Histopathology 2016 DOI: 10.1111/his.12933 Overexpression of eukaryotic initiation factor 5A2 (EIF5A2) is associated with cancer progression and poor prognosis in patients with early-stage cervical cancer Shan-Shan Yang, Ying Gao, 1 De-Ying Wang, 2 Bai-Rong Xia, 1 Yun-Duo Liu, 1 Yu Qin, 3 Xiao-Ming Ning, 3 Gen-Ying Li, Li-Xiao Hao, Min Xiao 4 & Yun-Yan Zhang Department of Gynaecological Radiotherapy, 1 Department of Gynaecology, The Affiliated Tumour Hospital of Harbin Medical University, 2 Department of Gynecology and Obstetrics, The Second Affiliated Hospital of Harbin Medical University, 3 Department of Pathology, and 4 Department of Breast Surgery, The Affiliated Tumour Hospital of Harbin Medical University, Harbin, China Date of submission 11 December 2015 Accepted for publication 16 January 2016 Published online Article Accepted 22 January 2016 Yang S-S, Gao Y, Wang D-Y, Xia B-R, Liu Y-D, Qin Y, Ning X-M, Li G-Y, Hao L-X, Xiao M & Zhang Y-Y (2016) Histopathology. DOI: 10.1111/his.12933 Overexpression of eukaryotic initiation factor 5A2 (EIF5A2) is associated with cancer progression and poor prognosis in patients with early-stage cervical cancer Aims: As one of the only two isoforms of the eukaryotic initiation factor (EIF)5A family, EIF5A2 plays an important role in tumour progression and prognosis evaluation. The aim of this study was to investigate EIF5A2 expression in International Federation of Gynecology and Obstetrics (FIGO) stage I II cervical cancer and to evaluate its clinical significance. Methods and results: The mrna and protein expression levels of EIF5A2 were analysed in 20 tissue samples of FIGO stage I II cervical cancer and paired surrounding non-tumour cervical tissues by real-time polymerase chain reaction and western blot analysis. Immunohistochemistry was performed to examine EIF5A2 protein expression in paraffin-embedded tissues from 314 patients with cervical cancer. The mrna and protein expression levels of EIF5A2 were significantly elevated in tumour tissues. The increased EIF5A2 expression was correlated with higher Keywords: cervical cancer, eukaryotic initiation factor 5A2, prognosis, progression FIGO stage (P < 01), deep cervical stromal invasion (P = 26), lymphovascular space involvement (P = 02), pelvic lymph node metastasis (P < 01) and postoperative recurrence (P < 01) in patients with cervical cancer. Patients with tumours showing high EIF5A2 expression had a poorer survival time than those with normal EIF5A2 expression, especially the patients with negative pelvic lymph nodes and FIGO stage II. In addition, multivariate Cox analysis showed that high EIF5A2 expression was an independent prognostic factor for overall survival [hazard ratio 1.949; 95% confidence interval (CI) 1.116 3.404; P = 19] and disease-free survival (hazard ratio 1.980; 95% CI 1.189 3.297; P = 09). Conclusions: may contribute to cancer progression and poor prognosis. Therefore, EIF5A2 could be a novel potential prognostic marker for FIGO stage I II cervical cancer. Address for correspondence: Dr Y-Y Zhang, Department of Gynaecological Radiotherapy, The Affiliated Tumour Hospital of Harbin Medical University, Baojian Road 6, Nangang District, Harbin 150081, China. e-mail: zhangyunyan_1972@163.com; and Dr M Xiao, Department of Breast Surgery, The Affiliated Tumour Hospital of Harbin Medical University, Baojian Road 6, Nangang District, Harbin 150081, China. e-mail: doctor_xm@126.com Shan-Shan Yang and Ying Gao contributed equally to this work. Introduction Cervical cancer is the fourth most common malignancy and the fourth most frequent cause of cancerrelated mortality in women worldwide. 1 With its constantly increasing frequency, especially in developing countries, an estimated 500 000 new cases of cervical cancer are diagnosed annually, accounting for 2016 John Wiley & Sons Ltd.
