Clinical Characteristics
Between 2011 and 2014, 503 patients with cervical cancer were clinically staged as 2018 FIGO stage III, in accordance with the inclusion criteria. We excluded 85 patients (27 for unfinised treatment, 3 combined with other malignant tumors, 32 without imaging finding of prompting lymph node or tumor size, 23 no treatment was given after recurrence), leaving 418 patients that were appropriate for analysis (Fig. 1). These patients were diagnosed with stages IIIA (n = 42, 10.0%), IIIB (n = 120, 28.7%), IIIC1 (n = 190, 45.5%), or IIIC2 (n = 66, 15.8%). Patients with stage IIIC1 were younger than those with stage IIIA (P < 0.001). The histological type, cell differentiation, tumor size, type of radiotherapy, and administration of consolidation chemotherapy among the different stages were balanced (each, P > 0.05). The differences between patients with stage IIIC1 or IIIC2 who underwent radical hysterectomy, minimally invasive surgery, or neoadjuvant chemotherapy were balanced (each P > 0.05) (Table 1). Pelvic lymphadenectomy was exclusively administered to 106 patients, and the median number of pelvic lymph nodes removed was 25 (range, 11–42). Sixty-two patients underwent pelvic and para-aortic lymphadenectomy, and the median number of para-aortic lymph nodes removed was 5 (range, 2–11).
Table 1
Patient demographics of 2018FIGO stage Ⅲ cervical cancer
Characteristic
|
Stage ⅢA
|
Stage ⅢB
|
Stage ⅢC1
|
Stage ⅢC2
|
p-value
|
N = 42
(10.0%)
|
N = 120
(28.7%)
|
N = 190
(45.5%)
|
N = 66
(15.8%)
|
Age(years)
|
|
|
|
|
0.003
|
< 65
|
30 (71.4%)
|
98 (81.7%)
|
174 (91.6%)
|
56 (84.8%)
|
|
≥ 65
|
12 (28.6%)
|
22 (18.3%)
|
16 (8.4%)
|
10 (15.2%)
|
|
Histology
|
|
|
|
|
0.680
|
Squamous
|
29 (69.0%)
|
96 (80.0%)
|
152 (80.0%)
|
49 (74.3%)
|
|
Adenocarcinoma
|
7 (16.7%)
|
13 (10.8%)
|
23 (12.1%)
|
8 (12.1%)
|
|
Adenosquamous Carcinoma
|
6 (14.3%)
|
11 (9.2%)
|
15 (7.9%)
|
9 (13.6%)
|
|
Grade
|
|
|
|
|
0.279
|
3
|
6 (14.3%)
|
27 (22.5%)
|
37 (19.4%)
|
11 (16.7%)
|
|
2
|
26 (61.9%)
|
68 (56.7%)
|
126 (66.3%)
|
46 (69.6%)
|
|
1
|
5 (11.9%)
|
5 (4.2%)
|
6 (3.2%)
|
4 (6.1%)
|
|
Unknown
|
5 (11.9%)
|
20 (16.6%)
|
21 (11.1%)
|
5 (7.6%)
|
|
Tumor size(cm)
|
|
|
|
|
0.727
|
< 4
|
21 (50.0%)
|
49 (40.8%)
|
87 (45.8%)
|
30 (45.5%)
|
|
≥ 4
|
21 (50.0%)
|
71 (59.2%)
|
103 (54.2%)
|
36 (54.5%)
|
|
Neoadjuvant chemotherapy
|
|
|
|
|
|
Yes
|
0
|
0
|
51 (26.8%)
|
21 (31.8%)
|
0.439 #
|
No
|
42
|
120
|
139 (73.2%)
|
45 (68.2%)
|
|
Radical Hysterectomy
