In Pancreatic Cancer, Does New PAXG Beat Out mFOLFIRINOX?

CHICAGO — For patients with pancreatic resectable stage I-Ill pancreatic ductal adenocarcinoma (PDAC), neoadjuvant chemotherapy can help improve overall survival compared with upfront surgery.

But what chemotherapy regimen provides the best survival outcomes? 

Findings from the phase 3 CASSANDRA PACT-21 trial, presented at the American Society of Clinical Oncology (ASCO) 2025 Annual Meeting, revealed that a newer regimen — PAXG chemotherapy — significantly prolonged event-free survival compared with standard-of-care chemotherapy with mFOLFIRINOX. 

Although immature, the overall survival data suggest that PAXG — capecitabine, cisplatin, nab-paclitaxel, and gemcitabine — may beat out mFOLFIRINOX — oxaliplatin, irinotecan, leucovorin, and 5-fluorouracil. PAXG also led to a significant improvement in key secondary outcomes, including the disease control rate, pathological stage, and detection of intraoperative or early postoperative metastases.

Should these findings change practice?

PAXG appears to be the “most suitable option” for neoadjuvant treatment of these patients, reported Michele Reni, MD, with the Department of Medical Oncology, IRCCS San Raffaele Scientific Institute and Vita-Salute San Raffaele University, Milan, Italy, who presented the findings at ASCO.

However, study discussant Brian Wolpin, MD, MPH, is not yet ready to call PAXG the new standard of care. While preoperative PAXG is a “very promising approach with the potential to change standard of care, more follow-up is needed, particularly overall survival data, to decide whether it should change our standard of care,” said Wolpin, chair of pancreatic cancer at Dana-Farber Cancer Institute in Boston, Massachusetts. 

In other words, the data are “not quite sufficient” at this point to change standard of care, Wolpin said. 

The CASSANDRA PACT-21 trial enrolled 261 patients aged 75 and younger with treatment-naive stage I-III resectable or borderline resectable PDAC and a Karnofsky performance status (KPS) greater than 60%. 

After CA19-9 stratification, 133 were randomly allocated to PAXG and 128 to mFOLFIRINOX administered every 14 days for 4 months. After this, a second randomization was performed, with patients assigned to receive either 2 additional months of the same chemotherapy regimen followed by surgery, or surgery followed by the same chemotherapy. 

The primary endpoint was event-free survival, defined as absence of progression, recurrence, two consecutive CA19-9 increases ≥ 20% separated by at least 4 weeks, unresectability, intraoperative metastasis, or death. Median follow-up was 24.5 months in the PAXG group and 26 months in the mFOLFIRINOX group. 

In the final analysis, with 179 events, median event-free survival was 16 months with PAXG vs 10 months with mFOLFIRINOX. The 3-year event-free survival rate was 31% with PAXG vs 13% with mFOLFIRINOX (hazard ratio [HR], 0.64; P = .003). 

Subgroup analyses showed a doubling of 3-year event-free survival with PAXG over mFOLFIRINOX in the resectable group (41% vs 22%) and borderline resectable group (19% vs 9%), Reni noted. 

Reni said the overall survival data are immature, but favor PAXG over mFOLFIRINOX, with median overall survival of 37 months vs 26 months and a 3-year overall survival rate of 51% vs 40% (HR, 0.70; P = .07). 

Regarding key secondary endpoints, PAXG significantly improved the disease control rate (98% vs 91%), CA19-9 response ≥ 50% rate (88% vs 64%), pathological complete response rate (3% vs 0%), N0 resection rate (36% vs 23%), and detection of intra- or postoperative metastases (5% vs 12%). 

The secondary endpoints “add some additional evidence” suggesting that PAXG may be an effective regimen compared to mFOLFIRINOX, Wolpin told ASCO attendees.

In terms of adverse events, no statistically significant difference was detected, except for a higher rate of grade 3-4 neutropenia in the PAXG group (42% vs 29%), Reni reported. 

There was a trend toward worse score in nausea/vomiting scale in the mFOLFIRINOX group at 4 months compared with baseline. At 4 months, both groups had clinically meaningful deterioration in fatigue, taste, weight loss concern, and hepatic symptoms. 

In his discussion, Wolpin noted that even though PAXG looks like it may be a more effective therapy in preventing recurrence, the event-free survival rate was still only 31% at 3 years. 

“This really shows that we may hit a plateau with what we can do with chemotherapy and surgery, and there has to be some additional thought and implementation of biologically driven therapies,” he explained.

Looking to the future, “I eagerly await further survival data from the CASSANDRA trial,” alongside other relevant trial data from PREOPANC-3 and ALLIANCE A021806, to “help define the appropriate therapy for these patients,” Wolpin said.

The CASSANDRA PACT-21 trial was fully funded by patient associations. Reni has disclosed consulting or advisory roles for AstraZeneca, LEO Pharma, MSD, Oncoscience, Panavance Therapeutics, SERVIER, Sotio, and Viatris, and travel expenses from Advanced Accelerator Application and Daiichi Sankyo. Wolpin has disclosed consulting or advisory roles with Agenus, BeiGene, EcoR1 Capital, Harbinger Health, Ipsen, Mirati Therapeutics, Revolution Medicines, Tango Therapeutics, and Third Rock Ventures.