Abstract:INTRODUCTION: Immunotherapies provide promising new treatment options for lung cancer and melanoma, however success in glioblastoma (GBM) patients has been much rarer and uneven. Identifying candidates with GBM that are most likely to benefit from immunotherapy remains a challenge. The abundance of tumor-associated macrophages (TAMs) correlates response to immunotherapy in GBM patients. Here we leverage a new imaging technique for quantifying TAMs via tandem stimulated Raman histology (SRH) and two-photon excitation fluorescence microscopy (TPEF) to assess the impact of genetic classification on the abundance of TAMs. METHODS: We conducted tandem SRH and TPEF on specimens from 79 patients undergoing high-grade glioma resection with 5-ALA. We trained a MaskR-CNN deep neural network to identify TAMs and compute their density in each specimen. Corresponding clinical, demographic, genomic and methylomic data were collected for each patient. Independent sample t-tests were used to identify significant correlations between genetic markers, treatment, and TAM density. RESULTS: We analyzed 3,121,524 300 x 300 micron fields of view, from 151 fresh specimens, enabling segmentation of 611,789 TAMs. IDH-mutant tumors had significantly lower TAM density than IDH-wildtype tumors (p < 0.01). The absence of MGMT promoter hypermethylation was associated with significantly higher TAM density (p = 0.03). Tumors with EGFR amplification averaged significantly lower TAM density than EGFR wildtype specimens (p < 0.01). Recurrent tumors possessed 31.5% lower TAM density than primary resections. CONCLUSIONS: Tandem SRH and TPEF can estimate variations in TAM density in human GBMs. We found TAM density to be associated with IDH status, MGMT promoter methylation, and EGFR amplification. This imaging technique creates a new avenue for forecasting response to immunotherapy in GBM patients.

293 The Impact of Tumor Mutational Status on Protoporphyrin IX Accumulation in Tumor-Associated Macrophages in Glioblastoma Patients