Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Predictive value of GGN and CAG repeat polymorphisms of androgen receptors in testicular cancer: a meta-analysis

Predictive value of GGN and CAG repeat polymorphisms of androgen receptors in testicular cancer:... www.impactjournals.com/oncotarget/ Oncotarget, Vol. 7, No. 12 Predictive value of GGN and CAG repeat polymorphisms of androgen receptors in testicular cancer: a meta-analysis 1 1 1 1 1 1 Weijun Jiang , Jing Zhang , Qing Zhou , Shuaimei Liu , Mengxia Ni , Peiran Zhu , 1 1 1 1 Qiuyue Wu , Weiwei Li , Mingchao Zhang , Xinyi Xia Department of Reproduction and Genetics, Institute of Laboratory Medicine, Jinling Hospital, Nanjing University School of Medicine, Nanjing 210002, P.R. China Correspondence to: Xin yi Xia, e-mail: xiaxynju@163.com Keywords: androgen receptor, GGN repeat, CAG repeat, testicular cancer, polymorphism Received: October 22, 2015 Accepted: January 29, 2016 Published: February 12, 2016 ABSTRACT The risk of testicular cancer (TC) is markedly increased in subjects with androgen insensitivity, and previous studies have proposed that GGN and CAG repeats in androgen receptors (AR) could be related to the risk of TC. To evaluate the association between the length of GGN and CAG repeats in AR and TC, a meta-analysis involving 3255 TC cases and 2804 controls was performed. The results suggested that long GGN repeats are associated with an increased risk of TC compared with those < 23 [odds ratio (OR) = 1.22, 95% confidence interval (CI) = 1.05–1.41]; similarly, a subgroup analysis revealed that this association occurred in studies with case sizes > 200, and in the mid-latitude, and seminoma subgroups. The subgroup analysis based on populations, high-latitude, and seminomas/non-seminomas suggested that AR CAG repeat polymorphisms with > 25 and < 21 + > 25 repeats might confer a protective effect to the patients with TC (in the high-latitude subgroup analysis, for > 25 vs. 21–25: OR = 0.54, 95% CI = 0.41–0.70). In contrast, an increased risk of TC was observed for AR CAG repeat polymorphisms with > 25 and < 21 + > 25 repeats in the mid-latitude subgroup (for > 25 vs. 21–25: OR = 1.65, 95% CI = 1.09–2.50). In addition, no associations between the remaining subgroups and male infertility were observed. In short, this meta-analysis suggested that AR GGN and CAG repeat polymorphisms may be involved in the etiology of TC. encoded by a (CAG) CAA stretch, as well as a INTRODUCTION (GGT) GGG(GGT) (GGC) repeat, and these repeats 3 2 n are designated CAG and GGN repeats, respectively [9]. Testicular cancer (TC) is an malignancy, accounting The extreme variability of the number of these repeats for 1%–2% of all tumors among men worldwide [1], determines the different lengths of the polyglutamine and and affects primarily young men in the age group 15–44 polyglycine segments in the N-terminal transactivation years. The incidence of TC is increasing worldwide and domain of the AR [10]. In men, the number of CAG has steeply increased in the past 40 years in almost all repeats can vary from 8 to 37, with an average of 20–22, Western countries [2–4]. Clinical studies reported that depending on the ethnic origin. Africans and Asians have 95% of all TCs are testicular germ cell tumors (TGCT), a lower number of repeats than Caucasians and a reduced with an approximately equal division between seminomas risk of TGCT [10]. Changes in the length of the CAG and non-seminomas, and epidemiological studies have polymorphic trinucleotide repeat in the AR gene may suggested that environmental factors, including endocrine lead to the altered transactivation of the AR gene and disrupting agents, which act as either weak estrogen have been implicated to play a role in the pathogenesis of agonists or androgen antagonists, are