Cost-Benefit Implication of Gene Expression Profiling and Adjuvant Therapy in Stage IIIA Melanoma

Abstract

Indiscriminate use of adjuvant therapy in stage IIIA melanoma is controversial. We sought to model the clinical impact and cost of implementing a gene expression profile (GEP) test to guide adjuvant therapy. A Markov decision-analysis model was created to represent resected stage IIIA melanoma with 3 treatment options: observation (OBS), adjuvant pembrolizumab for all patients (ALL), and selective adjuvant therapy (SEL). In the SEL option, only high-risk patients based on GEP stratification were treated with pembrolizumab. Cost of adjuvant therapy was normalized to reflect Medicare reimbursement schedules. The primary outcome was cost per mortality avoided at 10 years. Model projections for 10-year overall survival were 68% for OBS, 73% for SEL, and 76% for ALL. The estimated incremental cost-per-mortality-avoided (compared to OBS) was $2.1 million for SEL and $2.4 million for ALL. These translate to costs of $583.0K and $697.1K per life-year for the SEL and ALL strategies, respectively. Routine adjuvant pembrolizumab for stage IIIA melanoma is costly, and risk-stratification by GEP only marginally improves the value of therapy.