Table 4.
Multivariable analysis of variables to predict systemic recurrence in TNBC.
| Variables | Clinical model | Rad model | |||
|---|---|---|---|---|---|
| Hazard ratio (95% CI) | P value | Hazard ratio (95% CI) | P value | ||
| Pathologic invasive cancer size | 1.014 (0.98, 1.049) | 0.438 | 1.016 (0.968, 1.067) | 0.516 | |
| Lymphovascular invasion | Yes | 7.875 (2.679, 23.153) | <0.001 | 4.414 (1.331, 14.637) | 0.015 |
| No | |||||
| Surgery type | Brest conserving surgery | 0.231 (0.078, 0.679) | 0.008 | 0.335 (0.084, 1.337) | 0.121 |
| Total mastectomy | |||||
| Number of metastatic axillary lymph nodes after surgery | 1.06 (1.002, 1.121) | 0.043 | 1.002 (0.92, 1.091) | 0.964 | |
| Rad score | 39.302 (11.351, 136.078) | <0.001 | |||
| C-index of the training set (95% CI) | 0.879 (0.744, 1) | 0.97 (0.835, 1) | |||
| *P value | 0.009 | ||||
| C-index of the validation set (95% CI) | 0.939 (0.604, 1) | 0.848 (0.513, 1) | |||
| *P value | 0.100 | ||||
TNBC = triple-negative breast cancer, CI = confidence interval, *P value was calculated by comparing the C-index between the Clinical and Rad models.