Arch Gynecol Obstet DOI 10.1007/s00404-015-3985-6
GYNECOLOGIC ONCOLOGY
CPE overexpression is correlated with pelvic lymph node metastasis and poor prognosis in patients with early-stage cervical cancer Hong-wei Shen1 • Jin-feng Tan1 • Jian-hong Shang1 • Min-zhi Hou1 Jun Liu1 • Li He1 • Shu-zhong Yao1 • Shan-yang He1
•
Received: 5 October 2015 / Accepted: 7 December 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose Elevated carboxypeptidase E (CPE) levels play crucial roles in tumorigenesis and metastasis. This study investigated the expression and clinicopathological significance of CPE in early-stage cervical cancer. Methods Elevated carboxypeptidase E expression was analyzed using quantitative polymerase chain reaction and western blotting in normal cervical tissue, cervical cancer cell lines, and in cervical cancer tissues and adjacent noncancerous tissues (ANTs) from the same patient. Immunohistochemistry (IHC) was used to examine CPE expression in tissue samples from 112 patients with early-stage cervical cancer (FIGO stages Ia2–IIa2), 60 patients with cervical intraepithelial neoplasia, and 19 patients with normal cervical tissues (NCTs). Associations between CPE expression and prognostic and diagnostic factors were evaluated statistically. Results CPE expression was significantly higher in cervical cancer cell lines and tissues than in normal tissues and ANTs. Semi-quantitative analysis of IHC indicated that CPE gradually increased from CIN I to cervical cancer, but was
absent in NCTs. CPE expression was seen in 40.2 % (45/ 112) of the cervical cancer samples. CPE expression was significantly associated with FIGO stage (P = 0.003), tumor size (P = 0.012), stromal invasion (P \ 0.001), lymphovascular space invasion (P = 0.016), parametrial infiltration (P = 0.027), vaginal involvement (P = 0.007), postoperative adjuvant therapy (P = 0.024), recurrence (P \ 0.001), survival (P \ 0.001), and pelvic lymph node metastasis (PLNM) (P \ 0.001), and it was significantly higher in tissues from patients with PLNM than without PLNM. Logistic regression analysis identified high-level CPE expression as an independent risk factor for PLNM (P = 0.001). Patients with higher CPE expression had shorter overall survival duration than patients with lower CPE expression. Univariate and multivariate Cox-regression analyses suggested that high-level CPE expression is an independent prognostic factor for overall survival in early-stage cervical cancer. Conclusions High-level CPE expression was associated with a poor prognosis in early-stage cervical cancer. CPE may serve as a biomarker for predicting PLNM and survival in these patients.
H. Shen and J. Tan were joint first authors of this article.
Keywords CPE Cervical cancer Pelvic lymph node metastasis Prognosis
& Shu-zhong Yao
[email protected] & Shan-yang He
[email protected] Hong-wei Shen
[email protected] Jin-feng Tan
[email protected] 1
Department of Obstetrics and Gynecology, The First Affiliated Hospital, Sun Yat-sen University, No.58, Zhongshan 2 Road, Guangzhou 510080, People’s Republic of China
Introduction Cervical cancer is the third most frequently diagnosed cancer in women, with a recent estimate of 529,800 new cases and 275,100 deaths annually [1]. Pelvic lymph node metastasis (PLNM) is a major cause of cancer-related mortality [2, 3]. The 5-year survival of patients with earlystage cervical cancer and PLNM is lower than that of patients who are PLNM negative [3–5]. The preoperative
123
Arch Gynecol Obstet
determination of PLNM plays a significant role in treatment planning [6–8]. However, traditional imaging techniques and sentinel lymph node biopsy are not sufficiently reliable to predict PLNM or to guide therapeutic decision making [9–12]. FABP5 [13], CISD2 [14], GPX3 [15], maspin [16], and HOTAIR [17] were shown to be potentially useful markers for PLNM and prognostic indicators in cervical cancer, but novel biomarkers able to provide practical information regarding prognosis and therapeutic strategies are still needed. Carboxypeptidase E (CPE), also known as carboxypeptidase H or enkephalin invertase, is a multifunctional protein [18–22]. As a key carboxypeptidase, it is involved in the biosynthesis of peptide hormones and neurotransmitters in the neuroendocrine system [19]. CPE also plays an important role in the modulation of metabolic and glucose homeostasis, obesity, fertility, bone remodeling, neuroprotection, stress, sexual behavior, mood and emotional responses, learning, and memory [20–22]. A role for CPE in promoting tumorigenesis and metastasis in several cancer types has also been reported [23–26]. Bachtiary et al. analyzed the expression of CPE mRNA in 33 biopsies from 11 cervical cancer patients and found aberrant overexpression in cervical cancer, whereas expression was virtually absent in normal cervical tissues [27]. However, that study did not include immunohistochemistry (IHC) to confirm CPE expression at the protein level, and the clinical/prognostic significance of CPE in cervical cancer was not determined. In the present study, we evaluated CPE expression in cervical cancer cells and tissues. Our results suggest the potential of CPE as a prognostic biomarker based on its significant association with PLNM in patients with earlystage cervical cancer.