|
|
|
|
|
0.319 #
|
Yes
|
0
|
0
|
128 (67.4%)
|
40 (60.6%)
|
|
No
|
42
|
120
|
62 (32.6%)
|
26 (39.4%)
|
|
Minimally invasive surgery
|
|
|
|
|
0.524 #
|
Yes
|
0
|
0
|
23 (18.0%)
|
9 (22.5%)
|
|
No
|
42
|
120
|
105 (82.0%)
|
31 (77.5%)
|
|
Radiotherapy
|
|
|
|
|
0.555
|
3-DCRT
|
19 (45.2 %)
|
48 (40.0%)
|
88 (46.3 %)
|
25 (37.9%)
|
|
IMRT
|
23 (54.8 %)
|
72 (60.0%)
|
102 (53.7%)
|
41 (62.1%)
|
|
Consolidation chemotherapy
|
|
|
|
|
0.765
|
Yes
|
31 (73.8%)
|
89 (74.2%)
|
147 (77.4%)
|
47 (71.2%)
|
|
No
|
11 (26.2%)
|
31 (25.8%)
|
43 (22.6%)
|
19 (28.8%)
|
|
3- DCRT, 3-Dimensional Conformal Radiotherapy; IMRT, Intensity Modulated Radiotherapy; # Stage ⅢC1 vs ⅢC2 |
Survival Outcomes
The 5-year OS rates were 54.1% for stage IIIC1 vs 43.3% for stage IIIA vs 40.6% for stage IIIB vs 23.1% for stage IIIC2 (P < 0.001) (Fig. 2a). The 5-year OS rate in patients with stage IIIC1 was higher, but not statistically significant compared with stages IIIA (P = 0.484) and IIIB (P = 0.042). Stage IIIC2 cervical cancer patients experienced significantly shorter OS compared with those with stage IIIC1, IIIA, or IIIB (each, P < 0.001). Increasing tumor size was associated with shorter 5-year OS rates as follows: 69.4%, tumor diameter < 4 cm; 36.5%, tumor diameter ≥ 4 cm (P < 0.001). Patients undergoing radical hysterectomy had a significantly higher 5-year OS rate (72.1%) compared with those who did not (48.3%) (P < 0.001); However, the histological type, cell differentiation, neoadjuvant chemotherapy, type of radiotherapy, minimally invasive surgery, or consolidation chemotherapy did not affect the 5-year OS rate (each, P > 0.05) (Table 2). Variables with P ≤ 0.1 in the univariate analysis were included in the multivariable analysis, the results of which are as follows: 2018 FIGO stage III was an independent risk factor. Stage IIIC2 was significantly associated with an increased risk of mortality compared with stage IIIC1 ((hazard ratio HR = 2.066; 95% confidence interval [CI]: 1.438 to 2.969; P < 0.001), stage IIIB (HR = 2.606; 95% CI: 1.752 to 3.877; P < 0.001), and stage IIIA (HR = 2.958; 95% CI: 1.757 to 4.983; P < 0.001). Stage IIIC1 was not significantly associated with an increased risk of mortality compared with stage IIIA (HR = 1.432; 95% CI: 0.867 to 2.366; P = 0.161), and stage IIIB (HR = 1.261; 95% CI: 0.871 to 1.827; P = 0.219) (Table 2).
Table 2
Univariate and multivariable analysis for survival in stage III cohort
Characteristic
|
No.
