Disparities in Epidermal Growth Factor Receptor Mutation Testing for Non-Small Cell Lung Cancer: Insights From the National Cancer Database.

INTRODUCTION: Epidermal growth factor receptor (EGFR) mutation testing is central to metastatic and adjuvant non-small cell lung cancer (NSCLC) management. Although National Comprehensive Cancer Network guidelines have recommended EGFR testing since 2011 and the Food and Drug Administration approved adjuvant osimertinib in December 2020, real-world implementation across stages and practice settings remains poorly characterized. We used the National Cancer Database (NCDB) to quantify testing rates and assess disparities at patient and facility levels in the first full years following guideline expansion to resectable disease. METHODS: We analyzed the 2022 NCDB, restricting to adults (≥18 y) with NSCLC diagnosed in 2021-2022. EGFR testing status was categorized as "tested," "not tested," or "missing." Bivariate associations were assessed using chi-squared tests, and multilevel logistic regression with facility-level random effects was performed to identify independent predictors of EGFR mutation testing. To assess whether patient-level disparities were independent of facility case mix, a fixed effects logistic regression model incorporating facility indicator variables was conducted as a sensitivity analysis. Stratified multivariable models were additionally run separately for early-stage, locally advanced, and metastatic disease. RESULTS: Of the 144,017 eligible patients, 36.9% had EGFR testing reported to the NCDB. Testing rates varied by race/ethnicity, increased with median household income, and were higher at academic/research programs (P < 0.001). Testing rates by facility type ranged from 33.4% at community cancer programs to 39.2% at academic/research centers. In adjusted analyses, Asian/Pacific Islander (adjusted odds ratio [aOR] 1.60, 95% confidence interval [CI] 1.28-2.00), Hispanic (aOR 1.20, 95% CI 1.06-1.36), female sex (aOR 1.04, 95% CI 1.02-1.07), private insurance (aOR 1.21, 95% CI 1.07-1.36), and highest income quartile (aOR 1.22, 95% CI 1.11-1.34) were associated with greater odds of testing. Lower odds were observed in patients with a comorbidity score ≥3 (aOR 0.92, 95% CI 0.85-1.00) and those living in the most rural areas (aOR 0.71, 95% CI 0.61-0.84). Facility-level variation remained significant (P < 0.001), with an intraclass correlation coefficient of 0.94 indicating that the majority of unexplained variation in testing resides at the facility level. In the fixed effects model, insurance status and Asian/Pacific Islander race remained robust patient-level predictors, whereas the income effect attenuated toward the null. Stage-stratified analyses revealed that geographic barriers predominated in early-stage disease, socioeconomic factors drove disparities in locally advanced disease, and institutional factors were most consequential in metastatic disease. CONCLUSIONS: Despite guideline recommendations, the majority of NSCLC patients did not have EGFR testing reported to the NCDB in 2021-2022, and significant disparities exist by race/ethnicity, insurance status, socioeconomic status, geography, comorbidity burden, and facility type. Facility-level factors account for the majority of variation in testing, and the determinants of disparities differ meaningfully by disease stage. Standardized reflex testing, provider education, payer incentives, and investment in diagnostic infrastructure are needed to address these inequities and fully realize the promise of precision oncology.