2202 - Fractionated Stereotactic Radiosurgery for Brain Metastases: Effect of Lesion Size and Dose on Local Failure and Radionecrosis

03:00pm - 04:00pm ET

Poster Hall - Exhibit Hall A

Screen: 5

POSTER

Presenter(s)

Daniel Schep, MD, MS - Juravinski Cancer Centre, Hamilton, ON

D. Schep¹, S. Tran¹, G. J. Li¹, H. Chen¹, J. Detsky¹, D. Dinakaran², C. Heyn³, L. Holden², P. Maralani³, A. Sahgal¹, C. L. Tseng², K. L. Zeng¹, and H. Soliman¹; ¹Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada, ²Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, Toronto, ON, Canada, ³Department of Medical Imaging, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada

Purpose/Objective(s):

While stereotactic radiosurgery (SRS) for brain metastases has traditionally been delivered in a single fraction, fractionated SRS (FSRS) is increasingly used especially for large lesions. The aim of this study was to assess the outcomes of brain metastases treated with FSRS using the Gamma Knife ICON system at our institution.

Materials/Methods:

Consecutive patients with intact brain metastases who completed FSRS between 2017 and 2021 were included for retrospective analysis if they had at least one follow-up MRI brain available. Lesions were initially measured by the maximum dimension on axial slice, and were assessed for progression using RANO-BM criteria. Necrosis was considered present if pathologically confirmed, or if deemed present by the reading radiologist and treating radiation oncologist. Descriptive statistics were generated to summarize demographic and dosimetric characteristics. Cumulative incidence functions were used to calculate the incidence of local failure (LF), radionecrosis (RN) and symptomatic radionecrosis (SRN) with death as a competing risk, and Gray’s test used to compare between groups. Kaplan-Meier estimators were used to calculate progression free survival and overall survival (OS).

Results:

Two-hundred and sixteen patients with 511 brain metastases were available for analysis. The most common histologies were lung (172, 34%), breast (139, 27%), and melanoma (57, 11%). The median lesion diameter was 16mm (range: 2-41mm), and 166 (32%) were =20mm. The most common doses/fractionations were 27.5Gy/5 fractions (189, 37%), 25Gy/5 fractions (181, 35.4%), and 24Gy/3 fractions (49, 9.6%). The median follow-up was 12.8 months. The 1-year and median overall survival was 57% and 14.3 months, respectively. The 1-year incidence of LF, RN, and SRN was 15%, 7.8%, and 4%, respectively. Compared to those <20mm, lesions measuring =20mm had no significant difference in LF (1y LF 15% in both groups, NS), RN (1-year RN 8% vs 7.7%, NS), or SRN (1-year SRN 6.8% vs 2.7%, NS). Compared to those treated with BED₁₀ <42.6Gy (n=230, equivalent to 27.5Gy/5), lesions treated with BED₁₀ =42.6Gy (n=281) had significantly lower LF (1-year LF 13% vs 18%, p=0.005), as well as significantly higher RN (1-year RN 9.5% vs 5.7%, p<0.001) and SRN (1-year SRN 5.1% vs 2.6%, p=0.002).

Conclusion:

FSRS for brain metastases results in excellent rates of LF with acceptable toxicity. LF, RN, and SRN are similar in smaller and larger lesions. Increased dose results in better control at the cost of higher rates of RN.