2048 - Incidence of Radiation Necrosis and Local Failure in Small Brain Metastases Treated with Single-Fraction vs. Multi-Fraction Stereotactic Radiosurgery
Presenter(s)
L. A. del Balzo1,2, L. A. Boe3, A. T. Ilica4,5, A. Meng5, D. Gutman5, H. Walch3, B. A. Mueller2, A. J. Khan2, M. Zinovoy2, Y. Yu2, J. Y. Shin2, Y. Yamada2, A. R. Barsky2, C. Kinslow2, G. Cederquist2, C. B. Jackson2, A. Ballangrud6, M. Aristophanous6, B. S. Imber2, and L. R. G. Pike2; 1Department of Radiation Oncology, Winship Cancer Institute of Emory University, Atlanta, GA, 2Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, 3Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, 4Department of Neuroradiology, The University of Miami Miller School of Medicine, Miami, FL, 5Department of Neuroradiology, Memorial Sloan Kettering Cancer Center, New York, NY, 6Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY
Purpose/Objective(s): Radiation necrosis (RN) is a critical treatment toxicity that can develop following stereotactic radiosurgery (SRS) for brain metastases (BM). While multi-fraction SRS (MF-SRS) has been associated with reduced rates of RN and local failure (LF) in larger lesions, its role in small (<2 cm) BMs remains unclear, as they are often treated with single-fraction SRS (SF-SRS) per landmark studies. We sought to rigorously evaluate the radiographic incidence of RN and LF between SF-SRS and MF-SRS in a contemporary cohort of RT-naïve patients with BMs <2 cm.
Materials/Methods: In a single-institution retrospective study, we evaluated patients with BMs <2 cm in diameter that were treated with LINAC-based SF- or MF-SRS during 2017-2020. Clinical, radiographic, and radiation dosimetry data were systematically collected. BMs were measured on longitudinal, contrast-enhanced, axial T1-weighted MRIs; if they increased = 25% in diameter relative to the pre-SRS baseline, they were independently graded by three neuroradiologists with access to perfusion and PET imaging sequences to differentiate RN from LF. Clustered statistical analyses were independently performed by the institution’s biostatistics department.
Results: In a cohort of 135 patients, the median age at BM diagnosis was 64 years. Most patients were female (n=80, 59%), and the most common histologies were NSCLC (n=56, 41%), breast cancer (n=16, 12%), and melanoma (n=16, 12%). Median overall survival was 20 months (95% CI: 13–23). Patients received SF-SRS (n=66), MF-SRS (n=38), or both fractionation schemes (n=31) for independent lesions. Of their 460 BMs, 277 BMs were treated with SF (median: 21 Gy x 1 fraction) to a median PTV of 1.05 cc, and 183 BMs were treated with MF (median: 27 Gy x 3 fractions) to a median PTV of 1.98 cc (p= <0.001). Lesion-level histology did not differ between fractionation groups (p=0.4). The overall cumulative incidence of RN was 5.2% at 12 months and 9.9% at 24 months. The rate of RN was 6.1% vs 3.8% at 12 months and 11% vs 7.8% at 24 months for SF vs MF, respectively (p=0.03). 25 of the 60 total RN events were symptomatic. Three instances of LF occurred; the cumulative incidence of LF at 6 months after SRS was 0.65%. In subgroup analysis, LF rates were 0.72% for SF vs 0.55% for MF (p=0.8). In multivariable clustered analysis adjusting for histology, KPS, fractionation, and lesion volume, MF was associated with a reduced risk of RN (HR 0.45, CI 0.22-0.92, p=0.03), and lesions greater than the cohort’s median volume of 1.27 cm3 were associated with a higher risk of RN (HR 1.91, CI 1.06-3.44, p=0.03).
Conclusion: We present a large cohort of patients with small BMs treated with SF- or MF-SRS. SRS demonstrated a robust rate of local control; however, RN rates were significantly lower in lesions treated with MF-SRS than in those treated with SF-SRS, even though the MF-SRS lesions were significantly larger. Given that the risk of RN appears higher than that of LF, further consideration must be given to fractionation selection to mitigate toxicity.