primarily responsible several forms of endocrine cancer and certain reproductive for the increased incidence of TC [5, 6]. disorders [11]. Subjects with reproductive disorders that The AR gene, located on Xq11-12, has 8 exons are associated with a relative deficiency in androgen and 7 introns, and in exon 1, this gene contains function have an increased risk of TC [12, 13]. In the two important polymorphic trinucleotide repeats of past decade, some studies have attempted to evaluate the polyglutamine and polyglycine tracts [7, 8], which are www.impactjournals.com/oncotarget 13754 Oncotarget association between CAG and GGN repeat number and the 95% CI = 1.02– 1.41, P = 0.028; for > 23 vs. ≤ 23: OR = risk of TC [6, 9, 10, 12– 15]; however, the results appear 1.24, 95% CI = 1.00–1.54, P = 0.050). contradictory because of differences in the sources of the study participants and inconsistencies in the inclusion Association between CAG repeat length and the criteria in case and control subjects among the studies risk of TC [9, 16]. To the best of our knowledge, to date, no meta- analysis has analyzed the results of all the studies that The association between CAG repeat in AR and evaluated this association. Therefore, this meta-analysis the risk of TC is summarized in Figures 3 and 4 and was conducted to investigate the association between in Table 4. In brief, the overall analysis indicated no CAG and GGN repeat polymorphisms and the risk of significant association between CAG repeats and the risk TC, as well as the genetic heterogeneity across different of TC for the models evaluated; however, in the subgroup control sources and study designs. Herein, seven reports analysis based on latitude, sample size, control source, involving 3255 TC cases and 2804 controls were identified and histology, significant associations were found between according to the inclusion criteria for the pooled analysis. CAG repeats and TC in the population-based (PB) subgroup (for < 21 + > 25 vs. 21–25: OR = 0.81, 95% CI = 0.68–0.96, P = 0.017), mid-latitude subgroup (for > 25 vs. RESULTS 21–25: OR = 1.65, 95% CI = 1.09–2.50, P = 0.017; for < 21 + > 25 vs. 21–25: OR = 1.38, 95% CI = 1.05–1.82, Study characteristics P = 0.021), high-latitude subgroup (for > 25 vs. 21–25: OR = 0.54, 95% CI = 0.41–0.70, P = 0.000; for < 21 + > 25 Because not all of the studies evaluated provided vs. 21–25: OR = 0.76, 95% CI= 0.64–0.90, P = 0.002), specific distributions of AR CAG or GGN repeat counts, seminoma subgroup (for > 25 vs. 21–25: OR = 0.47, 95% we used a CAG repeat length of 21–25 as reference to CI = 0.33–0.68, P = 0.000; for < 21 + > 25 vs. 21–25: evaluate dichotomous comparisons (< 21 CAG repeats OR = 0.73, 95% CI = 0.57–0.92, P = 0.008), and non- vs. the reference, > 25 CAG repeats vs. the reference, and seminoma subgroup (for > 25 vs. 21–25: OR = 0.52, 95% < 21 + > 25 CAG repeats vs. the reference). Similarly, the CI = 0.37–0.74, P = 0.000; for < 21 + > 25 vs. 21–25: OR GGN genotype of ≤ 23 repeats was used as reference to = 0.78, 95% CI = 0.62–0.98, P = 0.032). assess the association between > 23 repeats and the risk of TC. Through literature search and selection based on the Publication bias and small-study effects inclusion criteria, 7 articles published between 2002 and 2015 were identified after reviewing potentially relevant To assess the publication bias of the studies, Begg’s articles (Figure 1). The characteristics of the enrolled funnel plot and Egger’s test were performed. The shapes studies are summarized in Tables 1 and 2. of the funnel plots revealed no evidence of obvious For the GGN repeats, seven studies involving 1636 asymmetry. Egger’s test was used to provide statistical cases (range of 74–635, average of 272.67 ± 180.5) and evidence of funnel plot symmetry (Supplementary 1519 