Early Stopping Saves $18M
PCSK9 inhibitor for major adverse cardiovascular events (MACE) in secondary prevention
$18.1M
Cost Savings (24%)
12 mo
Duration Reduction
$2.4B
NPV from Earlier Approval
Scenario
A pharmaceutical company was conducting a Phase III cardiovascular outcomes trial (CVOT) for a novel PCSK9 inhibitor in patients with established atherosclerotic cardiovascular disease. The primary endpoint was time to first major adverse cardiovascular event (MACE): cardiovascular death, myocardial infarction, or stroke.
The trial was designed with a fixed sample size of 2,400 patients, targeting 430 MACE events over 48 months with an estimated cost of $76.8 million.
The Group Sequential Design
4-Look O'Brien-Fleming Design
Instead of waiting for all 430 events, the trial was designed with 4 interim analyses using O'Brien-Fleming spending function for efficacy and non-binding Hwang-Shih-DeCani boundaries for futility.
Information Fractions
[0.25, 0.50, 0.75, 1.00]
Events at Each Look
[110, 220, 330, 442]
The slightly inflated maximum sample size (442 vs 430 events) is the “cost” of interim monitoring, but the expected sample size under the alternative hypothesis is significantly lower.
Stopping Boundaries
| Look | Events | Efficacy Z | Futility Z | Result |
|---|---|---|---|---|
| 1 | 110 | 3.750 | -1.465 | Continue (Z=1.89) |
| 2 | 220 | 2.540 | -1.155 | Stop for Efficacy (Z=2.89) |
| 3 | 330 | 2.021 | -0.972 | Not reached |
| 4 | 442 | 1.718 | 1.718 | Not reached |
Trial Outcome
At Interim Analysis 2 (Month 36, 220 events), the observed hazard ratio was HR = 0.68 (32% risk reduction), corresponding to a Z-statistic of Z = 2.89.
This exceeded the efficacy boundary of Z = 2.540, and the trial was stopped early for overwhelming efficacy. The remaining 12 months of planned enrollment was avoided.
Observed HR
0.68
32% risk reduction
Z-statistic
2.89
vs boundary 2.540
Design Comparison
| Metric | Fixed Sample | Group Sequential |
|---|---|---|
| Maximum Events | 430 | 442 |
| Actual Events | 430 | 220 |
| Duration | 48 months | 36 months |
| Cost | $76.8M | $58.7M |
Impact Summary
$18.1M cost savings (24% reduction)
Avoided 12 months of additional enrollment and follow-up costs
12-month duration reduction (25% faster)
Trial completed at 36 months instead of planned 48 months
BLA submission 12 months earlier
$2.4 billion NPV gain from accelerated market entry
Earlier patient access
Effective therapy reached patients 12 months sooner
Regulatory Support
“Group sequential designs represent the simplest and most well-established adaptive designs... Group sequential designs have a well-developed statistical framework and extensive operating experience.”
— FDA Guidance: Adaptive Designs for Clinical Trials of Drugs and Biologics (November 2019)
Design your own GSD trial
Calculate stopping boundaries for your interim analysis schedule.