Where cells sit and how tumours move is reshaping cancer and immune research

How a tumour moves

In a preprint, researchers estimated a glioblastoma migration-to-proliferation ratio, written as D over k, from single time-point MRI data, and the measure turned out to carry prognostic weight. Elevated D over k was significantly associated with reduced overall survival, with p equal to 0.01. This suggests that how fast tumour cells spread relative to how fast they divide matters for outcomes, not only the size or presence of a tumour at one moment. The same preprint reports that mesenchymal-like glioblastoma phenotypes exhibit higher migration-to-proliferation ratios, linking a cell state to a more movement-weighted pattern. m:…3133

The preprint also points to two routes for reading that ratio without repeated measurement. A radiomic factor derived from T2 and FLAIR smoothness of non-enhancing tumour regions correlates with D over k, with r equal to 0.57. In parallel, glycosylation-related pathways are enriched in genes associated with D over k variability, with p equal to 0.0123. Together they suggest that imaging patterns and gene pathways may both carry information about movement, so a single imaging time point may hold more information about tumour behaviour than a static picture implies. m:…3133

A liver barrier that can fail

Location matters just as directly in breast cancer spread to the liver, where a preprint describes liver-resident type 1 innate lymphoid cells, known as ILC1s, as an essential immunosurveillance barrier against liver metastasis. The barrier is not fixed. Heterozygous copy number deletion at TNFRSF10B, which encodes DR5, reduces DR5 expression in human tumours and predisposes to liver metastasis. This suggests metastasis risk may depend on the interaction between tumour genetics and immune cells resident in a specific organ, rather than on tumour aggressiveness in general. m:…9111

The same preprint traces a second route to barrier failure through the liver itself. Hepatosteatosis reduces IL-15 expression in macrophages, leading to depletion of hepatic ILC1s. That chain connects a liver condition to the loss of the very cells described as a barrier, which helps explain why the same cancer might reach the liver in one setting and not another. The preprint also reports that GLP-1 receptor agonists can reverse hepatosteatosis-induced depletion of hepatic ILC1s. That reversal does not by itself show a change in metastasis outcomes, but it gives researchers a concrete mechanism to test. m:…9111

One cell type across three tissues

A preprint on systemic sclerosis extends the location theme from single organs to shared biology across tissues. Its multi-organ single-cell atlas reports a shared myofibroblast subpopulation expressing SFRP2, COL1A1, POSTN, ACTA2 and COMP across lung, skin and synovium in SSc patients. The presence of the same subpopulation in three tissues suggests that scarring and tissue change in this disease may follow a common cellular thread, even when symptoms appear in different parts of the body. Findings in one tissue may therefore be relevant to another. m:…2182

The atlas also reports shared signalling alongside tissue-specific patterns. Interferon signalling is a recurrent pathogenic signature across skin, synovium and peripheral blood mononuclear cells in SSc. Perivascular and mitotic fibroblasts in SSc exhibit both profibrotic and inflammatory signatures, including interferon- and IL-2-related signalling. Together, these findings suggest a common inflammatory and scar-forming programme running through several tissues, alongside tissue-specific features. m:…2182

Growth control beyond one mutation

A preprint on NSCLC cells offers a different kind of location lesson: the relevant target may not be the mutation most often used to sort patients. SF3b inhibitors, namely pladienolide B and herboxidiene, suppress proliferation of NSCLC cells regardless of EGFR mutation status. That result suggests that grouping patients by EGFR status alone may miss a broader sensitivity. The preprint also reports that EGFR intron 2 retention is a marker of SF3b target engagement in NSCLC cells, giving researchers a measurable sign that the inhibitors are acting on their target. A marker of engagement is useful precisely because it does not depend on assuming which downstream effect matters most. m:…0424

The preprint adds a caution against a simple explanation for the effect. Growth inhibition by SF3b inhibitors in NSCLC is not solely dependent on EGFR protein loss. That finding suggests that the inhibitors act through more than one route, and that tracking EGFR protein alone would understate their activity. The implication is to watch for additional mechanisms rather than treating one protein change as the whole story. m:…0424

Mutation patterns shift with counting

The final preprint in this set is a reminder that where a finding sits depends on how it is counted. In ESCC, mutational signature associations are sensitive to compositional representation, meaning that absolute and proportional views of the same data can give different associations. Researchers comparing studies therefore need to know which representation was used, since method choices are part of the finding. m:…5793

Within that caution, two associations are reported. SBS18 is consistently inversely associated with APOBEC-associated mutagenesis in ESCC, and ZNF750-mutant ESCC tumours exhibit higher APOBEC burden compared with wild-type tumours. The first finding held consistently across the representations examined, which gives it more weight than an association seen in only one view. The second links a specific tumour genotype to a higher mutation burden. Taken together, the preprint suggests a field in which genuine biological patterns exist, but in which counting choices can obscure them. The work ahead is to report representations plainly so that patterns can be compared across studies. m:…5793