Validating the SIR: a better prognostic score index for patients with brain metastases treated with stereotactic radiosurgery

Objective: The aim of this paper is to validate the score index for survival in patients treated with stereotactic radiosurgery, using a classification prepared to better evaluate the prognosis of patients with brain metastasis submitted to stereotactic surgery, re-evaluating survival of patients and reviewing the medical literature. Methods: Data from 100 patients with brain metastases treated with stereotactic radiosurgery at a single institution, between July 1993 and February 2000, were retrospectively analyzed. The prognostic factors and scores studied were age, Karnofsky performance status, extracranial disease status, number of brain lesions, volume of the largest lesion, primary tumor type, treated or not with whole brain radiation therapy, SIR, and RPA. Kaplan-Meier actuarial survival curves for subsets were calculated and compared by log-rank test. Complete and backward elimination Cox models were utilized to identify the prognostic factors and scores independently associated with survival. Results: Karnofsky performance status, extracranial disease status, volume of the largest brain lesion, RPA, and SIR were significantly correlated with prognosis in Kaplan-Meier survival analysis. Applying Cox models, significance was observed for KPS and volume of the largest lesion (p < 0.0001 and p = 0.0182, respectively), as well as for SIR and RPA when tested individually (p < 0.0001 and p = 0.0002, respectively). However, when testing SIR and RPA together, only SIR reached independent statistical significance (p < 0.0001). Conclusion: SIR classification demonstrated a better accuracy in predicting survival time than RPA. SIR was tested in other centers, showing superior accuracy and applicability than the RPA, thus validating this score.