controls (range of 115–576, average of 304 ± 154) Figure S1) and indicated no evidence of publication bias were included in the meta-analysis. or small-study effects across the studies (for GGN repeats For the CAG repeats, 6 studies involving 1609 > 23 vs. ≤ 23, P = 0.840; for CAG repeat < 21 vs. 21–25, cases (range of 83–635, average of 230 ± 185) and 1285 P = 0.371; for CAG repeats > 25 vs. 21–25, P = 0.671; for controls (range of 110–322, average of 214 ± 82.1) met the CAG repeats < 21 + > 25 vs. 21–25, P = 0.941). inclusion criteria and were selected for the meta-analysis. Sensitivity analysis Association between GGN repeat length and the To confirm the results of the current study, the I risk of TC statistics calculated for the overall analysis of the CAG The association between GGN repeats and the risk repeats > 25 vs. 21–25 and CAG repeats < 21 + > 25 vs. of TC is summarized in Figure 2 and Table 3. The overall 21–25 were 73.4% and 56.6%, respectively, indicating analysis indicated a significant association between GGN that more than 50% of the abnormal CAG repeats may repeats and TC [odds ratio (OR) = 1.22, 95% confidence be due to between-study heterogeneity. We then evaluated interval (CI) = 1.05–1.41, P = 0.010]. To clarify the the source of heterogeneity in these comparisons by potential effect of latitude, sample size, and histological sample size, latitude, and histology stratifications, and differences, a subgroup analysis of study populations we observed no heterogeneity in latitude and histology was also conducted, and a significant association stratifications (Figure 4). Sensitivity analyses were was found between GGN repeats and TC in studies conducted to determine whether modifications in the with a sample size > 200 and in the mid-latitude and inclusion criteria of the meta-analysis affected the results. seminoma subgroups (for > 23 vs. ≤ 23: OR = 1.23, 95% Our results indicated that the studies by Grassetti et al. CI = 1.00– 1.51, P = 0.050; for > 23 vs. ≤ 23: OR = 1.20, [10] and Garolla et al. [17] caused this heterogeneity. www.impactjournals.com/oncotarget 13755 Oncotarget seminoma, and non-seminoma subgroups. However, AR DISCUSSION CAG repeat polymorphism with > 25 and < 21 + > 25 repeats in the mid-latitude subgroup were associated with The present meta-analysis, including 3255 TC an increased risk of TC. cases and 2804 controls from seven case-control studies, TC is a very common disease and its incidence has explored the association between GGN and CAG repeat increased worldwide in recent decades. TGCT makes up polymorphisms in AR and the risk of TC. Our results 95% of all TCs and is the most common solid tumor in indicated that long GGN repeats were associated with men aged 15–39 years [2, 10]. Although there has been an increased risk of TC, compared with repeats < 23; enormous progress in the clinical treatment of TC and similarly, sample size > 200 and the mid-latitude and preservation of fertility through sperm banks in recent seminoma subgroups were associated with an increased years, the main causes of the disease remain unclear. risk of TC. In contrast, AR CAG repeat polymorphism with However, important risk factors include work, lifestyle, > 25 and < 21 + > 25 repeats may confer a protective effect diet, familial history, environmental conditions, and genetic to the TC patients in the analysis of the PB, high-latitude, Figure 1: Flow diagram of the study selection process. www.impactjournals.com/oncotarget 13756 Oncotarget www.impactjournals.com/oncotarget 13757 Oncotarget Table 1: Main characteristics of the studies on the GGN repeats included in the meta-analysis Case Control Control Case Cases/ Author (year) Country Ethnicity Method Latitude Group source size Controls ≤ 23 > 23 ≤ 23 > 23 Grassetti D (2015) Italy Caucasian