(FIGO) criteria. The last follow-up was carried out in December 2013, and the mean observation period was 69.06 ± 19.524 months (17–95 months). During that time there were 24 cancer-related deaths within the study population. As controls, 19 samples of normal cervical tissues (NCTs) from patients undergoing simple hysterectomy because of uterine leiomyomata were obtained. Additional controls consisted of samples from 20 patients with cervical intraepithelial neoplasia (CIN) I, 17 patients with CIN II who underwent biopsy, and 23 with CIN III who underwent hysterectomy. Eight fresh cervical SCC specimens and corresponding adjacent noncancerous tissues (ANTs) were collected from the cervical cancer patients without preoperative radiotherapy or chemotherapy. Both the cancer specimens and the ANTs were used for quantitative polymerase chain reaction (qPCR) and western blotting analysis. Prior written consent was obtained from the patients for the use of their clinical materials for research purposes. The study was approved by the Institutional Ethics Board of the First Affiliated Hospital of Sun Yat-sen University. The clinicopathological parameters are summarized in Table 1. Cell lines The cervical cancer cell lines HeLa, SiHa, C33A, and HCC94, expressing low levels of CPE, were cultured in Eagle’s minimum essential medium (Gibco BRL, Rockville, MD, USA). Caski cells, expressing high levels of CPE, were cultured in RPMI-1640 medium (Gibco BRL). The five cell lines were purchased from the American Type Culture Collection (ATCC, MD, USA). The media were supplemented with 10 % fetal bovine serum (HyClone Laboratories, Logan, UT, USA) and 1 % antibiotics (100 U penicillin/mL and 100 lg streptomycin/mL) in a 5 % CO2humidified atmosphere at 37 °C.
Materials and methods RNA extraction and quantitative RT-PCR Samples and patients This retrospective study enrolled 112 patients diagnosed with early-stage cervical squamous cell carcinoma (SCC) who underwent radical hysterectomy and lymphadenectomy in the Department of Obstetrics and Gynecology, the First Affiliated Hospital, Sun Yat-sen University (Guangdong, China), from January 2006 to December 2008. All enrolled patients had Ia2–IIa2 stage disease and available clinical follow-up data; none had undergone preoperative radiotherapy or chemotherapy. The mean patient age was 43.88 ± 8.613 years (range 27–68 years). Clinical stages were determined according to the International Federation of Obstetrics and Gynecology, 2009
123
Total RNA from cultured cells and fresh tissues was extracted using the Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. CDNA was synthesized using 2 lg of RNA from each sample and random hexamer primers. qPCR was used to measure the amounts of CPE mRNA. Each sample was tested in triplicate. Expression data were normalized to the geometric mean with reference to the housekeeping gene GAPDH. The CPE primer sequences were 50 -CCAT CAGCAGGATTTACACG-30 (forward) and 50 -TAAATT CAGGCTCACCAGGC-30 (reverse), and those for GAPDH were 50 -ACCACAGTCCATGCCATCAC-30 (forward) and 50 -TCCACCACCCTGTTGCTGTA-30 (reverse).