|
Survival rate (%)
|
|
Multivariable
(vs stage ⅢA)
|
Multivariable
(vs stage ⅢB)
|
Multivariable
(vs stage ⅢC1)
|
5-year
|
P-value
|
HR
(95%CI)
|
P-value
|
HR
(95%CI)
|
P-value
|
HR
(95%CI)
|
P-value
|
Age(years)
|
|
|
0.806
|
|
|
|
|
|
|
< 65
|
358
|
46.4
|
|
|
|
|
|
|
|
≥ 65
|
60
|
39.6
|
|
|
|
|
|
|
|
Histology
|
|
|
0.402
|
|
|
|
|
|
|
Squamous
|
326
|
45.8
|
|
|
|
|
|
|
|
Adenocarcinoma
|
51
|
42.9
|
|
|
|
|
|
|
|
Adenosquamous
|
41
|
34.6
|
|
|
|
|
|
|
|
Grade
|
|
|
0.481
|
|
|
|
|
|
|
3
|
81
|
47.4
|
|
|
|
|
|
|
|
2
|
266
|
46.0
|
|
|
|
|
|
|
|
1
|
20
|
48.0
|
|
|
|
|
|
|
|
Unknown
|
51
|
44.4
|
|
|
|
|
|
|
|
FIGO stage
|
|
|
< 0.001
|
|
|
|
|
|
|
ⅢA
|
42
|
43.3
|
|
1.000
|
|
0.881
(0.551–1.408)
|
0.532
|
0.698
(0.423–1.154)
|
0.082
|
ⅢB
|
120
|
40.6
|
|
1.135
(0.710–1.814)
|
0.596
|
1.000
|
|
0.793
(0.547–1.148)
|
0.121
|
ⅢC1
|
190
|
54.1
|
|
1.432
(0.867–2.366)
|
0.161
|
1.261
(0.871–1.827)
|
0.219
|
1.000
|
|
ⅢC2
|
66
|
23.1
|
|
2.958
(1.757–4.983)
|
< 0.001
|
2.606
(1.752–3.877)
|
< 0.001
|
2.066
(1.438–2.969)
|
< 0.001
|
Tumor size (cm)
|
|
|
< 0.001
|
|
|
|
|
|
|
< 4
|
187
|
69.4
|
|
1
|
0.001
|
1
|
0.001
|
1
|
0.001
|
≥ 4
|
231
|
36.5
|
|
1.554
(1.191–2.029)
|
|
1.554
(1.191–2.029)
|
|
1.554
(1.191–2.029)
|
|
Neoadjuvant chemotherapy
|
|
|
0.271
|
|
|
|
|
|
|
Yes
|
72
|
62.3
|
|
|
|
|
|
|
|
No
|
346
|
53.9
|
|
|
|
|
|
|
|
Radical hysterectomy
|
|
|
< 0.001
|
|
|
|
|
|
|
Yes
|
168
|
72.1
|
|
1
|
< 0.001
|
1
|
< 0.001
|
1
|
< 0.001
|
No
|
250
|
48.3
|
|
0.450
(0.319–0.635)
|
|
0.450
(0.319–0.635)
|
|
0.450
(0.319–0.635)
|
|
Minimally invasive surgery
|
|
|
0.752
|
|
|
|
|
|
|
Yes
|
32
|
70.9
|
|
|
|
|
|
|
|
No
|
136
|
73.5
|
|
|
|
|
|
|
|
Radiotherapy
|
|
|
0.083
|
|
|
|
|
|
|
3-DCRT
|
180
|
46.5
|
|
1
|
0.654
|
1
|
0.654
|
1
|
0.654
|
IMRT
|
238
|
54.1
|
|
1.071
(0.793–1.447)
|
|
1.071
(0.793–1.447)
|
|
1.071
(0.793–1.447)
|
|
Consolidation chemotherapy
|
|
|
0.852
|
|
|
|
|
|
|
Yes
|
314
|
45.9
|
|
|
|
|
|
|
|
No
|
104
|
45.0
|
|
|
|
|
|
|
|
3-DCRT, 3-Dimensional Conformal Radiotherapy; IMRT, Intensity Modulated Radiotherap |
We therefore believed it reasonable to conclude that tumor size and local invasion affected prognosis. To test this hypothesis, we stratified patients diagnosed with 2018 FIGO stage IIIC1 on the basis of T stage (T1/T2/T3) and compared their survival outcomes. Tumor staging for cervical cancer was carried out according to the American Joint Committee on Cancer (Version 9). T1, carcinoma is strictly confined to the cervix; T2, carcinoma invades beyond the uterus but has not extended onto the lower one-third of the vagina or to the pelvic wall; T3, carcinoma involves the lower one-third of the vagina and/or extends to the pelvic wall and/or causes hydronephrosis or nonfun-tcioning kidney [13]. The 5-year OS rates of T1, T2, and T3 were 72.2%, 54.1%, and 18.6%, respectively (P < 0.001) (Fig. 2b). There was no significant difference between the 5-year OS rates associated with age, histological type, cell differentiation, type of radiotherapy, minimally invasive surgery, or consolidation chemotherapy (each, P > 0.05). Factors with P ≤ 0.1 on univariate analysis were included for further multivariate analysis, which revealed that T-stage was an independent prognostic factor associated with survival of patients with stage IIIC1 (P < 0.001) (Table 3). Patients with stage IIIC1(T1) experienced longer 5-year OS compared with those with stages IIIA (P = 0.004) or IIIB (P < 0.001). Survival of patients with stage IIIC1(T2) was similar compared with patients with stage IIIA (P = 0.522) or IIIB (P = 0.133). Patients with stage IIIC1(T3) experienced shorter 5-year OS compared with those with stage IIIA or IIIB (each, P = 0.001) (Fig. 2c).
