| BackgroundPenile squamous cell carcinoma (PSCC), accounting for the vast majority of the penile malignancy, was popular in40to60years old patients with redundant prepuce or phimosis. The exact etiology and mechanism have not been fully understood, Hepatoma derived growth factor (HDGF) has been verified to serve as a credible prognostic marker for several types of cancers, but its role in PSCC remains undetermined.PurposeIn this study, we tested the expression of HDGF, as well as its relative factors such as vascular endothelial growth factor-A (VEGF-A) and Ki-67, in PSCC by the method of immunohistochemistry (IHC). Then, we analyzed the association between the expression of HDGF, VEGF-A, Ki-67and the clinicopathological characteristics of PSCC patients. In addition, we evaluated the association between the expression of HDGF and the expression of VEGF-A, Ki-67. Still, we analyzed the significance of HDGF, VEGF-A and Ki-67in the development and prognosis of PSCC.MethodsFormalin-fixed paraffin-embedded PSCC samples from54patients receiving surgery at Qilu Hospital of Shandong University between June2004and February 2011were included in the retrospective study. For each PSCC sample, three successive slides with4μm thickness were prepared for immunohistochemical analysis of HDGF, VEGF-A and Ki-67expression by a non-biotin polymerized horseradish peroxidase method. Tissue sections were deparaffinized through a series of xylene baths and then rehydrated through a series of graded alcohol baths. Afterwards, antigen retrieval was implemented by heating specimens in microwaves for15minutes in10mM citrate buffer (pH=6.0). After blocking endogenous peroxidase activity by methanol containing3.0%hydrogen peroxide (H2O2) for15min, tissue slides were incubated with10%goat serum to block non-specific binding of immunoreagents. Subsequently, tissues slides were incubated at4℃overnight with diluted primary antibodies against HDGF (1:100), VEGF-A (1:100) and Ki-67(1:100). Then, the detections of the primary antibodies were realized with a secondary detection kit (PV-9001) and DAB kit (ZLI-9031). Finally, tissue samples were counterstained with hematoxylin, treated with the solution of Hydrochloric acid and ethanol, dehydrated with ethanol, and mounted with glycerol gelatin under a coverslip. To evaluate the immunostaining of HDGF, VEGF-A and Ki-67, more than five independent fields (400x magnifications) of each tissue were randomly captured under microscopy. And stained samples were assessed in a blind manner without prior knowledge of clinical data. HDGF and Ki-67were mainly expressed in the nucleus area and assessed as previously described, while VEGF-A was foremostly accumulated in the cytoplasm area and evaluated with previous method. The percentage of positive cells (number of positive cells/that of total cells) in each field was recorded and the expressions of HDGF, VEGF-A, Ki-67for each sample were calculated accordingly. The relationships between the expressions of HDGF and VEGF-A, Ki-67were assessed. Moreover, their correlations with clinical as well as pathologic characteristics and disease prognosis were respectively evaluated. The Statistical Package for Social Sciences version16.0(SPSS16.0, SPSS Inc, Chicago, IL) was adopted to analyze the database. The Pearson’s Chi-Squared test was used to compare the differences among various groups. Also, a logistic regression model was constructed to discriminate prognostic parameters by transforming characteristics, such as HDGF expression and symptom interval, into categorical variables. A two-tailed P<0.05was considered to be statistically significant.ResultsImmunohistochemical results indicated that HDGF and Ki-67predominantly expressed in nucleus, while VEGF-A mainly accumulated in cytoplasm. HDGF, VEGF-A, Ki-67expression was categorized as high (positive, more than the median value), while HDGF, VEGF-A, Ki-67expression was categorized as low (negative, less than the median value). The HDGF analysis was sorted to be high in28(51.9%) and low in26(48.1%) of all PSCC patients for further comparison. The relationships of HDGF expression with clinical and pathological characteristics were compared. There were no significant differences between HDGF level and parameters, such as age (P=0.284), symptom interval (P=0.180), tumor size (P=0.984), PSCC cell type (P=0.645), grade (P=0.884), N stage (P=0.513), and smoking status (P=0.150). VEGF-A was positively expressed in29(53.7%) and negatively expressed in25(46.3%) of total patients. The relationships of VEGF-A expression with clinical and pathological characteristics were compared. VEGF-A expression correlated closely with cell types of PSCC (P=0.030). There were no significant differences between VEGF-A level and parameters, such as age (P=0.184), symptom interval (P=0.951), tumor size (P=0.816), grade (P=0.109), N stage (P=0.492), and smoking status (P=0.991). Ki-67analysis was strong in26patients (48.1%) and weak in28patients (51.9%). The relationships of Ki-67expression with clinical and pathological characteristics were compared. The expression of Ki-67also showed a close association with PSCC type (P=0.020). There were no significant differences between Ki-67level and parameters, such as age (P=0.226), symptom interval (P=0.761),tumor size (P=0.984), grade (P=0.145), N stage (P=0.627), and smoking status (P=0.422). In PSCC tissues, the expression of VEGF-A in the HDGF positive carcinomas was higher than that in the HDGF negative carcinomas, which confirmed a significant association between the two (P=0.030). However, the expression of a marker for proliferation, Ki-67, in the HDGF positive cancers was more or less the same as that in the HDGF negative carcinomas, which indicated an insignificant association between the two (P=0.176). First, we detected the association between clinical and pathological characteristics and the survival rate of all PSCC patients. Patients with symptom interval more than six months had a significantly poorer survival rate than those with symptom interval less than six months (43.3%vs.70.8%, P=0.043). Patients with positive HDGF expression also showed a dramatically poorer survival rate than those with negative HDGF expression (39.3%vs.73.1%, P=0.013). Other variables, such as age, tumor size, grade, lymphatic involvement, cell subtype, smoking status, VEGF-A and Ki-67expressions, were insignificantly related with survival rate of PSCC patients. To further determine the association of clinical and pathological characteristics with survival rate, a logistic regression equation was performed to evaluate potential and independent prognostic factors. The significant factors, such as HDGF expression and symptom interval, were recruited into the logistic regression. The final result confirmed that the high level of HDGF expression was an independent predictor for the prognosis of postoperative PSCC patients (P=0.028).Conclusions1. In the current study, overexpression of HDGF correlated closely with the prognosis of PSCC (P<0.05). HDGF may serve as a reliable and promising indicator of prognosis and therapeutic target.2. The expression of VEGF-A correlated with cell type of PSCC (P<0.05), but not the prognosis of PSCC. HDGF regulated the expression of VEGF-A in PSCC.3. The expression of Ki-67correlated with cell type of PSCC (P<0.05), but not the prognosis of PSCC. HDGF expression was not correlated with the expression of Ki-67in PSCC.4. The time course may be an indicator of the prognosis of PSCC. PurposeThe hypoxia-inducible factor-1alpha (HIF1A) plays a vital role in cancer initiation and progression. HIF1A gene polymorphisms have been reported to contribute to the risk of various cancers. Previous studies have confirmed the existence of HIFl A P582S and A588T missense polymorphisms in renal, urothelial, and prostatic carcinomas, however, the effects remain conflicting. We therefore performed a comprehensive meta-analysis to assess the association between these sites and the susceptibility of urinary cancers (UC).MethodsIn the current study, we performed a systematic article search in the electric PubMed database without limits on language until Nov25,2012for studies exploring the relationship of HIF1A P582S and A588T polymorphisms and the risk of urinary cancers. The following keywords were adopted:(hypoxia-inducible factor-1OR hypoxia-inducible factor OR HIF-1OR HIFl A OR HIF),(polymorphism OR variant OR SNP OR mutation) and (kidney OR renal OR urothelial OR transitional cell carcinoma OR bladder OR prostatic OR prostate). On screening titles, abstracts and full texts, eleven eligible studies conformed to inclusion criteria were finally included. Additionally, article search was supplemented by screening the references of retrieved studies manually. For each study, we extracted data through a standard form. The following characteristics were respectively extracted from the included studies:name of first author, year of publications, country of origin, ethnicity, gender of recruited subjects, cancer types, numbers of various genotypes in case and control groups, methods for detecting HDF1A P582S and/or A588T polymorphisms, Hardy-Weinberg equilibrium (HWE). In case of disagreement, discrepancies of included studies were finally resolved by discussion.The genotype and allele differences of HIF1A P582S and A588T polymorphisms in Caucasian and Asian populations were calculated by Pearson’s chi-squared test. We evaluated the contribution of HIF1A P582S and A588T polymorphisms to the risk of urinary cancers by adopting the RevMan5.0software, which is developed by Cochrane Collaboration. For HIF1A P582S polymorphism, we evaluated the risk in the dominant model (TT+CT vs CC), the recessive model (TT vs CT+CC and the allele model (T vs C respectively. For HIF1A A588T polymorphism, we only evaluated the risk in the dominant model (AA+AG vs GG) and the allele model (A vs G) due to few frequencies of AA genotype in subjects. Then, we performed the subgroup meta-analysis according to the status of HWE, cancer type and ethnicity. The strength of association was estimated by calculating ORs and the corresponding95%CIs. Still, a P-value less than0.05was considered to be statistically significant. Heterogeneity assumption was assessed by the chi-squared based Q test, and P<0.10was regarded to be statistically significant. The random-effects model (the Dersimonian-Laird method) would be used if the test of heterogeneity was significant; otherwise the fixed-effects model (the Mantel-Haenszel method) would be applied in the analyses. Sensitivity analyses were carried out to assess the stability of the results by conducting subgroup meta-analysis of studies with controls in HWE. The potential publication bias was primarily appraised by the funnel plot. An asymmetric plot suggests a possible publication bias. The asymmetry of funnel plot was further evalued by Begg’s test with STATA12.1software. A P-value less than0.05was considered to be statistically significant.ResultsOverall, eleven individual case-control studies with5195cases and5786controls for P582S polymorphism, and nine studies with3482cases and4304controls for A588T polymorphism were respectively included in the final meta-analysis.As for HIF1A P582S polymorphism,1106controls of Caucasian population and1803controls of Asian population were included in the meta-analysis. The frequencies of the C and T alleles for Caucasian were80.74%,19.26%, while those for Asian were95.90%and4.10%, respectively. The frequencies of the CC, CT, and TT genotypes for Caucasian were69.17%,23.15%,7.69%respectively, while those for Asian were91.90%,7.99%, and0.11%. The frequency distributions of the alleles and genotypes for HIF1A P582S polymorphism were obviously different between Caucasian and Asian groups. Firstly, we conducted meta-analysis of the effect of HIFIA P582S polymorphism on the susceptibility of urinary cancers based on eleven case-control studies. The results showed no significant association between the two in the dominant model (TT+CT vs CC:OR=1.10,95%CI=0.83-1.45, Pheterogeneity=0.00, P=0.52), the recessive model (TT vs CT+CC:OR=1.17,95%CI=0.67-2.05, Pheterogeneity=0.02, P=0.57) and the allele model (T vs C:OR=1.13,95%CI=0.90-1.41, Pheterogeneity=0.00, P=0.30). Secondly, we performed subgroup meta-analysis based on the difference of ethnicity, cancer type, and HWE status. We found that subjects with TT genotype had1.60fold higher risk than those with CC or CT genotype in Caucasian population (TT vs CT+CC:OR=1.60,95%CI=1.09-2.33, Pheterogeneity=0.11, P=0.02). The remaining subgroup pooled ORs from the meta-analysis were insignificant (all P>0.05).As for HDF1A A588T polymorphism, the frequencies of the AA+AG and GG genotypes for Caucasian were6.72%,93.28%respectively, while those for Asian were6.11%and93.89%. The frequency distributions of the genotypes for HIF1A A588T polymorphism were statistically insignificant between the Caucasian and Asian groups. The frequencies of the A and G alleles for Caucasian were3.84%,96.16%, while those for Asian were2.61%and97.39%, respectively. The frequency distributions of the alleles for HIF1A A588T polymorphism were obviously different between the Caucasian and Asian groups. In the beginning, we conducted meta-analysis of the effect of HIF1A A588T polymorphism on the susceptibility of urinary cancers based on nine case-control studies. The results showed no significant association between the two in the dominant model (AA+AG vs GG:OR=1.40,95%CI=0.76-2.58,Pheterogeneity=0.00, P=0.28) and the allele model (A vs G:OR=1.57,95%CI=0.89-2.76, Pheterogeneity=0.00, P=0.12). Subsequently, we performed subgroup meta-analysis based on the difference of ethnicity, cancer type, and HWE status. We found that subjects carrying A allele had1.45fold higher risk than those with GG genotype in prostate cancer (AA+AG vs GG:OR=1.45,95%CI=1.00-2.12,Pheterogeneity=0.50, P=0.05). Still, significant associations were found in the allele model in prostate cancer (A vs G; OR=1.41,95%CI=1.03-1.93,^heterogeneity=0.22, P=0.03), prostate cancer in HWE (A vs G:OR=1.45,95%CI=1.00-2.11,Pheterogeneity=0.33, P=0.05) and Asian population (A vs G:OR=1.46,95%CI=1.01-2.12, Pheterogeneity=0.49, P=0.04). The remaining subgroup pooled ORs from the analyses were insignificant (all P>0.05).Significant heterogeneity was observed in some comparisons (P<0.10), and results were listed. Sensitivity analyses were carried out to assess the stability of the results by conducting subgroup meta-analysis of studies with controls in HWE. The result of the recessive model comparison showed no evidence that HIFIA P582S polymorphism conferred to an increased urinary cancers risk in Caucasian population (control group conforming to HWE)(TTvs CT+CC:OR=1.57,95%CI=0.22-11.14,Pheterogeneity=0-04, P=0.65). The other results of subgroup analyses showed no difference. The potential publication bias was firstly appraised by the funnel plot, which showed no apparently asymmetric. Still, the results of Begg’s test revealed no publication bias (P>0.05). Conclusions1. The frequency distributions of the alleles and genotypes for HIF1A P582S polymorphism were obviously different between Caucasian and Asian groups. The differences of the genotype distribution for HIF1A A588T polymorphism were insignificant between the Caucasian and Asian groups. The frequency distributions of the alleles for HIF1A A588T polymorphism were obviously different between the Caucasian and Asian groups.2. HIF1A P582S polymorphism correlates with urinary cancers risk in Caucasian population.3. HIF1A A588T polymorphism increases the risk of urinary cancers in Asian population.4. HIF1A A588T polymorphism increases the risk of prostate cancer. |
PDF Full Text Request