Arch Gynecol Obstet Table 1 Correlation between CPE expression and clinicopathologic features of early-stage cervical cancer (N = 112) Variable
N
Percentage (%)
CPE expression None or low (n = 67)
High (n = 45)
Chi-square value
P value
Age (years) \42
49
43.8
30
19
63
56.2
37
26
Ia2
12
10.7
10
2
Ib1
59
52.7
42
17
Ib2
19
17.0
8
11
IIa1
14
12.5
5
9
IIa2
8
7.1
2
6
\4
88
78.6
58
30
C4
24
21.4
9
15
G1 or G2
52
46.4
32
20
G3
60
53.6
35
25
C42 FIGO stage
0.071
0.789
15.691
0.003a
6.332
0.012
0.119
0.730
12.304
\0.001
Tumor size (cm)
Differentiation grade
Stromal invasion \1/2
55
49.1
42
13
C1/2 57 Lymphovascular space invasion
50.9
25
32
9
8.0
2
7
103
92.0
65
38
Yes No
5.756
0.016a
12.614
\0.001a
4.912
0.027a
7.244
0.007a
1.851
0.174a
5.085
0.024
22.895
\0.001
23.668
\0.001a
Pelvic lymph node metastasis Yes
18
16.1
4
14
No
94
83.9
63
31
6
5.4
1
5
106
94.6
66
40
Yes
10
8.9
2
8
No
102
91.1
65
37
6
5.4
2
4
106
94.6
65
41
Postoperative adjuvant therapy Yes 76
67.9
40
36
32.1
27
9
Parametrial infiltration Yes No Vaginal involvement
Positive surgical margins Yes No
No
36
Recurrence Yes
28
25.0
6
22
No
84
75.0
61
23
Alive
88
78.6
63
25
Death from SCC
24
21.4
4
20
Vital status at follow-up
a
Fisher’s exact test
123
Arch Gynecol Obstet
Western blotting Cell lysates from cell lines and fresh tissue were prepared using cold RIPA buffer. Total proteins (50 lg) were fractionated using SDS-PAGE and transferred onto a nitrocellulose membrane for western blotting [20] using an antiCPE antibody (1:1000, BD, Franklin Lakes, NJ, USA).
compared using the log-rank test of GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA, USA). Cox-regression analysis was carried out to assess factors related to survival. All statistical analyses were performed with the SPSS 13.0 software package (SPSS, Chicago, IL, USA). A P value \0.05 was considered to indicate statistical significance.
Immunohistochemical assay
Results Altered protein expression in samples from 112 human cervical cancer tissues, 60 patients with CIN, and 19 patients with NCTs was evaluated using IHC. Briefly, 4-lm-thick paraffin sections of cervical cancer tissue were heated at 65 °C for 30 min, deparaffinized with xylene, and then rehydrated. Antigen retrieval was achieved by submerging the sections in EDTA antigenic retrieval buffer. The samples were treated with 3 % hydrogen peroxide in methanol to quench endogenous peroxidase activity, followed by incubation with 1 % bovine serum albumin to block nonspecific binding. The sections were then incubated with an anti-CPE antibody (1:150, BD) overnight at 4 °C. Normal goat serum was used as a negative control. After a washing step, the tissue sections were incubated with a biotinylated anti-mouse secondary antibody followed by further incubation with streptavidin-horseradish peroxidase complex (both from Abcam). The tissue sections were then immersed in 3-amino-9-ethyl carbazole, counterstained with 10 % Mayer’s hematoxylin, dehydrated, and mounted in Crystal Mount. Two observers blinded to the histopathological features of the samples and the patient data independently evaluated the degree of immunostaining of the formalin-fixed paraffin-embedded sections. Scoring was based on both the proportion of positively stained tumor cells and the intensity of staining according to the following criteria: (a) percentage of positive tumor cells in the tumor tissue: 1 (\10 %), 2 (10–50 %), 3 (50–75 %), and 4 ([75 %); (b) staining intensity: 0 (none), 1 (weak), 2 (moderate), and 3 (strong). The staining index was calculated as: staining intensity score 9 proportion of positive tumor cells (range 0–12). The scores assigned by the two independent investigators were averaged. A final score C5 indicated high-level expression.
