OBJECTIVE High-resolution double-dose gadolinium-enhanced Gamma Knife (GK) radiosurgery-planning MRI (GK MRI) on the day of GK treatment can detect additional brain metastases undiagnosed on the prior diagnostic MRI scan (dMRI), revealing increased intracranial disease burden on the day of radiosurgery, and potentially necessitating a reevaluation of appropriate management. The authors identified factors associated with detecting additional metastases on GK MRI and investigated the relationship between detection of additional metastases and postradiosurgery patient outcomes. METHODS The authors identified 326 patients who received GK radiosurgery at their institution from 2010 through 2013 and had a prior dMRI available for comparison of numbers of brain metastases. Factors predictive of additional brain metastases on GK MRI were investigated using logistic regression analysis. Overall survival was estimated by Kaplan-Meier method, and postradiosurgery distant intracranial failure was estimated by cumulative incidence measures. Multivariable Cox proportional hazards model and Fine-Gray regression modeling assessed potential risk factors of overall survival and distant intracranial failure, respectively. RESULTS The mean numbers of brain metastases (SD) on dMRI and GK MRI were 3.4 (4.2) and 5.8 (7.7), respectively, and additional brain metastases were found on GK MRI in 48.9% of patients. Frequencies of detecting additional metastases for patients with 1, 2, 3-4, and more than 4 brain metastases on dMRI were 29.5%, 47.9%, 55.9%, and 79.4%, respectively (p < 0.001). An index brain metastasis with a diameter greater than 1 cm on dMRI was inversely associated with detecting additional brain metastases, with an adjusted odds ratio of 0.57 (95% CI 0.4-0.9, p = 0.02). The median time between dMRI and GK MRI was 22 days (range 1-88 days), and time between scans was not associated with detecting additional metastases. Patients with additional brain metastases did not have larger total radiosurgery target volumes, and they rarely had an immediate change in management (abortion of radiosurgery or addition of whole-brain radiation therapy) due to detection of additional metastases. Patients with additional metastases had a higher incidence of distant intracranial failure than those without additional metastases (p = 0.004), with an adjusted subdistribution hazard ratio of 1.4 (95% CI 1.0-2.0, p = 0.04). Significantly worse overall survival was not detected for patients with additional brain metastases on GK MRI (log-rank p = 0.07), with the relative adjusted hazard ratio of 1.07, (95% CI 0.81-1.41, p = 0.65). CONCLUSIONS Detecting additional brain metastases on GK MRI is strongly associated with the number of brain metastases on dMRI and inversely associated with the size of the index brain metastasis. The discovery of additional brain metastases at time of GK radiosurgery is very unlikely to lead to aborting radiosurgery but is associated with a higher incidence of distant intracranial failure. However, there is not a significant difference in survival. ▪ CLASSIFICATION OF EVIDENCE Type of question: prognostic; study design: retrospective cohort trial; evidence: Class IV.
Keywords: GK = Gamma Knife; GK MRI = GK radiosurgery planning MRI; HR = hazard ratio; IRSPGR = inversion recovery spoiled gradient recalled; MRI; OR = odds ratio; SHR = subdistribution HR; SRS = stereotactic radiosurgery; UCSF = University of California, San Francisco; WBRT = whole-brain radiation therapy; brain metastases; dMRI = diagnostic MRI; distant intracranial failure; gadolinium; oncology; stereotactic radiosurgery; survival.