Neoadjuvant PAXG Significantly Improves EFS Over mFOLFIRINOX in Resectable PDAC

Cisplatin, nab-paclitaxel, gemcitabine, and capecitabine significantly improved EFS vs mFOLFIRINOX in resectable or borderline resectable pancreatic ductal adenocarcinoma.

The neoadjuvant chemotherapy regimen PAXG (cisplatin, nab-paclitaxel, gemcitabine, and capecitabine) has demonstrated superior efficacy compared with modified FOLFIRINOX (mFOLFIRINOX) in patients with resectable (R) or borderline resectable (BR) pancreatic ductal adenocarcinoma (PDAC). Final results from the phase 3 CASSANDRA trial (NCT04793932) that were presented at the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting revealed a statistically significant improvement in event-free survival (EFS) with PAXG.¹

With a median follow-up of 24.5 months for PAXG and 26.0 months for mFOLFIRINOX and 179 events observed, PAXG demonstrated a statistically significant improvement in EFS. The 3-year EFS rate was 31% (95% CI, 25%-36%) in the PAXG arm compared with 13% (95% CI, 9%-17%) in the mFOLFIRINOX arm, corresponding to a hazard ratio (HR) of 0.64 (95% CI, 0.48-0.86; P =.003). The median EFS was 16.0 in the PAXG arm (95% CI, 12.4-19.8) vs 10.2 (95% CI, 8.7-13.0) in the mFOLFIRINOX arm.

“PAXG significantly improved the median event-free survival by almost 6 months, with an increase in the 3-year event-free survival rate from 13% to 31%,” Reni, a specialist in medical oncology at the Department of Medical Oncology at the San Raffaele Research Hospital in Milan, Italy, said during the presentation at ASCO 2025.

About the Phase 3 CASSANDRA Trial

The trial enrolled 260 eligible patients across 17 Italian centers between November 2020 and April 2024. Patients ≤75 years of age with treatment-naive, histologically confirmed PDAC and Karnofsky performance status ≥70 were randomly assigned in a 2x2 factorial design.

The first randomization assigned patients to receive either PAXG (n = 133) or mFOLFIRINOX (n = 128), and the second randomization determined whether chemotherapy would be administered for 6 months preoperatively or split into 4 months preoperative and 2 months postoperative.

Patients in the PAXG arm were a median of 65 years of age (range, 42-76) compared with 63 (range, 41-76) in the mFOLFIRINOX. Across these arms, 93% and 91% had a Karnofsky performance score of 90 to 100, and 52% and 51% were borderline resectable while the rest were resectable, respectively.

The primary end point was EFS, defined broadly to capture multiple clinical failure points, including radiologic progression, recurrence, intraoperative discovery of metastasis, unresectability, death, or 2 consecutive increases in CA19-9 ≥20% separated by at least 4 weeks.

Additional Safety and Efficacy Data

Secondary end points further reinforced the superiority of PAXG. The disease control rate (DCR) was significantly higher with PAXG (98% vs 91%; P =.009), as was the proportion of patients achieving a CA19-9 reduction greater than 50% (88% vs 64%; P <.001). Although the resection rate did not differ significantly between groups (75% vs 67%; P =.165), PAXG resulted in more favorable pathological outcomes, including a greater percentage of patients with stage <II disease (35% vs 23%; P =.03) and a higher rate of node-negative resections (R0/N0).

While overall survival (OS) data remain immature, early trends mirror the EFS curves, suggesting a likely downstream OS benefit. The median OS in the PAXG arm was 37.3% (95% CI, 26.9-not reached) vs 26.0 (95% CI, 23.6-39.4) in the mFOLFIRINOX arm. The HR was 0.70 (95% CI, 0.47-1.04; P =.07).

“Regarding the other secondary end points—assessed in the intention-to-treat population—PAXG significantly improved the disease control rate, the CA19-9 response rate, the pathological complete response rate, the N0 resection rate, and consistently reduced the detection of intraoperative and postoperative metastases,” Reni added.

Specifically, the DCR in the PAXG and mFOLFIRINOX arms were 98% and 91% (95% CI, 1.02-1.14; P =.01), CA19-9 response rates were 88% and 64% (95% CI, 1.15-1.65; P <.001), and pathological complete response rates were 4 and 0 (P =.047). Notably, PAXG was also associated with a lower rate of intra- and early postoperative metastasis detection (5% vs 12%), a clinically relevant marker of preoperative efficacy.

In terms of safety, both regimens were generally tolerable, though grade 3 to 4 neutropenia was more common in the PAXG arm (42% vs 29%). Other grade 3 to 4 toxicities, including fatigue, diarrhea, nausea/vomiting, neuropathy, and hepatic enzyme elevation, were comparable between arms, with only modest numerical differences.

The mFOLFIRINOX arm experienced higher attrition rates (47% vs 28%), primarily driven by toxicity and disease progression, which led to fewer patients completing both the preoperative and postoperative phases of treatment. While overall quality of life (QoL) scores did not differ significantly between the 2 treatment groups, patients receiving mFOLFIRINOX exhibited greater deterioration across more QoL domains by the fourth month. This was most notable in areas such as nausea, vomiting, and general symptom burden.

Potential Limitations

Reni noted that this trial was initially designed as a phase 2 study to identify the better neoadjuvant regimen but was upgraded to a phase 3 trial after the PREOPANC-1 and -2 studies established neoadjuvant therapy, specifically mFOLFIRINOX, as a standard of care in resectable and borderline resectable PDAC. This made CASSANDRA’s comparison both timely and clinically relevant.

The use of EFS as the primary end point aligns with FDA standards for neoadjuvant trials, similar to disease-free survival in adjuvant studies like PRODIGE-24 and CONKO-001. While the inclusion of CA19-9 kinetics as a failure criterion may be debated, it was applied equally across arms, with a higher rate of baseline CA19-9 normalization in the mFOLFIRINOX group, if anything, biasing against PAXG.

Conclusions

The CASSANDRA trial delivers compelling evidence that PAXG should be considered the preferred neoadjuvant regimen for patients with resectable or borderline resectable PDAC. The significant improvement in EFS, enhanced tumor response metrics, and manageable toxicity profile position PAXG as a strong alternative to mFOLFIRINOX.

These findings, if corroborated by maturing OS data, could reshape standard treatment algorithms for patients with PDAC.

"PAXG appears to be the most suitable option for neoadjuvant therapy in patients with resectable or borderline resectable PDAC," concluded Reni.

Reference

Reni M, Macchini M, Orsi G, et al. Results of a randomized phase III trial of pre-operative chemotherapy with mFOLFIRINOX or PAXG regimen for stage I-III pancreatic ductal adenocarcinoma. J Clin Oncol. 2025;43(suppl 17):LBA4004. doi:10.1200/JCO.2025.43.17_suppl.LBA4004