Abstract
PURPOSE: To compare outcomes between gastrointestinal and nongastrointestinal patients with brain metastases after radiosurgery. METHODS AND MATERIALS: Retrospective cohort study identifying patients completing an initial course of radiosurgery between January 2015 and December 2020, with follow-up data collected through November 2022. Multi-institutional, academic referral centers. The primary outcomes were overall survival and intracranial progression-free survival, calculated by the Kaplan-Meier method. Progression was defined as concern on postradiosurgery imaging for recurrence determined by clinical multidisciplinary consensus. Cox proportional hazard models were used to assess associations between outcomes and covariates. RESULTS: This study included 1281 nongastrointestinal patients and 102 gastrointestinal patients, of which 45.1% were colorectal, 33.3% esophageal, and the remaining 21.6% comprising other sites. Gastrointestinal patients were more likely to be younger (mean 59.1 vs 63.5 years, P = .001), male (56.9% vs 44.3%, P = 0.014), have received systemic therapy (73.5% vs 63.9%, P = .049), and have resection of brain metastases (45.1% vs 25.0%, P < .001) prior to radiosurgery. Median overall survival was lower for gastrointestinal patients at 5.4 months (95% CI, 3.8-7.7) versus nongastrointestinal patients at 10.6 months (95% CI, 9.3-11.6, P < 0.0001). In a multivariate model, gastrointestinal patients had worse overall survival compared to nongastrointestinal patients (hazard ratio, 1.92; P < .0001; 95% CI, 1.53-2.41). Median intracranial progression-free survival was lower for gastrointestinal patients at 6.2 months (95% CI, 4.0-9.6) versus nongastrointestinal patients at 12.3 months (95% CI, 10.8-13.9; P = 0.0002). In a multivariate model, gastrointestinal patients had worse intracranial progression-free survival compared to nongastrointestinal patients (hazard ratio, 1.60; 95% CI, 1.20-2.14; P = 0.0013). There were no significant differences between colorectal primary patient or esophageal primary patient outcomes compared to all other gastrointestinal primary patients. CONCLUSIONS: Across a multi-institutional stereotactic radiosurgery cohort, brain metastases of gastrointestinal origin demonstrated inferior overall survival and intracranial progression-free survival to those of nongastrointestinal origin. These data may help inform treatment decisions and postradiosurgery surveillance.