Table 3.
Actual and alternative biomarker-stratified trial design of KEYNOTE-119 (6)
| Patient population | Actual designa |
Biomarker-stratified designb |
||||
|---|---|---|---|---|---|---|
| Actual sample size | Target HR (power, %) | Formally tested | Target sample size | Target HR (power, %) | Formally tested | |
| CPS ≥ 10 | 194 | 0.6 (85) | Yes | — | — | No |
| CPS ≥ 1 | 405 | 0.7 (90) | Yes | — | — | No |
| CPS ≥ 20 | — | — | No | 194 | 0.6 (85) | Yes |
| 20 > CPS ≥ 1 | — | — | No | 420 | 0.7 (90) | Yes |
| CPS < 1 | — | — | No | 360 | 0.7 (85) | Yes |
| Overall | 622 | 0.78 (80) | Yes | 974 | — | No |
Sequential biomarker-positive/overall design: initial α allocation 0.017 and 0.008 to CPS≥10 and CPS≥1 subgroups, respectively, with sequential testing of the overall population if the CPS≥1 subgroup is statistically significant. (This procedure controls the overall type I error of the design at level α). CPS = PD-L1 combined positive score; HR = hazard ratio.
Sequential biomarker-stratified design; fully sequential testing with α = 0.025: first test CPS≥20 subgroup; if statistically significant, then test the 20 > CPS ≥ 1 subgroup, and if statistically significant, then test the CPS<1 subgroup. (This procedure controls the overall type I error of the design at level α).