2 S-S Yang et al. 275 000 deaths worldwide. 2 Infection with oncogenic human papillomavirus (HPV) contributes to the development of the precursors of cervical cancer, 3,4 Although HPV infection is necessary, viral infection alone is insufficient to cause cervical cancer, 5 and the molecular mechanisms underlying cervical carcinogenesis remain elusive. Therefore, the identification of potential markers that may be involved in carcinogenesis and the progression of cervical cancer might play a significant role in understanding the underlying molecular mechanisms and providing a valuable therapeutic target. Eukaryotic initiation factor (EIF)5A2 (EIF5A2) is a member of the EIF5A family, and the EIF5A2 gene is located at human chromosome 3q26.2. 6 EIF5A2 was first discovered in a primary ovarian cancer cell line in 2001, and has been classified as an oncogene. 7 Several studies have demonstrated that EIF5A2 is upregulated in various human cancers and has clinical relevance for prognosis. 8 13 However, the available information on EIF5A2 expression and its prognostic significance in human cervical cancer is limited. The aim of the current study was to investigate EIF5A2 expression in surgically treated early-stage cervical cancer, and to analyse the relevance of EIF5A2 to the clinicopathological factors and prognosis of cervical cancer patients. Materials and methods PATIENTS AND TISSUE SPECIMENS In our study, formalin-fixed paraffin-embedded samples were obtained from 314 patients with FIGO stage I II cervical cancer who underwent surgical resection (radical hysterectomy and pelvic and/or paraaortic lymphadenectomy) at the Department of Gynaecology, the Affiliated Tumour Hospital of Harbin Medical University, China, between 1 January 2008 and 1 January 2010. None of the included patients received chemotherapy, radiotherapy, or immunotherapy before surgery. Patients with paraaortic lymph node metastasis were excluded. Another 20 pairs of the fresh cervical tumour tissues (which met the abovementioned criteria) and their corresponding nontumour cervical tissues were obtained (between 1 January 2014 and 31 May 2014) and frozen in liquid nitrogen for quantitative real-time reverse transcription polymerase chain reaction (qrt-pcr) and western blot analysis. Data regarding stage were assessed by vaginorectoabdominal examination under general anaesthesia according to the International Federation of Gynecology and Obstetrics (FIGO) staging system. 14 The histopathological subtype and histopathological grade were classified according to the World Health Organization criteria (2004). The largest primary tumour diameter in each sample was determined on the basis of the pathology reports: a tumour size of 40 mm was used as the cutoff, based on the FIGO staging system. 14 Postoperative radiotherapy was given to patients with high-risk pathological factors according to the pathological results. A total dose of 45 5 Gy was delivered at 1.8 Gy per daily fraction. All 314 patients were followed up for an interval of 5 84 months (median, 68.50 months). Overall survival (OS) was defined as the period from the date of surgery until death or until the time of the most recent follow-up. Patients who died of other diseases were excluded. Disease-free survival (DFS) was defined as the period from the date of surgery to recurrence. The study was approved by the Institute Research Medical Ethics Committee of Harbin Medical University. Written informed consent was obtained from each patient. qrt-pcr Total RNA samples from primary tissues were extracted with TRIzol reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer s instructions. The extracted RNA was pretreated with RNasefree DNase, and 1 lg of RNA from each sample was used for cdna synthesis with random hexamers. The cdna products were subjected to qrt-pcr with Super M-MLV Reverse Transcriptase (PR6502; BioTeke, Beijing, China). The following primer sequences were used: EIF5A2-F, 5 0 -AAGATGGTTACCTTTCCCTG-3 0 ; EIF5A2-R, 5 0 -TACAGCATATTCTTCACTCATTG-3 0 ; b- actin-f, 5 0 -CTTAGTTGCGTTACACCCTTTCTTG-3 0 ; and b-actin-r, 5 0 -CTGTCACCTTCACCGTTCCAGTTT-3 0. The initial polymerase chain reaction (PCR) amplification was performed as follows: 95 C for 10 min, followed by 40 cycles of 95 C for 10 s, 60 C for 20 s, and 72 C for 30 s. The PCR amplification was performed in triplicate. The qrt-pcr results were calculated with the 2 DDC(t) method. 15 WESTERN BLOT ANALYSIS Total protein was extracted and separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis, and then transferred onto poly(vinylidene difluoride) membranes (Millipore, Billerica, MA, USA). The blotted membranes were incubated with the EIF5A2
