Using an Aurora prognostic signature to concentrate CV-death events and shrink outcomes trials. UK Biobank (N ≈ 500,000)

The Enrichment Problem

Cardiovascular outcomes trials are powered on events, not patients. When the per-patient event rate is low, sponsors must enrol large cohorts and wait years to accrue enough endpoints. Prognostic enrichment—enrolling patients more likely to have an event—is the standard lever, and the FDA explicitly endorses it. Its value depends entirely on how cleanly the marker separates high- from low-risk patients.

Aurora Concentrates the Events That Matter

 

Stratifying heart failure patients by the Aurora proteomic signature places roughly 52% of the high-risk tertile into a CV-death event over 15 years, against about 6% in the low-risk tertile (from approximately 48% versus 94% survival; HR 3.50 per SD, 95% CI 3.18–3.85, log-rank p = 1.3e-86). A single canonical marker, NT-proBNP, gives only HR 1.77 and far weaker separation, so a panel is what makes enrichment worthwhile.

What That Does to Trial Size (Illustrative)

 

Because an event-driven trial needs a fixed number of endpoints, required enrolment scales inversely with the event rate. Enrolling the Aurora high-risk tertile (approximately 52% event rate) instead of an unselected HF population sharply reduces the patients needed for the same statistical power.

Population enrolled	Assumed 15-year event rate	Patients for 100 events
Unselected HF (conservative)	15%	667
Unselected HF (mid)	20%	500
Unselected HF (higher)	25%	400
Aurora high-risk tertile	~52%	192

Illustrative only: assumes event-driven powering and the labelled all-comers event rates; the enriched rate (~52%) is the observed high-risk tertile. Against a 20% all-comers assumption, enrichment cuts required enrolment by roughly 60% (500 to 192 patients).

 

Why It Matters for Your Programme

Fewer patients, fewer sites, and a faster readout translate enrichment directly into trial cost and timeline. Aurora discovers and validates these signatures on population-scale data, then ports them to a new indication by editing the cohort definition, so the same enrichment approach extends across cardiometabolic, renal, and other outcome-driven programmes. Cohorts are auditable by design, supporting a defensible enrichment strategy in regulatory interactions.