OnCo
ideasIdea

Shared concurrent control arms across sponsors' trials in the same setting

When five companies each run a trial against the same standard treatment in the same patients, let them pool the standard-treatment patients so fewer people are randomised to the old drug.

Concurrent trials in the same indication each recruit their own control arm on identical standard of care. A pre-agreed common data model, shared eligibility core and a neutral data custodian would let sponsors borrow concurrent (not historical) controls with regulatory pre-agreement, as pioneered by the Alzheimer's Disease EPOCH and the Bayesian borrowing frameworks in paediatric oncology.

Hypothesis
A shared-control network in first-line metastatic urothelial or pancreatic cancer reduces total patients randomised to control by at least 40% across participating trials without inflating type I error beyond pre-specified bounds.
Rationale
Concurrent controls avoid the drift and selection bias that make historical controls unreliable. FDA's complex innovative design programme and EMA have both accepted Bayesian borrowing when exchangeability is demonstrated.
What would test it
Pilot with three sponsors in one indication under FDA's complex innovative trial design meeting programme; simulate operating characteristics, then run and compare control-arm outcomes across sponsors for exchangeability.
Maturity
early clinical
Who has to act
regulator
Cost to try
Medium ($1M to $50M)
Years to first evidence
4
Bottlenecks it attacks

Connected

7top

Pages like this

not linked directly; found by shared links