Gamma Knife treatment of brain metastases using high versus standard isodose lines has equivalent efficacy

Keywords: brain metastasis, radiosurgery, gamma knife, outcome, dose escalation

Interactive Manuscript

Ask Questions of this Manuscript:



     Comparisons between different dose planning concepts for brain metastases are few.
     To determine the relationship of local control compared to heterogeneity (HI) and conformality (CI) indices for patients treated with Gamma Knife (GK) for 1-3 brain metastases (bMet).
     112 patients were studied. Median clinical follow-up time by lesion was 5.4 months (range: 0-42.2).
     From an IRB-approved database of patients treated with GK, 112 patients with 183 total bMet were identified who were treated for 1-3 bMet between July 2007 and April 2011. Patients were treated to a median dose of 24 Gy (range: 15-24) in a single session, with a median of 2 bMet treated per patient. The median HI was 1.894 (range: 1.088-2.042); the median CI was 1.63 (range: 1.15-4.2). Lesions were divided into two categories: “radioresistant” included melanoma and renal cell; “radiosensitive” included all other histologies. Cox proportional hazards analysis was performed to identify any association of the following with local control: HI, CI, isodose line (IDL), dose, tumor size, recursive partitioning analysis (RPA) class, and tumor radioresistance.
      The median survival time for the series was 15.1 months. Of the 183 lesions, 9.8% failed (defined as progression based on RECIST criteria). The 6-, 12-, and 18-month local control rates were 94%, 92%, and 80%, respectively. The only variable associated with local control was RPA class 3 vs 2 (HR 4.74, 95% CI=1.28-17.54, p=0.02). All other variables, including HI (HR 0.18, 95% CI=0.02-2.01) and CI (HR 1.11, 95% CI=0.41-3.02), were non-significant.
     This was a retrospective review.
     Our single-institution study of patients with 1-3 bMet treated with GK demonstrated that there is no difference in local control with varying HI. Significantly, this result suggests that physicians can comfortably treat patients to lower HI (higher IDL, e.g. 80% IDL and above) without concern of increased local failure rates, thereby reducing treatment times without sacrificing outcome.
     Interestingly, the Conformity index had no bearing on local outcome, although correlation with toxicity rates has not yet been analysed. Our study also supported the results of previous studies in demonstrating excellent local control rates (92% at 12 months). A larger pool of patients is being studied. Updated results will be presented at the 16th Leksell Gamma Knife Society Meeting.


Project Roles: