An easy-to-use survival score compared to existing tools for older patients with cerebral metastases from colorectal cancer

Abstract

An easy-to-use survival score was developed specifically for older patients with cerebral metastases from colorectal cancer, and was compared to existing tools regarding the accuracy of identifying patients who die in ≤6 months and those who survive for ≥6 months. The new score was built from 57 patients receiving whole-brain irradiation. It included three groups identified from 6-month survival rates based on two independent predictors (performance status and absence/presence of non-cerebral metastases), with 6-month survival rates of 0% (0 points), 26% (1 point), and 75% (2 points), respectively. This score was compared to diagnosis-specific scores, namely the diagnosis-specific graded prognostic assessment (DS-GPA), the Dziggel-Score and the WBRT-30-CRC (whole-brain radiotherapy with 30 Gy in 10 fractions for cerebral metastases from colorectal cancer) score and to a non-diagnosis-specific score for older persons (Evers-Score). Positive predictive values were 100% (new score), 87% (DS-GPA), 86% (Dziggel-Score), 91% (WBRT-30-CRC), and 100% (Evers-Score), respectively, for patients dying ≤6 months, and 75%, 33%, 75%, 60%, and 45%, respectively, for survivors ≥6 months. Of the five tools, the new score and the Evers-Score were most precise in identifying patients dying ≤6 months. The new score and the Dziggel-Scores were best at identifying patients surviving ≥6 months. When combining the results, the new score appeared preferable to the existing tools. The score appears not necessary for patients with additional liver metastases, since their 6-month survival rate was 0%.