Table 3
Univariate and multivariable analysis for survival in stage IIIC1 cohort
Characteristic
|
No.
|
Survival rate (%)
|
|
Multivariable
|
5- year
|
P-value
|
HR(95%CI)
|
P-value
|
Age(years)
|
|
|
0.311
|
|
|
< 65
|
174
|
52.9
|
|
|
|
≥ 65
|
16
|
64.3
|
|
|
|
Histology
|
|
|
0.228
|
|
|
Squamous
|
152
|
58.6
|
|
|
|
Adenocarcinoma
|
23
|
55.2
|
|
|
|
Adenosquamous
|
15
|
33.3
|
|
|
|
Grade
|
|
|
0.745
|
|
|
3
|
37
|
48.9
|
|
|
|
2
|
126
|
54.7
|
|
|
|
1
|
6
|
55.9
|
|
|
|
Unknown
|
21
|
52.4
|
|
|
|
T stage
|
|
|
< 0.001
|
|
< 0.001
|
T1
|
87
|
72.2
|
|
1
|
|
T2
|
58
|
54.1
|
|
2.189 (1.197–4.005)
|
0.011
|
T3
|
45
|
18.6
|
|
5.085 (2.827–9.147)
|
< 0.001
|
Tumor size (cm)
|
|
|
0.003
|
|
0.015
|
< 4
|
87
|
65.9
|
|
1
|
|
≥ 4
|
103
|
44.1
|
|
1.735 (1.112-2.70)
|
|
Neoadjuvant chemotherapy
|
|
|
|
|
|
Yes
|
51
|
61.3
|
|
|
|
No
|
139
|
54.2
|
|
|
|
Radical hysterectomy
|
|
|
< 0.001
|
|
0.054
|
No
|
62
|
34.9
|
|
1
|
|
Yes
|
128
|
63.5
|
|
0.353 (0.122–1.018)
|
|
Minimally invasive surgery
|
|
|
0.159
|
|
|
Yes
|
23
|
67.2
|
|
|
|
No
|
167
|
52.4
|
|
|
|
Radiotherapy
|
|
|
0.084
|
|
0.216
|
3-DCRT
|
88
|
47.8
|
|
1
|
|
IMRT
|
102
|
56.2
|
|
0.751 (0.478–1.182)
|
|
Consolidation chemotherapy
|
|
|
0.469
|
|
|
Yes
|
147
|
54.9
|
|
|
|
No
|
43
|
51.5
|
|
|
|
3-DCRT, 3-Dimensional Conformal Radiotherapy; IMRT, Intensity Modulated Radiotherapy |
We further investigated the influence of the number of pelvic lymph node metastases (PLNMs) on the prognosis of patients with stage IIIC1p (T1/T2a) undergoing radical hysterectomy and pelvic lymphadenectomy with or without paraortic lymphadenectomy. Metastatic pelvic lymph nodes were evaluated using histopathology. The number of PLNMs per patient ranged between 1 and 9. The 25th, 50th, and 75th percentiles represented 1, 2, and 3 PLNMs, respectively. The median value of the number of PLNMs represents the optimal prognostic cut-off value. The 5-year OS rate of patients with 1 PLNM was 89.1% vs 47.5% for those with > 1 PLNM (P < 0.001). Patients with > 2 PLNMs experienced shorter 5-year OS rates compared with patients harboring 1 or 2 PLNMs (5-year OS rates, 47.9% vs 76%, respectively, P < 0.001). The 5-year OS rates of patients with > 3 PLNMs were not significantly different from those with 1 to 3 PLNMs (59.3% vs 65.1%, respectively, P = 0.337). Univariate analysis revealed that the survival of