ideasIdea
Bank yearly blood from cancer survivors so future tests can be validated
To prove a leftover-cancer test works you need blood taken years before relapse. Collecting and freezing yearly samples now makes every future test testable.
Every new MRD or early detection assay needs pre-diagnostic or pre-relapse serial specimens, which take years to accumulate and cannot be created retrospectively. A prospective survivor cohort of 50,000 people banking annual plasma, with registry-linked outcomes and open access rules, would become the shared validation substrate for the whole field.
Hypothesis
A serial survivor plasma bank with linked outcomes shortens the validation cycle for a new MRD assay from five years to under one, and is used by at least ten independent developers within five years of opening.
Rationale
The UK Biobank and PLCO-style cohorts demonstrate that pre-event specimen collections become disproportionately valuable over time and are impossible to replicate quickly. Survivors are highly motivated donors and are already in follow-up pathways.
What would test it
Pilot in 2,000 survivors at high recurrence risk with annual draws for three years; measure retention, cost per sample-year, and number of external validation studies enabled.
Maturity
speculative
Who has to act
data
Cost to try
Medium ($1M to $50M)
Years to first evidence
8
Bottlenecks it attacks
- Dormant cells and minimal residual disease · After a 'successful' treatment, cells can sleep for years then relapse. We can barely detect them and cannot target them.
- Data silos · Records, scans, genomes and outcomes sit in separate systems that cannot talk. Every patient's experience is lost to the next.
- Most lethal cancers are found late · Screening exists for only a few cancers. Pancreatic, ovarian, liver, oesophageal and most lung cancers are found when cure is unlikely.