Presentation (CTI)

OBJECTIVES: Emerging extrapolation techniques using time-varying network meta-analyses (NMAs) can more precisely compare survival improvements of novel therapies against legacy treatments, especially in immune-oncology. Time-varying NMA can mitigate statistical limitations when the proportional hazards (PH) assumption for the treatment survival curves is violated. However, these advanced methods remain underutilized in North American regulatory and reimbursement environments compared with UK and EU practice, despite their potential to maximally capture long-term benefits of a new therapy. We investigated a recent multiple technology appraisal (ID3760) from the UK’s National Institute for Health and Care Excellence (NICE) in advanced renal cell carcinoma, where PH violations for progression-free and overall survival (PFS/OS) curves were identified. The appraisers chose not to apply time-varying NMA results in the cost-effectiveness analysis (CEA), stating difficulty in interpretating their unintuitive results. We aimed to demonstrate how different time-varying NMA methods can impact the CEA, represented by associated incremental cost-effectiveness ratios (ICERs).
METHODS: We enhanced NICE’s decision problem, creating different CEA scenarios by only changing the type of NMA model estimating PFS/OS hazard ratios. Available updated clinical trial data were applied. Time-varying NMA included multivariate parametric, fractional polynomial, restricted cubic spline, and piecewise NMA. Their resultant ICERs were compared with using Cox PH (time-constant) NMA. For each model type, we shared key technical and clinical assumptions, advantages, and limitations.
RESULTS: We demonstrated that cost-effectiveness results are sensitive to NMA model type, underscoring the importance of appropriately capturing changes in comparative effectiveness over time.
CONCLUSIONS: With emerging therapies requiring more nuanced survival estimation, the anticipated uptake of time-varying NMA methods is expected to strongly influence the comparative effectiveness results informing North America regulatory and reimbursement agencies. Increased education including transparency of these advanced methods can help reduce uncertainties for appraisers and improve decision-making, ultimately impacting patient access.