Outcomes After Surgical Resection of Melanoma Brain Metastases in the Age of Checkpoint Inhibitor Treatment. (16th November 2020)

• Record Type: Journal Article
• Authors: Morshed, Ramin; Morshed, Ramin A; Chung, Jason; Sudhakar, Vivek; Cummins, Daniel; Young, Jacob S; Daras, Mariza; Aghi, Manish K
• Is Part Of: Neurosurgery. Volume 67(2010)Supplement 1
• Publication Date: 2020-11-16
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1093/neuros/nyaa447_910
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract: INTRODUCTION: Metastasis of melanoma to the brain is associated with poor outcomes. Recent trials demonstrate improved survival after treatment with immune checkpoint inhibitors. METHODS: This retrospective, single-center study included patients undergoing first-time surgery for melanoma brain metastases. A multivariate Cox proportional model was used to estimate the association of patient and treatment factors with OS and CNS progression. Kaplan-Meier curves were used to plot survival and progression time observations. RESULTS: 85 patients (mean age 60) underwent first-time resection of 97 melanoma brain metastases with a median follow-up of 9.5 months. Checkpoint inhibitors (Pembrolizumab, Ipilimumab, and/or Nivolumab) were used in 55.1% of cases (19 pre-op; 47 post-op; median 9 cycles). Patients treated with checkpoint inhibitors had similar peri-op systemic disease status and KPS but had been treated with more systemic agents and had more instances of CNS progression prior to surgery. Median OS and time to CNS progression for the cohort were 1 year and 237 days, respectively. In a multivariate Cox regression model, age (HR 1.03 by decade; P = .02), treatment with a checkpoint inhibitor (HR 0.27; P < .0001), prior radiotherapy (HR 2.44; P = .007), and number of brain metastases at the time of surgery (HR 1.05 per metastasis; P = .04) were significant predictors of OS. Checkpoint inhibitor treatment was associated with longer OS from surgery (median 3 vs 0.5 yrs, log-rank P = .004). However, patients who underwent craniotomy after prior checkpoint inhibitor treatment had poor OS (median 0.56 yrs). Prior radiotherapy was associated with poor OS (median 0.53 yrs). A threshold of 10 brain metastases differentiated OS in the subgroup of checkpoint inhibitor treated patients ( P = .03). Although extent of resection predicted OS in the entire cohort ( P = .02), it did not impact OS in the subgroup of checkpoint inhibitor treated patients (p = 0.57). CONCLUSION: While checkpoint inhibitor treatment was associated with improved survival in this surgical cohort of melanoma brain metastases, patients who require resection after checkpoint inhibitor treatment or radiotherapy are poor surgical candidates.