2218 - The Effect of Non-Enhancing Peri-Tumoral Hemorrhage on Local Control for Melanoma Brain Metastases Treated with Stereotactic Radiosurgery

Purpose/Objective(s): Hemorrhagic brain metastases are associated with higher rates of local recurrence following stereotactic radiosurgery (SRS). Here, we investigate if peri-tumoral hemorrhage (PTH), as defined on susceptibility weighted angiography (SWAN) MR sequence, represents an area at risk by examining the correlation between its volume and marginal recurrence following Gamma Knife SRS for patients with hemorrhagic melanoma brain metastases.

Materials/Methods: Our institutional database was queried for consecutively treated patients with brain metastases of melanoma primary spanning 2014-2025. Further inclusion/exclusion criteria then applied were presence of SWAN MRI sequence included in treatment-planning MRI, no surgical resection, presence of follow-up MRI imaging >2 months beyond date of SRS completion, and the presence of peri-tumoral hemorrhage beyond contrast-enhancing tumor. PTH volume was contoured as defined on SWAN sequence, and multiple dosimetric indices were recorded including tumor minimal dose (D_min), PTH D_min and mean dose (D_mean), and PTH volume. Marginal recurrence was defined as progressive nodular enhancement centered away from the treated enhancing tumor, but within 1cm radius. Cox proportional hazards analysis was done to analyze potential risk factors.

Results: 900 lesions in 146 patients were screened, yielding a total of 27 lesions in 12 patients treated over 18 courses of SRS, which met our inclusion criteria. All lesions were treated in single-stage SRS aside from two cases, one with two-fraction treatment separated by ~1 week, and one with staged two-fraction treatment ~1 month apart. The median age was 65 at the time of SRS. The median prescription dose was 18 Gy (range 18-20 Gy) to a median isodose line of 65% (range 48% - 90%). The mean enhancing disease volume was 2.05cc, while the mean PTH volume was 4.34 cc. The median coverage was 100%, the mean D_min to enhancing disease was 17.2 Gy, and the mean D_min to the PTH volume was 6.8 Gy. Eight lesions experienced marginal recurrence (29.6%) at a median interval of 4 months. PTH volume was significantly associated with the marginal recurrence (HR 1.165; P=0.0196) while the enhancing disease size showed no significant correlation, with a trend towards lower risk (HR 0.221; P=0.078), and neither did D_min to enhancing disease or PTH. Further details are in the table.

Conclusion: In our cohort, we found a high rate of marginal recurrence with significant correlation between PTH volume independent of the enhancing disease size or its D_min which is suggestive that it is an area at risk prompting further investigation. The low D_min on average to PTH could explain the lack of correlation between D_min to PTH and marginal recurrence.