Research Article
Prognostic Models for Nonmetastatic Triple-Negative Breast Cancer Based on the Pretreatment Serum Tumor Markers with Machine Learning
Table 2
The optimal prognostic cut-off value of each tumor marker determined by X-tile.
| Tumor marker | Cut-off | No. of patients N (%) |
| CEA/(ng/mL) | 2.15 | <2.15 | 167 (64.7) | ≥2.15 | 91 (35.3) | CA19-9/(U/mL) | 17.30 | <17.30 | 202 (78.3) | ≥17.30 | 56 (21.7) | CA125/(U/mL) | 9.05 | <9.05 | 64 (24.8) | ≥9.05 | 194 (75.2) | CA242/(U/mL) | 8.85 | <8.85 | 217 (84.1) | ≥8.85 | 41 (15.9) | CA211/(ng/mL) | 1.15 | <1.15 | 155 (60.1) | ≥1.15 | 103 (39.9) | CA15-3/(U/mL) | 16.00 | <16.00 | 215 (83.3) | ≥16.00 | 43 (16.7) |
|
|
CEA: carcinoembryonic antigen; CA: cancer antigen.
|