Jashodeep Datta, MD

Pancreatic ductal adenocarcinoma (PDAC) is a devastating malignancy, owing in part to the fact that most patients present with advanced disease, with only 20% of patients eligible for surgical resection. The dismal oncologic outcomes with surgery alone, and the near universal distant failure, in patients with operable PDAC have mandated a multimodality approach for this disease.

The advent of increasingly effective combination chemotherapy regimens, perioperative safety and technical mastery of pancreatectomy, and growing appreciation of the molecular underpinnings of PDAC have resulted in gradual but consistent improvements in disease-specific survival and OS over the past decade, although much work lies ahead.

Paramount to the continued progress in the field is the pressing issue of multimodality treatment sequencing in operable PDAC. Adjuvant therapy with modified FOLFIRINOX was established as the new benchmark in the PRODIGE-24 trial, but perhaps applicable to only the fittest patients completing the Darwinian selection process of tolerating major pancreatectomy, avoiding major perioperative complications, and being eligible for timely adjuvant therapy initiation.

In parallel, strategies to deliver these systemic regimens, with or without radiation, in the neoadjuvant setting are increasingly popular. The most compelling rationale for this approach is the growing understanding that PDAC, even when appearing localized on best-available imaging, is a systemic disease at presentation.

Additional benefits of the neoadjuvant approach include the ability to deliver cytotoxic therapies when patients are less debilitated (vs. following major pancreatectomy), assess in vivo chemoresponsiveness, potentially increase the ability to achieve a margin-negative resection and, importantly, avoid nontherapeutic pancreatectomy in patients who rapidly progress on neoadjuvant treatments.

While the neoadjuvant strategy is standard for patients who present with borderline resectable or locally advanced disease, many centers — including ours — also utilize neoadjuvant treatment sequencing in patients with resectable anatomy, despite the paucity of level 1 data supporting the latter approach. Potentially supportive data include subgroup analyses from phase 2/phase 3 studies with conflicting results (eg, Prep-02/JSAP05, PACT-15, NEONAX, PANACHE01-PRODIGE48), extrapolation of benefit for this strategy from several trials in borderline resectable disease (PREOPANC, ESPAC-5, NCT01458717), and retrospective analyses from national registries. In fact, a prespecified subgroup analysis of resectable patients in the largest of these trials, PREOPANC, failed to demonstrate survival benefit. However, proponents of the neoadjuvant approach in resectable PDAC cite the lack of equipoise exhibited by excluding a surgery-first arm in the Southwest Oncology Group S1505 phase 2 trial. Given the pressing need to answer this question, there are two ongoing large-scale phase 3 studies currently randomly assigning patients with resectable PDAC to neoadjuvant chemotherapy versus upfront surgery — the PREOPANC-2 and ALLIANCE-A021806 trials.

Results of the phase 2 NORPACT-1 trial, presented at ASCO, appear to be a victory for the surgery-first proponents, and social media erupted with calls for a moratorium on neoadjuvant chemotherapy for resectable PDAC. But the devil is always in the details, and there are ample reasons for skepticism of these results until the full manuscript is presented and long-term data mature.

First, only half of patients randomly assigned to the neoadjuvant arm actually received their full complement of 4 cycles—we and others have shown that upwards of 7 to 8 cycles in the neoadjuvant setting may be the threshold needed to realize the true benefit of upfront chemotherapy. Second, dose reductions were observed in over two-thirds of patients in the neoadjuvant arm — not typical of contemporary practice at experienced centers routinely delivering these therapies in the neoadjuvant setting. These data are particularly striking because dose density of neoadjuvant therapy is an important contributor to oncologic outcomes.

Third, the bizarre association between higher rates of margin-/node-negative resection with worse survival in the neoadjuvant arm of this trial is difficult to reconcile with multiple lines of prior evidence. Fourth, lack of long-term data and knowledge of post-protocol treatment and/or salvage patterns makes interpretation of these short-term data challenging, if not impossible.

This study poses more questions than answers, and centers favoring neoadjuvant therapy in resectable PDAC should feel sufficiently uneasy about reflexively changing their practice based on these data. These data, however, are reflective of the inescapable fact that “one size” cannot and must not fit all in this disease. Ongoing efforts to develop biomarkers of response to systemic therapies, immunologically targeted approaches (e.g., neoantigen mRNA vaccines, KRAS vaccines, etc.), and fundamental knowledge of the hostile pancreatic tumor microenvironment that underpins these heterogeneous outcomes are paramount to make nuanced clinical decisions in the future.