Generation of a New Disease-specific Prognostic Score for Patients With Brain Metastases From Small-cell Lung Cancer Treated With Whole Brain Radiotherapy (BMS-Score) and Validation of Two Other Indices

Patients with small-cell lung cancer (SCLC) demonstrate an exception in the treatment of brain metastases (BM), because in patients with SCLC whole brain radiotherapy (WBRT) only is the preferred treatment modality. The purpose of this study was to develop a prognostic score for patients with brain metastases from SCLC treated with WBRT.The present study was conducted utilizing a single-institution, previously described, retrospective database of patients with SCLC who were treated with WBRT (n = 221). Univariate and multivariate analyses were performed to generate the "brain metastases from SCLC score" (BMS score) based on favorable prognostic factors: Karnofsky performance status (KPS70), extracerebral disease status (stable disease/controlled), and time of appearance of BM (synchronous). Furthermore, the disease-specific graded prognostic assessment score as well as the recursive partitioning analysis (RPA) were performed and compared with the new BMS score by using the log-rank (Mantel-Cox) test.BMS score and RPA showed the most significant differences between classes (P .001). BMS score revealed a mean overall survival (OS) of 2.62 months in group I (0-1 points), 6.61 months in group II (2-3 points), and 12.31 months in group III (4 points). The BMS score also identified the group with the shortest survival (2.62 months in group I), and the numbers of patients in each group were most equally distributed with the BMS score.The new BMS score was more prognostic than the RPA and disease-specific graded prognostic assessment scores. The BMS score is easy to use and reflects known prognostic factors in contemporary patients with SCLC treated with WBRT. Future studies are necessary to validate these findings.