GeoMx and RNAscope : A Comparative Assessment of Their Utility for Spatial mRNA Expression Profiling in Formalin‐Fixed Breast Cancer Tissue

ABSTRACT mRNA expression analysis in formalin‐fixed tissue is essential for many biomarker and tumor microenvironment studies. NanoString GeoMx Digital Spatial Profiler is a recent technique that offers high‐plex spatial transcriptomics (up to 18,000 genes), while RNAscope (1–4 plex) is a well‐established mRNA in situ hybridization method. This study compares quantitative expression estimates obtained by GeoMX and RNAscope. Serial sections from two TMAs containing mammary cancers, including triple‐negative ( n = 48) and varying ER/HER2 status ( n = 45), were analyzed using GeoMx Cancer Transcriptomics Atlas (CTA) covering approximately 1,800 genes and RNAscope probes for GATA3, SOX10, and PD‐L1 mRNAs. Expression was quantified as counts/cell (GeoMx, geometric mean of five probes per gene) and average dots/cell (RNAscope, QuPath). Positivity was defined as above the limit of quantification (GeoMx) or > 0.1 dots/cell (RNAscope). High correlation was observed for GATA3 ( R = 0.87) and SOX10 ( R = 0.77) between methods. PD‐L1 expression was high in only one core, precluding correlation analysis. RNAscope demonstrated higher sensitivity and a broader dynamic range than GeoMx. In conclusion, GeoMx CTA and RNAscope exhibit a strong correlation. GeoMx enabled highly multiplexed gene expression analysis, whereas RNAscope provided better sensitivity and single‐cell resolution. The choice of method should be guided by study objectives.