Cost-effectiveness analysis of molecular reflex testing-based targeted therapy in non-small cell lung cancer from a German healthcare payer perspective using a decision tree model

BACKGROUND: Reflex testing (RT) - pathologist-initiated molecular testing performed at non-small cell lung cancer (NSCLC) diagnosis - enables rapid identification of actionable genomic alterations for targeted therapy (TT) selection. Despite national and international guideline recommendations, approximately 15% of patients with advanced NSCLC in Germany remain untested for driver mutations. The portion of untested patients in early stages is likely to be even higher. OBJECTIVE: This study simulated the clinical and economic impact of introducing RT for untested NSCLC patients under different implementation scenarios, using a decision-tree model from the German statutory health insurance (SHI) perspective. METHODS: A decision-tree model simulated four RT cases (C): (C1) no molecular testing, (C2) testing per German S3 guideline recommendations, (C3) testing according to current TT approvals, and (C4) comprehensive panel testing (standard in Austria). PD-L1 expression was tested in all cases. The untested NSCLC population was projected for 2025 based on SHI demographics. Model inputs comprised RT sensitivity, mutation and stage distribution, drug acquisition costs (Lauer-Taxe, April 2025), and median overall/progression-free survival (mOS/mPFS) per therapeutic strategy. A scenario analysis was conducted assuming NGS test costs of €3,000 per panel to assess their potential influence on cost-effectiveness versus no testing. RESULTS: All RT cases (C2-C4) substantially increased actionable mutation detection, enabling earlier TT initiation and lower mean treatment costs versus no molecular testing. C2-C4 were clinically dominant, driven by increased mOS by 5.48-5.47 months and mPFS by 4.89-5.42 months. Model yielded annual cost savings versus no testing of €740 (C2) and €535 (C3/C4) per patient (annual cost savings of €1,175 - €1,621 per life-year gained). When €3,000 next-generation sequencing (NGS) panel costs were included, annual costs per patient increased modestly (C2: €2,260; C3/C4: €2,465 versus no testing), at annual incremental costs of €4,946 - €5,409 per life-year gained. CONCLUSIONS: Reflex testing across implementation cases, by enabling earlier identification of actionable biomarkers and receipt of matched targeted therapy, improves clinical outcomes and is cost-effective for untested NSCLC patients from the German SHI perspective. Structured initiatives, such as national Network for Genomic Medicine (nNGM), could further expand equitable access to precision oncology in Germany.