Elevated expression of CPE in cervical cancer cell lines and early-stage cervical cancer To investigate the characteristic of CPE expression in cervical carcinoma, we compared the CPE mRNA and protein profiles in different cervical cancer cell lines and samples. Western blotting and qPCR analyses revealed higher expression of CPE protein and mRNA, respectively, in all 5 cervical carcinoma cell lines than in NCT (Fig. 1). Furthermore, comparative analysis showed that CPE mRNA and protein levels were differentially up-regulated in all eight cervical carcinoma samples compared with the
Statistical analysis The relationship between CPE expression and the clinicopathological characteristics was assessed using the v2 test or Fisher’s exact test. The risk factors for PLNM in earlystage cervical cancer were established by logistic regression analysis. Disease-free survival and overall survival were estimated with the Kaplan–Meier method and
123
Fig. 1 Overexpression of CPE mRNA and protein in cervical cancer cell lines. The expression of CPE mRNA and protein in cervical cancer cell lines (Hela, SiHa, CasKi, C33A, HCC94) and NCT was examined using qPCR (a) and western blotting (b). Expression levels were normalized against GAPDH. Error bars represent the standard deviation of the mean (SD) calculated from three parallel experiments. *P \ 0.05
Arch Gynecol Obstet
cervical tissue and CIN and cervical cancer (P \ 0.001). (Fig. 3; Table 2) (P \ 0.001). Furthermore, the clinicopathologic analysis exhibited that CPE expression was strongly correlated with FIGO stage (P = 0.003), tumor size (P = 0.012), stromal invasion (P \ 0.001), LVSI (P = 0.016), parametrial infiltration (P = 0.027), vaginal involvement (P = 0.007), postoperative adjuvant therapy (P = 0.024), recurrence (P \ 0.001), survival (P \ 0.001), and PLNM (P \ 0.001). However, there were no significant associations between high-level CPE expression and other clinical features, including age, differentiation grade, and positive surgical margins (Table 1). Correlation of CPE expression with PLNM
Fig. 2 Overexpression of CPE mRNA and protein in cervical cancer tissues. a Average T/N ratios of CPE mRNA expression in paired cervical cancer (T) and adjacent non-tumor cervical tissues (ANTs) were quantified using qPCR and normalized against GAPDH. Error bars represent the SD calculated from three parallel experiments. b Western blotting of CPE protein in eight pairs of matched cervical cancer (T) and ANTs. *P \ 0.05
respective ANTs (Fig. 2). These results demonstrated that CPE is up-regulated in cervical cancer tissues. Association between CPE expression and clinical features in cervical carcinoma To further investigate the clinical significance of the above findings, CPE expression was examined by IHC in 112 paraffin-embedded, archived cervical cancer tissues, and in tissues from 20 cases of CIN I, 17 cases of CIN II, 23 cases of CIN III, and 19 NCTs. As shown in Figs. 3 and 4, the subcellular location of CPE was mainly in the nucleus and partly in the cytoplasm of the cancer cells. IHC revealed that negative or weak immunohistochemical staining was detected in the normal cervical tissues. The positive expression rate of CPE in CIN I, CIN II, and CIN III were 0 % (0/20), 11.8 % (2/17), and 13.0 % (3/23), the positive expression rate of CPE was 45/112 (40.2 %) in cervical cancer. CPE expression gradually increased in normal
Univariate logistic regression analysis revealed that tumor size (P = 0.013), stroma invasion (P = 0.006), LVSI (P = 0.026) vaginal involvement (P = 0.006), and CPE high expression (P = 0.001) were significant risk factors for PLNM. In a stepwise multivariate analysis, the independent risk factors for PLNM were LVSI (P = 0.046), CPE high expression (P = 0.010) (Table 3). Furthermore, CPE expression was significantly higher in the PLNM group than in the non-PLNM group (Table 1; Fig. 4). Thus, according to logistic regression analysis, high CPE expression was an independent risk factor for PLNM. Uni- and Multivariable analyses of clinicopathologic factors associated with 5-year OS and DFS rates of SCC patients Patient follow-up data was available for all 112 individuals of SCC. Duration of follow-up ranged from 17 to 95 months, with a median follow-up period of 69.06 months, and by the time of last contact 24 total deaths had occurred due to persistent or recurrent local tumors and distant metastases. The 5-year overall survival (OS) rate and disease-free survival (DFS) rate for the SCC patients were 78.6 % (88/112) and 75 % (84/112), respectively. To determine the prognostic significance of high CPE expression in patients with cervical cancer, the Kaplan– Meier analysis and the log-rank test were used to calculate the effect of CPE expression on DFS and OS. The log-rank test showed that DFS and OS were longer in patients with low-CPE-expressing tumors and shorter in those with high CPE-expressing tumors (Fig. 5) (P \ 0.0001). The cumulative 5-year DFS and OS rates for the patients with high levels of CPE expression were 51.1 and 55.6 %, respectively, whereas the rates were 91.0 and 94.0 %, respectively, for the patients with low or no CPE expression.
