OnCo
ideasIdea

Match therapy to the type of scar-forming cell in the tumour

The support cells that build a tumour's scaffolding come in several types: some protect the tumour, others restrain it. Treating all of them the same way explains past failures.

Attempts to strip tumour stroma wholesale, notably hyaluronidase in pancreatic cancer and hedgehog inhibition, failed or worsened outcomes, consistent with the finding that some fibroblast subsets restrain tumour growth. Human single-cell and spatial atlases now define myofibroblastic, inflammatory and antigen-presenting fibroblast states. A classifier that assigns patients to stroma-directed treatment by subtype, rather than by disease, is the logical next step.

Hypothesis
A validated fibroblast-subtype classifier predicts which patients benefit from stroma-directed therapy, and pooling all subtypes together dilutes an effect that exists only in the myofibroblastic-dominant subgroup.
Rationale
Hedgehog inhibition accelerated pancreatic cancer in mice by depleting restraining stroma, a clear demonstration that stromal identity determines direction of effect. Subtype-guided assignment turned negative results into positive ones in breast cancer and in colorectal consensus subtypes.
What would test it
Reanalyse tissue from failed stroma-targeting trials by fibroblast subtype for a retrospective interaction test; if positive, run a prospective subtype-selected trial.
Maturity
speculative
Who has to act
research
Cost to try
Medium ($1M to $50M)
Years to first evidence
7
Bottlenecks it attacks

Connected

12top

Pages like this

not linked directly; found by shared links