Clinicopathological characteristics among the BC, TC2BC and BC2TC cohorts
The clinicopathological features of patients in BC, TC2BC and BC2TC cohorts were shown in Table 1. Among these three cohorts, we found significant differences in age, race, lymph node (LN) status, tumor grade, TNM stage, ER status and PR status (P<0.05). Compared with BC cohort, patients in TC2BC cohort were older and patients in BC2TC were younger when diagnosed with BC. Regarding race, TC2BC and BC2TC cohorts had fewer black women and more from other race including Asian. More patients in BC cohort had more than 10 lymph nodes involved and advanced TNM stage accordingly. More patients in TC2BC had better well differentiated tumor and hormone receptor positive compared to patients in other cohorts.
Table 1. Clinicopathological characteristics among the BC, TC2BC and BC2TC cohorts
Characteristic
|
BC (n=345236)
|
TC2BC (n=606)
|
BC2TC (n=1269)
|
P value
|
n
|
%
|
n
|
%
|
n
|
%
|
Age diagnosed with BC
|
|
|
|
|
|
|
<0.001
|
≤40
|
22100
|
6.4
|
25
|
4.1
|
109
|
8.6
|
|
40-60
|
164968
|
47.8
|
287
|
47.4
|
753
|
59.3
|
|
>60
|
158168
|
45.8
|
294
|
48.5
|
407
|
32.1
|
|
Race
|
|
|
|
|
|
|
0.014
|
White
|
281666
|
81.6
|
500
|
82.5
|
1038
|
81.8
|
|
Black
|
34141
|
9.9
|
49
|
8.1
|
100
|
7.9
|
|
Other
|
29429
|
8.5
|
57
|
9.4
|
131
|
10.3
|
|
Tumor size
|
|
|
|
|
|
|
0.777
|
≤20mm
|
210571
|
61.0
|
374
|
61.7
|
795
|
62.6
|
|
20-50mm
|
110122
|
31.9
|
189
|
31.2
|
384
|
30.3
|
|
>50mm
|
24543
|
7.1
|
43
|
7.1
|
90
|
7.1
|
|
Positive LN
|
|
|
|
|
|
|
<0.001
|
0
|
223459
|
64.7
|
386
|
63.7
|
820
|
64.6
|
|
1-3
|
82817
|
24.0
|
160
|
26.4
|
423
|
33.3
|
|
4-9
|
26400
|
7.6
|
40
|
6.6
|
89
|
7.0
|
|
10
|
12560
|
3.6
|
20
|
3.3
|
36
|
2.8
|
|
Histology
|
|
|
|
|
|
|
0.103
|
Duct carcinoma
|
278515
|
80.7
|
475
|
78.4
|
1020
|
80.4
|
|
Lobular carcinoma
|
25048
|
7.3
|
57
|
9.4
|
85
|
6.7
|
|
Mixed
|
38137
|
11.0
|
73
|
12.0
|
151
|
11.9
|
|
Other
|
3536
|
1.0
|
1
|
0.2
|
13
|
1.0
|
|
Tumor grade
|
|
|
|
|
|
|
0.012
|
Ⅰ (well differentiated)
|
68499
|
19.8
|
125
|
20.6
|
235
|
18.5
|
|
Ⅱ (moderately differentiated)
|
148149
|
42.9
|
296
|
48.8
|
551
|
43.4
|
|
Ⅲ (poorly differentiated)
|
125030
|
36.2
|
180
|
29.7
|
464
|
36.6
|
|
Ⅳ (undifferentiated)
|
3558
|
1.0
|
5
|
0.8
|
19
|
1.5
|
|
TNM stage
|
|
|
|
|
|
|
0.028
|
0
|
61
|
0.0
|
0
|
0.0
|
2
|
0.2
|
|
Ⅰ
|
164212
|
47.6
|
292
|
48.2
|
606
|
47.8
|
|
Ⅱ
|
133354
|
38.6
|
236
|
38.9
|
499
|
39.3
|
|
Ⅲ
|
43154
|
12.5
|
70
|
11.6
|
152
|
12.0
|
|
Ⅳ
|
4455
|
1.3
|
8
|
1.3
|
10
|
0.8
|
|
ER
|
|
|
|
|
|
|
0.013
|
Positive
|
273760
|
79.3
|
512
|
84.5
|
1023
|
80.6
|
|
Negative
|
70916
|
20.5
|
93
|
15.3
|
244
|
19.2
|
|
Borderline
|
560
|
0.2
|
1
|
0.2
|
2
|
0.2
|
|
PR
|
|
|
|
|
|
|
0.006
|
Positive
|
235316
|
68.2
|
445
|
73.4
|
900
|
70.9
|
|
Negative
|
108150
|
31.3
|
158
|
26.1
|
360
|
28.4
|
|
Borderline
|
1770
|
0.5
|
3
|
0.5
|
9
|
0.7
|
|
HER-2
|
|
|
|
|
|
|
0.257
|
Positive
|
13991
|
4.1
