1093 - Maximum Tolerated Dose of Neoadjuvant Stereotactic Radiosurgery and Post-Treatment Immune Profiling for Metastatic Brain Tumors: Results of a Phase I Dose Escalation Trial

S. C. Zhang¹, E. J. Yoshida², M. A. Diniz³, C. S. Patil⁴, J. S. Yu⁴, R. M. Chu⁴, J. K. Jang¹, K. Anderson¹, B. Hakimian¹, A. J. Mirhadi¹, J. L. Hu⁵, J. Rudnick⁵, K. M. Atkins¹, and S. L. Shiao¹; ¹Department of Radiation Oncology, Cedars-Sinai Medical Center, Los Angeles, CA, ²Department of Radiation Oncology, University of California San Francisco,, San Francisco, CA, ³Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai, New York City, NY, ⁴Department of Neurosurgery, Cedars-Sinai Medical Center, Los Angeles, CA, ⁵Department of Neurology, Cedars-Sinai Medical Center, Los Angeles, CA

Purpose/Objective(s): Neoadjuvant stereotactic radiosurgery (SRS) for brain metastases can be safe and effective. However, the maximum tolerated dose (MTD) in this setting is unknown. Prior neoadjuvant series reported de-escalated SRS doses relative to definitive (i.e., RTOG 9005) doses, possibly due to concerns for combined toxicity. We hypothesized that definitive SRS doses could be used in the neoadjuvant setting without significant toxicity or post-operative complications. Materials/

Methods: This is a single institution phase I trial. Patients with at least one brain metastasis =4 cm in diameter planned for resection and up to 4 brain metastases total were enrolled. We treated a single lesion with single fraction SRS prior to resection. Initial doses were 18Gy for tumors =2cm, 14Gy for tumors 2.1-3cm, and 13Gy for tumors 3.1-4cm. Dose was escalated using the Bayesian optimized adaptive Escalation with Overdose Control design with an upper bound of 24Gy. Following resection, MRI was obtained every 3 months. Dose limiting toxicities (DLT) were defined as significant CNS events (e.g., =g2 stroke, =g3 seizure) within 30 days of SRS. The primary endpoint was MTD. Secondary endpoints included local control, distant brain control, and overall survival. Exploratory endpoints included changes in T-cell and myeloid cell populations in peripheral blood and in the resected irradiated tumor prior to and after SRS analyzed with flow cytometry (FC).
Results: 25 patients were enrolled. Median follow up was 12.3 months (interquartile range 2.5-27.7 months). Median tumor size was 2.2cm (range 1.2-4.0cm). 9 patients were enrolled onto the =2cm dose strata, 10 onto the 2.1-3cm strata, and 6 onto the 3.1-4cm strata. Median doses by stratum were 23Gy (max 24Gy), 19Gy (max 23Gy), and 16.5Gy (max 18Gy), respectively. Overall cohort median dose was 20Gy (range 13-24Gy). No DLTs or significant post-operative complications occurred. SRS was generally well tolerated with only one grade 3 CNS event unrelated to treatment. 15/25 (60%) patients died. Overall survival at 1 and 2 years was 54.2% and 37.5%, respectively. Local control at 1 and 2 years was 93.3% and 84.9%, respectively. Univariable Cox regression did not show a significant association between dose and local control (p=0.68). Distant brain control at 1 and 2 years was 68.1% and 61.3%, respectively. One patient developed leptomeningeal disease and 2 patients developed radiation necrosis at 7.5 and 11.9 months post resection. FC results on peripheral blood showed alterations in immune cell populations after SRS and resection, and we present FC characterization of the immune cells in the radiated tumors.
Conclusion: Neoadjuvant SRS was safely escalated up to 24Gy for tumors =2cm, 23Gy for tumors 2.1-3cm, and 18Gy for tumors 3.1-4cm. These findings provide rationale for use of definitive or higher doses even in metastases planned for resection. We did not find evidence of a dose response relationship with local control though our analysis is limited by a low number of recurrences.