192P - A multicenter crossover analysis of first and second-line FOLFIRINOX or gemcitabine plus nab-paclitaxel administered to pancreatic cancer patients: Results from the NAPOLEON study

Background

FOLFIRINOX (FFX) and gemcitabine plus nab-paclitaxel (GnP) are considered standard 1st-line chemotherapy (CTx) for patients with metastatic pancreatic cancer, and reports about the effects of FFX and GnP as 2nd-line CTx have been accumulating. However, compared data of two crossover sequences and those of the comparison of the crossover sequences with others are unknown.

Methods

We investigated the efficacies of the two crossover sequences using the data of the multicenter observational study conducted in patients with CTx-naive advanced or unresectable pancreatic cancer (AUPC) treated with FFX or GnP from 14 hospitals in Japan during the period from December 2013 to June 2018. Patient characteristics and clinical outcomes including overall survival (OS), time-to-first and -second progression (TTFP/TTSP) and overall response rate (ORR) were analyzed between the two crossover groups [CG] firstly. Then, the efficacies of them were compared with non-crossover groups [NCG].

Results

Of 318 AUPC patients, 118 and 200 patients received FFX and GnP as 1st-line CTx, respectively. Of these, 91 and 100 patients received 2nd-line CTx, of which 72 (79%) and 17 (17%) patients were shifted to GnP (F-to-G) and FFX (G-to-F), respectively (p<0.01). The F-to-G group comprised more patients with higher body mass index, biliary drainage, peritoneal metastasis, and maximal tumor size of 20 mm or larger at baseline. In the two crossover sequences, there was no significant difference in the median OS (11.5 vs. 16.4 months; hazard ratio [HR] 1.25; p=0.45) between the F-to-G and G-to-F groups, respectively. No significant differences in the ORR of 1st-line (p=0.12) and 2nd-line (p=0.74) were seen. The median TTFP and TTSP were also comparable between the two groups (5.7 vs. 5.0 months; HR 0.99; p=0.97, 8.8 vs. 12.2 months; HR 1.12; 95% CI 0.67-2.02; p=0.58), respectively. Then, the median OS in the CG (n=89) was not significantly longer than that in the NCG (n=99) (11.6 vs. 12.8 months; HR 1.24; p=0.19).

Conclusions

There were no significant efficacy differences between the two CGs. The OS of CG was not statistically superior to NCG in our study.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.