|
51
|
8.4
|
52
|
4.1
|
|
Negative
|
79810
|
23.1
|
244
|
40.3
|
223
|
17.6
|
|
Borderline*
|
2000
|
0.6
|
8
|
1.3
|
5
|
0.4
|
|
Unknow
|
249435
|
72.3
|
303
|
50.0
|
989
|
77.9
|
|
Subtype
|
|
|
|
|
|
|
0.096
|
HR+/HER2- (Luminal A)
|
68791
|
19.9
|
215
|
35.5
|
192
|
15.1
|
|
HR+/HER2+ (Luminal B)
|
9834
|
2.8
|
41
|
6.8
|
32
|
2.5
|
|
HR-/HER2+ (HER2 enriched)
|
4157
|
1.2
|
10
|
1.7
|
20
|
1.6
|
|
HR-/HER2- (Triple Negative)
|
11019
|
3.2
|
29
|
4.8
|
31
|
2.4
|
|
Unknow
|
251435
|
72.8
|
311
|
51.3
|
994
|
78.3
|
|
History of breast surgery
|
|
|
|
|
|
|
0.097
|
Yes
|
343656
|
99.5
|
604
|
99.7
|
1268
|
99.9
|
|
No
|
1580
|
0.5
|
2
|
0.3
|
1
|
0.1
|
|
*The algorithm used to derive HER-2 information is detailed in the following link (https://seer.cancer.gov/seerstat/databases/ssf/her2-derived.html).
Abbreviation: BC, breast cancer; TC, thyroid cancer; LN, lymph node; HER2, human epidermal growth factor receptor 2; ER, estrogen receptor; PR, progesterone receptor.
|
|
|
|
|
|
|
|
|
|
Impact of clinicopathological characteristics on survival
Multivariate Cox regression analyses were performed to determine independent prognostic factors for OS and DSS of breast cancer (Table 2). We found that the patients in BC2TC cohort had better OS (HR: 0.483 (0.412-0.567), P<0.001) and DSS (HR: 0.390 (0.307-0.495), P<0.001) than those in BC cohort. Furthermore, it is also demonstrated that all these clinicopathological characteristics, including age, race, tumor size, LN status, tumor grade, histologic type, TNM stage, surgical treatment for breast cancer, ER/PR status, Her-2 status and breast cancer molecular subtype, were significantly associated with OS and DSS(P<0.05).
Table 2. Multivariate Cox regression analyses on OS and DSS of breast cancer
Characteristic
|
OS (overall survival)
|
DSS (disease-specific survival)
|
HR (95%CI)
|
P value
|
HR (95%CI)
|
P value
|
Cohort
|
|
<0.001
|
|
<0.001
|
BC
|
Reference
|
|
Reference
|
|
TC2BC
|
1.087 (0.909-1.300)
|
0.361
|
1.125 (0.878-1.440)
|
0.352
|
BC2TC
|
0.483 (0.412-0.567)
|
<0.001
|
0.390 (0.307-0.495)
|
<0.001
|
Age at diagnosis
|
|
<0.001
|
|
<0.001
|
≤40
|
Reference
|
|
Reference
|
|
40-60
|
0.876 (0.849-0905)
|
<0.001
|
0.861 (0.832-0.890)
|
<0.001
|
>60
|
2.812 (2.726-2.900)
|
<0.001
|
1.323 (1.279-1.369)
|
<0.001
|
Race
|
|
<0.001
|
|
<0.001
|
White
|
Reference
|
|
Reference
|
|
Black
|
1.301 (1.275-1.329)
|
<0.001
|
1.319 (1.284-1.354)
|
<0.001
|
Other
|
0.732 (0.711-0.754)
|
<0.001
|
0.828 (0.797-0.859)
|
<0.001
|
Tumor size
|
|
<0.001
|
|
<0.001
|
≤20mm
|
Reference
|
|
Reference
|
|
20-50mm
|
1.582 (1.548-1.617)
|
<0.001
|
1.703 (1.655-1.752)
|
<0.001
|
>50mm
|
1.884 (1.827-1.942)
|
<0.001
|
2.227 (2.145-2.312)
|
<0.001
|
Positive LN
|
|
<0.001
|
|
<0.001
|
0
|
Reference
|
|
Reference
|
|
1-3
|
1.428 (1.396-1.460)
|
<0.001
|
1.817 (1.764-1.872)
|
<0.001
|
4-9
|
2.051 (1.993-2.111)
|
<0.001
|
3.034 (2.925-3.147)
|
<0.001
|
10
|
3.071 (2.975-3.171)
|
<0.001
|
