Screening process and demographic characteristics of patients
During the 2004–2015 time period, a total of 145,445 patients with endometrial cancer were identified from the SEER database. Of these, 63,510 patients were enrolled in this study according to the inclusion criteria and exclusion criteria listed in Fig. 1. These eligible patients were randomly divided into a training cohort (n = 42,340) and a validation cohort (n = 21,170). The demographic characteristics of the patients were listed in Table 1. In the training cohort, the mean age of the patients was 62.43 ± 11.66 years old. The mean tumor size was 6.56 ± 15.60 cm. For the marital status, a total of 22,726 (53.68%) patients were married, while 8,168 (19.29%) patients were single. Among these patients, 37,041 (87.48%) patients were diagnosed with EEA, 3,185 (7.52%) patients with SEA, and 2,114 (4.99%) patients were diagnosed with other cancer types. The myometrial invasion of most patients (27,721, 65.74%) was less than 50%. Most people were negative for cervical stromal invasion (33,797, 79.82%) and lymph node metastasis (37,500, 88.57%). As far as the clinical stage aspect, 70.46% people were at stage I, 8.17% were at stage II, 15.64% were at stage III, and 5.73 were at stage IV. The median survival time was 61.17 ± 40.99 months. The rest of the clinicopathologic characteristics of the patients in the training and validation group are listed in Table 1. There was no significant difference regarding all the enrolled variables between the two cohorts.
Table 1
Demographic characteristics of patients in training and validation cohorts
|
Training cohort
|
Validation cohort
|
P
|
Total number
|
42340
|
21170
|
|
|
Mean + SD exp
|
Mean + SD val
|
|
Age (years)
|
62.43 ± 11.66
|
62.37 ± 11.54
|
|
Tumor size (cm)
|
6.56 ± 15.60
|
6.62 ± 15.76
|
|
Survival months
|
61.17 ± 40.99
|
61.17 ± 40.98
|
|
|
N (%)
|
N (%)
|
|
Age (years)
|
|
|
0.711
|
≤ 40
|
1497 (3.54)
|
719 (3.40)
|
|
41–50
|
4241 (10.02)
|
2164 (10.22)
|
|
51–60
|
13001 (30.71)
|
6553 (30.95)
|
|
61–70
|
13533 (31.96)
|
6719 (31.74)
|
|
> 70
|
10068 (23.78)
|
5015 (23.69)
|
|
Year of diagnosis
|
|
|
0.921
|
2004–2007
|
10840 (25.60)
|
5410 (24.28)
|
|
2008–2011
|
14395 (34.00)
|
7193 (33.98)
|
|
2012–2015
|
17105 (40.40)
|
8567 (41.47)
|
|
Marital status
|
|
|
0.105
|
Married
|
22726 (53.68)
|
11454 (54.10)
|
|
Single
|
8168 (19.29)
|
4046 (19.11)
|
|
Others
|
11446 (27.03)
|
5670 (26.78)
|
|
Tumor size (cm)
|
|
|
0.243
|
≤ 2
|
10437 (24.65)
|
5171 (24.43)
|
|
2.1-5
|
20917 (49.40)
|
10463 (49.42)
|
|
5.1–10
|
8571 (20.24)
|
4321 (20.41)
|
|
> 10
|
2415 (5.70)
|
1215 (5.74)
|
|
Histological type
|
|
|
0.418
|
EEA
|
37041 (87.48)
|
18468 (87.24)
|
|
SEA
|
3185 (7.52)
|
1621 (7.66)
|
|
Other types
|
2114 (4.99)
|
1081 (5.11)
|
|
Myometrial invasion
|
|
|
0.682
|
< 50
|
27721 (65.47)
|
13792 (65.14)
|
|
≥ 50
|
14619 (34.53)
|
7378 (34.86)
|
|
Cervical stromal invasion
|
|
|
0.959
|
Negative
|
33797 (79.82)
|
16884 (79.75)
|
|
Positive
|
8543 (20.18)
|
4286 (20.25)
|
|
LNM
|
|
|
0.327
|
Negative
|
37500 (88.57)
|
18775 (88.69)
|
|
Positive
|
4840 (11.43)
|
2395 (11.31)
|
