Correlation of fluorescence optomics method classification performance to varying expression level of epidermal growth factor receptor

Fluorescence molecular imaging using ABY-029, an epidermal growth factor receptor (EGFR)-targeted synthetic Affibody peptide labeled with a near-infrared fluorophore, is under investigation for surgical guidance during head and neck squamous cell carcinoma (HNSCC) resection. However, tumor-to-normal tissue contrast is confounded by intrinsic physiological limitations of heterogeneous EGFR expression. In this study, a machine learning-based optomics analysis, which interprets the textural pattern differences in EGFR expression conveyed by fluorescence, was applied to optical ABY-029 fluorescence image data of HNSCC surgical specimens. The study objective was to determine the correlations between optomics method classification performance and tissue inherent EGFR expression level. Fluorescence image data were collected through a Phase 0 clinical trial of ABY-029, which involved a total of 20,073 sub-image patches (size of 1.8x1.8 mm(2)) extracted from 24 bread-loafed slices of HNSCC surgical resections from 12 patients who were stratified into three dose groups (30, 90, and 171 nanomoles). The optomics approach utilized a supervised machine learning pipeline. Each dose group was randomly partitioned on the specimen-level 75%/25% into training/testing sets, then all training and testing sets were aggregated. A total of 1,472 standardized optomic features were extracted from each patch and evaluated by minimum redundancy maximum relevance feature selection, and 25 top-ranked features were used to train a support vector machine classifier. A conceptual framework of correlation analysis to evaluate the relationship between optomics tumor classification performance and underlying EGFR expression level was provided, but the present results are underpowered. Some generalized conclusions about the ABY-029 fluorescence optomics method correlating to varied levels of EGFR expression were summarized, suggesting that optomics method using fluorescence molecular imaging data offers a potentially stable image analysis technique for cancer detection for fluorescence-guided surgery applications; however, further study with additional samples is needed to validate this conclusion.