Cost-Effectiveness Analysis of an AI Biomarker

A cost-effectiveness analysis published by investigators at UCLA suggest that ArteraAI-guided androgen deprivation therapy (ADT) decision-making in intermediate-risk prostate cancer lowered costs compared with other strategies. In this segment, Amar U. Kishan, MD, describes the study, which compared 3 strategies: ADT for all intermediate-risk patients, National Comprehensive Cancer Network (NCCN) risk group-guided ADT, and ArteraAI-guided ADT. The investigators used Markov modeling, generating an incremental cost-effectiveness ratio (ICER) expressed as cost per quality-adjusted life year (QALY). The commonly cited willingness-to-pay threshold of $100,000 per QALY served as the benchmark for cost-effectiveness.

The results established an economic hierarchy among the 3 strategies. The NCCN strategy dominated the ADT-for-all strategy, and the ArteraAI-guided strategy then dominated the NCCN strategy. In cost-effectiveness methodology, domination reflects the condition of being more effective and less costly. Kishan emphasizes that achieving domination at both steps of the comparison reflects the compounding value of progressively more precise patient selection.

The robustness of these findings across sensitivity analyses is a key feature of the analysis. Results were consistent regardless of the proportion of patients classified as ArteraAI-negative, the ADT regimen modeled (including 4 to 6 months of leuprolide or relugolix, and probabilistic stress-testing of the full model. Most notably, the model was insensitive to the price of the ArteraAI test itself under both Medicare and commercial insurance reimbursement scenarios. Kishan explains that this finding reveals that the economic driver of the ArteraAI strategy is not the test but the downstream savings from avoiding ADT administration.