A multicenter retrospective study of gemcitabine plus nab-paclitaxel or FOLFIRINOX in metastatic pancreatic cancer: NAPOLEON study

Abstract

Introduction: Both gemcitabine plus nab-paclitaxel (GnP) and FOLFIRINOX (FFX) have been established as standard first-line chemotherapy in metastatic pancreatic cancer (mPC) patients with good performance status. However, it is still unclear which regimen is preferable due to a lack of prospective studies comparing GnP with FFX. Therefore, we conducted a retrospective study to investigate the clinical features of the two regimens in the real world. Methods: This retrospective study collected data from chemotherapy naïve mPC patients treated with GnP or FFX as first-line chemotherapy from 14 hospitals in Kyushu, Japan during the period from December 2013 to June 2018. The FFX group included both original FFX and modified FFX (mFFX) that was allowed a dose reduction of irinotecan (150mg/m2) and eliminated bolus 5-FU from FFX. Patient characteristics, treatment efficacy, and adverse events were analyzed and compared between GnP group and FFX group. Overall survival (OS) was estimated by Kaplan-Meier analysis, and a Cox proportional hazard model was employed to compare OS in FFX with in GnP. Results: GnP or FFX was administered to 92 (43.6%) and 119 (56.4%) patients, respectively. The median follow-up period was 14.9 (range, 0.7-30.9) months in GnP group and 13.1 (1.3-40.8) months in the FFX group. The statistically significant differences between the two groups on baseline patient characteristics were age (68y in GnP group vs. 61y in FFX group), ECOG PS 0 (57% vs. 71%), family history with malignant tumors (23% vs. 39%), liver metastasis (66% vs. 82%), peritoneal metastasis (42% vs. 13%), and ascites (30% vs. 14%). The median OS was 10.2 months in the GnP group as compared with 11.1 months in the FFX group (HR 0.98; 95% CI, 0.72-1.34; P=0.91). The median PFS was 5.5 months in the GnP group as compared with 5.7 months in the FFX group (HR 1.05; 95% CI, 0.79-1.40; P=0.75). In the GnP and FFX groups, the overall response rates were 29% and 27% (P= .72). Grade 3/4 anorexia (6% vs. 20%), diarrhea (0% vs. 11%), and nausea (3% vs. 11%) were more frequent in FFX group, whereas grade 3/4 fatigue (4% vs. 0%) and peripheral sensory neuropathy (12% vs. 3%) were more frequent in GnP group. Grade 3/4 hematological adverse events and febrile neutropenia (10% in GnP group vs. 13% in FFX group) were comparable between the groups. Multivariate analyses of OS revealed that age≧65 years old, PS ≧1, primary site of pancreatic head, ascites LDH >240IU/L, and CRP >0.3mg/dL were significant indicators of poor prognosis and the regimen was not associated with OS. In the FFX group, no statistically significant difference was observed between original FFX and mFFX in efficacy, but significant differences in toxicities were observed, with grade 3/4 thrombocytopenia (18% vs. 2%) and biliary infection (0% vs. 10%). Conclusion: In the real world, patients treated with FFX had more favorable conditions than patients treated with GnP. There was no statistically significant difference in OS between the two groups without adjustment of patient backgrounds. Both regimens were well-tolerated as first-line chemotherapy in mPC, however, the toxicity profiles of the two regimens differed.