123
Arch Gynecol Obstet
Fig. 3 CPE protein expression as determined in an immunohistochemical assay of tissues from patients with cervical cancer. Among 112 patients, tissues from 36 showed positive CPE staining mainly in the nucleus and partly in the cytoplasm of cancer cells. CPE
expression gradually increased from cervical intraepithelial neoplasia (CIN) through to cervical cancer. CPE expression was undetectable in NCTs, marginal in CIN I, moderate in CIN II and CIN III, and strong in cervical cancer tissues
Fig. 4 Immunohistochemical validation of CPE protein expression in early-stage cervical cancer with and without PLNM. CPE expression was significantly higher in the PLNM group than in the NPLNM group
Table 2 CPE expression in normal cervical tissue (NCT), CIN I, CIN II, CIN III, and cervical cancer
N
None or low
High
Positive expression rate (%)
NCT
19
19
0
0
CIN I
20
20
0
0
CIN II
17
15
2
11.8
CIN III
23
20
3
13.0
112
67
45
40.2
Cervical cancer
123
CPE expression
Chi-square value
P value
Trend value
P value
28.918
\0.001
25.662
\0.001
Arch Gynecol Obstet Table 3 Univariate and multivariate logistic regression analyses of factors associated with PLNM (N = 112) Variables
HR (95 % CI)
of potential value as a biomarker for predicting the prognosis of patients with cervical cancer.
P value
Univariate analysis, N = 112
Discussion
Tumor size (cm) B4
1
[4
3.900 (1.332–11.417)
CPE expression None or low High Stromal invasion
0.013 0.001
1 7.113 (2.161–23.416) 0.006
B1/2
1
[1/2
6.190 (1.679–22.819)
Lymphovascular space invasion
0.026
No
1
Yes
5.086 (1.216–21.261)
Vaginal involvements
0.006
No
1
Yes
6.846 (1.741–26.926)
Multivariate analysis, N = 112 CPE expression
0.010
None or low
1
High
5.032 (1.464–17.294)
Stromal invasion B1/2 [1/2
0.046 1 4.004 (1.026–15.625)
Backward LR method was used for multivariate logistic regression analysis to drop insignificant variables. Variables entered for analysis were: FIGO stage, age, tumor size, differentiation grade, stromal invasion, vaginal involvements, parametrial infiltration, lymphovascular space invasion and CPE expression Univariate and multivariate logistic
Univariable analysis using log-rank tests showed that CPE expression (P \ 0.001), FIGO stage (P = 0.002), tumor size (P = 0.002), stromal invasion (P = 0.011), LVSI (P = 0.002), PLNM (P \ 0.001), parametrial infiltration (P = 0.001), vaginal involvement (P \ 0.001), and positive surgical margins (P = 0.014) as factors significantly associated with the 5-year DFS rate for 112 SCC patients. Multivariable analysis showed that CPE expression (P = 0.009), FIGO stage (P = 0.023), LVSI (P = 0.046), PLNM (P = 0.010) were factors significantly associated with the 5-year DFS rate for these patients. Both analyses determined CPE expression as being insignificant in affecting the 5-year DFS rate (Table 4). In addition, univariate and multivariate Cox-regression analyses identified CPE expression (P = 0.001), LVSI (P = 0.043) and PLNM (P = 0.045) as the independent prognostic factors significantly associated with the 5-year OS for 112 SCC patients (Table 4). Thus, according to our findings CPE is
The results of this study demonstrated the overexpression of CPE in cervical cancer cell lines and tissues at both the mRNA and protein levels. Immunostaining analysis showed a significant correlation between the level of CPE protein expression in histological sections and both PLNM and the reduced survival of patients with early-stage cervical cancer. In a logistic regression analysis, high CPE expression was an independent risk factor for PLNM, whereas in a multivariate analysis CPE expression was an independent prognostic indicator for poor outcome in patients with early-stage cervical cancer. Elevated levels of CPE mRNA have been demonstrated in several non-endocrine cancers, including cervical, colon, rectal, and renal cancers, based on data from the GEO database. In agreement with our findings, high levels of CPE mRNA were detected in cervical cancer, whereas expression was virtually absent in normal cervical tissues [27]. However, these studies did not examine the clinical/prognostic significance of CPE expression in cervical cancer. Thus, to our knowledge, ours is the first study to determine an association between high-level CPE expression and both a high rate of PLNM and a poor prognosis, suggesting the oncogene-like activity of CPE in early-stage cervical cancer. PLNM is associated with mortality in cervical cancer patients and its detection is important in determining the treatment approach [28]. In previous studies, LVSI, depth of stromal invasion, parametrial involvement, and serum SCCAg (squamous cell carcinoma antigen) were independent risk factors for PLNM [29, 30]. In the present work, high-level CPE expression and LVSI