Accounting for Cured Patients in Cost-Effectiveness Analysis.

Area of Science:

• Oncology
• Health Economics
• Biostatistics

Background:

• Economic evaluations frequently use mean overall survival (OS) to assess cancer interventions.
• Emerging cancer therapies can lead to long-term survival, potentially
• curing
• patients, a scenario not adequately captured by a single mean OS value.

Purpose of the Study:

• To demonstrate the incorporation of patient heterogeneity, specifically the
• cured
• proportion, into health economic evaluations.
• To provide a more accurate assessment of cancer interventions where long-term survival is possible.

Main Methods:

• Analysis of clinical trial data for advanced melanoma patients treated with ipilimumab (Ipi) versus glycoprotein 100 (gp100).
• Application of statistical mixture cure models to differentiate between cured and non-cured patient cohorts.
• Consideration of background mortality unrelated to cancer for both patient groups.

Main Results:

• Ignoring cure proportions overestimated ipilimumab's OS benefit.
• Mixture cure modeling revealed a higher cured proportion (21%) with ipilimumab compared to gp100 (6%).
• The incremental cost-effectiveness ratio (ICER) for ipilimumab versus gp100 decreased from $324,000/QALY to $113,000/QALY when using cure models.

Conclusions:

• Cure modeling is essential for health economic evaluations in advanced melanoma.
• Accurate estimation of OS and economic measures like QALYs and ICERs is improved by accounting for a cured patient proportion.
• This approach mitigates bias in assessing therapies with potential for long-term patient survival.