Prognostic Models for Nonmetastatic Triple-Negative Breast Cancer Based on the Pretreatment Serum Tumor Markers with Machine Learning
Table 1
Baseline clinicopathological characteristics of the study population (n = 258).
Characteristics
No. of patients
Percent/%
Age at diagnosis/years old (median 51.5)#
<40
22
8.5
40∼50
87
33.7
50∼60
89
34.5
≥60
60
23.3
Side
Left
131
50.8
Right
126
48.8
Bilateral
1
0.4
Histology
Nonspecific invasive cancer
203
78.7
Mixed
30
11.6
Other
25
9.7
Grade
I
6
2.3
II
77
29.9
III
110
42.6
Unknown
65
25.2
Ki-67
<30%
49
19.0
≥30%
193
74.8
Unknown
16
6.2
T-stage
1
128
49.6
2
99
38.4
3
16
6.2
4
3
1.2
Unknown
12
4.6
N-stage
0
183
70.9
1
46
17.8
2
18
7.0
3
11
4.3
Stage
I
100
38.8
II
111
43.0
III
36
14.0
Unknown
11
4.3
Surgery
Lumpectomy
80
31.0
Mastectomy
178
69.0
Chemotherapy
Yes
236
91.5
No
14
5.4
Unknown
8
3.1
Radiotherapy
Yes
114
44.2
No
124
48.1
Unknown
20
7.7
No. of events total: n = 53 (20.5%)
Locoregional recurrence
16
6.2
Distant metastasis##
31
12.0
Second primary cancer
3
1.2
Death because of other reasons
3
1.2
No. of deaths total: n = 28 (10.8%)
Death because of breast cancer
23
8.9
Death because of other reasons
5
1.9
#The age range takes the lower limit but not the upper limit. ## Some patients with distant metastasis were accompanied by locoregional recurrence, all of which were counted as distant metastasis.