|
Grade
|
|
|
0.545
|
1
|
17748 (41.92)
|
8838 (41.75)
|
|
2
|
12691 (29.97)
|
6262 (29.58)
|
|
3
|
11901 (28.11)
|
6070 (28.67)
|
|
Stage
|
|
|
0.594
|
I
|
29831 (70.46)
|
14935 (70.55)
|
|
II
|
3458 (8.17)
|
1677 (7.92)
|
|
III
|
6623 (15.64)
|
3336 (15.76)
|
|
IV
|
2428 (5.73)
|
1222 (5.77)
|
|
Chemotherapy
|
|
|
0.082
|
No
|
34276 (80.95)
|
17026 (80.43)
|
|
Yes
|
8064 (19.05)
|
4144 (19.57)
|
|
Radiation
|
|
|
0.052
|
No
|
30202 (71.33)
|
15044 (71.06)
|
|
Yes
|
12138 (28.67)
|
6126 (28.94)
|
|
Overall survival
|
|
|
0.778
|
Alive
|
32905 (77.72)
|
16435 (77.63)
|
|
Death
|
9435 (22.28)
|
4735 (22.37)
|
|
Cancer-specific survival
|
|
|
0.972
|
Alive
|
35695 (84.31)
|
17791 (84.04)
|
|
Death
|
6645 (15.69)
|
3379 (15.96)
|
|
EEA, endometrial endometrioid adenocarcinoma; SEA, serous endometrioid adenocarcinoma; LNM, lymph node metastasis |
Independent prognostic factors for OS and CSS
The univariate and multivariate analyses of potential predictors for the OS in the training group and the validation group are shown in Table 2. Age at diagnosis, marital status, tumor size, histological type, myometrial invasion, cervical stromal invasion, lymph node metastasis, tumor grade, clinical stage, chemotherapy, radiotherapy were significantly associated as risk factors for the OS according to univariate analysis for both the groups. For this reason, these risk factors were included in the multivariate analysis. Multivariate analysis identified that age at diagnosis (51–60, OR = 1.37, 95%CI: 1.15–1.63, p = 0.004; 61–70, OR = 2.17, 95%CI: 1.82–2.58, p < 0.001; >70, OR = 4.01, 95%CI: 3.37–4.77, p < 0.001), marital status (single, OR = 1.39, 95%CI:1.31–1.47, p < 0.001; others, OR = 1.46, 95%CI:1.39–1.52, p < 0.001), tumor size (2.1-5cm, OR = 1.13, 95%CI: ,1.06–1.20 p < 0.001; 5.1-10cm, OR = 1.39, 95%CI: ,1.30–1.49 p < 0.001; >10 cm, OR = 1.75, 95%CI: ,1.60–1.92 p < 0.001;), histological type (SEA, OR = 1.13, 95%CI: 1.06–1.21, p < 0.001; other types, OR = 1.53, 95%CI: 1.43–1.64, p < 0.001), lymph node metastasis (OR = 1.36, 95%CI: 1.28–1.44, p < 0.001), tumor grade (G2, OR = 1.45, 95%CI: 1.36–1.54, p < 0.001; G3, OR = 2.62, 95%CI: 2.46–2.79, p < 0.001), clinical stage (stage II, OR = 1.24, 95%CI: 1.15–1.35, p < 0.001; stage III, OR = 1.73, 95%CI: 1.61–1.86, p < 0.001; stage IV, OR = 4.92, 95%CI: 4.54–5.34, p < 0.001) were independent risk factors to predict OS in the training group. The analysis of the validation group showed similar results (Table 2). These variables remained statistically significant (p < 0.01) according to the multivariate analysis for CSS (Table 3). For this reason, age at diagnosis, marital status, tumor size, histological type, lymph node metastasis, tumor grade, and clinical stage were identified as independent prognostic factors for both OS and CSS for the EC patients. Independent prognostic factors were used to develop the nomograms for 3-year and 5-year OS and CSS (Fig. 2A and Fig. 2B). Detailed points of each predictor in the nomograms were listed in Table 4. By including these scores to the total on the bottom scale, the 3- and 5-year OS and CSS can be predicted.
