Impact of Single and Dual Immune Checkpoint Blockade on Risk of Radiation Necrosis among Patients with Brain Metastases Treated with Stereotactic Radiosurgery

Abstract

<h3>Purpose/Objective(s)</h3> While stereotactic radiosurgery (SRS) is often an efficacious treatment for brain metastases, it carries a significant risk of radionecrosis (RN). Single and dual immune checkpoint inhibition (ICPI) have emerged as common treatment options for many patients, particularly those with melanoma and non-small cell lung cancer (NSCLC). Frequently, patients with brain metastases treated with SRS will also receive ICPI. While data suggest a cancer control benefit of combining SRS and ICPI, we hypothesized that concurrent receipt of dual ICPI with SRS increases the risk for RN. <h3>Materials/Methods</h3> We performed a retrospective review of serial patients with metastatic melanoma or NSCLC treated with SRS for intact brain metastases from 2014-2020 at our single institution. Patients were stratified by receipt of dual vs. single ICPI vs. SRS alone. RN was biopsy confirmed or determined radiographically, in combination with clinical assessment and steroid use. Kaplan-Meier estimates were used to compare rates of RN between cohorts. <h3>Results</h3> 673 brain lesions from 93 patients met inclusion criteria [median (Q1, Q3): 5.0 (2.0-10.0) lesions per patient]. Median follow-up of lesions was 8.1 months (95% CI: 7.3, 8.7). Median age and Karnofsky Performance Status of patients were 64.5 years and 90, respectively. Most (82.8%) lesions were supratentorial and histologies included melanoma (53.5%), adenocarcinoma NSCLC (27.3%), squamous cell NSCLC (6.1%), and NSCLC NOS (6.1%). In the entire cohort, 88 lesions from 25 patients (27%) developed RN. 77 (87%) lesions were diagnosed clinico-radiographically and 11 (13%) were biopsy-proven. ICPI use was highly enriched among lesions that developed RN (85.2%) versus those that did not (19.8%). Freedom from RN at 6 months was 80% for dual ICPI, 82% for single ICPI, and 97% for SRS alone; 12-month rates were 78% in each of the ICPI cohorts and 95% with SRS alone (P=0.0002). Rates of symptomatic RN, as measured by frequency of neurologic complications and hospitalization, were not significantly increased with the addition of ICPI to SRS. <h3>Conclusion</h3> In a large cohort of SRS-treated brain metastases, we observed an increased risk of RN among patients who received either dual or single ICPI concurrently with SRS. These results may guide clinical decision-making when performing radiographic surveillance of brain metastases following SRS. Prospective studies to validate these findings are warranted.