Comprehensive Prognostic Index (CPI): A New Prognostic Index for Patients Treated with Stereotactic Robotic Radiotherapy for Brain Metastases

Abstract

The goal of this study was to analyze potential prognostic factors in a population of patients treated with robotic radiosurgery for brain metastases irrespective of the location of the primary tumor, to create a simple prognostic score that can be used even in patients without full diagnostic workup. The indices currently used are either created based on population of patients with limited number of metastases or are diagnosis-specific and thus, cannot be used for certain populations of patients or require results of sophisticated test, not always available at the time of clinical decision making. A retrospective analysis of 142 patients with brain metastases treated with stereotactic robotic radiosurgery or hypofractionated radiotherapy was performed. The patients had 1 – 9 metastases (single metastasis – 56.4%, two – 22.1%, three or more – 21.5%) and were treated with single fraction or with hypofractionated (2-5 fractions) stereotactic radiosurgery with doses ranging between 5 Gy and 30 Gy. Volume of the lesions was calculated using linear dimensions of the tumors (CC, LR, AP) and 4/3*pi *(CC/2)*(LR/2)*(AP/2) formula for volume of a spheroid, to avoid the need for volumetric analysis of diagnostic images. Kaplan-Meier method and log-rank test were used to analyze survival. Variables significantly associated with overall survival in univariate analysis were included into Cox multivariate analysis. The validity of the model was tested with the bootstrap method and the variables from the final model were used to construct a new prognostic index by assigning points according to the impact of a certain variable on overall survival. Prognostic scores existing to date were also tested. In the multivariate analysis four factors: Karnofsky Performance Status (KPS), number of brain metastases, volume of the largest metastasis and presence of extracerebral metastases, were independent predictors of survival. Total scores ranged from 0 to 12 points, and patients were divided into four groups based on median survival of each subgroup. The median survival was: 18.8 months for 0-1 points, 16.9 months for 2-3 points, 5.6 months for 3-4 points, and 4.9 months for ≥6 points (p < 0.001). The prognostic value of all indices was confirmed. The most significant differences between classes were observed in the Score Index for Radiosurgery (SIR) and our new Comprehensive Prognostic Index (CPI) – p = 0.00021 and p = 0.00033 respectively. The new prognostic index is simple to calculate and has strong prognostic value in heterogeneous population of patients with various number of brain metastasis, but its value requires confirmation in another cohort.Abstract 3570; TablePrognostic factorHR95% CIp valueKPS0.960.93 – 0.980.000068Number of brain metastases1.191.03 – 1.370.019Volume of the largest lesion1.021.01. – 1.040.0037Extracerebral metastases0.510.34 – 0.780.0017 Open table in a new tab