REGORAFENIB USE IN METASTATIC COLORECTAL CANCER AFTER FAILING ALL APPROVED LINES OF THERAPY: ADEQUATE QUALITY OF LIFE AND COST-EFFECTIVE?

Wednesday, October 22, 2014
Poster Board # PS4-18

Melissa Robins, MS1, Wen Wee Ma, MD1 and James G. Dolan, MD2, (1)Roswell Park Cancer Institute, Buffalo, NY, (2)University of Rochester, Rochester, NY
Purpose: To determine if treatment with Regorafenib is the best management strategy for metastatic colorectal cancer after failing all other approved therapies, based on quality of life and cost of treatment, versus supportive care alone.

Method: Three decision tree models were constructed using Precision Tree 6, to assess the objective of this analysis.  Probabilities for each model were abstracted from the phase III CORRECT trial.  Health utility was the outcome of interest for Model 1.  The health utilities were determined using the average of the EQ-5D baseline and end of treatment utilities reported in the CORRECT trial.  Adverse event utilities were derived from the Common Terminology Criteria for Adverse Events Version 4.0.  Quality adjusted life years (QALY’s) were the outcome of interest in Model 2, to take into account quality and quantity of life.  Overall survival and progression free survival reported in the CORRECT trial was used to calculate QALY’s.  The outcome of interest in Model 3 was total treatment cost of each branch of the decision tree.  A one-way sensitivity analysis was run to determine which uncertainties had an effect on each model.  Two-way sensitivity analysis was then run to determine the decision thresholds for each model.

Result: The strategy of choice for Model 1 was the comparator group of supportive care only.  The health utility was 0.621 in the supportive care group versus 0.507 in the regorafenib group. Model 2 showed regorafenib as the strategy of choice.  Total QALY’s for the regorafenib group was 0.31 versus 0.27 in the supportive care group.  Model 3 analyzed cost differences found between the regorafenib and supportive care group.  The clear strategy of choice was supportive care, showing the average cost per patient in the regorafenib group as $42,944, and $16,718 in the supportive care group, with an incremental cost difference of $26,226. 

Conclusion: Treatment with regorafenib depends on overall treatment objective. If modest increase in overall survival is weighted more heavily regorafenib is a viable treatment option.  However, if quality of life, less time spent at the treatment facility, and a reduction in cost is most valuable, the use of regorafenib should be reconsidered.