Impact of Recurrence Modelling on Cost-Effectiveness of New Treatments for Early Cancers
Author(s)
Asad Faraz, BSc1, Yunni Yi, PhD2, Grace Mountain, MSc2, Alex Hirst, BSc, MSc2, Louise Heron, BA, BSc, MSc2.
1University of Sheffield, Sheffield, United Kingdom, 2Adelphi Values PROVE, Bollington, United Kingdom.
1University of Sheffield, Sheffield, United Kingdom, 2Adelphi Values PROVE, Bollington, United Kingdom.
Presentation Documents
OBJECTIVES: Cancer recurrence significantly impacts patient outcomes and healthcare costs, particularly in early-stage cancers where recurrence free survival (RFS) data are often immature, posing challenges in evaluating the cost-effectiveness of novel treatments which may prevent or delay recurrence. This study addressed these challenges using a cost-effectiveness analysis of alectinib as an adjuvant therapy for resected ALK-positive non-small-cell lung cancer as an example.
METHODS: A three-state partitioned survival model (recurrence free, progressive disease, and death) was developed using digitized trial data from ALINA for RFS and overall survival (OS) data from ANITA for chemotherapy. A hazard ratio (HR) relative to the chemotherapy OS was assumed for alectinib. Parametric survival models were used to extrapolate RFS and OS to lifetime. Cost-effectiveness was assessed from United States payer perspective. Scenario analyses explored the impact of time horizons, alternative survival models, and exclusion of subsequent treatments for recurrence. Deterministic and probabilistic sensitivity analyses were performed to assess the robustness of the results.
RESULTS: Alternative RFS distributions versus Weibull in the base-case had a substantial impact on ICERs, with KMs increasing the ICER by 234.63%, Gomperz by 40.40% and log-normal decreasing it by 54.90%. ICERs increased greatly with shorter time horizons of 1, 2, 5 or 10 years; excluding subsequent treatment costs increased the ICER by 12.35%. While ±20% changing in HR for OS changed ICER by around ±20%, alternative OS distributions had relatively small impact on ICERs, with KM decreasing it by 20.58% and exponential increasing it by 3.41%.
CONCLUSIONS: Accounting for recurrence and related impacts over long-term is important for evaluating early cancer treatments. Cost-effectiveness results are significantly influenced by recurrence survival modeling, time horizons and subsequent treatment costs. Future economic evaluations of early cancer treatments should address these uncertainties with more mature survival data and consider all relevant costs over patient’s lifetime.
METHODS: A three-state partitioned survival model (recurrence free, progressive disease, and death) was developed using digitized trial data from ALINA for RFS and overall survival (OS) data from ANITA for chemotherapy. A hazard ratio (HR) relative to the chemotherapy OS was assumed for alectinib. Parametric survival models were used to extrapolate RFS and OS to lifetime. Cost-effectiveness was assessed from United States payer perspective. Scenario analyses explored the impact of time horizons, alternative survival models, and exclusion of subsequent treatments for recurrence. Deterministic and probabilistic sensitivity analyses were performed to assess the robustness of the results.
RESULTS: Alternative RFS distributions versus Weibull in the base-case had a substantial impact on ICERs, with KMs increasing the ICER by 234.63%, Gomperz by 40.40% and log-normal decreasing it by 54.90%. ICERs increased greatly with shorter time horizons of 1, 2, 5 or 10 years; excluding subsequent treatment costs increased the ICER by 12.35%. While ±20% changing in HR for OS changed ICER by around ±20%, alternative OS distributions had relatively small impact on ICERs, with KM decreasing it by 20.58% and exponential increasing it by 3.41%.
CONCLUSIONS: Accounting for recurrence and related impacts over long-term is important for evaluating early cancer treatments. Cost-effectiveness results are significantly influenced by recurrence survival modeling, time horizons and subsequent treatment costs. Future economic evaluations of early cancer treatments should address these uncertainties with more mature survival data and consider all relevant costs over patient’s lifetime.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
EE526
Topic
Economic Evaluation
Disease
SDC: Oncology