The development of new oncology drugs is greatly hampered by tumour heterogeneity. Finding novel targets and useful model systems requires an understanding of the spatial tumour landscape. Here, by profiling 40 tissue slices and 80,024 capture sites across a variety of tissue types, sample formats, and RNA capture chemistries, we investigate the usefulness of Spatial Transcriptomics (ST) for oncology discovery. By utilising matched pathology analysis, which gives a ground truth for tissue section composition, we validate the precision and integrity of ST. Then, using spatial data, we show how important tumour depth parameters such as hypoxia, necrosis, vasculature, and extracellular matrix change may be captured. In syngeneic cancer models, we also use spatial context to pinpoint relative cell-type locations that demonstrate the anti-correlation of tumour and immune cells. In clinical pancreatic adenocarcinoma samples, we demonstrate target identification methods and highlight tumour intrinsic indicators and paracrine signalling.
