Predictors of biomarker testing among patients (pts) with metastatic non-small cell lung cancer (mNSCLC).

9130 Background: The MYLUNG (Molecularly Informed Lung Cancer Treatment in a Community Cancer Network) Consortium pragmatic study assessed real-world biomarker testing rates for mNSCLC within The US Oncology Network of over 1,000 providers. We previously reported that 90% of pts had at least one test performed for ALK, BRAF, EGFR, ROS1, or PD-L1, but only 46% had testing for all 5. We examined patient factors associated with rates of biomarker testing. Methods: Retrospective observational chart review study of pts with mNSCLC initiating first-line systemic therapy between 04/01/2018 and 03/31/2020. Patient characteristics including age, sex, race, geographic location, BMI, laboratory values, histology, performance status, stage at diagnosis, and comorbidities were obtained from iKnowMed electronic health records. Predictors of biomarker testing were identified using multivariable logistic regression. Results: Of the 3474 adults included in the study, median age was 69 years (range 23-90), 51% female, 65% White, 8% Black or African American, and 46% treated in the South US census region. Histology was adenocarcinoma in 74% and squamous cell carcinoma in 19%. Modeling for any biomarker testing resulted in histology and geographic region as significantly associated with increased likelihood of testing: adenocarcinoma vs squamous cell, odds ratio (OR) 1.76 (CI 1.33-3.23, p<0.0001) and Midwest vs South, OR 2.04 (CI 1.24-3.34, p=0.005). For comprehensive testing of all 5 biomarkers, increased albumin, adenocarcinoma or not otherwise specified histology, Midwest, Northeast or West geographic regions, and larger practice size were all significantly associated with increased likelihood of testing, while Black or African American race was associated with significantly decreased likelihood of testing (Table). Conclusions: Black or African American race, smaller practice size, Southern practice, and squamous cell histology were significantly associated with lower comprehensive biomarker testing rates. Understanding clinical and social determinants of health will be important when evaluating interventions to improve testing rates in the prospective phases of the MYLUNG study. [Table: see text]