patients with IIIC1p(T1/T2a) differed significantly as a function of the number of PLNMs. Thus, the 5-year OS rate of patients with stage IIIC1pN1-2 group was significantly higher compared with IIIC1pN > 2 group (76% vs 47.9%, respectively, P < 0.001) (Fig. 2d). However, age, histology, cell differentiation, type of radiotherapy, and minimally invasive surgery did not have a significant effect on survival (each, P > 0.05) (Table 4). Analysis of multiple factors revealed that the number of PLNMs was an independent risk factor of prognosis. The mortality risk of patients with stage IIIC1pN > 2 was 2.75-fold higher compared with stage IIIC1pN1-2 (HR = 2.753; 95% CI: 1.527 to 4.965; P = 0.001) (Table 4). Patients with stage IIIC1pN1-2 experienced longer 5-year OS rate compared with those with stages IIIA (P = 0.01) and IIIB (P < 0.001). The 5-year survival rates of patients with stage IIIC1pN > 2 was not significantly different compared with patients with stage IIIA (P = 0.836) or IIIB (P = 0.668) (Fig. 2e).
Table 4
Univariate and multivariable analysis for survival in stage IIIC1p (T1/T2a) cohort
Characteristic
|
No.
|
Survival rate (%)
|
Multivariable
|
5-year
|
P-value
|
HR (95% CI)
|
P-value
|
Age (years)
|
|
|
0.266
|
|
|
< 65
|
116
|
62.7
|
|
|
|
≥ 65
|
12
|
79.1
|
|
|
|
Histology
|
|
|
0.665
|
|
|
squamous
|
109
|
65.8
|
|
|
|
other
|
19
|
63.2
|
|
|
|
Grade
|
|
|
0.827
|
|
|
G1-2
|
94
|
65.2
|
|
|
|
G3
|
23
|
60.9
|
|
|
|
Unknown
|
11
|
71.4
|
|
|
|
T stage
|
|
|
< 0.001
|
|
0.001
|
T1
|
87
|
75.1
|
|
1
|
|
T2a
|
41
|
46.5
|
|
2.659 (1.487–4.755)
|
|
Lymph node metastasis
|
|
|
< 0.001
|
|
0.001
|
N1-2
|
80
|
76
|
|
1
|
|
N > 2
|
48
|
47.9
|
|
2.753 (1.527–4.965)
|
|
Bilateral lymph node metastasis
|
|
|
0.825
|
|
|
Yes
|
27
|
69.6
|
|
|
|
No
|
101
|
64.4
|
|
|
|
Tumor size(cm)
|
|
|
0.012
|
|
0.059
|
< 4
|
66
|
77.4
|
|
1
|
|
≥ 4
|
62
|
54.1
|
|
1.804 (0.979–3.323)
|
|
Neoadjuvant chemotherapy
|
|
|
0.218
|
|
|
Yes
|
51
|
59.5
|
|
|
|
No
|
77
|
66.5
|
|
|
|
Minimally invasive surgery
|
|
|
0.364
|
|
|
Yes
|
23
|
63.6
|
|
|
|
No
|
105
|
73.9
|
|
|
|
Radiotherapy
|
|
|
0.937
|
|
|
3-DCRT
|
23
|
65.2
|
|
|
|
IMRT
|
105
|
65.5
|
|
|
|
Consolidation chemotherapy
|
|
|
0.027
|
|
0.071
|
Yes
|
111
|
68.3
|
|
0.507(0.242–1.060)
|
|
No
|
17
|
47.1
|
|
1
|
|
3-DCRT, 3-Dimensional Conformal Radiotherapy; IMRT, Intensity Modulated Radiotherapy |