Robustness of the neurological prognostic score in brain metastasis patients treated with Gamma Knife radiosurgery.

OBJECTIVE The neurological prognostic score (NPS) was recently proposed as a means for predicting neurological outcomes, such as the preservation of neurological function and the prevention of neurological death, in brain metastasis patients treated with Gamma Knife radiosurgery (GKRS). NPS consists of 2 groups: Group A patients were expected to have better neurological outcomes, and Group B patients were expected to have poorer outcomes. NPS robustness was tested in various situations. METHODS In total, 3040 patients with brain metastases that were treated with GKRS were analyzed. The cumulative incidence of the loss of neurological function independence (i.e., neurological deterioration) was estimated using competing risk analysis, and NPS was compared between Groups A and B by employing Gray's model. NPS was tested to determine if it can be applied to 5 cancer categories-non-small cell lung cancer, small cell lung cancer, gastrointestinal tract cancer, breast cancer, and other cancers-as well as if it can be incorporated into the 5 major grading systems: recursive partitioning analysis (RPA), score index for stereotactic radiosurgery (SIR), basic score for brain metastases (BSBM), graded prognostic assessment (GPA), and modified-RPA (M-RPA). RESULTS There were 2263 patients in NPS Group A and 777 patients in Group B. Neurological deterioration was observed in 586 patients (19.2%). The cumulative incidences of neurological deterioration were 9.5% versus 21.0%, 14.1% versus 25.4%, and 17.6% versus 27.8% in NPS Groups A and B at 1, 2, and 5 years, respectively. Significant differences were detected between the NPS groups in all cancer categories. There were significant differences between NPS Groups A and B for all classes in terms of the BSBM, GPA, and M-RPA systems, but the differences failed to reach statistical significance in terms of RPA Class I and SIR Class 0 to 3. CONCLUSIONS The NPS was verified as being highly applicable to all cancer categories and almost all classes for the 5 grading systems in terms of neurological function independence. This NPS system appears to be quite robust in various situations for brain metastasis patients treated with GKRS.