Abstract
Abstract
Purpose
Existing prognostic models for brain metastases from lung cancer rely on clinical variables and do not capture the biological heterogeneity of host or tumor factors underlying survival outcomes. We evaluated whether MRI-derived radiomic biomarkers improve prognostic stratification beyond clinical variables. We also investigated their biological correlates using matched histopathology.
Experimental Design
This retrospective study included 103 patients with brain metastases from lung cancer and matched pre-operative MRI and histopathology. Radiomic features from T1-weighted contrast-enhanced and T2-weighted fluid-attenuated inversion recovery sequences were combined with clinical variables to train Cox proportional hazards, random survival forest, and DeepSurv models for overall survival (OS), CNS progression, and leptomeningeal (LM) progression. Risk groups from an imaging-only model were compared for clinical and histopathologic differences.
Results
Imaging features improved prognostic performance beyond clinical variables across all modeling frameworks, with the combined clinical-imaging DeepSurv model achieving the highest concordance index for OS (0.728). Risk groups from the best-performing model for each endpoint differed significantly in OS (median 26.7 vs. 9.1 months,
p
p
= 0.003), but not in CNS progression. Imaging-only risk groups did not differ significantly in any clinical variable, but high-risk patients were less likely to have tumor-infiltrating lymphocytes (46% vs. 77%,
p
= 0.010) and PD-L1 expression ≥ 50% (10% vs. 32%,
p
= 0.049).
Conclusions
MRI radiomic biomarkers provide prognostic information beyond established clinical variables in lung cancer brain metastases and correspond to features of the tumor immune microenvironment, supporting their potential to refine risk stratification.
Statement of Translational Relevance
Brain metastases affect 30-40% of patients with lung cancer. Existing prognostic models rely on clinical variables, which do not fully capture the biological heterogeneity that drives outcomes. We show that machine learning models integrating MRI-derived radiomic features with clinical variables improve prognostic discrimination beyond clinical variables alone. Because these biomarkers are derived from standard-of-care pre-operative imaging, they can be generated without additional testing or delay to treatment planning. Using matched histopathology from resected lesions, we further show that imaging-defined risk groups do not simply reproduce established clinical risk factors but instead correspond to differences in the tumor immune microenvironment, including tumor-infiltrating lymphocytes and PD-L1 expression. These findings support the use of quantitative imaging biomarkers as a non-invasive method to refine risk stratification at diagnosis, informing treatment selection and clinical trial enrollment for patients with brain metastases from lung cancer.