Table 2
Univariate and multivariate analysis of overall survival in training and validation cohorts
|
Training cohort
|
Validation cohort
|
|
Univariate
|
Multivariate
|
Univariate
|
Multivariate
|
Age at diagnosis
|
|
|
|
|
≤ 40
|
1.0
|
1.0
|
1.0
|
1.0
|
41–50
|
1.13 (0.93, 1.37) 0.2093
|
1.08 (0.89, 1.31) 0.4095
|
1.00 (0.77, 1.29) 0.9711
|
1.04 (0.80, 1.36) 0.7544
|
51–60
|
1.47 (1.23, 1.74) < 0.001
|
1.37 (1.15, 1.63) 0.004
|
1.35 (1.06, 1.71) 0.0140
|
1.36 (1.07, 1.73) 0.0115
|
61–70
|
2.56 (2.16, 3.04) < 0.001
|
2.17 (1.82, 2.58) < 0.001
|
2.22 (1.76, 2.81) < 0.001
|
2.01 (1.58, 2.54) < 0.001
|
> 70
|
5.83 (4.92, 6.92) < 0.001
|
4.01 (3.37, 4.77) < 0.001
|
4.95 (3.92, 6.24) < 0.001
|
3.78 (2.98, 4.79) < 0.001
|
Marital status
|
|
|
|
|
Married
|
1.0
|
1.0
|
1.0
|
1.0
|
Single
|
1.28 (1.21, 1.36) < 0.001
|
1.39 (1.31, 1.47) < 0.001
|
1.23 (1.14, 1.34) < 0.001
|
1.27 (1.17, 1.37) < 0.001
|
Others
|
2.07 (1.98, 2.17) < 0.001
|
1.46 (1.39, 1.52) < 0.001
|
2.06 (1.94, 2.20) < 0.001
|
1.39 (1.30, 1.48) < 0.001
|
Tumor size (cm)
|
|
|
|
|
≤ 2
|
1.0
|
1.0
|
1.0
|
1.0
|
2.1-5
|
1.66 (1.56, 1.76) < 0.001
|
1.13 (1.06, 1.20) 0.001
|
1.67 (1.53, 1.82) < 0.001
|
1.14 (1.05, 1.25) 0.0026
|
5.1–10
|
3.10 (2.91, 3.31) < 0.001
|
1.39 (1.30, 1.49) < 0.001
|
3.23 (2.95, 3.54) < 0.001
|
1.54 (1.40, 1.70) < 0.001
|
> 10
|
3.24 (2.98, 3.53) < 0.001
|
1.75 (1.60, 1.92) < 0.001
|
3.28 (2.91, 3.71) < 0.001
|
1.90 (1.67, 2.16) < 0.001
|
Histological type
|
|
|
|
|
EEA
|
1.0
|
1.0
|
1.0
|
1.0
|
SEA
|
3.22 (3.04, 3.41) < 0.001
|
1.13 (1.06, 1.21) 0.001
|
3.20 (2.95, 3.47) < 0.001
|
1.12 (1.03, 1.22) 0.0101
|
Other types
|
3.14 (2.94, 3.35) < 0.001
|
1.53 (1.43, 1.64) < 0.001
|
2.98 (2.71, 3.27) < 0.001
|
1.43 (1.29, 1.57) < 0.001
|
Myometrial invasion
|
|
|
|
|
< 50
|
1.0
|
|
1.0
|
|
≥ 50
|
3.18 (2.96, 3.43) < 0.001
|
|
2.83 (2.56, 3.12) < 0.001
|
|
Cervical invasion
|
|
|
|
|
Negative
|
1.0
|
|
1.0
|
|
Positive
|
3.74 (3.58, 3.91) < 0.001
|
|
3.76 (3.53, 4.00) < 0.001
|
|
LNM
|
|
|
|
|
Negative
|
1.0
|
1.0
|
1.0
|
1.0
|
Positive
|
4.08 (3.90, 4.27) < 0.001
|
1.36 (1.28, 1.44) < 0.001
|
4.12 (3.86, 4.39) < 0.001
|
1.27 (1.17, 1.38) < 0.001
|
Tumor grade
|
|
|
|
|
1
|
1.0
|
1.0
|
1.0
|
1.0
|
2
|
1.98 (1.86, 2.10) < 0.001
|
1.45 (1.36, 1.54) < 0.001
|
1.85 (1.69, 2.01) < 0.001
|
1.36 (1.25, 1.49) < 0.001
|
3
|
5.88 (5.57, 6.20) < 0.001
|
2.62 (2.46, 2.79) < 0.001
|
5.85 (5.42, 6.31) < 0.001
|
2.45 (2.24, 2.67) < 0.001
|
Stage
|
|
|
|
|
I
|
1.0
|
1.0
|
1.0
|
1.0
|
II
|
