SURG-48. Predictors of clinical and oncologic outcomes after salvage resection for recurrent brain metastases: a multivariable analysis

Salvage treatment is increasingly requested for recurrent brain metastases to achieve local control and improve neurologic symptoms. However, optimal criteria for patient selection and surgical planning remain unclear. This study aimed to identify predictors of radiographic failure, need for additional treatment, overall survival, symptom improvement, and steroid independence following salvage resection. Patients who underwent salvage surgery for locally recurrent brain metastases between January 2004 and September 2023 at a single institution were identified. Multivariable logistic regression and Cox proportional hazards models were used to identify independent predictors. A total of 207 patients with 234 recurrent brain metastases were included. Before salvage surgery, 73.5% of patients were symptomatic, 51.7% were steroid-dependent, and 81.2% had good performance status. Gross total resection (GTR) was achieved in 67.1% of procedures. Among symptomatic patients, 71.5% experienced symptom improvement, and 55.4% of steroid-dependent patients achieved steroid independence. Radiographic and treated failure rates were 37.6% and 28.6%, respectively. GTR was independently associated with reduced hazard of radiographic failure (HR 0.16, p < 0.001), progression requiring additional treatment (HR 0.17, p < 0.001), and mortality (HR 0.70, p = 0.024), as well as increased odds of symptom improvement (OR 2.88, p = 0.011). Poor performance status was associated with increased hazard of radiographic failure (HR 2.34, p = 0.002), progression requiring additional treatment (HR 2.56, p = 0.007), and mortality (HR 1.97, p < 0.001), along with lower odds of achieving steroid independence (OR 0.26, p = 0.021) and symptom improvement (OR 0.24, p < 0.001). Therefore, a resection strategy aimed at GTR significantly improves radiographic control, symptom relief, and survival even in the salvage setting for recurrent brain metastases. This supports a proactive surgical approach in appropriately selected patients, with performance status as a key determinant of benefit in the salvage setting.