Rethinking resectability: the case for neoadjuvant treatment in resectable pancreatic cancer

We read with interest the paper by Schwarz et al. entitled “Neoadjuvant FOLF(IRIN)OX Chemotherapy for Resectable Pancreatic Adenocarcinoma: A Multicenter Randomized Noncomparative Phase II Trial (PANACHE01 FRENCH08 PRODIGE48 study)” (1). This was a multicentre randomised, non-comparative, phase II trial where 153 patients with upfront resectable pancreatic adenocarcinoma (rPDAC) were randomised in a (2:2:1) fashion to neoadjuvant mFOLFIRINOX (leucovorin, fluorouracil, irinotecan and oxaliplatin), neoadjuvant FOLFOX (leucovorin, fluorouracil and oxaliplatin) or conventional up-front surgery followed by adjuvant chemotherapy.

There is growing interest in the role of neoadjuvant chemotherapy in the management of rPDAC, as well as borderline resectable and locally advanced pancreatic cancer. Empirically, neoadjuvant chemotherapy offers several advantages over up front surgery, including improved treatment completion rates—given that approximately one-third of patients do not receive or complete adjuvant chemotherapy postoperatively for a variety of reasons (2-4). Furthermore, one side of the debate is that delivering systemic therapy upfront allows for early assessment of tumour biology and behaviour, potentially sparing patients from non-beneficial surgery in biologically aggressive tumours.

The primary objectives of the PANACHE01 study were to assess the feasibility, safety, and efficacy of neoadjuvant treatment in rPDAC. The trial was conducted from February 2017 to July 2020, with a median follow-up of 37.7 months. Other neoadjuvant chemotherapy trials have utilised longer regimens and indeed our departmental protocol for borderline and locally advanced PDAC is to deliver all systemic anti-cancer treatment in a total neo-adjuvant approach prior to surgery if feasible (5).

In this study, despite a relatively short course of four cycles of neoadjuvant chemotherapy, the irinotecan dose required reduction during the trial, and Grade 3 and 4 toxicities were reported in 47% and 37.8% of patients, respectively. Nevertheless, neoadjuvant chemotherapy was completed in 88.6% of patients, with no significant dose reduction required in 59.4%—a notable improvement compared to the NORPACT-1 trial (6), which used classic FOLFIRINOX rather than the modified regimen (mFOLFIRINOX). These findings raise the question of whether extending neoadjuvant mFOLFIRINOX to six or more cycles could further improve overall survival (OS) in rPDAC, as suggested by retrospective studies indicating potential survival benefits. Interestingly, 7% of patients in this trial were found to have venous involvement, although the extent of involvement was not detailed. In many centres, such findings would prompt the use of more prolonged neoadjuvant regimens, and possibly the addition of neoadjuvant chemoradiation, prior to surgical resection.

Initially, the study aimed to recruit 64 patients in each chemotherapy arm and 32 in the upfront surgery arm, with an interim analysis planned after 27 evaluable patients. During the recruitment phase, 30 patients were excluded due to screening failure or not undergoing randomisation. As a result, the intention-to-treat (ITT) cohort consisted of 142 patients, and the per-protocol cohort included 132 patients. Following interim analysis, Arm 2 (FOLFOX) was discontinued due to the death of eight patients and failure to complete two or more chemotherapy cycles; however, the specific causes of these deaths were not discussed in detail.

In this study, 82% of patients had rPDAC located in the head of the pancreas, and 41% had baseline carbohydrate antigen 19-9 (CA19-9) levels exceeding 200 U/mL. Elevated CA19-9 has been associated with an increased risk of inferior outcomes and locoregional recurrence (7), and it would be valuable to assess CA19-9 trends following neoadjuvant treatment in future analyses. Of the 61 patients who received mFOLFIRINOX, 51 proceeded to surgical resection, achieving an R0 resection rate of 74.5%, which is consistent with rates reported in the literature. One-year OS in this group was 84.3%, with a median OS of 31.3 months and a median event-free survival (EFS) of 12.4 months. These outcomes compare favourably with those in the upfront surgery group, where 1-year OS was 82.1% and median EFS was only 6.8 months. The mFOLFIRINOX arm also demonstrated higher R0 resection rates and lower pathological lymph node involvement than the upfront surgery group. Surgery was associated with 20.4% rate of severe postoperative complications. However, this did not appear to adversely affect the delivery of adjuvant chemotherapy, which was completed by 63% and 60% of patients in Arms 1 and 2, respectively.

The recently published Prep-02/JSAP05 trial from Japan demonstrated neoadjuvant chemotherapy with gemcitabine plus S-1 followed by surgical resection conferred a significant survival advantage of median OS 37.0 vs. 26.6 months compared to upfront surgery in patients with resectable PDAC (8). The CASSANDRA trial also reported PAXG significantly improved EFS compared to mFOLFIRINOX (9) but this cohort included both resectable and borderline resectable PDAC. An up-to-date systemic review published this year which included 8 randomized controlled trials (RCTs) reported rPDAC patients treated with neoadjuvant therapy (NAT) conferred better disease-free survival (DFS) and high R0 and pN0 rates. However, this interestingly did not translate to any significant OS benefit [hazard ratio (HR) 0.81, P=0.06] (10).

In line the above, this study supports the use of neoadjuvant chemotherapy in rPDAC, demonstrating improved survival outcomes and reinforcing the oncological benefit of a neoadjuvant approach. However, the question of the optimal duration of neoadjuvant treatment and precise regimen are yet to be fully determined. We eagerly await the results of ongoing trials such as those quoted in the paper, including PREOPANCE-3 and ALLIANCE A0211806. Furthermore, with follow-up stopping at 3 years the long-term outcomes are not known especially as there may well be a survival advantage beyond this timepoint. Longer term survival benefit in patients receiving neoadjuvant treatment is absolutely essential in future studies. The individualised approach to rPDAC in terms of transcriptomics will be partly sought from data from ESPAC-6.

A final key point to discuss is that the current classification of pancreatic cancer based solely on radiological appearances (11) is perhaps outdated. The definition of resectable, borderline resectable and locally advanced disease should incorporate biologic and conditional factors as proposed by the Trans-Atlantic Pancreatic Surgery Consortium (12). Accurate staging of the disease is paramount to the appropriate selection of treatment modality and sequences for the best chance of a good outcome. Further technologies such as detection of circulating DNA using next generation sequencing has shown promise as an additional biomarker of treatment and to aid assessment of response following neoadjuvant treatment (13).

In summary, pancreatic cancer is a heterogenous disease, and thus there is no one size fits all treatment. The reported study demonstrates that mFOLFIRINOX is a safe regimen in the neoadjuvant setting for rPDAC and at minimum, gives equivalent patient outcomes to upfront surgery. There is no doubt that some patients with up front resectable pancreatic cancer could benefit from neoadjuvant treatment. However, the optimal treatment strategies, agents and duration are yet to be determined.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.