A randomised phase II study of modified FOLFIRINOX versus gemcitabine plus nab-paclitaxel for locally advanced pancreatic cancer (JCOG1407).

Aim: We compared the efficacy of modified 5-fluorouracil, leucovorin, irinotecan, and oxaliplatin (mFOLFIRINOX) with that of gemcitabine plus nab-paclitaxel (GnP) for locally advanced pancreatic cancer (LAPC).

Methods: Patients with untreated LAPC were randomly assigned (1:1) to receive mFOLFIRINOX or GnP. One-year overall survival (OS) was the primary endpoint. The major secondary end-points included progression-free survival (PFS), response rate (RR), carbohydrate antigen 19-9 (CA19-9) response, and adverse events. The sample size was 124 patients to select a more effective regimen with a minimum probability of 0.85 and to examine the null hypothesis of the 1-year OS <53%.

Results: Of the 126 patients enrolled from 29 institutions, 125 were deemed eligible. The 1-year OS was 77.4% (95% CI, 64.9-86.0) and 82.5% (95% CI, 70.7-89.9) in the mFOLFIRINOX and GnP arms, respectively. The median PFS was 11.2 (95% CI, 9.9-15.9) and 9.4 months (95% CI, 7.4-12.8) in the mFOLFIRINOX and GnP arms, respectively. The RR and CA19-9 response rate were 30.9% (95% CI, 19.1-44.8) and 57.1% (95% CI, 41.0-72.3) and 42.1% (95% CI 29.1-55.9) and 85.0% (95% CI, 70.2-94.3) in the mFOLFIRINOX and GnP arms, respectively. Grade 3-4 diarrhoea and anorexia were predominant in the mFOLFIRINOX arm.

Conclusion: GnP was considered the candidate for a subsequent phase III trial because of its better RR, CA19-9 response, and mild gastrointestinal toxicities. Both regimens displayed higher efficacy in the 1-year survival than in the historical data of gemcitabine monotherapy.