The Avastin in Glioblastoma trial has shown that patients newly diagnosed with glioblastoma multiforme (GBM) treated with bevacizumab plus radiotherapy and temozolomide versus radiotherapy and temozolomide alone showed improvement in progression-free survival, possibly leading to a new indication for first-line use of bevacizumab in GBM. The cost-utility of this new intervention remains unknown; therefore, we developed a Markov model estimating the incremental cost-utility ratio (ICUR) from a Canadian public payer perspective.
We incorporated trial data for state transitions and treatment effects from the Avastin in Glioblastoma trial, costs and resource use data from Canadian published studies and databases, and utility parameters from published literature. We addressed uncertainty through one-way deterministic and probabilistic sensitivity analyses, extended the model to lifetime horizon and by another arm to compare first-line versus second-line use of bevacizumab on progression, performed value of information analysis, and performed US costing sensitivity analysis.
Adding bevacizumab to radiotherapy and temozolomide resulted in increases of 0.13 quality-adjusted life-years (QALYs) and $80,000 per patient over 2-year time horizon at the base case analysis. The ICUR was $607,966/QALY (95% CI, $305,000/QALY to $2,550,000/QALY), with 0% chance of being cost effective at the $100,000/QALY willingness-to-pay threshold and never going below $450,000/QALY in the one-way sensitivity analysis. The ICUR using the US costing data was $787,519/QALY. The lifetime ICUR was $439,764/QALY (95% CI, $235,000/QALY to $1,520,000/QALY), never going below $350,000/QALY in the sensitivity analysis. Second-line use of bevacizumab on progression is more effective and less expensive than its first-line use. Value of information analysis revealed that future research is unwarranted.
Bevacizumab has only limited effectiveness and is therefore not likely to be cost effective in treating adult patients with newly diagnosed GBM.