4.777 (4.594-4.967)
|
<0.001
|
Tumor grade
|
|
<0.001
|
|
<0.001
|
Ⅰ
|
Reference
|
|
Reference
|
|
Ⅱ
|
1.164 (1.139-1.189)
|
<0.001
|
1.808 (1.733-1.887)
|
<0.001
|
Ⅲ
|
1.372 (1.340-1.404)
|
<0.001
|
2.534 (2.425-2.647)
|
<0.001
|
Ⅳ
|
1.365 (1.283-1.451)
|
<0.001
|
2.543 (2.339-2.764)
|
<0.001
|
Histologic type
|
|
<0.001
|
|
<0.001
|
Duct carcinoma
|
Reference
|
|
Reference
|
|
Lobular carcinoma
|
0.957 (0.932-0.982)
|
<0.001
|
1.029 (0.990-1.070)
|
0.141
|
Mixed
|
0.951 (0.930-0.972)
|
<0.001
|
0.973 (0.942-1.005)
|
0.092
|
Other
|
0.903 (0.850-0.960)
|
0.001
|
0.849 (0.787-0.916)
|
<0.001
|
TNM stage
|
|
<0.001
|
|
<0.001
|
0
|
Reference
|
|
Reference
|
|
Ⅰ
|
1.526 (0.762-3.057)
|
0.233
|
1.768 (0.441-7.083)
|
0.421
|
Ⅱ
|
1.524 (0.760-3.055)
|
0.235
|
2.584 (0.645-10.351)
|
0.180
|
Ⅲ
|
1.898 (0.947-3.806)
|
0.071
|
3.145 (0.785-12.605)
|
0.106
|
Ⅳ
|
6.031 (3.006-12.100)
|
<0.001
|
11.878 (2.963-47.622)
|
<0.001
|
History of breast surgery (Yes vs. No)
|
0.562 (0.522-0.606)
|
<0.001
|
0.481 (0.442-0.523)
|
<0.001
|
ER Status (Negative vs. Positive)
|
1.101 (1.076-1.127)
|
<0.001
|
1.228 (1.191-1.266)
|
<0.001
|
PR Status (Negative vs. Positive)
|
1.201 (1.179-1.225)
|
<0.001
|
1.382 (1.345-1.420)
|
<0.001
|
HER2 (Negative vs. Positive)
|
0.894 (0.799-1.000)
|
0.050
|
0.858 (0.738-0.999)
|
0.048
|
Breast cancer molecular subtype
|
|
<0.001
|
|
<0.001
|
HR+/HER2- (Luminal A)
|
Reference
|
|
Reference
|
|
HR+/HER2+ (Luminal B)
|
0.899 (0.845-0.957)
|
0.001
|
0.804 (0.742-0.870)
|
<0.001
|
HR-/HER2+ (HER2 enriched)
|
0.953 (0.881-1.031)
|
0.230
|
0.800 (0.729-0.879)
|
<0.001
|
HR-/HER2- (Triple Negative)
|
1.690 (1.613-1.771)
|
<0.001
|
1.676 (1.584-1.774)
|
<0.001
|
Abbreviation: OS, overall survival; DSS, disease-specific survival; BC, breast cancer; TC, thyroid cancer; LN, lymph node; HER2, human epidermal growth factor receptor 2; ER, estrogen receptor; PR, progesterone receptor.
Survival analysis of BC patients among three cohorts
The KM curves for OS and DSS among the BC, TC, TC2BC and BC2TC cohorts were shown in Figure 2. The 10-year OS rate of patients in TC cohort was 93%, compared to 89%, 73%, and 74%, respectively, for patients in BC2TC, TC2BC and BC cohorts (P<0.001). OS was much better in the BC2TC cohort than in the BC cohort (Figure 2A). OS showed no statistical difference between the TC2BC cohort and BC cohort (Figure 2A), and a similar trend was observed in terms of DSS (Figure 2B).
In order to validate the prognostic effect of thyroid cancer, a 1:1 PSM was performed to balance the baseline characteristics of patients in BC cohort and BC2TC cohort. After matching, 1,260 pairs of patients were selected for the analysis (Table 3), and the survival curves of data after PSM were presented in Figure 3. The 10-year OS rate were 89% vs. 78%, respectively. (HR: 0.511(0.421-0.620), P<0.001, Figure 3A). The 10-year DSS rate were 94% vs. 88%, respectively. (HR: 0.458(0.346-0.606), P<0.001, Figure 3B).
