FOLFIRINOX Plus Imaging-Based Biomarkers Achieve Resection Rates in Pancreatic Cancer Similar to Standard of Care

December 30, 2019
Darcy Lewis

Results from a phase II prospective trial indicate that fluorouracil-, leucovorin-, irinotecan-, and oxaliplatin-based treatment and imaging-based biomarkers for borderline resectable pancreatic ductal adenocarcinoma achieved a resection rate of 45% and a median overall survival of 2 years.<br /> &nbsp;

Results from a phase II prospective trial indicate that fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX)-based treatment and imaging-based biomarkers for borderline resectable pancreatic ductal adenocarcinoma (PDAC) achieved a resection rate of 45% and a median overall survival (OS) of 2 years.1

The trial, which was published in the online-onlyJCO Precision Oncology, was powered to differentiate a resection rate of 60% from 40%, so the resection rate of 45% was lower than expected.

For patients who received pancreatectomy, the median OS was 42 months (95% Confidence Interval [CI], 17.7 months to not estimable). Among patients who were not able to undergo pancreatectomy, the median OS was 14 months (95% CI, 9.0 to 24.8 months).

“Our data suggest that modified FOLFIRINOX followed by chemoradiation has a resection rate similar to historical controls for patients with borderline resectable PDAC,” wrote the authors, led by Eugene J. Koay, MD PhD, of the University of Texas MD Anderson Cancer Center in Houston. “Our data indicate that the novel imaging-based delta classification and interface response of PDAC warrant additional investigation as predictive biomarkers for surgical benefit.”

This trial builds on previous work that found that, following a baseline standard-of-care pancreas protocol CT, so-called high-delta PDAC tumors show an abrupt change between the tumor and normal pancreas tissue, whereas low-delta tumors do not. Following preoperative therapy, tumors that show a type I response are clearly defined at the interface of tumor and parenchyma, while type II responses become less defined. “These CT-based biomarkers associate with pathologic features of PDAC, such as the extent of stromal reaction and pathologic response to therapy,” the authors wrote.2,3

Thirty-three patients began treatment in the current trial. Inclusion criteria included newly diagnosed, histology-confirmed PDAC with at least one high-risk feature, ECOG performance status of 0 or 1, and adequate hematologic status.

Patients received modified FOLFIRINOX, including oxaliplatin (75 mg/m2) delivered in 2-hour infusions, followed by 90-minute infusions of irinotecan (150 mg/m2), and a continuous infusion of fluorouracil (2,000 mg/m2) over 46 hours. Treatment was given every 2 weeks for a total of 6 doses.

Within 4 to 6 weeks after completing the modified FOLFIRINOX regimen, patients were restaged via imaging. Qualifying patients then received radiation therapy at a total dose of 50.4 Gy in 28 fractions and gemcitabine (350 mg/m2) once weekly for five doses. After restaging with CT or MRI 4 to 6 weeks after the last dose of gemcitabine, patients whose disease had not progressed received surgery.

Of the 33 patients who began the modified FOLFIRINOX regimen, 6 did not complete it. Four of these experienced excessive toxicity or a change in performance status. Of those who completed FOLFIRINOX treatment, 2 patients each experienced local progression and progression of disease with metastasis.

Of the 23 patients who began the radiation/gemcitabine regimen, 5 patients experienced progression of disease with metastasis. One patient each developed local metastasis, comorbidities or local tumor factors that prevented resection. That left 15 patients who were eligible and completed surgery.

In the total trial population, 10 patients were found to have low-delta disease, while 23 had high-delta disease. High-delta tumors were associated with a lower rate of resection. Seventeen patients had a type I response with clear interface after FOLFIRINOX, while 16 patients showed a type II response with a less visually clear interface. Koay et al found that patients who exhibited a type II interface response were less likely to undergo pancreatectomy compared with those who exhibited a type I response.

At the most recent follow-up, nearly three quarters of the patients had died of their disease (n = 24, 73%). Of those who died, 16 patients were among the 18 patients (89%) who were not able to undergo surgery on protocol. Eight of the 15 patients (53%) who had undergone surgery had died. Two thirds of the surgical patients had a distant recurrence at last follow-up (n = 10, 67%).

In addition to the median OS of 24 months (95% CI, 16.2 to 29.6 months), Koay et al also recorded median OS duration of 42.1 (95% CI, 17.7 to not estimable) months in the surgical group and 14 (95% CI, 9.0 to 24.8) months in the group with unresectable disease.

In patients with high-delta tumors, the median OS was 17 months (95% CI, 12.0 to 25.0 months), while patients with low-delta tumors had not reached the median OS as of data cutoff (95% CI, 9.3 months to not estimable. “Patients with high-delta tumors had significantly lower 3-year PFS ([progression-free survival] 4% v 40%) and 3-year OS rates (20% v 60%) than those with low-delta tumors (bothP< .05),” Koay et al wrote.

In patients with type I interface response, the median OS was 30 months (95% CI, 18 months to not evaluable). This compared with a median OS of 14 months for patients with type II response (95% CI, 10 to 25 months). According to Koay et al, patients with a type II interface response had significantly lower 3-year PFS (0% v 29%) and 3-year OS rates (16% v 47%) than those with type I response (bothP< .001).

Koay et al believe that their baseline imaging-based delta classification can identifya priorithe patients at high risk for metastatic progression during preoperative therapy. Following therapy, the interface response may further refine prognostic groups. “A combined approach of baseline and post-treatment imaging-based biomarkers using an early interim look may help inform preoperative and adaptive treatment recommendations for medical and surgical therapies, given both the demonstrated prognostic associations of the imaging-based biomarkers and the association of margin status with interface response,” they wrote.

“[Additionally,] treatment intensification or selection for new combination systemic agents may be most appropriately considered in patients with high-delta tumors or type II responses. Our results argue for a more selective, personalized approach to preoperative therapy for patients with borderline resectable disease, driven by biomarkers.”


  1. Koay EJ, Katz MHG, Wang H, et al. Computed Tomography—Based Biomarker Outcomes in a Prospective Trial of Preoperative FOLFIRINOX and Chemoradiation for Borderline Resectable Pancreatic Cancer.JCO Precis Oncol. Published online August 23, 2019: DOI
  2. Koay EJ, Truty MJ, Cristini V, et al. Transport properties of pancreatic cancer describe gemcitabine delivery and response.J Clin Invest. 2014;124:1525-1536.
  3. Amer AM, Zaid M, Chaudhury B, et al. Imaging-based biomarkers: Changes in the tumor interface of pancreatic ductal adenocarcinoma on computed tomography scans indicate response to cytotoxic therapy.Cancer. 2018;124:1701-1709.