Source of evidence | Treatment effect compared to BSC | Comments | |
---|---|---|---|
Estimand 1. All comers – OS HR (95% CI) | Estimand 2. WT KRAS – OS HR (95% CI) | ||
Study 20020408 Panitumumab + BSC (n = 231) KRAS assessable (n = 208) WT KRAS (n = 124) MT KRAS (n = 84) BSC alone (n = 232) KRAS assessable (n = 219) WT KRAS (n = 119) MT KRAS (n = 100) Study affected by treatment switching from BSC onto panitumumab. Conducted analysis of KRAS groups to ‘adjust’ for switching [15] | Compared WT and MT KRAS patients randomised to panitumumab, to MT KRAS patients randomised to BSC HR = 0.76 (0.60–0.98) | Compared WT KRAS patients randomised to panitumumab, to MT KRAS patients randomised to BSC HR = 0.66 (0.49–0.87) | Assumption that KRAS status is not prognostic for survival, caution required due to breaking of randomisation |
Study 20100007 [25] Panitumumab + BSC (n = 189) BSC alone (n = 188) Study only included patients with WT KRAS. Switching from BSC onto panitumumab not allowed | - | Intention-to-treat analysis HR = 0.73 (0.57–0.93) | Newer study, BSC survival may have improved |
CO.17 study [16] Cetuximab + BSC (n = 287) BSC alone (n = 285) Switching from BSC onto cetuximab not allowed | Intention-to-treat analysis HR = 0.77 (0.64–0.92) | Compared WT KRAS patients randomised to cetuximab, to WT patients randomised to BSC HR = 0.55 (0.41–0.74) [0.62 (0.44–0.87) when adjusted for potentially prognostic covariates] | Assume similar effectiveness for cetuximab and panitumumab, based on ASPECCT study [24] |