Volume of Disease as a Predictor for Clinical Outcomes in Patients With Melanoma Brain Metastases Treated With Stereotactic Radiosurgery and Immune Checkpoint Therapy

Document Type

Journal Article

Publication Date



Frontiers in oncology






SRS; brain metastases; immunotherapy; melanoma; radiation necrosis (RN)


Purpose/Objectives: Clinical trials of anti-Programmed cell death protein 1 (PD-1) and cytotoxic T-lymphocyte-associated protein (CTLA-4) therapies have demonstrated a clinical benefit with low rates of neurologic adverse events in patients with melanoma brain metastases (MBMs). While the combined effect of these immunotherapies (ITs) and stereotactic radiosurgery (SRS) has yielded impressive results with regard to local control (LC) and overall survival (OS), it has also been associated with increased rates of radiation necrosis (RN) compared to historical series of SRS alone. We retrospectively reviewed patients treated with IT in combination with SRS to report on predictors of clinical outcomes. Materials and Methods: Patients were included if they had MBMs treated with SRS within 1 year of receiving anti-PD-1 and/or CTLA-4 therapy. Clinical outcomes including OS, LC, intracranial death (ID), and RN were correlated with type and timing of IT with SRS, radiation dose, total volume, and size and number of lesions treated. Results: Twenty-nine patients with 171 MBMs were treated between May 2012 and May 2018. Patients had a median of 5 lesions treated (median volume of 6.5 cm) over a median of 2 courses of SRS. The median dose was 21 Gy. Most patients were treated with ipilimumab (n = 13) or nivolumab-ipilimumab (n = 10). Most patients underwent SRS concurrently or within 3 months of receiving immunotherapy (n = 21). Two-year OS and LC were 54.4% and 85.5%, respectively. In addition, 14% of patients developed RN; however, only 4.7% of the total treated lesions developed RN. The median time to development of RN was 9.5 months. Patients with an aggregate tumor volume >6.5 cm were found to be at increased risk of ID (p = 0.05) and RN (p = 0.03). There was no difference in OS, ID, or RN with regard to type of IT, timing of SRS and IT, number of SRS courses, SRS dose, or number of cumulative lesions treated. Conclusions: In our series, patients treated with SRS and IT for MBMs had excellent rates of OS and LC; however, patients with an aggregate tumor volume >6.5 cm were found to be at increased risk of ID and RN. Given the efficacy of combined anti-PD-1/CTLA-4 therapy for MBM management, further study of optimal selection criteria for the addition of SRS is warranted.


Neurological Surgery