The role of EIF5A2 in early-stage cervical cancer 3 rabbit monoclonal antibody (ab150439; Abcam, Cambridge, MA, USA) at a dilution of 1:1000, and this was followed by incubation with the horseradish peroxidase-labelled secondary antibody. b-actin protein expression was also determined with a specific antibody (WL0001; Wanleibio, Shenyang, China) as a loading control. IMMUNOHISTOCHEMICAL STAINING AND EVALUATION A total of 314 pairs of formalin-fixed paraffinembedded tumour tissues and 20 corresponding samples of non-tumour tissues were used for EIF5A2 immunohistochemical staining. After deparaffinization and blocking, the antigen antibody reaction was incubated overnight at 4 C. Diaminobenzidine was applied to detect the signal from the antigen antibody reaction. All sections were counterstained with haematoxylin. The primary anti-eif5a2 rabbit monoclonal antibody (ab150439; Abcam) was used at a dilution of 1:100. As a negative staining control, the primary antibody was replaced with phosphate-buffered saline; ovarian tumour tissue with positive EIF5A2 expression was used as the positive control. 8 Two independent pathologists who were blinded to the clinicopathological information and corresponding haematoxylin and eosin-stained slides of patients evalued the immunohistochemical staining of EIF5A2. A semiquantitative scoring method was used, according to previous studies, 9,13 which was based on the staining intensity (0, negative staining; 1, weak staining; 2, moderate staining; 3, strong staining) and the proportion of immunopositive cells (1, <25%; 2, 25 50%; 3, 50 75%; 4, 75%). The final score for EIF5A2 expression was the product of the two above-mentioned scores, ranging from 0 to 12. For the statistical analysis, a final staining index of 0 3 represented normal EIF5A2 expression, whereas a staining index of 4 12 represented EIF5A2 overexpression. Any score discrepancies were reviewed by the original two pathologists and a senior pathologist until a consensus was reached. STATISTICAL ANALYSIS All analyses were performed with SPSS 13.0 (SPSS, Chicago, IL, USA). Numerical variables were recorded as means standard deviations (SDs), and analysed with Student s t-test. Categorical variables were analysed with the v 2 test. Survival curves were constructed with the Kaplan Meier method, and evaluated with the log-rank test. A Cox proportional hazard regression model was used to identify factors that were independently associated with OS and DFS. All P-values were two-tailed, and P-values of 5 were considered to indicate statistical significance. Results EIF5A2 EXPRESSION IN CERVICAL CANCER For the 20 paired tumour tissues and corresponding non-tumour tissues, EIF5A2 mrna expression in tumour tissues was significantly higher than that in non-tumour tissues. The mean expression value of EIF5A2 mrna in tumour tissues was 2.5308 0.9444 (mean SD, normalized by b-actin gene expression), which was significantly higher than that in the corresponding adjacent non-tumour tissues (1.2340 0.5952, P < 01, Student s t-test); a 2.05-fold difference between these mean values. Variations in EIF5A2 mrna expression are shown in Figure 1A,B. EIF5A2 was detected as an ~17-kDa band by western blot analysis (Figure 2A). Consistent with the mrna expression, EIF5A2 protein expression in tumour tissues was also significantly higher than that in corresponding non-tumour tissues (0.3566 931 versus 378 639, P = 001; Figure 2B). These results suggested that EIF5A2 is overexpressed in cervical tumour tissues, and this finding was further confirmed through immunohistochemical analysis. RELATIONSHIP BETWEEN CLINICOPATHOLOGICAL VARIABLES AND EIF5A2 IN PATIENTS WITH FIGO STAGE I II CERVICAL CANCER Representative immunohistochemical staining of EIF5A2 in tumour tissues and adjacent non-tumour tissues is shown in Figure 3A F. Positive EIF5A2 staining in tumour tissues was mainly localized within the cytoplasm of tumour cells. EIF5A2 expression was elevated in tumour tissues as compared with adjacent non-tumour tissues, and the difference was significant (P < 01; Table 1). was detected in 201 tumour tissue samples (64.0%, 201/314), but in only three adjacent nontumour tissue samples (15.0%, 3/20). As shown in Table 2, was correlated with higher FIGO stage (P < 01), deep cervical stromal invasion (P = 26), lymphovascular space involvement (P = 02), pelvic lymph node metastasis (P < 01), and postoperative recurrence (P < 01). However, no significant correlation was
4 S-S Yang et al. A Relative EIF5A2 mrna expression B 6.00 5.00 4.00 3.00 2.00 0 0 EIF5A2 mrna expression relative to β-action 1 2 5 4 3 2 1 0 Normal 3 4 5 6 7 8 9 10 11 Normal Tumour 12 13 14 15 16 17 18 19 20 * Tumour Figure 1. Eukaryotic initiation factor (EIF)5A2 mrna expression in early-stage cervical cancer. A, EIF5A2 mrna levels were increased in 19 of 20 cervical cancer cases. The relative EIF5A2 mrna expression is indicated in the histogram (Tumour, cervical tumour tissues; Normal, the paired surrounding nontumour cervical tissues). B, EIF5A2 mrna levels were significantly increased in earlystage cervical cancer, as determined with a matched paired Students t-test. The levels of b-actin were used as an internal control, and EIF5A2 mrna expression was calculated with the 2 DDC(t) method. The data are presented as mean standard deviation, *P value <01, Normal versus Tumour. A 17-kDa EIF5A2 43-kDa β-actin B EIF5A2 protein expression N 0.5000 500 000 0.3500 0.3000 500 000 0.1500 0.1000 500 000 T N T N T N T Normal * Tumour Figure 2. Elevated expression of eukaryotic initiation factor (EIF)5A2 in early-stage cervical cancer (T, primary cervical tumour tissues; N, the paired surrounding non-tumour cervical tissues). A, Representative western blot analysis of EIF5A2 expression in 20 paired tumour cervical tissues. Protein samples were analysed with sodium dodecyl sulphate polyacrylamide gel electrophoresis followed by immunoblotting with antibody against EIF5A2. The levels of b-actin were used as an internal control. B, Histogram of pooled data from representative cervical tumour tissues and matched adjacent non-tumour tissues. The expression of EIF5A2 protein was significantly increased in cervical tumour tissues. The data are presented as mean standard deviation, *P = 001.
