Preoperative frailty and 30-day mortality after resection of brain metastases: a multicenter surgical registry analysis of 11,038 patients

Abstract

OBJECTIVE Surgery plays a key role in the management of brain metastases. Stratifying surgical risk and individualizing treatment will help optimize outcomes because there is clinical equipoise between radiation and resection as treatment options for many patients. Here, the authors used a multicenter database to assess the prognostic utility of baseline frailty, calculated with the Risk Analysis Index (RAI), for prediction of mortality within 30 days after surgery for brain metastasis. METHODS The authors pooled patients who had been surgically treated for brain metastasis from the American College of Surgeons National Surgical Quality Improvement Program database (2012–2020). The authors studied the relationship between preoperative calculated RAI score and 30-day mortality after surgery for brain metastasis by using linear-by-linear proportional trend tests and binary logistic regression. The authors calculated C-statistics (with 95% CIs) in receiver operating characteristic (ROC) curve analysis to assess discriminative accuracy. RESULTS The authors identified 11,038 patients who underwent brain metastasis resection with a median (interquartile range) age of 62 (54–69) years. The authors categorized patients into four groups on the basis of RAI: robust (RAI 0–20), 8.1% of patients; normal (RAI 21–30), 9.2%; frail (RAI 31–40), 75%; and severely frail (RAI ≥ 41), 8.1%. The authors found a positive correlation between 30-day mortality and frailty. RAI demonstrated superior predictive discrimination for 30-day mortality as compared with the 5-factor modified frailty index (mFI-5) on ROC analysis (C-statistic 0.65, 95% CI 0.65–0.66). CONCLUSIONS The RAI frailty score accurately estimates 30-day mortality after brain metastasis resection and can be calculated online with an open-access software tool: https://nsgyfrailtyoutcomeslab.shinyapps.io/BrainMetsResection/. Accordingly, RAI can be utilized to measure surgical risk, guide treatment options, and optimize outcomes for patients with brain metastases. RAI has superior discrimination for predicting 30-day mortality compared with mFI-5.