RESUMEN
Background and purpose: We performed a validity test of a recently-reported, small cell lung cancer (SCLC) graded prognostic assessment (GPA) system for SCLC patients with brain metastases (BMs). Thereafter, we created a new prognostic index, the SCLC Grade, for such patients. Materials and methods: We studied 508 SCLC patients selected from among nearly 7000 consecutive patients undergoing gamma knife SRS for BMs since 1998. Results: In the SCLC GPA, there were no median survival time (MST) differences among pairs of the neighboring subgroups. Therefore, the 508 patients were randomly divided into the two series, i.e., a test (340 patients) and a validity (168) series. In the test series, five factors were identified by univariable analyses as favoring longer survival (rounded lower 95 % CI of the HR was at least 1.3): Sex, Karnofsky Performance Status, tumor numbers, primary tumor status and extracerebral metastases. This new index is the sum of scores (0 and 1) of these five factors: SCLC-Grade 4-6 (score of 4, 5 or 6), 2-3 (2 or 3), and 0-1 (0 or 1). This new system showed highly statistically significant MST differences among subclasses. Next, this SCLC-Grade was applied to the verification series. Consistent results were obtained, i.e., there were highly statistically significant MST differences among subclasses. Conclusions: Our validity test results for the SCLC GPA demonstrated this system to not precisely reflect the outcomes of SCLC patients with BMs. Our results suggest the herein-proposed SCLC-Grade to have superior prognostic value.