated a favorable overall survival (OS) rate vs mFOLFIRINOX plus hypofractionated radiotherapy in patients with borderline resectable pancreatic cancer, according to findings from the phase 2 A021501 trial (NCT02839343). However, the results do not support the complete omission of radiation therapy from standard therapy in this population, according Matthew H. G. Katz, MD, CMQ, FACS, FASCO.

This trial was designed to investigate the efficacy of neoadjuvant mFOLFIRINOX, a chemotherapy regimen consisting of oxaliplatin, irinotecan, leucovorin, and fluorouracil, and neoadjuvant mFOLFIRINOX followed by hypofractionated radiation therapy in patients with borderline resectable pancreatic cancer. The primary end point was the 18-month OS rate, which was compared with a historical control rate of 50%.

Patients who received neoadjuvant mFOLFIRINOX alone displayed a Kaplan-Meier–estimated 18-month survival rate of 66.7% (95% CI, 56.1%-79.4%), exceeding the preselected historical OS control rate of 50%. Accrual to the second arm closed early due to futility. At the time of the interim futility analysis, the Kaplan-Meier–estimated 18-month survival rate was 47.3% (95% CI, 35.8%-62.5%) in this population.¹

“Demonstrating a survival duration [with chemotherapy] that’s different from that from a local treatment is a heavy lift, but we see it in this study,” Katz said. “However, to say that radiation therapy was harmful, [or that] this is the end of radiation therapy, is an overstatement. We simply need to figure out who benefits [from that modality].”

In an interview with OncLive^®, Katz, a professor and chair in the Department of Surgical Oncology in the Division of Surgery at The University of Texas MD Anderson Cancer Center, explained how the results of this study support the use of neoadjuvant mFOLFIRINOX in patients with borderline resectable pancreatic cancer. He also emphasized that, although the efficacy of radiation therapy could not be determined in this study, this treatment option is tolerable and should be explored further.

OncLive^®: Could you summarize the findings from the A021501 clinical trial?

Katz: This study compared 2 preoperative treatment regimens for patients with anatomically advanced pancreatic cancer. [A total of] 126 patients were randomized to 1 of 2 arms; the first was 4 months of systemic chemotherapy with mFOLFIRINOX prior to anticipated pancreatectomy. The other was a little under 4 months of systemic chemotherapy with mFOLFIRINOX followed by stereotactic body radiation therapy prior to anticipated surgery. Ultimately, the objective was not to compare the arms but to independently compare each with a historical 18-month OS control rate of 50%.

The second arm closed for futility at the interim analysis after 56 patients were accrued, as an insufficient number of patients had undergone R0 resection. The systemic chemotherapy-alone arm completed to full accrual and ultimately, that arm, the preoperative therapy arm, was associated with a median OS duration of 29.8 months.

We concluded from this study that the survival rate of patients in the chemotherapy-only arm was superior to the historical control rate. We declared that arm efficacious and established that this regimen should be built upon in future studies of pancreatic cancer, such as in combination with novel therapeutics.

We couldn’t say much about the radiation arm; based on the protocol, we could have only evaluated the efficacy of the arm statistically if [it had fully] accrued, which it did not. The role of radiation therapy in this context is still a bit ambiguous.

What do these findings indicate about the potential future for radiation therapy in patients with pancreatic cancer?

Radiation therapy did not harm the patients; they tolerated the radiation therapy well. There’s no objective evidence to say that the radiation therapy was harmful. Did it help? It would be an overstatement to say that the radiation therapy was helpful.

Frankly, there still does remain a role for radiation therapy, though not routinely. In my interpretation of the existing data provided by this trial, routine radiation therapy is not indicated for any stage of localized pancreatic cancer. However, there may be a role for radiation therapy for localized pancreatic cancer in certain patients. The problem is, we haven’t figured out precisely who those patients are, [as the patients enrolled in this trial would probably not benefit]. The key for us moving forward is to figure out: Who are those patients, and when do they need it?

How did the baseline characteristics of the trial population influence the results?

[These findings] were about what I expected. Most people would have thought that either arm would do better because there are single-institution series of patients with advanced disease who have higher resection rates, OS rates, and overall duration rates than were exhibited by either arm on this trial.

