Characterizing the tumor microenvironment (TME) of archived clinical tissues requires reliable gene expression profiling (GEP) of formalin-fixed paraffin-embedded (FFPE) samples. The EdgeSeq Precision Immuno-oncology Panel (PIP) is a targeted GEP assay designed for TME characterization but lacks widespread technical validation on a large cohort of clinical samples. Here, we evaluated its performance by exploring its concordance with multiple orthogonal platforms using 1,220 FFPE samples across various cancer types. Quantitative comparisons with RNA-seq and NanoString showed strong correlations at the sample level (median ρ = 0.73 and 0.81) and moderate correlations at the single-gene level (median ρ = 0.49 and 0.57). Gene signature analysis revealed high concordance with RNA-seq on widely used signatures for TME characterization and immune checkpoint inhibitor (ICI) efficacy prediction, though some genes in these signatures are not targeted by EdgeSeq PIP. From a histopathological viewpoint, the tumor/immune abundances derived from hematoxylin and eosin (H & E) staining were well recapitulated by the transcriptomic profiles assessed by EdgeSeq PIP. Furthermore, the mRNA level of PD-L1 assessed by EdgeSeq PIP was moderately correlated with the PD-L1 score (ρ = 0.65) estimated by immunohistochemistry (IHC); the mRNA level of CD8A aligned well (ρ = 0.55) with the IHC-derived abundance of CD8+ T cells. Overall, our results showed that EdgeSeq PIP generated well-correlated data with independent approaches at mRNA, protein, and histological levels, thus providing strong technical support for further using EdgeSeq PIP in biomarker studies and companion diagnostic (CDx) development.
Keywords: edgeseq PIP; gene expression profile; immunotherapy; platform evaluation; tumor microenvironment.
Copyright © 2022 Shi, Ma, Shen, Liu, Liang, Liu, Shen, Zhang and Zhang.