The role of EIF5A2 in early-stage cervical cancer 5 A B C D Figure 3. Representative photomicrographs of eukaryotic initiation factor (EIF)5A2 immunohistochemical staining. A, in cervical squamous cell carcinoma. B, EIF5A2 normal expression in cervical squamous cell carcinoma. C, in cervical adenocarcinoma. D, Low in cervical adenocarcinoma. E, in matched adjacent non-tumour tissues. F, EIF5A2 normal expression in matched adjacent non-tumour tissues. E F Table 1. Expression of eukaryotic initiation factor 5A2 in International Federation of Gynecology and Obstetrics stage I and II cervical cancer specimens Group No. of patients High EIF5A2 expression n % Adjacent non-tumour tissues 20 3 15.0 <01 Cervical tumour tissues 314 201 64.0 *Cervical tumour tissues versus adjacent non-tumour tissues; v 2 test. observed between and other clinicopathological factors, including age (P = 63), histological type (P = 11), histological grade (P = 0.500), tumour size (P = 69), and surgical margin (P = 40). P* ASSOCIATION OF EIF5A2 EXPRESSION WITH SURVIVAL OF PATIENTS WITH FIGO STAGE I II CERVICAL CANCER Among all of the patients with cervical cancer, the 5- year OS rate was 75.2% and the 5-year DFS rate was 65.7%. Patients with were likely to have significantly shorter OS (P < 01; Figure 4A) and DFS (P < 01, Figure 4B). We performed a FIGO stage-stratified analysis of all patients according to the level of EIF5A2 expression. The results demonstrated that had substantial effects on OS and DFS in FIGO stage II patients (P = 02 and P < 01, respectively); the corresponding survival curves are shown in Figure 5B,D. However, this expression pattern was not a prognostic predictor of OS and DFS in FIGO stage I patients (P = 24 and P = 0.127, respectively; Figure 5A,C).
6 S-S Yang et al. Table 2. Association analysis between the expression level of eukaryotic initiation factor (EIF)5A2 and clinicopathological factors of International Federation of Gynecology and Obstetrics (FIGO) stage I and II cervical cancer Variables Age (years) No. of patients (N = 314) EIF5A2 expression Normal, no. (%) (N = 113, 36.0%) Overexpression, no. (%) (N = 201, 64.0%) P* <45 188 75 (39.9) 113 (60.1) 78 45 126 38 (3) 88 (69.8) FIGO stage I 136 64 (47.1) 72 (52.9) <01 II 178 49 (27.5) 129 (72.5) Histological type SCC 293 107 (36.5) 186 (63.5) 64 AC 21 6 (28.6) 15 (71.4) Differentiation grade 1 78 24 (3) 54 (69.2) 0.526 2 127 47 (37.0) 80 (63.0) 3 109 42 (38.5) 67 (61.5) Tumour size (mm) 40 213 76 (35.7) 137 (64.3) 69 >40 101 37 (36.6) 64 (63.4) Depth of cervical stromal invasion <1/2 111 49 (44.1) 62 (55.9) 26 1/2 203 64 (31.5) 139 (68.5) Lymphovascular space involvement No 228 94 (41.2) 134 (58.8) 02 Yes 86 19 (22.1) 67 (77.9) Surgical margin Negative 304 108 (35.5) 196 (64.5) 0.348 Positive 10 5 (5) 5 (5) Pelvic lymph node metastasis No 221 97 (43.9) 124 (56.1) <01 Yes 93 16 (17.2) 77 (82.8) Table 2. (Continued) Variables No. of patients (N = 314) Postoperative recurrence EIF5A2 expression Normal, no. (%) (N = 113, 36.0%) Overexpression, no. (%) (N = 201, 64.0%) P* No 204 93 (45.6) 111 (54.4) <01 Yes 110 20 (18.2) 90 (81.8) AC, Adenocarcinoma; SCC, Squamous cell carcinoma. Bold font highlights P value <5. *v 2 test. Survival analysis of pelvic lymph node status was performed. This showed that was correlated with shorter OS and DFS in patients with a negative pelvic lymph node status (P = 23 and P = 01, respectively; Figure 6A,C). However, this expression pattern was not a good prognostic predictor of OS and DFS for patients with pelvic lymph node metastasis (P = 62 and P = 0.157, respectively; Figure 6B,D). UNIVARIATE AND MULTIVARIATE ANALYSES OF PROGNOSTIC VARIABLES IN PATIENTS WITH FIGO STAGE I II CERVICAL CANCER Univariate and multivariate analyses were conducted to determine the predictors for OS and DFS (Table 3). The univariate analysis revealed that advanced FIGO stage (P < 01), adenocarcinoma (P = 01), high histological grade (P < 01), large tumour size (P = 28), deep stromal invasion (P = 25), lymphovascular space involvement (P < 01), pelvic lymph node metastasis (P < 01) and (P < 01) were significant indicators of poor OS. Moreover, advanced FIGO stage (P < 01), high histological grade (P = 23), deep stromal invasion (P = 25), lymphovascular space involvement (P < 01), pelvic lymph node metastasis (P < 01) and (P < 01) were significant indicators of poor DFS. All of the variables in the univariate analysis performed with the Kaplan Meier method were included in the Cox regression model in a stepwise manner (backward, likelihood ratio) to evaluate the independent prognostic significance of those variables for confounding. As shown in Table 3, multivariate
The role of EIF5A2 in early-stage cervical cancer 7 A Overall survival ratio P < 01 B P < 01 Overall survival ratio 0 20 40 60 80 100 0 20 40 60 80 100 Figure 4. Kaplan Meier curves for prognosis of survival in 314 patients with early-stage cervical cancer according to the categories of high and normal expression of eukaryotic initiation factor 5A2 (analysed with the log-rank test). A, Overall survival. B, Disease-free survival. A B P = 24 P = 02 Overall survival ratio Overall survival ratio FIGOstage: I FIGOstage: II 0 20 C Disease-free survival ratio FIGOstage: I 40 60 80 100 P = 0.127 D Disease-free survival ratio 0 20 40 60 80 100 FIGOstage: II P < 01 0 20 40 60 80 100 0 20 40 60 80 100 Figure 5. Kaplan Meier curves showing the correlation of eukaryotic initiation factor 5A2 expression with overall survival (OS) and disease-free survival (DFS) in cervical cancer patients in different FIGO stages. A, OS in patients with FIGO stage I cervical cancer. B, OS in patients with FIGO stage II cervical cancer. C, DFS in patients with FIGO stage I cervical cancer. D, DFS in patients with FIGO stage II cervical cancer.
8 S-S Yang et al. A Overall survival ratio P = 23 B P = 62 Negative pelvic lymph node Overall survival ratio Pelvic lymph node metastasis 0 20 40 60 80 100 0 20 40 60 80 100 C P = 01 D P = 0.157 Disease-free survival ratio Negative pelvic lymph node Disease-free survival ratio Pelvic lymph node metastasis 0 20 40 60 80 100 0 20 40 60 80 100 Figure 6. Kaplan Meier curves showing the correlation of eukaryotic initiation factor 5A2 expression with overall survival (OS) and diseasefree survival (DFS) in cervical cancer patients with different pelvic lymph node statuses. A, OS in patients with negative pelvic lymph node status. B, OS in patients with pelvic lymph node metastasis. C, DFS in patients with negative pelvic lymph node status. D, DFS in patients with pelvic lymph node metastasis. analysis revealed that was an independent factor that influenced the OS [hazard ratio (HR) 1.949, 95% CI 1.116 3.404, P = 19] and DFS (HR 1.980, 95% CI 1.189 3.297, P = 09) of patients with FIGO stage I II cervical cancer. Discussion Cervical cancer is one of the leading causes of death among women in the developing world. 