The cells left behind. The question still ahead.
A tumour cell surviving an experimental exposure is an observation. Showing that it contributes to renewed growth is a different achievement. NEUVIOR’s survivor-cell workstream asks how to connect a defined malignant population with what happens next, rather than treating every surviving cell as the cause of recurrence. We are seeking a research collaboration around that distinction: one model, one identifiable population and a comparison strong enough to change the next experiment.
Begin with the identity of the survivor.
Before asking which cells matter, we would establish what the assay calls a malignant cell and what it calls a survivor. The proposed design would distinguish malignant identity, cell state and continued capacity to divide. A fluorescent label, expression signature or positive viability signal would each answer a different question. The collaboration would document how the chosen identity measurement behaves before exposure, during recovery and at regrowth, including signal loss and changes introduced by the measurement itself.
Published work supplies a specific starting point. Banu and colleagues used genetically engineered murine gliomas with p53 deletion, with or without constitutively active Notch signalling, to investigate cell-state-specific GPX4 dependence. This is an experimental mouse context, not evidence that every human glioblastoma survivor shares the same dependency. Banu et al., 2024
Make contribution measurable.
The proposed comparison is the later contribution of a pre-defined surviving population against an appropriate control population under matched recovery conditions. Depending on the agreed model, readouts could include the fraction of regrowing malignant cells descended from the tracked population, time to renewed growth and the distribution of contributing clones. We would select one primary readout before the experiment, alongside a measurement of whether the tracking system itself changes growth. Absolute cell counts would accompany proportions so that a population can increase in share without being mistaken for an increase in number.
Separate selection from a change of state.
A population becoming more visible after exposure could reflect preferential survival, a transition into a detectable state or the disappearance of other cells. The proposed work would examine those competing explanations rather than choosing the most appealing one. Baseline sampling, matched untreated recovery and repeated observations would help interpret the sequence. An endpoint image alone would be insufficient for this question. If the chosen assay cannot distinguish persistent identity from a newly acquired state, that limitation would determine the next method before any mechanism claim is made.
Challenge the explanation with a second comparison.
Where ethically and technically justified, a second stage would ask whether selectively changing the tracked population alters regrowth while leaving relevant comparison populations measurable. The partner would help choose a perturbation with an interpretable specificity profile and an orthogonal way to examine its effect. A fall in total growth would not automatically establish that the tracked cells caused it. General injury, altered support from neighbouring cells and changed starting numbers would remain competing accounts that the experiment needs to address.
A negative result can redirect the programme.
We would agree in advance what result would justify further work. Reproducible contribution from a defined population could support a more demanding causal experiment; unstable labels or inconsistent contribution could redirect the programme towards a different population or model. The study would report uncertainty across independent biological preparations rather than treating many images from one preparation as independent confirmation. Sample size, exclusions and the minimum effect worth pursuing would be agreed with the scientific partner before the data are examined.
A collaboration with a defined output.
The proposed first output is a jointly scoped survivor-contribution protocol and a feasibility report, supported by an annotated account of the model, identity measurements and competing interpretations. A partner with lineage-tracing, glioma-model or quantitative imaging expertise could help determine whether the experiment is technically credible before a larger study is commissioned. Access to materials, analysis responsibility, publication rights and the treatment of an unhelpful result would be agreed at the outset. NEUVIOR is describing an exploratory research direction; these are proposed investigations rather than results we have generated.
