Main Session

Sep 28

PQA 02 - Lung Cancer/Thoracic Malignancies, Patient Reported Outcomes/QoL/Survivorship, Pediatric Cancer

2410 - Genomic Predictors of Brain Metastases in Non-Small Cell Lung Cancer: Implications for Early Detection and Targeted Surveillance

04:45pm - 06:00pm PT

Hall F

Screen: 8

POSTER

Presenter(s)

Victor Lee, MD, BS - Yale New Haven Hospital, Middletown, CT

V. Lee¹, P. Oh¹, A. Rybkin¹, K. Jani², A. A. Patel³, T. J. Hayman¹, S. Y. Kim⁴, and H. S. M. Park⁵; ¹Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT, ²Rutgers University Robert Wood Johnson Medical School, New Brunswick, NJ, ³Yale University School of Medicine, New Haven, CT, ⁴Yale School of Medicine, New Haven, CT, ⁵Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, CT

Purpose/Objective(s): Brain metastases (BMs) are a common and devastating complication of non-small cell lung cancer (NSCLC), with significant implications for prognosis and treatment. Identifying patients at higher risk for developing BMs could enable earlier detection and intervention, reducing the risk of symptomatic BM. This study aims to identify genomic alterations associated with BMs at diagnosis and those that develop after diagnosis using the Flatiron Health database.

Materials/Methods: From the nationwide (US-based) deidentified Flatiron Health-Foundation Medicine NSCLC Clinico-Genomic Database, containing data originated from ~280 US cancer clinics, we analyzed patients with advanced NSCLC. We categorized patients into two groups: (1) those with BMs at diagnosis and (2) those who developed BMs after diagnosis. Genomic mutations were analyzed for their association with BMs using univariable logistic regression. P value for significance threshold was set to 0.001. Statistical analyses were performed using Python 3.7, utilizing the scipy.stats, statsmodels, and sklearn libraries.

Results: There were a total of 20,821 patients (mean age 68.0) in the cohort with 2,840 patients (mean age 64.4) that had BM at initial advanced diagnosis and 3,197 patients (mean age 65.4) had BM after initial advanced diagnosis. Notable mutations associated with BM at initial advanced diagnosis included: CDH2 (OR: 1.88, CI:1.41-2.5), CDH20 (OR: 1.85, CI:1.36-2.52), EGFR (OR: 1.53, CI:1.39-1.68), FAT3 (OR: 1.49, CI:1.29-1.73), PTPRD (OR: 1.46, CI:1.19-1.78), KEAP1 (OR: 1.37, CI:1.24-1.51), LRP1B (OR: 1.37, CI:1.2-1.55), TP53 (OR: 1.34, CI:1.23-1.46), STK11 (OR: 1.30, CI:1.17-1.44), NTRK3 (OR: 1.27, CI:1.11-1.46), and KRAS (OR: 1.25, CI:1.15-1.36). Notable mutations associated with BM after initial advanced diagnosis included: ARID1B (Odds ratio [OR]: 1.73, 95% confidence interval [CI]: 1.42-2.11), EPHA5 (OR: 1.51, CI:1.21-1.88), EGFR (OR: 1.44, CI:1.31-1.57), FAT3 (OR: 1.43, CI:1.24-1.64), FAT1 (OR: 1.42, CI:1.18-1.71), and STK11 (OR: 1.31, CI:1.18-1.45).

Conclusion: This study identified specific genomic alterations associated with brain metastases at and after initial advanced diagnosis in NSCLC, highlighting potential molecular pathways that facilitate tumor spread to the brain. Many of these genes, particularly CDH2, FAT1, FAT3, EPHA5, NTRK3, and LRP1B, have direct roles in neuronal adhesion, signaling, and plasticity. Identifying genomic alterations associated with BMs in NSCLC may enable earlier surveillance with brain MRI, potentially allowing for timely SRS and reducing the need for WBRT. By stratifying patients based on molecular risk factors, this study provides a framework for personalized surveillance strategies to improve outcomes and preserve neurocognitive function in NSCLC patients at risk for BMs.