MMDA-16 ASSESSING THE RELATIONSHIP OF NON-HORMONAL SYSTEMIC THERAPY TIMING AND OUTCOMES IN TRIMODAL BRAIN METASTASIS TREATMENT: A RETROSPECTIVE, MULTI-FACILITY HEALTH SYSTEM ANALYSIS

Abstract

Inter-modality timing influences complication, progression, and survival rates in brain metastasis treatment. However, little evidence exists regarding appropriate timing for chemotherapy initiation following brain metastasis resection and stereotactic radiotherapy (SRS), with only one prior study examining the topic indirectly through secondary analysis. This study examines the relationship between outcomes and postoperative non-hormonal systemic therapy (NHST) timing. Patients undergoing brain metastasis resection followed by SRS and NHST within 120 days postoperatively were retrospectively identified. Individuals were grouped according to SRS-to-NHST timing: NHST-initiation prior to SRS (rapid initiation), initiation ≤30 days after SRS (early post-SRS), and initiation >30 days after SRS (late post-SRS). Local and distant intracranial progression-free survival (PFS), extracranial PFS, global PFS, and overall survival (OS) were assessed using Kaplan-Meier, cumulative incidence, and Cox proportional hazards analyses. 193 patients were included: 28, 112, and 53 patients in the rapid initiation, early post-SRS, and late post-SRS groups, respectively. Kaplan-Meier curves demonstrated similar local intracranial PFS (p = 0.086), distant intracranial PFS (p = 0.19), extracranial PFS (p = 0.2), global PFS (p = 0.17), and OS (p = 0.12) across the pan-cancer timing groups. Extracranial disease status at resection was the dominant predictor of outcomes for the pan-cancer multivariable analysis. Similar rates of radiation necrosis, pseudomeningocele, wound dehiscence, and surgical site infection were present across groups. Both the gross total resection (GTR) non-small cell lung cancer (NSCLC) and the melanoma subgroups demonstrated statistical significance on Kaplan-Meier analysis. For both subgroups, rapid initiation demonstrated least favorable outcomes. On multivariable analysis, rapid initiation was associated with worse global PFS and OS for GTR-treated NSCLC (HR 6.64, 95% CI 2.17-20.28, p < 0.001). Selection bias involving the tendency for patients with more aggressive disease to receive earlier NHST after brain metastasis resection likely contributes to these findings. In total, these results highlight the importance of stratifying by histology and resection extent when evaluating treatment sequencing.