SURG-34. Real-world predictors of resection prior to stereotactic radiosurgery for brain metastases: a multi-center study

Abstract Stereotactic radiosurgery (SRS) is standard of care for brain metastases, yet the selection criteria for performing upfront craniotomy are inconsistently defined. Among 1172 patients from two institutions, we identified clinical and radiotherapeutic predictors of surgical resection prior to linear accelerator or Gamma Knife-based SRS. One institution’s data were randomly split into training and validation datasets. Multivariable logistic regression with stepwise selection showed that patients with a single brain metastasis (OR=3.18; 95% CI: 1.63–6.24; p=0.0007), lesion diameter ≥2.5 cm (OR=33.3; 95% CI: 12.2–90.9; p<0.0001), higher KPS (40–70 vs. 80–100; OR=0.46; 95% CI: 0.22–0.99; p=0.048) or with a neurologic deficit on presentation (OR=0.40; 95% CI: 0.20–0.80; p=0.0094) were more likely to receive resection prior to SRS. The model demonstrated strong predictive accuracy (Somers’ D=0.77 in training; 0.825 internal validation cohort; 0.809 external validation with data from second institution). Patients with single brain metastases who underwent resection experienced significantly longer survival, p=0.016. With available data, Cox proportional hazards modeling identified younger age (HR per 10-year decrease: 0.87; 95% CI: 0.81–0.93; p=0.0002), controlled extracranial disease (HR=0.73; 95% CI: 0.62–0.87; p=0.0003), and absence of cerebellar (HR=0.79; 95% CI: 0.67–0.93; p=0.0051) or brainstem involvement (HR=0.67; 95% CI: 0.52–0.86; p=0.0019) as significant predictors of prolonged survival when controlling for primary disease site. An extended Cox proportional hazards model that integrated radiation parameters also found that larger maximum target volume (HR=0.956; 95% CI: 0.931–0.982; p=0.0009) and lower total target volume (HR=1.048; 95% CI: 1.025–1.071; p<0.0001) were associated with improved outcomes. This analysis highlights predictors of surgical intervention prior to SRS for brain metastases. These findings reflect intuitive surgical decision making—namely the need for craniotomy in large, symptomatic lesions—and can thereby inform patient selection prior to SRS.