|| Checking for direct PDF access through Ovid
A reliable prognostic score for cancer patients with brain metastases submitting to stereotactic radiosurgery (SR) was established. Between July 1993 and December 1995, 34 cancer patients (36 treatments and 69 lesions) with brain metastases were treated with SR at the Hospital Israelita Albert Einstein of São Paulo, Brazil. The exclusion criteria were: more than 5 lesions, any lesion larger than 30 cm3, Karnofsky performance status (KS) less than 50, clinical evidence urging neurosurgical intervention or very poor overall prognosis due to fast progressive systemic disease (SD). The score index for stereotactic radiosurgery of brain metastases (SIR) was obtained by summing up the points of the five variables: age, KS, SD, size (LS), and number of lesions (NL). The Kaplan–Meier method together with log rank was used to estimate and test correlation between survival and all prognostic factors. Applying the SIR definition to our patients' survival data through the Kaplan–Meier method we found out that the expected median survival for patients with 6 points or less was 3.67 months and for those with more than 6 points it was 16.16 months (p = 0.0005). Age, SD, KS, and NL were also significantly correlated with the prognosis (p-values at log-rank test of 0.0020, 0.0071, 0.0181, and 0.0454, respectively). Analyzing our data we concluded that the SIR is a reliable prognostic score for cancer patients with brain metastases submitting to SR. This index should be tested for larger series.