were independent risk factors for PLNM. Support for this relationship came from the higher rate of PLNM in patients whose tumors expressed high vs. low levels of CPE. High CPE expression characterized the tumors of 77.78 % (14/18) of patients in the PLNM group compared with 32.98 % (31/94) of those with low-level CPE expression. Therefore, determination of CPE expression levels in cervical tissues obtained by transvaginal biopsy can provide important information regarding the risk of PLNM and the preferred treatment modality for patients with treatment-naive early-stage cervical cancer. The potential of CPE as a biomarker for predicting PLNM and poor prognosis in early-stage cervical cancer is supported by previous studies showing the aberrant expression of CPE in several types of cancer [23–26]. In 2011, a bioinformatics analysis identified an N-terminal truncated splice variant of CPE (CPE-DN) [31]. CPE-DN was shown to promote the invasion and metastasis of tumor
123
Arch Gynecol Obstet
Fig. 5 Prognostic assessment of positive CPE expression in early-stage cervical cancer patients. The patients with higher CPE expression had a significantly shorter disease-free survival (a) and overall survival than those with high expression of CPE Table 4 Univariate and Multivariate analyses of different prognostic parameters in cervical cancer patients by Cox-regression analysis of the various factors associated with DFS and OS in SCC patients
N
Univariate analyses
Multivariate analyses
P
Regression coefficient (SE)
P
HR (95 % CI)
\0.001
6.884 (0.462)
0.009
3.804 (1.403–10.316)
0.002
3.317 (0.388)
0.023
2.715 (1.019–7.233)
0.002
3.357 (0.384)
0.154
1.086 (0.801–4.073)
0.011
2.882 (0.419)
0.835
1.104 (0.434–2.810)
0.002
4.366 (0.465)
0.046
2.715 (1.019–7.233)
\0.001
5.444 (0.386)
0.010
2.893 (1.292–6.480)
0.001
5.634 (0.500)
0.450
1.687 (0.435–6.550)
\0.001
4.805 (0.439)
0.885
0.888 (0.176–4.470)
0.014
3.771 (0.542)
0.929
1.059 (0.300–3.732)
Disease-free survival (DFS) CPE High
45
None or low
67
FIGO stage I stage II stage Tumor size (cm)
90 22
\4
88
C4
24
Stromal invasion B1/2
55
[1/2
57
Lymphovascular space invasion Yes No
9 103
Pelvic lymph node metastasis Yes
18
No
94
Parametrial infiltration Yes
6
No 106 Vaginal involvement Yes
10
No
102
Positive surgical margins
123
Yes
76
No
36
Arch Gynecol Obstet Table 4 continued
N
Univariate analyses
Multivariate analyses
P
Regression coefficient (SE)
P
HR (95 % CI)
\0.001
9.345 (0.548)
0.001
6.542 (2.128–20.117)
0.010
2.980 (0.422)
0.221
1.717 (0.722–4.082)
0.002
3.597 (0.411)
0.120
1.948 (0.841–4.515)
0.008
3.472 (0.472)
0.355
1.616 (0.585–4.464)
\0.001
5.387 (0.474)
0.043
2.736 (1.033–7.247)
\0.001
5.064 (0.416)
0.045
2.431 (1.020–5.793)
0.004
4.800 (0.550)
0.648
1.424 (0.313–6.478)
0.008
3.779 (0.504)
0.835
0.843 (0.169–4.213)
Overall survival (OS) CPE High
67
None or low
45
FIGO stage I stage
90
II stage
22
Tumor size (cm) \4
88
C4
24
Stromal invasion B1/2
55
[1/2
57
Lymphovascular space invasion Yes No
9 103
Pelvic lymph node metastasis Yes
18
No
94
Parametrial infiltration Yes No
6 106
Vaginal involvement Yes
10
No
102
Backward LR method was used for Cox-Regression analysis and total cases were 112. Variables entered for analysis were: FIGO stage, age, tumor size, differentiation grade, stromal invasion, vaginal involvements, parametrial infiltration, pelvic lymph node metastasis, lymphovascular space invasion, postoperative adjuvant therapy, recurrence, and CPE expression
cells by regulating the canonical Wnt signaling pathway or by up-regulating genes associated with metastatic progression, such as NEDD9 [32, 33] and NF-jB [34, 35]. In our study, we used generic primers that amplified both F-CPE and CPE-DN cDNA, rather than CPE-DN-specific primers. Furthermore, the anti-CPE used in immunostaining detects both F-CPE and CPE-DN. Thus, we were unable to determine the form of CPE involved in cervical cancer, which is the main limitation of our study. Thus, a prospective study comparing the expression profiles and functional roles of F-CPE vs. CPE-DN in patients with non-metastatic vs. metastatic cervical cancer is required. Topics of future research include whether the down-regulation of CPE inhibits cell invasion and elucidation of the role of CPE in promoting PLNM metastasis in cervical cancer. In summary, we found that in early-stage cervical cancer the overexpression of CPE was closely correlated with PLNM and poor prognosis. Thus, CPE may be a new
therapeutic target and a clinically novel indicator for predicting PLNM and survival. Acknowledgments This project was supported by the Guangdong Natural Science Foundation, China (Nos. S2013010015599; S2012010008270). Compliance with ethical standards Conflict of interest The authors declare that they have no conflicts of interest.
