Cost-effectiveness of first-line osimertinib informed by electronic medical records via text-mining: a real-world Italian case study of EGFR-mutated advanced NSCLC patients

Background: Tyrosine kinase inhibitors (TKIs) have significantly changed the therapeutic landscape of non-small-cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) gene mutation. The primary objective is to assess the use of real-world data (RWD) extracted from electronic health records (EHRs) via a text-mining technique. Then, RWD are used to inform a cost-effectiveness model of first-line TKI third-generation osimertinib compared with standard EGFR inhibitors. Materials and methods: A cohort of 190 NSCLC EGFR-mutated patients has been enrolled from an Italian cancer institute. A rule-based named-entity recognition algorithm is carried out to extract RWD from medical records stored in the data warehouse. A Bayesian cost-effectiveness analysis is developed from the Italian health care system perspective. Survival curves are extrapolated over an observed follow-up horizon of a real-world population to estimate the time to treatment discontinuation (TTD) and overall survival (OS) distributions. Results: An incremental cost-effectiveness ratio (ICER) of 48398/quality-adjusted life year (QALY) is found for the base-case scenario, while 51779/QALY is found for a second scenario. While higher total annual costs are observed, effectiveness analysis confirms greater QALYs associated with osimertinib versus non-osimertinib. This result is also supported by significantly longer TTD (median 15 months) and OS (median 27 months) for osimertinib. Sensitivity and scenario analyses are used to validate the results. Conclusions: This study represents a pilot assessment for encouraging the adoption of text-mining approaches for RWD generation and informing health economic evaluations in oncology. RWD analysis confirms that osimertinib is cost-effective as a first-line treatment compared with previous generations of TKIs for advanced EGFR-mutated NSCLC patients.