Fig. 6: Monitoring the response to treatment in PDX models. | Nature Communications

Fig. 6: Monitoring the response to treatment in PDX models.

From: Spatio-temporal transcriptomics of chromothriptic SHH-medulloblastoma identifies multiple genetic clones that resist treatment and drive relapse

Fig. 6

A Mapping of genetic clones as identified from the patient tumour (c.f. Fig. 4B) to spatial transcriptomics locations (n = 1147 spots) from xenografted mouse brain tissue, which identifies both major clones in the xenografts. More progenitor-like cells, more cycling cells and more pro-tumour macrophages are within or close to clone B Spatial segmentation of a relapsed LFS medulloblastoma sample (n = 3983 spots). C UMAPs of an integrated dataset across all 11 PDX samples, across time points and treatment groups (n = 9989 spots in total). Left: Colour-coded by treatment group. Middle: Colour-coded by time point. Right: Colour-coded by the most abundant cell type as inferred by cell2location, where cell types are defined by tissue regions of C. D Projections of the tissue region signatures from the original patient tumour on the combined PDX dataset. Violin plots show the evolution over the course of the treatment of cell types from all regions of the original patient tumour. First, second, and third quartiles are indicated. Statistical significance was assessed by a regression t-test on the medians (n = 11 samples, control, n = 1193 spots; early effect, n = 3391 spots; MRD, n = 547 spots; regrown, n = 4858, spots). E Representative samples across the time course coloured by the estimated dominant cell type at each Visium location (PDX B, n = 1166 spots; PDX C2, n = 1610 spots; PDX D1, n = 416 spots, PDX E3, n = 1741 spots).

Back to article page