2.11 (1.97, 2.27) < 0.001
|
1.24 (1.15, 1.35) < 0.001
|
2.10 (1.90, 2.34) < 0.001
|
1.24 (1.11, 1.40) 0.003
|
III
|
3.67 (3.49, 3.86) < 0.001
|
1.73 (1.61, 1.86) < 0.001
|
3.88 (3.62, 4.15) < 0.001
|
1.71 (1.55, 1.90) < 0.001
|
IV
|
13.17 (12.45, 13.93) < 0.001
|
4.92 (4.54, 5.34) < 0.001
|
14.07 (13.00, 15.23) < 0.001
|
5.04 (4.52, 5.61) < 0.001
|
Chemotherapy
|
|
|
|
|
0
|
1.0
|
|
1.0
|
|
1
|
2.82 (2.70, 2.94) < 0.001
|
|
2.71 (2.55, 2.88) < 0.001
|
|
Radiation
|
|
|
|
|
0
|
1.0
|
|
1.0
|
|
1
|
1.34 (1.28, 1.39) < 0.001
|
|
1.32 (1.24, 1.40) < 0.001
|
|
EEA, endometrial endometrioid adenocarcinoma; SEA, serous endometrioid adenocarcinoma; LNM, lymph node metastasis |
Table 3
Univariate and multivariate analysis of cancer-specific survival in training and validation cohorts
|
Training cohort
|
Validation cohort
|
|
Univariate
|
Multivariate
|
Univariate
|
Multivariate
|
Age (year)
|
|
|
|
|
≤ 40
|
1.0
|
1.0
|
1.0
|
1.0
|
41–50
|
1.12 (0.91, 1.39) 0.2910
|
1.09 (0.90, 1.32) 0.3924
|
1.03 (0.76, 1.38) 0.8671
|
1.04 (0.77, 1.40) 0.7927
|
51–60
|
1.43 (1.17, 1.74) 0.004
|
1.36 (1.14, 1.62) 0.006
|
1.33 (1.01, 1.74) 0.0407
|
1.26 (0.96, 1.66) 0.0964
|
61–70
|
2.31 (1.90, 2.80) < 0.001
|
2.17 (1.82, 2.57) < 0.001
|
2.12 (1.62, 2.76) < 0.001
|
1.71 (1.30, 2.24) 0.001
|
> 70
|
4.57 (3.77, 5.54) < 0.001
|
4.15 (3.49, 4.93) < 0.001
|
4.05 (3.11, 5.28) < 0.001
|
2.68 (2.04, 3.52) < 0.001
|
Marital status
|
|
|
|
|
Married
|
1.0
|
1.0
|
1.0
|
1.0
|
Single
|
1.27 (1.19, 1.36) < 0.001
|
1.41 (1.33, 1.49) < 0.001
|
1.16 (1.05, 1.28) 0.0023
|
1.12 (1.02, 1.24) 0.0206
|
Others
|
1.86 (1.77, 1.97) < 0.001
|
1.48 (1.41, 1.54) < 0.001
|
1.91 (1.77, 2.05) < 0.001
|
1.30 (1.21, 1.41) < 0.001
|
Tumor size (cm)
|
|
|
|
|
≤ 2
|
1.0
|
1.0
|
1.0
|
1.0
|
2.1-5
|
1.82 (1.69, 1.97) < 0.001
|
1.11 (1.04, 1.18) 0.0010
|
1.75 (1.57, 1.95) < 0.001
|
1.10 (0.98, 1.22) 0.1043
|
5.1–10
|
3.99 (3.68, 4.32) < 0.001
|
1.34 (1.26, 1.44) < 0.001
|
3.98 (3.57, 4.45) < 0.001
|
1.51 (1.35, 1.70) < 0.001
|
> 10
|
4.52 (4.09, 5.00) < 0.001
|
1.75 (1.60, 1.91) < 0.001
|
4.41 (3.83, 5.08) < 0.001
|
1.92 (1.65, 2.22) < 0.001
|
Histological type
|
|
|
|
|
EEA
|
1.0
|
1.0
|
1.0
|
1.0
|
SEA
|
4.05 (3.80, 4.31) < 0.001
|
1.10 (1.04, 1.17) 0.0021
|
4.01 (3.67, 4.38) < 0.001
|
1.17 (1.06, 1.29) 0.0014
|
Histological type
|
4.05 (3.76, 4.36) < 0.001
|
1.52 (1.42, 1.63) < 0.001
|
3.73 (3.36, 4.15) < 0.001
|
1.50 (1.35, 1.67) < 0.001
|
Myometrial invasion
|
|
|
|
|
< 50
|
1.0
|
|
1.0
|
|
≥ 50
|