16 However, the available methods for evaluating prognosis, such as parameters based on clinical and histopathological variables, do not fully represent the varied clinical course of cervical cancer. Cervical tumorigenesis is a multistep process: an initial benign stage is followed by cervical intraepithelial neoplasia, which progresses into in-situ carcinomas and then into invasive carcinomas, culminating in metastatic disease. The initiation and progression of cervical cancer constitute a complex process involving genetic and epigenetic changes. 17 Consequently, the identification of tumour-associated antigens, which are specifically expressed in cancer tissues, is of utmost importance in the prognosis evaluation and individualized treatment strategies of cervical cancer. EIF5A2 is a candidate oncogene that was isolated from a frequently amplified region at chromosome 3q
The role of EIF5A2 in early-stage cervical cancer 9 Table 3. Univariate Kaplan Meier analysis and multivariate cox regression model of prognostic covariates in 314 patients with International Federation of Gynecology and Obstetrics (FIGO) stage I and II cervical cancer OS (months) DFS (months) Univariate Multivariate Univariate Multivariate Variables N = 314 Mean SE 95% CI P* HR 95% CI P Mean SE 95% CI P* HR 95% CI P** Age (years) <45 188 69 2 66 73 0.946 0 (reference) 64 2 60 69 64 45 126 68 2 64 73 54 08 49 78 63 3 58 68 FIGO stage I 136 78 2 74 81 <01 0 (reference) 73 2 69 77 <01 0 (reference) II 178 62 2 58 67 2.631 1.507 4.595 01 57 2 52 61 1.664 45 2.648 32 Histological type SCC 293 70 1 67 73 01 0 (reference) 65 2 61 68 0.114 0 (reference) AC 21 49 7 36 62 3.338 1.675 6.653 01 52 7 38 66 1.965 0.971 3.975 60 Differentiation grade 1 78 77 2 73 82 <01 0 (reference) 68 3 62 74 23 0 (reference) 2 127 69 2 65 74 3.984 1.886 8.416 <01 67 2 62 72 1.270 0.760 2.122 0.362 3 109 62 3 57 68 4.948 2.341 159 <01 57 3 51 63 1.725 41 2.859 34 Tumour size (mm) 40 213 71 2 68 75 19 65 2 61 69 80 >40 101 64 3 58 69 62 3 56 68 Depth of cervical stromal invasion <1/2 111 74 2 69 78 25 70 2 65 74 25 1/2 203 66 2 63 70 61 2 57 65 Lymphovascular space involvement No 228 75 1 72 78 <01 0 (reference) 73 2 70 76 <01 0 (reference) Yes 86 53 3 46 59 2.899 1.819 4.618 <01 41 3 34 47 3.547 2.344 5.368 <01
10 S-S Yang et al. Table 3. (Continued) OS (months) DFS (months) Univariate Multivariate Univariate Multivariate Variables N = 314 Mean SE 95% CI P* HR 95% CI P Mean SE 95% CI P* HR 95% CI P** Surgical margin Negative 304 69 1 66 72 44 0 (reference) 64 2 61 68 53 0 (reference) Positive 10 55 10 36 74 2.934 45 8.238 41 49 9 31 67 3.477 1.478 8.181 04 Pelvic lymph node metastasis No 221 73 1 71 76 <01 0 (reference) 71 2 68 74 <01 0 (reference) Yes 93 57 3 51 63 1.862 1.187 2.923 07 45 3 39 52 2.342 1.564 3.508 <01 EIF5A2 expression Normal 113 76 2 72 80 <01 0 (reference) 73 2 69 78 <01 0 (reference) Overexpression 201 65 2 61 69 1.949 1.116 3.404 19 59 2 54 63 1.980 1.189 3.297 09 AC, Adenocarcinoma; CI, Confidence interval; DFS, Disease-free survival; EIF, Eukaryotic initiation factor; HR, Hazard ratio; OS, Overall survival; SCC, Squamous cell carcinoma; SE, Standard error. Bold font highlights P value <5. *Log-rank test. Cox regression test.