Table 3. Clinical characteristics of patients in BC group and BC2TC group before and after PSM
Characteristic
|
before PSM
|
P value
|
after PSM
|
P value
|
BC (n=345236)
|
BC2TC (n=1269)
|
BC
(n=1260)
|
BC2TC
(n=1260)
|
Age
|
|
|
<0.001
|
|
|
0.999
|
≤40
|
22100
|
109
|
|
108
|
108
|
|
40-60
|
164968
|
753
|
|
746
|
747
|
|
>60
|
158168
|
407
|
|
406
|
405
|
|
Race
|
|
|
0.007
|
|
|
0.997
|
White
|
281666
|
1038
|
|
1037
|
1036
|
|
Black
|
34141
|
100
|
|
100
|
100
|
|
Other
|
29429
|
131
|
|
123
|
124
|
|
Tumor size
|
|
|
0.442
|
|
|
0.963
|
≤20mm
|
210571
|
795
|
|
795
|
790
|
|
20-50mm
|
110122
|
384
|
|
381
|
383
|
|
>50mm
|
24543
|
90
|
|
84
|
87
|
|
Positive lymph node
|
|
|
<0.001
|
|
|
0.976
|
0
|
223459
|
820
|
|
816
|
816
|
|
1-3
|
82817
|
423
|
|
328
|
322
|
|
4-9
|
26400
|
89
|
|
82
|
86
|
|
10
|
12560
|
36
|
|
34
|
36
|
|
Histology
|
|
|
0.708
|
|
|
0.976
|
Duct carcinoma
|
278515
|
1020
|
|
1014
|
1012
|
|
Lobular carcinoma
|
25048
|
85
|
|
86
|
84
|
|
Mixed
|
38137
|
151
|
|
149
|
151
|
|
Other
|
3536
|
13
|
|
11
|
13
|
|
Grade
|
|
|
0.272
|
|
|
0.838
|
Ⅰ
|
68499
|
235
|
|
237
|
235
|
|
Ⅱ
|
148149
|
551
|
|
547
|
549
|
|
Ⅲ
|
125030
|
464
|
|
463
|
458
|
|
Ⅳ
|
3558
|
19
|
|
13
|
18
|
|
TNM stage
|
|
|
0.002
|
|
|
0.855
|
0
|
61
|
2
|
|
0
|
1
|
|
Ⅰ
|
164212
|
606
|
|
606
|
603
|
|
Ⅱ
|
133354
|
499
|
|
499
|
497
|
|
Ⅲ
|
43154
|
152
|
|
144
|
150
|
|
Ⅳ
|
4455
|
10
|
|
11
|
9
|
|
ER
|
|
|
0.511
|
|
|
0.954
|
Positive
|
273760
|
1023
|
|
1023
|
1017
|
|
Negative
|
70916
|
244
|
|
236
|
242
|
|
Borderline
|
560
|
2
|
|
1
|
1
|
|
PR
|
|
|
0.052
|
|
|
0.965
|
Positive
|
235316
|
900
|
|
899
|
896
|
|
Negative
|
108150
|
360
|
|
353
|
355
|
|
Borderline
|
1770
|
9
|
|
8
|
9
|
|
HER-2
|
|
|
0.167
|
|
|
0.972
|
Positive
|
13991
|
52
|
|
50
|
51
|
|
Negative
|
79810
|
223
|
|
225
|
222
|
|
Borderline
|
2000
|
5
|
|
2
|
3
|
|
Unknow
|
249435
|
989
|
|
983
|
984
|
|
Subtype
|
|
|
0.119
|
|
|
0.986
|
HR+/HER2- (Luminal A)
|
68791
|
192
|
|
194
|
191
|
|
HR+/HER2+ (Luminal B)
|
9834
|
32
|
|
34
|
32
|
|
HR-/HER2+ (HER2 enriched)
|
4157
|
20
|
|
16
|
19
|
|
HR-/HER2- (Triple Negative)
|
11019
|
31
|
|
31
|
31
|
|
Unknow
|
251435
|
994
|
|
985
|
987
|
|
History of breast surgery
|
|
|
0.045
|
|
|
|
Yes
|
343656
|
1268
|
|
1260
|
1260
|
|
No
|
1580
|
1
|
|
0
|
0
|
|
Abbreviation: PSM, propensity score matching; BC, breast cancer; TC, thyroid cancer; LN, lymph node; HER2, human epidermal growth factor receptor 2; ER, estrogen receptor; PR, progesterone receptor.
Subgroup survival analysis of BC patients in BC2TC group and BC group
To explore the effect by tumor characteristics, we performed subgroup analysis for the OS and DSS in patients between BC and BC2TC cohorts based on all the clinicopathological characteristics (Figure 4, S1 and S2). Patients in BC2TC cohort had longer OS than those in BC cohort in all subgroups except for the Her2-positive subgroup (Figure 4A). DSS was better in BC2TC group except for the other race, Her2-positive and Her2-negative subgroup (Figure 4B).
Interestingly, in the subgroup analysis for BC subtypes (Figure 5), luminal BC patients with SPTC may yield longer survival than luminal BC patients (Figure 5A and 5B), but no such trend was found in Her-2 enriched or triple negative breast cancer patients (Figure 5C and 5D).