4.11 (3.72, 4.54) < 0.001
|
|
4.03 (3.51, 4.63) < 0.001
|
|
Cervical invasion
|
|
|
|
|
Negative
|
1.0
|
|
1.0
|
|
Positive
|
5.26 (4.97, 5.57) < 0.001
|
|
5.35 (4.93, 5.80) < 0.001
|
|
LNM
|
|
|
|
|
Negative
|
1.0
|
1.0
|
1.0
|
1.0
|
Positive
|
5.42 (5.15, 5.71) < 0.001
|
1.31 (1.23, 1.39) < 0.001
|
5.42 (5.04, 5.82) < 0.001
|
1.29 (1.18, 1.41) < 0.001
|
Grade
|
|
|
|
|
1
|
1.0
|
1.0
|
1.0
|
1.0
|
2
|
2.65 (2.43, 2.88) < 0.001
|
1.39 (1.31, 1.48) < 0.001
|
2.34 (2.08, 2.63) < 0.001
|
1.64 (1.46, 1.85) < 0.001
|
3
|
10.16 (9.43, 10.95) < 0.001
|
2.39 (2.25, 2.55) < 0.001
|
9.76 (8.81, 10.82) < 0.001
|
3.50 (3.12, 3.93) < 0.001
|
Stage
|
|
|
|
|
I
|
1.0
|
1.0
|
1.0
|
1.0
|
II
|
2.60 (2.38, 2.85) < 0.001
|
1.19 (1.10, 1.29) < 0.001
|
2.67 (2.35, 3.03) < 0.001
|
1.42 (1.23, 1.64) < 0.001
|
III
|
5.57 (5.24, 5.91) < 0.001
|
1.59 (1.48, 1.71) < 0.001
|
5.88 (5.41, 6.39) < 0.001
|
2.20 (1.95, 2.49) < 0.001
|
IV
|
21.67 (20.33, 23.11) < 0.001
|
4.68 (4.34, 5.06) < 0.001
|
22.45 (20.50, 24.58) < 0.001
|
6.55 (5.78, 7.43) < 0.001
|
Chemotherapy
|
|
|
|
|
0
|
1.0
|
|
1.0
|
|
1
|
3.94 (3.76, 4.14) < 0.001
|
|
3.75 (3.50, 4.01) < 0.001
|
|
Radiation
|
|
|
|
|
0
|
1.0
|
|
1.0
|
|
1
|
1.47 (1.39, 1.54) < 0.001
|
|
1.46 (1.36, 1.57) < 0.001
|
|
EEA, endometrial endometrioid adenocarcinoma; SEA, serous endometrioid adenocarcinoma; LNM, lymph node metastasis |
Table 4
Detailed score of each prognostic factor in overall and cancer-specific survival nomograms
Characteristics
|
OS nomogram
|
CSS nomogra
|
Age at diagnosis
|
|
|
≤ 40
|
0
|
0
|
41–50
|
5
|
2.5
|
51–60
|
18
|
12
|
61–70
|
45
|
28
|
> 70
|
82
|
52.5
|
Marital status
|
|
|
Married
|
0
|
0
|
Single
|
18
|
13
|
Others
|
21
|
15
|
Tumor size (cm)
|
|
|
≤ 2
|
0
|
0
|
2.1-5
|
10
|
8
|
5.1–10
|
20
|
20
|
> 10
|
33
|
30
|
Histological type
|
|
|
EEA
|
0
|
0
|
SEA
|
5
|
7.5
|
Other types
|
23
|
23
|
LNM
|
|
|
Negative
|
0
|
0
|
Positive
|
15
|
12.5
|
Tumor grade
|
|
|
1
|
0
|
0
|
2
|
20
|
28
|
3
|
50
|
60
|
Stage
|
|
|
I
|
0
|
0
|
II
|
25
|
28
|
III
|
40
|
47
|
IV
|
100
|
100
|
EEA, endometrial endometrioid adenocarcinoma; SEA, serous endometrioid adenocarcinoma; LNM, lymph node metastasis |
Calibration and discrimination of the nomograms
