AUTHOR=Zhu Li-Hua , Fan Xing-Wen , Sun Lu , Ni Ting-ting , Li Ya-qi , Wu Chao-Yang , Wu Kai-Liang TITLE=New prognostic system specific for epidermal growth factor receptor-mutated lung cancer brain metastasis JOURNAL=Frontiers in Oncology VOLUME=Volume 13 - 2023 YEAR=2023 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1093084 DOI=10.3389/fonc.2023.1093084 ISSN=2234-943X ABSTRACT=Brain metastases (BM) from lung cancer are heterogeneous, and accurate prognosis is required for effective treatment strategies. This study aimed to identify prognostic factors and develop a prognostic system exclusively for epidermal growth factor receptor (EGFR)-mutated lung cancer BM.In total, 173 patients with EGFR-mutated lung cancer from two hospitals who developed BM and received tyrosine kinase inhibitor (TKI) and brain radiation therapy were included. Univariate and multivariate analyses were performed to identify significant EGFR-mutated BM prognostic factors to construct a new EGFR recursive partitioning analysis (PRA) prognostic index. The predictive discrimination of five prognostic scoring systems including RPA, DS-GPA, BS-BM, lung-mol GPA, and EGFR-PRA were analyzed using log-rank test, C-index, and receiver operating characteristic curve. The potential predictive factors in the multivariable analysis obtained to construct a prognostic index included Karnofsky performance status, BM at initial lung cancer diagnosis, BM progression after TKI, EGFR mutation type, uncontrolled primary tumors, and number of BM. In the log-rank test, indices of RPA, DS-GPA, lung-mol GPA, BS-BM, and EGFR-PRA were all significant predictors of overall survival (OS) (p ≤ 0.05). The C-indices of each prognostic score were 0.603, 0.569, 0.613, 0.595, and 0.671, respectively; AUC values predicting 1-year OS were 0.565 (p=0.215), 0.572 (p=0.174), 0.641 (p=0.007), 0.585 (p=0.106), and 0.781 (p=0.000), respectively. Furthermore, EGFR-PRA performed well in terms of calibration.BM progression after TKI and EGFR mutation type were specific prognostic factors for EGFR-mutated lung cancer BM. EGFR-PRA was precise than other models, and useful for personal treatment.