2521 - Clinicogenomic Predictors of Survival and Intracranial Progression after Stereotactic Radiosurgery for Colorectal Cancer Brain Metastases

C. Gui¹, H. Walch², K. Mueller¹, L. A. Boe², A. T. Ilica³, J. Strong¹, J. Eichholz², K. Yu⁴, J. A. Wilcox⁵, P. Manca⁵, Y. Yu¹, Y. Yamada¹, B. S. Imber¹, S. Maron², M. B. Foote⁵, R. Yaeger⁵, N. Schultz², and L. R. G. Pike¹; ¹Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, ²Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, ³Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, ⁴Department of Neurological Surgery, Memorial Sloan Kettering Cancer Center, New York, NY, ⁵Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY

Purpose/Objective(s): Brain metastases (BM) from colorectal cancer (CRC) are associated with dismal prognosis. When considering BM-directed therapy, better methods are needed to identify patients at risk of poor oncologic outcomes, to optimize patient selection for closer surveillance or escalated therapy. In other cancer types, genomic features have been shown to predispose patients to specific patterns of intracranial progression. We sought to identify clinicogenomic predictors of survival and intracranial disease progression in patients with CRC BM treated with stereotactic radiosurgery (SRS). Materials/

Methods: Patients with newly diagnosed CRC BM who were treated with SRS between 2009 and 2022 and had next-generation sequencing of a selected panel of > 300 genes (MSK-IMPACT) were included. Frameless SRS was delivered in 1-5 fractions, alone or after resection. Outcomes were overall survival (OS) and intracranial progression (IP), assessed per patient, and local progression (LP), assessed per BM treated with SRS. IP was defined as diagnosis of a new BM, progression of a treated BM, or leptomeningeal progression. Associations between baseline clinical and genomic features and outcomes were evaluated with Cox regression and competing risk regression, with death as a competing risk.

Results: This study included 123 patients with 299 BM. At BM diagnosis, 111 patients (91%) had progressive extracranial disease, and 79 (64%) had = 3 extracranial metastatic sites. The median number of BM was 2 per patient (IQR: 1-3). The median prescribed biologically effective dose (BED) was 51.3 Gy (IQR: 51.3-65.1), corresponding to a prescription of 27 Gy in 3 fractions. OS, IP, and LP estimates at 1 year after SRS were 36%, 55%, and 12%, respectively. In a multivariable Cox regression model, worse OS was associated with progressive extracranial disease (HR 4.47, 95% CI 1.70-11.8, p = 0.002) and = 3 extracranial metastatic sites (HR 1.94, 95% CI 1.20-3.15, p = 0.007). In a multivariable competing risk regression model, lower risk of LP was associated with BED = 51.3 Gy (HR 0.27, 95% CI 0.08-0.86, p = 0.027), independent of BM diameter (HR 1.24/cm, 95% CI 0.83-1.88, p = 0.3). In a multivariable competing risk regression model, IP was independently associated with genomic alterations; TP53 driver alterations predicted higher risk of IP (HR 2.71, 95% CI 1.26-5.79, p = 0.01), whereas MYC pathway alterations predicted lower risk (HR 0.15, 95% CI 0.03-0.68, p = 0.01).

Conclusion: To our knowledge, this is the first study to incorporate genomic features in modeling of outcomes after local therapy for CRC BM. Advanced extracranial disease predicted worse survival. Insufficient SRS doses predicted local failure. TP53 and the MYC pathway alterations were associated with IP, with opposite effects. Patients at high risk of IP may be considered for more frequent surveillance or escalated therapy.