Lifetime cost-effectiveness simulation of once-weekly exenatide in type 2 diabetes: A cost-utility analysis based on the EXSCEL trial

Aims: The Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial assessed once-weekly exenatide (EQW) vs. placebo, added to usual care in 14,752 patients with type 2 diabetes mellitus (Clinicaltrials.gov: NCT01144338). We assessed the lifetime cost-effectiveness of adding EQW vs. usual care alone from a healthcare perspective. Methods: Medical resource use and EQ-5D utilities were collected throughout the study. Within-trial results were extrapolated to a lifetime horizon using the UK Prospective Diabetes Study Outcomes Model version 2 (UKPDS-OM2), predicting predict cardiovascular and microvascular events. Cost-effectiveness was evaluated separately for US and UK set-tings, with outcomes measured in quality-adjusted life-years (QALYs). Results: EQW plus usual care gained 0.162 QALYs at an additional cost of $41,545/patient, compared with usual care in a US setting. The incremental cost-effectiveness ratio (ICER) was $259,223/QALY. In a UK setting, the QALY gain was 0.151 at an additional cost of 6357 pound: an ICER of 42,589 pound/QALY. Sensitivity analyses ranged between $34,369-$269,571 and 3430- pound 46,560 pound per QALY gained. Conclusions: In a lifetime extrapolation, adding EQW to usual care increased QALYs and costs compared with usual care alone. The base-case ICERs exceeded the commonly-cited cost-effectiveness thresholds of $100,000/QALY and 20,000 pound/QALY. However, ICERs were considerably lower in some subgroups, and in sensitivity analyses. (c) 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Automated summary

The EXSCEL trial assessed the cost-effectiveness of adding once-weekly exenatide to usual care for patients with type 2 diabetes mellitus. The results showed a QALY gain but also an increased cost. ICERs exceeded commonly-cited thresholds in the US setting but were lower in some subgroups and sensitivity analyses.