However, [the patients in this trial] were a relatively unselected group of patients. They were selected only based on their anatomy, and they were very heterogeneous in terms of the extent of their disease when viewed in other ways. Additionally, they were enrolled from sites all over the United States, some of which were high-volume centers, some of which were not high-volume, [and received] operations performed by a variety of surgeons.

In that context, the data are impressive; a 29.8-month OS for an unselected group of patients with anatomically advanced disease operated on in centers across the United States is impressive. But I’m not particularly surprised. Patients who have localized disease should be treated with chemotherapy, but not all patients should receive radiation therapy; these data provide additional support for that.

The fact that the radiation therapy arm closed early was unfortunate. I wish it had done at least well enough to accrue fully. We need to run trials in the future that look at different disease populations that are selected on criteria other than anatomy. We need to stop trying to demonstrate differences in OS with radiation. Radiation is a local treatment, and we should probably be evaluating its success or failure on the basis of end points that are more reflective of local disease, such as local disease progression or symptoms related to the local tumor.

Although this study was not designed to reflect real-world practice, how much did the findings indicate how pancreatic cancer is treated in the clinic?

In many respects, this trial does reflect real-world practice, because we had accrual from such a diverse array of sites from across the United States. Though it was fairly real-world from that perspective, it was also well controlled. We included many quality control elements in this trial that have not been used in other trials, other than the trial we ran prior to this, which established those quality control measures.

A021501went to extra lengths to ensure that the patient population we were studying was homogenous, that the radiation therapy was applied using a consistent technique that met high standards, that surgery was performed to high standards, that the imaging studies were all interpreted the same way, and that the responses were all interpreted the same way. Although this study was conducted in many different institutions throughout the country, we went to great lengths to ensure that the care provided to each patient was of the highest quality possible and performed in such a way that the data were as solid as possible.

Do you think these results are practice changing?

They’re practice affirming, which is a bit different. The radiation debate has been going on for 30 years now, and this trial has not settled anything in that regard. Some providers, surgeons, and medical oncologists firmly believe in radiation, [and some] don’t believe in radiation. This trial is not going to lead either of those groups to change their opinion.

However, the wealth of the data suggests that radiation can be beneficial in certain patients. The problem is that, like other studies, this trial was trying to do too much in demonstrating that survival could be meaningfully prolonged in patients selected only based on their anatomic stage. Prior studies have shown that treatment doesn’t work in that way. This study showed that radiation therapy does not have that ability.

However, the answer may be somewhere in the middle. [Radiation therapy] almost certainly does help some patients, and future studies will hopefully do a better job of figuring out precisely who those patients are.

What will happen next with mFOLFIRINOX based on these results?

The phase 3 A021806 study [NCT04340141] is ongoing through the Alliance for Clinical Trials in Oncology and is looking at mFOLFIRINOX applied perioperatively; preoperative and postoperative mFOLFIRINOX vs de novo surgery followed by mFOLFIRINOX. It’s attempting to answer the question: Which is better, preoperative or postoperative mFOLFIRINOX?

All these studies, A021501 included, are missing an opportunity to evaluate pancreatic cancer on the basis of nonanatomic criteria. By limiting us to studying the disease in an anatomic way, [we] are restricting our ability to provide definitive answers. A021806 is an important trial that will provide additional data to help inform the treatment of patients with localized disease.

There are opportunities to study novel agents in the preoperative setting in combination with mFOLFIRINOX. How do you improve the results of systemic chemotherapy alone by adding to it novel agents?

There’s also a huge opportunity to use new measures of response to inform treatment in real time. For example, in the A021501 study, patients who were enrolled on the chemotherapy arm received 4 months of chemotherapy, without any attention given to whether that chemotherapy was effective. Though mFOLFIRINOX seems to be generally effective in a large number of patients, there are patients for whom it isn’t going to be effective. How do you figure out who those patients are, for whom the regimen is working early?

Looking at it the other way: How do you identify patients for whom the regimen is not working early, so you can switch gears and take a different approach? How can you use measures of disease response to modulate the duration of therapy within which the treatment is given? Some patients probably don’t need 4 months of therapy, maybe some need 2 months, maybe some need 8 months. How do you know, and how do you make those decisions quickly? Those are what we need to work on and what we are working on.

Reference

1. 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. Published online July 14, 2022. doi:10.1001/jamaoncol.2022.2319