Background
As the treatment landscape in patients with non‐small cell lung cancer (NSCLC) harboring mutations in the epidermal growth factor receptor (EGFRm) continues to evolve, real‐world health utility scores (HUS) become increasingly important for economic analyses.
Methods
In an observational cohort study, questionnaires were completed in EGFRm NSCLC outpatients, to include demographics, EQ‐5D‐based HUS and patient‐reported toxicity and symptoms. Clinical and radiologic characteristics together with outcomes were extracted from chart review. The impact of health states, treatment type, toxicities, and clinical variables on HUS were evaluated.
Results
Between 2014 and 2018, a total of 260 patients completed 994 encounters. Across treatment groups, patients with disease progression had lower HUS compared to controlled disease (0.771 vs 0.803; P = .01). Patients predominantly received gefitinib as the first‐line EGFR tyrosine kinase inhibitor (TKI) (n = 157, mean‐HUS = 0.798), whereas osimertinib (n = 62, mean‐HUS = 0.806) and chemotherapy (n = 38, mean‐HUS = 0.721) were more likely used in subsequent treatment lines. In longitudinal analysis, TKIs retained high HUS (>0.78) compared to chemotherapy (HUS < 0.74). There were no differences between the frequency or severity of toxicity scores in patients receiving gefitinib compared to osimertinib; however, TKI therapy resulted in fewer toxicities than chemotherapy (P < .05), with the exception of worse diarrhea and skin rash (P < .001). Severity in toxicities inversely correlated with HUS (P < .001). Clinico‐demographic factors significantly affecting HUS included age, Eastern Cooperative Oncology Group Performance Score (ECOG PS), disease state, treatment group, and metastatic burden.
Conclusions
In a real‐world EGFRm population, patients treated with gefitinib or osimertinib had similar HUS and toxicities, scores which were superior to chemotherapy. Health utility scores inversely correlated with patient‐reported toxicity scores. In the era of targeted therapies, future economic analyses should incorporate real‐world HUS.