OnCo
ideasIdea

Agree in advance how to borrow evidence between similar rare cancers

Statistical methods can combine information across similar rare cancers to reach an answer with fewer patients. Regulators need to say in advance when that is acceptable.

Hierarchical Bayesian models with borrowing across histologies or across related rare diseases can substantially reduce required sample size, and are used in basket trial analyses. Sponsors avoid them because acceptance is uncertain at review. A published position on acceptable borrowing structures, prior specification and pre-registration requirements would let sponsors design smaller trials with confidence.

Hypothesis
Regulatory clarity on borrowing methods reduces the median sample size of rare cancer registrational trials by a third without increasing the rate of subsequent effect reversals in confirmatory data.
Rationale
Similar clarity on adaptive designs and on external control arms rapidly changed practice once guidance existed. The statistical machinery is mature; the missing element is a pre-agreed acceptability boundary.
What would test it
Simulation study across historical rare cancer datasets to quantify type I error and bias under candidate borrowing structures, submitted jointly by academic statisticians and regulators as the basis for guidance.
Maturity
speculative
Who has to act
regulator
Cost to try
Small (under $1M)
Years to first evidence
4
Bottlenecks it attacks

Connected

7top

Pages like this

not linked directly; found by shared links