The Value of Targeted Therapies in Lung Cancer

Abstract

The goal of this dissertation was to examine the realized value of targeted therapies in routine care and to identify opportunities for improving the return on medical spending for these technologies. Chapter 1 investigated the value of targeted therapies in lung cancer patients who were treated in routine care. This observational, claims-based analysis used propensity score, and instrumental variable methods, combined with a Kaplan Meier Sample Average estimator to calculate lifetime costs and life expectancy. An incremental comparison showed that the realized value of targeted therapies in routine care was unfavorable relative to chemotherapy treatment. Subgroup analyses revealed that initial erlotinib therapy yielded effectiveness results that are substantially lower than efficacy survival outcomes in molecularly guided trials. Our results indicated that in routine care, chemotherapy was the most cost effective strategy. The unexpectedly low outcomes with first-line erlotinib suggested that some of the value of this treatment was not being realized in practice. Chapter 2 examined the practice patterns of targeted therapies and utilization of predictive biomarker testing in routine care to better understand the observed gaps between trial-based and `real-world' outcomes with these agents. In our nationally representative cohort of lung cancer patients, we found that the vast majority of patients did not undergo molecular testing to inform first-line therapy. Our prediction models for biomarker screening and first-line treatment suggested that phenotypic enrichment criteria guided selection for testing and initiation of erlotinib therapy. Since clinical characteristics do not adequately discriminate between mutation positive and wild type tumors, these practices signal the need for wider dissemination of biomarker screening to accurately target patients towards improving therapeutic gains with erlotinib. Chapter 3 assessed the cost-effectiveness of multiplexed predictive biomarker screening to inform treatment decisions in lung cancer patients. Using a micro-simulation model to evaluate the incremental value of molecularly guided therapy compared to chemotherapy in unselected patients, we found that personalized therapy is a cost effective strategy. Our results indicated that better value of targeted therapies in lung cancer is achievable through molecularly guided treatment.