The role of EIF5A2 in early-stage cervical cancer 11 of a primary ovarian cell line. 7 To date, numerous studies have demonstrated that is associated with the development and progression of diverse human malignancies, and that EIF5A2 may function as a highly promising prognostic marker. 8 13,18 However, no direct effects of EIF5A2 on the carcinogenesis of cervical cancer have been revealed. Therefore, the clinical significance of EIF5A2 activation in early-stage cervical cancer has been demonstrated for the first time in this study. The present study is the first to demonstrate that, unlike in normal cervical tissues, EIF5A2 expression is up-regulated at both the mrna level and protein level in FIGO stage I II cervical cancer tissues. The high expression of EIF5A2 was significantly associated with certain malignant tumour characteristics, such as advanced FIGO stage, deep cervical stromal invasion, lymphovascular space involvement, pelvic lymph node metastasis and postoperative recurrence in patients with early-stage cervical cancer. Most importantly, in early-stage cervical cancer can serve as an independent prognostic factor for poor OS and DFS, particularly in patients with FIGO stage II cervical cancer and patients with a negative pelvic lymph node status. These abovementioned results indicate that EIF5A2 might serve as a hallmark of tumour progression and poor prognosis in early-stage cervical cancer. EIF5A2 expression is tissue-specific and cell typespecific, particularly in the testis and brain, as well as in several cancer cell lines and tissues. 6 As a newly identified oncoprotein, EIF5A2 is overexpressed in several human cancers, and has been suspected to play an important role in carcinogenesis, metastasis, and chemoresistance. Previous in-vitro and in-vivo studies showed that EIF5A2 can regulate cancer cell proliferation and transformation. 7,19 22 In colorectal cancer, is significantly associated with the tumour cell proliferation rate detected by Ki67 staining. 21 Metastasis is one of the main causes of mortality for most solid tumours. The accumulated evidence demonstrates that EIF5A2 may promote invasion and metastasis by inducing epithelial mesenchymal transition, which is characterized by down-regulation of E-cadherin and up-regulation of fibronectin, N-cadherin, a-smooth muscle actin, and vimentin, 10,11,18 regulating hypoxia-inducible factor-1a and angiogenesis, 23 and stimulating cytoskeletal rearrangement via the activation of RhoA and Rac1. 18 In addition, EIF5A2 promotes cell invasion as a target of phosphoinositide 3-kinase/ AKT in melanomas. 11 Regarding chemosensitivity, some in-vitro studies have demonstrated that ablation or inhibition of EIF5A2 enhances the chemosensitivity of hepatocellular carcinoma cells to 5-fluorouracil (5-FU) or doxorubicin, and that of breast cancer cells or bladder cancer cells to doxorubicin. 24 27 Research on oesophageal squamous cell carcinoma (ESCC) revealed that in the ESCC cells increased chemoresistance to 5-FU, docetaxel, and taxol, and short hairpin RNA-mediated silencing of EIF5A2 in the ESCC cells increased their chemosensitivity towards these chemotherapeutic agents. 28 These findings suggest that EIF5A2 is a potential target for anticancer therapy. To the best of our knowledge, our findings provide the first evidence that EIF5A2 may be an important biomarker for predicting unfavourable biological behaviour and poor prognosis in patients with earlystage cervical cancer. Although the results are encouraging, this investigation has limitations. The current study is a retrospective analysis with a limited number of patients. Thus, a well-designed, prospective study involving a larger series of patients with detailed information is warranted to validate the role of EIF5A2 as a novel prognostic marker for cervical cancer. Further investigation of the cell biology and underlying mechanisms of EIF5A2, as well as its potential as a therapeutic target, are clearly warranted. Acknowledgements The authors thank all of the people and patients who participated in this study, especially Qi Huang and Xiang Ban (Department of Morphology, Basic Medical Science College, Harbin Medical University, Harbin, China) for technical assistance. This work was supported by the National Natural Science Foundation of China (nos 81372786 and 81502266), the HaiYan Foundation of the Affiliated Tumour Hospital of Harbin Medical University (no. JJZD2014-07), and the National Postdoctoral Science Foundation of China (2015M581482). Conflicts of interest The authors declare that they have no conflicts of interest. Author contributions Gen-Ying Li and Li-Xiao Hao performed the research. Shan-Shan Yang, Min Xiao and Yun-Yan Zhang
