Table 3 RFS analyses by the log-rank test according to the fifteen-mRNA signature in each set. Our result shown lower scores were associated with longer RFS, and higher scores were associated with shorter RFS in each set (P < 0.05).
Datasets | Risk group (n) | Recurrence-free survival | ||||
|---|---|---|---|---|---|---|
1-year | 3-year | 5-year | HR (95% CI) | P-value | ||
Training set (n = 226, GSE31210) | High-risk (n = 113) | 83.19% | 53.98% | 29.20% | 21.17 (5.073–88.31) | < 0.001 |
Low-risk (n = 113) | 99.12% | 84.07% | 54.87% | |||
Internal validation set (n = 226, GSE30219) | High-risk (n = 113) | 68.14% | 43.36% | 28.32% | 2.787 (1.962–3.957) | < 0.001 |
Low-risk (n = 113) | 86.73% | 76.11% | 61.95% | |||
External validation set (n = 181, GSE50081) | High-risk (n = 90) | 74.44% | 46.67% | 31.11% | 2.168 (1.356–3.466) | 0.001 |
Low-risk (n = 91) | 89.01% | 71.43% | 43.96% | |||