References 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D (2011) Global cancer statistics. CA Cancer J Clin 61:69–90 2. Gien LT, Covens A (2009) Lymph node assessment in cervical cancer: prognostic and therapeutic implications. J Surg Oncol 99:242–247
123
Arch Gynecol Obstet 3. Noguchi H, Shiozawa I, Sakai Y, Yamazaki T, Fukuta T (1987) Pelvic lymph node metastasis of uterine cervical cancer. Gynecol Oncol 27:150–158 4. Morice P, Castaigne D, Pautier P, Rey A, Haie-Meder C, Leblanc M, Duvillard P (1999) Interest of pelvic and paraaortic lymphadenectomy in patients with stage IB and II cervical carcinoma. Gynecol Oncol 73:106–110 5. Lai CH, Hong JH, Hsueh S, Ng KK, Chang TC, Tseng CJ, Chou HH, Huang KG (1999) Preoperative prognostic variables and the impact of postoperative adjuvant therapy on the outcomes of Stage IB or II cervical carcinoma patients with or without pelvic lymph node metastases: an analysis of 891 cases. Cancer 85:1537–1546 6. Cao DY, Yang JX, Wu XH et al (2013) Comparisons of vaginal and abdominal radical trachelectomy for early-stage cervical cancer: preliminary results of a multi-center research in China. Br J Cancer 109:2778–2782 7. Rob L, Pluta M, Strnad P, Hrehorcak M, Chmel R, Skapa P, Robova H (2008) A less radical treatment option to the fertilitysparing radical trachelectomy in patients with stage I cervical cancer. Gynecol Oncol 111:S116–S120 8. Landoni F, Maneo A, Colombo A et al (1997) Randomised study of radical surgery versus radiotherapy for stage Ib-IIa cervical cancer. Lancet 350:535–540 9. Selman TJ, Mann C, Zamora J, Appleyard TL, Khan K (2008) Diagnostic accuracy of tests for lymph node status in primary cervical cancer: a systematic review and meta-analysis. CMAJ 178:855–862 10. Mitchell DG, Snyder B, Coakley F et al (2009) Early invasive cervical cancer: MRI and CT predictors of lymphatic metastases in the ACRIN 6651/183 intergroup study. Gynecol Oncol 112:95–103 11. Lv K, Guo HM, Lu YJ, Wu ZX, Zhang K, Han JK (2014) Role of 18F-FDG PET/CT in detecting pelvic lymph-node metastases in patients with early-stage uterine cervical cancer: comparison with MRI findings. Nucl Med Commun 35:1204–1211 12. Altgassen C, Hertel H, Brandsta¨dt A, Ko¨hler C, Du¨rst M, Schneider A, AGO Study Group (2008) Multicenter validation study of the sentinel lymph node concept in cervical cancer: AGO Study Group. J Clin Oncol 26:2943–2951 13. Wang W, Jia HL, Huang JM, Liang YC, Tan H, Geng HZ, Guo LY, Yao SZ (2014) Identification of biomarkers for lymph node metastasis in early-stage cervical cancer by tissue-based proteomics. Br J Cancer 110:1748–1758 14. Liu L, Xia M, Wang J, Zhang W, Zhang Y, He M (2014) CISD2 expression is a novel marker correlating with pelvic lymph node metastasis and prognosis in patients with early-stage cervical cancer. Med Oncol 31:183 15. Zhang X, Zheng Z, Yingji S, Kim H, Jin R, Renshu L, Lee DY, Roh MR, Yang S (2014) Downregulation of glutathione peroxidase 3 is associated with lymph node metastasis and prognosis in cervical cancer. Oncol Rep 31:2587–2592 16. Liu Z, Shi Y, Meng W, Liu Y, Yang K, Wu S, Peng Z (2014) Expression and localization of maspin in cervical cancer and its role in tumor progression and lymphangiogenesis. Arch Gynecol Obstet 289:373–382 17. Huang L, Liao LM, Liu AW, Wu JB, Cheng XL, Lin JX, Zheng M (2014) Overexpression of long noncoding RNA HOTAIR predicts a poor prognosis in patients with cervical cancer. Arch Gynecol Obstet 290:717–723 18. Cawley NX, Wetsel WC, Murthy SR, Park JJ, Pacak K, Loh YP (2012) New roles of carboxypeptidase E in endocrine and neural function and cancer. Endocr Rev 33:216–253 19. Billova S, Galanopoulou AS, Seidah NG, Qiu X, Kumar U (2007) Immunohistochemical expression and colocalization of
