Influence of social determinants of health (SDoH) on metastatic non-small cell lung cancer (mNSCLC) testing in the community setting.

158 Background: Biomarker testing implementation in community settings remains hindered. We explored the influence of SDoH factors on NGS and PD-L1 testing uptake in patients diagnosed with mNSCLC in the community setting. Methods: Electronic health record data from Florida Cancer Specialists and the American Oncology Network were collected from 01/2020-04/2024 for patients ≥18 years old at time of mNSCLC diagnosis, with a confirmed adenocarcinoma histology and ≥30 days between diagnosis date and last clinical activity. Biomarker testing was confirmed using unstructured data abstraction via natural language processing. Patients were categorized into 4 cohorts (PD-L1+ NGS testing, only PD-L1, only NGS, neither tests). Chi-square tests, ANOVA, and multinomial logistic regression explored associations between SDoH factors and testing patterns. Multiple imputation was used to handle missing data for covariates with ≤35% missingness. Covariates with >35% missingness were excluded from adjusted analyses. A Fisher’s exact test assessed the association between race and stage. Results: Among 1,000 patients (567 NGS+PD-L1, 167 only PD-L1, 204 only NGS, 62 no testing), 51.2% identified as female, 75.9% as White, and 64.2% were initially diagnosed with mNSCLC. Data missingness ranged from 1.6%-55.2%. Patients initially diagnosed with earlier stage NSCLC were less likely to receive any molecular testing, after adjusting for demographics (Table). Patients initially diagnosed with stage I-IIIA or IIIB-IIIC were significantly less likely than those initially diagnosed with mNSCLC to receive NGS+PD-L1 testing (stage I-IIIA OR=0.2, stage IIIB-IIIC OR=0.2), only PD-L1 testing (stage I-IIIA OR=0.25, stage IIIB-IIIC OR=0.22), or only NGS testing (stage I-IIIA OR=0.15, stage IIIB-IIIC OR=0.13) as compared to no testing. Additionally, older age was associated with less likelihood of receiving only PD-L1 testing (OR=0.85) and those diagnosed after 2020 were more likely to receive only NGS testing (OR=1.71) as compared to no testing. Though Black patients were more likely than Whites to be initially diagnosed with mNSCLC (p<0.001), removal of stage from the logistic model did not reveal significant differences on testing uptake by race. Conclusions: Among patients with mNSCLC, an initial diagnosis of early-stage/locally advanced NSCLC and older age were associated with lower likelihood of receiving biomarker testing. Copious missing data underscores the need for consistent SDoH reporting to understanding its impact on NSCLC testing uptake. Impact of SDoH on molecular testing. Characteristic Adjusted OR Lower CI Upper CI p-value DV = PD-L1 & NGS Stage at Initial NSCLC Diagnosis I-IIIA 0.20 0.09 0.41 <.0001 IIIB or IIIC 0.20 0.07 0.56 0.0023 IV REF --- --- --- DV = NGS only Stage at Initial NSCLC Diagnosis I-IIIA 0.15 0.07 0.35 <.0001 IIIB or IIIC 0.13 0.03 0.58 0.009 IV REF --- --- ---