12 S-S Yang et al. designed the research study. Bai-Rong Xia and Yun- Duo Liu collected the tissues. De-Ying Wang and Min Xiao analysed the data. Xiao-Ming Ning and Yu Qin evaluated the immunohistochemical staining. Shan- Shan Yang wrote the paper. References 1. Torre LA, Bray F, Siegel RL, et al. Global cancer statistics, 2012. CA Cancer J. Clin. 2015; 65; 87 108. 2. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J. Clin. 2013; 63; 11 30. 3. Walboomers JM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 1999; 189; 12 19. 4. Munoz N, Bosch FX, de Sanjose S, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N. Engl. J. Med. 2003; 348; 518 527. 5. Syrjanen KJ. Spontaneous evolution of intraepithelial lesions according to the grade and type of the implicated human papillomavirus (HPV). Eur. J. Obstet. Gynecol. Reprod. Biol. 1996; 65; 45 53. 6. Jenkins ZA, Haag PG, Johansson HE. Human eif5a2 on chromosome 3q25 q27 is a phylogenetically conserved vertebrate variant of eukaryotic translation initiation factor 5A with tissue-specific expression. Genomics 2001; 71; 101 109. 7. Guan XY, Sham JS, Tang TC, et al. Isolation of a novel candidate oncogene within a frequently amplified region at 3q26 in ovarian cancer. Cancer Res. 2001; 61; 3806 3809. 8. Yang GF, Xie D, Liu JH, et al. Expression and amplification of eif-5a2 in human epithelial ovarian tumors and overexpression of EIF-5A2 is a new independent predictor of outcome in patients with ovarian carcinoma. Gynecol. Oncol. 2009; 112; 314 318. 9. He LR, Zhao HY, Li BK, et al. Overexpression of eif5a-2 is an adverse prognostic marker of survival in stage I non-small cell lung cancer patients. Int. J. Cancer 2011; 129; 143 150. 10. Zhu W, Cai MY, Tong ZT, et al. Overexpression of EIF5A2 promotes colorectal carcinoma cell aggressiveness by upregulating MTA1 through C-myc to induce epithelial mesenchymal transition. Gut 2012; 61; 562 575. 11. Khosravi S, Wong RP, Ardekani GS, et al. Role of EIF5A2, a downstream target of Akt, in promoting melanoma cell invasion. Br. J. Cancer 2014; 110; 399 408. 12. Meng QB, Kang WM, Yu JC, et al. Overexpression of eukaryotic translation initiation factor 5A2 (EIF5A2) correlates with cell aggressiveness and poor survival in gastric cancer. PLoS ONE 2015; 10; e0119229. 13. Wei JH, Cao JZ, Zhang D, et al. EIF5A2 predicts outcome in localised invasive bladder cancer and promotes bladder cancer cell aggressiveness in vitro and in vivo. Br. J. Cancer 2014; 110; 1767 1777. 14. Pecorelli S. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int. J. Gynaecol. Obstet. 2009; 105; 103 104. 15. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C (T)) method. Methods 2001; 25; 402 408. 16. Jones SB. Cancer in the developing world: a call to action. BMJ 1999; 319; 505 508. 17. Hildesheim A, Wang SS. Host and viral genetics and risk of cervical cancer: a review. Virus Res. 2002; 89; 229 240. 18. Tang DJ, Dong SS, Ma NF, et al. Overexpression of eukaryotic initiation factor 5A2 enhances cell motility and promotes tumor metastasis in hepatocellular carcinoma. Hepatology 2010; 51; 1255 1263. 19. Guan XY, Fung JM, Ma NF, et al. Oncogenic role of eif-5a2 in the development of ovarian cancer. Cancer Res. 2004; 64; 4197 4200. 20. Clement PM, Johansson HE, Wolff EC, et al. Differential expression of eif5a-1 and eif5a-2 in human cancer cells. FEBS J. 2006; 273; 1102 1114. 21. Xie D, Ma NF, Pan ZZ, et al. Overexpression of EIF-5A2 is associated with metastasis of human colorectal carcinoma. Hum. Pathol. 2008; 39; 80 86. 22. Zender L, Xue W, Zuber J, et al. An oncogenomics-based in vivo RNAi screen identifies tumor suppressors in liver cancer. Cell 2008; 135; 852 864. 23. Li Y, Fu L, Li JB, et al. Increased expression of EIF5A2, via hypoxia or gene amplification, contributes to metastasis and angiogenesis of esophageal squamous cell carcinoma. Gastroenterology 2014; 146; 1701 1713. 24. Wang FW, Cai MY, Mai SJ, et al. Ablation of EIF5A2 induces tumor vasculature remodeling and improves tumor response to chemotherapy via regulation of matrix metalloproteinase 2 expression. Oncotarget 2014; 5; 6716 6133. 25. Lou B, Fan J, Wang K, et al. N1-guanyl-1,7-diaminoheptane (GC7) enhances the therapeutic efficacy of doxorubicin by inhibiting activation of eukaryotic translation initiation factor 5A2 (eif5a2) and preventing the epithelial mesenchymal transition in hepatocellular carcinoma cells. Exp. Cell Res. 2013; 319; 2708 2717. 26. Yang J, Yu H, Shen M, et al. N1-guanyl-1,7-diaminoheptane sensitizes bladder cancer cells to doxorubicin by preventing epithelial mesenchymal transition through inhibition of eukaryotic translation initiation factor 5A2 activation. Cancer Sci. 2014; 105; 219 227. 27. Liu Y, Du F, Chen W, et al. EIF5A2 is a novel chemoresistance gene in breast cancer. Breast Cancer 2015; 22; 602 607. 28. Yang H, Li XD, Zhou Y, et al. Stemness and chemotherapeutic drug resistance induced by in esophageal squamous cell carcinoma. Oncotarget 2015; 6; 26079 26089.
本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 (www.xuebalib.com) 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具