123
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
somatostatin, carboxypeptidase-E and prohormone convertases 1 and 2 in rat brain. Neuroscience 147:403–418 Cawley NX, Zhou J, Hill JM et al (2004) The carboxypeptidase E knockout mouse exhibits endocrinological and behavioral deficits. Endocrinology 145:5807–5819 Cawley NX, Yanik T, Woronowicz A, Chang W, Marini JC, Loh YP (2010) Obese carboxypeptidase E knockout mice exhibit multiple defects in peptide hormone processing contributing to low bone mineral density. Am J Physiol Endocrinol Metab 299:E189–E197 Woronowicz A, Koshimizu H, Chang SY et al (2008) Absence of carboxypeptidase E leads to adult hippocampal neuronal degeneration and memory deficits. Hippocampus 18:1051–1063 Murthy SR, Pacak K, Loh YP (2010) Carboxypeptidase E: elevated expression correlated with tumor growth and metastasis in pheochromocytomas and other cancers. Cell Mol Neurobiol 30:1377–1381 Tang SS, Zhang JH, Liu HX, Li HZ (2009) PC2/CPE-mediated pro-protein processing in tumor cells and its differentiated cells or tissues. Mol Cell Endocrinol 303:43–49 Hong Y, Ho KS, Eu KW, Cheah PY (2007) A susceptibility gene set for early onset colorectal cancer that integrates diverse signaling pathways: implication for tumorigenesis. Clin Cancer Res 13:1107–1114 Krajnik M, Scha¨fer M, Sobanski P, Kowalewski J, Bloch-Boguslawska E, Zylicz Z, Mousa SA (2010) Enkephalin, its precursor, processing enzymes, and receptor as part of a local opioid network throughout the respiratory system of lung cancer patients. Hum Pathol 41:632–642 Bachtiary B, Boutros PC, Pintilie M et al (2006) Gene expression profiling in cervical cancer: an exploration of intratumor heterogeneity. Clin Cancer Res 12:5632–5640 Benedet JL, Bender H, Jones H 3rd, Ngan HY, Pecorelli S (2000) FIGO staging classifications and clinical practice guidelines in the management of gynecologic cancers. FIGO Committee on Gynecologic Oncology. Int J Gynaecol Obstet 70:209–262 Delgado G, Bundy B, Zaino R, Sevin BU, Creasman WT, Major F (1990) Prospective surgical-pathological study of disease-free interval in patients with stage IB squamous cell carcinomaof the cervix: a Gynecologic Oncology Group study. Gynecol Oncol 38:352–357 Sun JR, Zhang YN, Sun XM, Feng SY, Yan M (2011) Prediction model of pelvic lymph node metastasis in early stage cervical cancer and its clinical value. Minerva Chir 66:537–545 Lee TK, Murthy SR, Cawley NX et al (2011) An N-terminal truncated carboxypeptidase E splice isoform induces tumor growth and is a biomarker for predicting future metastasis in human cancers. J Clin Invest 121:880–892 Murthy SR, Dupart E, Al-Sweel N, Chen A, Cawley NX, Loh YP (2013) Carboxypeptidase E promotes cancer cell survival, but inhibits migration and invasion. Cancer Lett 341:204–213 Skalka N, Caspi M, Caspi E, Loh YP, Rosin-Arbesfeld R (2012) Carboxypeptidase E: a negative regulator of the canonical Wnt signaling pathway. Oncogene 32:2836–2847 Liu A, Shao C, Jin G, Liu R, Hao J, Shao Z, Liu Q, Hu X (2014) Downregulation of CPE regulates cell proliferation and chemosensitivity in pancreatic cancer. Tumour Biol 35:12459–12465 Sima N, Cheng X, Ye F, Ma D, Xie X, Lu¨ W (2013) The overexpression of scaffolding protein NEDD9 promotes migration and invasion in cervical cancer via tyrosine phosphorylated FAK and SRC. PLoS One 8:e74594
