Keywords:
CNS, MR, Radiation therapy / Oncology, Cancer
Authors:
M. D. Shanker, H. Foley, S. Crowley, E. Thomson, C. Bradhurst, M. Huo, V. Atkinson, M. Foote, M. Pinkham; Woolloongabba, QLD/AU
DOI:
10.26044/ranzcr2021/R-0020
Results
101 patients with 425 MBM were treated with SRS in the study period. Median follow-up was 29.2 months (IQR 19.7-39.8). Median dose was 20Gy (IQR 18-20). Median baseline volume and lesion diameter were 0.24cc (IQR 0.06-1.02) and 7.7mm (IQR 4.8-12.4) respectively. 53% of patients were BRAF mutant with 65% having failed BRAF inhibitors at time of SRS. 77% of patients received concurrent immunotherapy (2 weeks prior to 4 weeks post-SRS). Normalised to baseline, median interpolated lesion size at 3, 6, 9, 12- and 15-month was 73.6%, 55.4%, 49.7%, 54.2% and 42.5% respectively (Figure 1). There was statistically significant decrease in lesion size in the 0-3 and 3-6 month interval (p<0.0001). 89.7% of treated lesions had durable local control on MRI at last follow-up. Graphical representation of volumetric change stratified by variable are presented in Figures 2-4.
5% of patients experienced symptomatic radionecrosis. Every 1mm increase in maximal lesion diameter was associated with a 1.5% (95%CI 0.04-2.9, p=0.04) and 2% (95%CI 0.4-3.7%, p=0.016) greater reduction in volume at 3- and 6-months. Patients receiving concurrent immunotherapy had a significantly greater regression in tumor volume at 3-months (37% superior [95%CI 6.0-68.1%, p=0.02] and 6-months (48% superior [95%CI 7.4-89.5%, p=0.02] compared to those commencing >4 weeks post SRS however by 9-months and 12-months the velocity of regression was comparable.
Volumetric increase in size of lesions on initial 3 month MRI remained significantly associated with progression when adjusting for age and immunotherapy timing status. On multivariate analysis, patients demonstrating any increase in volume on initial post-SRS imaging were significantly less likely to achieve an ultimate CR or PR. Univariate and multivariate cox proportionate hazard regression analysis for volume change stratified by treatment response outcomes are presented in Table 1. Comparing changes to baseline imaging, on multivariate analysis 3-month MRI imaging was predictive of treatment response (complete or partial response) [HR=2.611, 95% CI 1.235-5.52, p=0.012] whereas 6-, 9- and 12- month MRI imaging was not predictive of ultimate best response to treatment. Waterfall plots of 3 month complete, partial response and progression of disease are represented in Figures 5-7.