2228 - Multi-Institutional Analysis of Dose and Factors Associated with All Cause Morbidity after Stereotactic Radiotherapy for Intact Melanoma Brain Metastases

Purpose/Objective(s): Survival has improved for melanoma brain metastases due to advances in systemic therapy. Patient and dose selection for stereotactic radiotherapy (SRT) should balance benefits of intracranial tumor control with the heightened synergistic toxicity with immune checkpoint inhibitors (ICI) and targeted therapy. We hypothesized that in larger melanoma metastases, higher SRT dose is associated with greater morbidity, inclusive of failure, hemorrhage, and necrosis.

Materials/Methods: Data from two academic institutions were pooled to review patients receiving SRT for intact melanoma brain metastases from 2012-2025. Characteristics for each patient and metastasis were extracted including receipt of ICI and BRAF inhibitors, metastasis size (largest dimension), and dose/fx, which was converted to equivalent dose at 2 Gy/fx using a/b=3 (EQD2). Local failure (LF), grade =2 hemorrhage, and grade =2 radiation necrosis (RN) were combined as a single local endpoint defined “all cause morbidity” given considerable overlap and significant clinical impact of all these events. Cox proportional hazards model analyzed freedom from morbidity, with separate analyses performed for metastases <1 cm vs =1 cm. The hypothesized covariates of size and EQD2 were included in all multivariate analyses, with ICI and BRAF included if trending toward significance.

Results: The final cohort included 89 pts with 302 intact metastases and median 11 mo. f/u after SRT. 77pts (87%) received ICI and 32 (36%) received BRAF inhibitors. Median size was 0.8 cm (range 0.1-5.6 cm) and distribution of <1, 1-2, and =2 cm metastases was 194 (64%), 77 (26%), and 31(10%), respectively. Mean EQD2 was 73 Gy. The most common dose/fx were: 20 Gy / 1 fx (EQD2 92 Gy; n=156), 25 Gy / 5 fx (EQD2 40 Gy; n=58), 30 Gy / 5 fx (EQD2 54 Gy; n=41), and 18 Gy / 1 fx (EQD2 76 Gy; n=25). All-cause morbidity was observed in 72 metastases (24%), with rates of LF, hemorrhage, and RN of 12%, 17%, and 12%, respectively. Mean freedom from morbidity for metastases <1, 1-2, and = 2 cm was 80, 39, and 18 mo., respectively. Factors associated with morbidity are shown in the Table. For metastases <1 cm, size and BRAF-inhibitors were associated with morbidity. For metastases =1 cm, both size and EQD2 were associated with increased morbidity. EQD2 was associated with decreased LF only in the <1cm cohort.

Conclusion: All-cause morbidity after SRT for melanoma brain metastases is greatly influenced by tumor size. For metastases =1cm, higher biologic doses were associated with greater morbidity, suggesting a potential benefit of dose de-intensification for patients with large (and numerous) metastases.