An Indirect Treatment Comparison and Cost-Effectiveness Analysis Comparing Folfirinox with Nab-Paclitaxel Plus Gemcitabine for First-Line Treatment for Patients with Metastatic Pancreatic Cancer

Abstract

Introduction: The management of metastatic pancreatic cancer (mPC) has recently changed to involve a greater use of combination chemotherapy such as FOLFIRINOX over gemcitabine monotherapy. The ACCORD study has demonstrated statistically significant improvements in overall survival (OS), progression‐free survival (PFS), objective response rate (ORR) and delayed time until quality of life deterioration compared to gemcitabine alone (Conroy et al 2011). Similarly, the MPACT trial showed combination nab‐Paclitaxel plus gemcitabine (NabP + G) compared to gemcitabine had a greater OS, PFS and ORR (Von Hoff 2013). Baseline patient characteristics and relative dose intensities in the FOLFIRINOX and NabP + G study groups were comparable. FOLFIRINOX significantly prolonged median OS by more than 4.3 months compared with gemcitabine (11.1 vs 6.8 months, Hazard ratio (HR) = 0.57; P < 0.001), while NabP + G provided 1.8 additional months over gemcitabine (8.5 vs 6.7 months, HR = 0.72; P < 0.001). Further follow‐up data presented at ASCO GI 2014 confirmed the HR and 95% CI for OS in the MPACT trial (Goldstein 2014). Clinical head‐to‐head data comparing FOLFIRINOX versus NabP + G is yet unavailable. Therefore, an indirect treatment comparison (ITC) was performed to facilitate the conduct of a Canadian cost‐effectiveness analysis (CEA) of these two treatments. Methods: Data from FOLFIRINOX and NabP + G phase III trials described above were used to inform the ITC. HRs were calculated for FOLFIRINOX versus NabP + G for OS and PFS (independent review). In addition, relative risks ratios (RRs) were calculated to compare the ORRs and the rates of adverse events (AEs) for FOLFIRINOX versus NabP + G. To conduct the CEA, a Markov model was used to estimate how patients progress through the following states: 'stable', 'progressed', 'dead'. OS and PFS data as well as RRs for ORRs and AEs were derived from the ITC results. Published utility data and Canadian cost data were applied based on time in each state and treatment related AEs. Costs included first and second‐line therapy, treatment administration, monitoring costs, and management costs of the AEs. Costs and outcomes were discounted at 5%. Sensitivity analyses were conducted to test the robustness of the model. Results: The ITC results comparing FOLFIRINOX with NabP + G for OS and PFS were HR 0.792 (95% CI, 0.597 to 1.05) and HR 0.681 (95% CI, 0.509 to 0.911) respectively. Most AEs were not statistically significantly different between treatments, however fatigue was significantly worse for NabP + G RR: 0.536 (95% CI, 0.294 to 0.978), and neutropenia was significantly worse for FOLFIRINOX RR: 1.535 (95% CI, 1.035 to 2.278). The results of the CEA indicated that FOLFIRINOX was cost effective when compared with NabP + G (ICER: $7,830/QALY). Due to longer treatment, and improved OS and PFS, there were slightly higher costs for FOLFIRINOX over a lifetime horizon when compared with NabP + G. The sensitivity analysis results demonstrated that FOLFIRINOX was associated with a low cost‐effectiveness ratio or was a dominant treatment strategy. Conclusion: The ITC results suggest that FOLFIRINOX had improved OS and PFS compared with NabP + G. Both had a comparable safety profile. Given the favorable costs per QALYs and increased efficacy, FOLFIRINOX remains an attractive cost‐effective first line mPC therapy. (Table Presented).