Internal and external validations of the prognostic nomograms were performed. Internal validation in the training cohort showed that the C-index values for nomogram predictions of OS and CSS were 0.816 (95%CI: 0.811–0.820) and 0.850 (95%CI: 0.846–0.855), respectively. Similarly, the corresponding C-index values in the external validation cohort were 0.812 (95%CI: 0.806–0.819) and 0.847 (95%CI: 0.841–0.854). These results confirmed that our prognostic nomograms were reasonably accurate. The calibration curves estimating the 3- and the 5-year OS and CSS rates showed accordance between the nomogram-predicted and observed values in both the training and the validation groups (Fig. 3). Furthermore, predictive accuracy and differentiation of patients were compared between the nomogram and the clinical stage system. In the training group, the AUC of the established nomogram (AUC = 0.83, 95%CI: 0.71–0.87) to predict OS was significantly higher than that of the clinical stage (AUC = 0.73, 95%CI: 0.67–0.80, p < 0.01; Fig. 4A). While for the CSS, the AUC of the nomogram (AUC = 0.87, 95%CI: 0.85–0.90) was also greater than that of the clinical stage (AUC = 0.79, 95%CI: 0.71–0.84, p < 0.01; Fig. 4B). The analysis of the validation group showed similar results (Fig. 4C and Fig. 4D). The nomograms for OS and CSS showed superior discrimination as compared to the clinical stage both in the training and the validation cohorts.
Performance of the nomogram in stratifying the risk for EC patients
We determined the cutoff values by dividing the patients in the training cohort evenly into 5 subgroups after sorting by the total OS score (score: 0–45, 46–73, 74–103, 104–143, 144–342) and the total CSS score (score: 0-35.5, 36-60.5, 61-88.5, 89-132.5, 133–293), and each group represented a distinct prognosis (Fig. 5). As shown in Fig. 5A, the mortality of patients in the low-score group was 3.98%, and reached the 59.19% value in the high-score group. After applying the cutoff values to the groups of patients in the different clinical staging cohorts, stratification into different risk subgroups in both the training and validation groups ensured a significant distinction between Kaplan-Meier curves for OS and the CSS outcomes within each clinical stage category (Fig. 6 and supplementary Fig. 1). For patients in stage IA, the overall survival was very different in the different subgroups. As an example, in patients in stage IA whose OS score was greater than 144, the mortality was as high as 44.44% (Fig. 6A), and patients in stage III-IV whose OS score was less than 73, the mortality was lower than 7.83% (Fig. 6D). Using this methodology, we could not only select high-risk patients at low stages, but also distinguish low-risk patients at high stages.