P13.03.A Radiomics for the non-invasive assessment of the PDL-1 expression in patients with non-small cell lung cancer brain metastases. (5th September 2022)

• Record Type: Journal Article
• Authors: Meißner, A; Gutsche, R; Galldiks, N; Kocher, M; Jünger, S; Eich, M; Nogova, L; Schmidt, N; Ruge, M; Goldbrunner, R; Proescholdt, M; Grau, S; Lohmann, P
• Is Part Of: Neuro-oncology. Volume 24(2022)Supplement 2
• Page Start: ii81
• Page End: ii81
• Publication Date: 2022-09-05
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noac174.283
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: BACKGROUND: The expression level of programmed cell death ligand 1 (PDL-1) might be an indicator for response to immunotherapy using checkpoint inhibitors in patients with non-small cell lung cancer (NSCLC). As intra-tumoral differences and discrepancies between the PDL-1 expression in the primary tumor and the brain metastases may occur, a method for a reliable non-invasive assessment of the intracranial PDL-1 expression would be of clinical value. We evaluated the potential of MRI radiomics for a non-invasive assessment of the PDL-1 expression in patients with NSCLC brain metastases. PATIENTS AND METHODS: Fifty-three patients with brain metastases from NSCLC from two university brain tumor centers (group 1, 36 patients; group 2, 17 patients) underwent tumor resection with subsequent immunohistochemical assessment of the PDL-1 expression. Brain metastases were manually segmented on preoperative T1-weighted contrast-enhanced MRI. Group 1 was used for model training and validation, group 2 for model testing. After image pre-processing and radiomics feature extraction from T1-weighted contrast-enhanced MRI, a test-retest analysis was performed to identify robust features prior to feature selection. The radiomics model was trained and validated using five-fold cross validation. Finally, the best performing radiomics model was applied to the test data. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analyses. RESULTS: An intracranial PDL-1 expression was found by immunohistochemistry in 18 of 36 patients (50%) in group 1, and 7 of 17 patients (41%) in group 2. Univariate analysis identified tumor volume as a significant clinical feature for PDL-1 expression (area under the ROC curve (AUC), 0.77). A random forest classifier using a four-parameter radiomics signature including tumor volume yielded an AUC of 0.83 ± 0.18 in the training data (group 1). Finally, the classifier achieved an AUC of 0.84 in the external test data (group 2). CONCLUSION: The developed radiomics classifiers allows a non-invasive assessment of the intracranial PD-L1 expression in patients with NSCLC brain metastases with a high diagnostic performance.