Sequence HB Mid-latitude > 200 302/322 Total 182 120 206 116 166 Seminomas 97 69 136 Non-seminomas 85 51 Kristiansen W (2012) Norway Caucasian Sequence PB High-latitude > 200 635/292 Total 391 244 196 96 315 Seminomas 191 124 320 Non-seminomas 200 120 Vastermark A (2011a) Sweden Caucasian Sequence PB High-latitude > 200 267/214 Total 175 92 141 73 Vastermark A (2011b) Denmark Caucasian Sequence PB High-latitude < 200 74/214 Total 43 31 141 73 158 Seminomas 101 57 178 Non-seminomas 114 64 Biggs ML (2008) USA Mixed Sequence PB Mid-latitude > 200 235/576 Total 133 102 359 217 Garolla A (2005) Italy Caucasian Sequence PB Mid-latitude < 200 123/115 Total 69 54 71 44 Abbreviations: HB, hospital-based; PB, population-based. www.impactjournals.com/oncotarget 13758 Oncotarget Table 2: Main characteristics of the studies on the CAG repeats included in the meta-analysis Case Control Author Control Case Cases/ Country Ethnicity Method Latitude Group (year) source size controls < 21 21–25 > 25 < 21 21–25 > 25 Grassetti D Italy Caucasian Sequence HB > 200 302/322 Total 83 171 48 74 211 37 (2015) Mid-latitude Kristiansen Norway Caucasian Sequence PB High-latitude > 200 635/304 Total 189 374 72 91 174 72 W (2012) 316 Seminomas 92 187 37 319 Non-seminomas 97 187 35 Vastermark Sweden Caucasian Sequence PB High-latitude > 200 275/214 Total 89 161 25 72 112 30 A (2011a) Vastermark Denmark Caucasian Sequence PB High-latitude < 200 89/214 Total 31 52 6 72 112 30 A (2011b) 172 Seminomas 59 97 16 186 Non-seminomas 60 111 15 Garolla A Italy Caucasian Sequence PB Mid-latitude < 200 123/115 Total 38 67 18 37 68 10 (2005) Giwercman Sweden Caucasian Sequence PB High-latitude < 200 83/220 Total 28 41 14 78 99 43 A (2004) 27 Seminomas 8 19 0 41 Non-seminomas 15 18 8 Rajpert-De Meyts E Denmark Caucasian Sequence HB High-latitude < 200 102/110 Total 32 61 9 43 55 12 (2002) Abbreviations: HB, hospital-based; PB, population-based. Table 3: Main results for the GGN repeats included in the meta-analysis Cases/Controls OR (95% CI) P P I Z P h b Overall 1636/1519 1.22 (1.05–1.41) 0.010 0.929 0.0% 2.59 0.840 Case size > 200 1439/1190 1.20 (1.02–1.41) 0.028 0.788 0.0% 2.19 < 200 197/329 1.32 (0.91–1.92) 0.143 0.798 0.0% 1.46 Latitude Mid-latitude 660/1013 1.23 (1.00–1.51) 0.050 0.934 0.0% 1.96 High-latitude 976/506 1.20 (0.97–1.49) 0.089 0.549 0.0% 1.7 Histology Seminomas 639/828 1.24 (1.00–1.54) 0.050 0.776 0.0% 1.96 Non-seminomas 634/828 1.14 (0.91–1.42) 0.249 0.845 0.0% 1.15 Note: P value of the Q-test for the heterogeneity test; P value of Egger’s test for publication bias. h b I : 0–25, absence of heterogeneity; 25–50, modest heterogeneity; 50, high heterogeneity. Bold numbers indicate significant differences. Figure 2: Forest plot of the GGN repeat polymorphism and the risk of testicular cancer. (A) Overall analysis. (B) Case size subgroup. (C) Latitude subgroup. (D) Histology subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). www.impactjournals.com/oncotarget 13759 Oncotarget Figure 3: Forest plot of the long CAG repeat polymorphism and the risk of TC in the subgroup analysis. (A) Latitude subgroup. (B) Histology subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). Figure 4: Forest plot of the abnormal CAG repeat polymorphism and the risk of testicular cancer in the subgroup analysis. (A). Control source subgroup. (B). Latitude subgroup. (C). Histology subgroup. (D). Case size subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). www.impactjournals.com/oncotarget 13760 Oncotarget www.impactjournals.com/oncotarget 13761 Oncotarget Table 4: Main results for the CAG repeats included in the meta-analysis < 21 vs. 21–25 > 25 vs. 21–25 < 21 + > 25 vs. 21–25 Cases/Controls OR (95% OR (95% OR (95% 2 2 2 P P I Z P P P I Z P P P I Z P h b h b h b CI) CI) CI) 0.98 0.78 0.90 Overall 1609/1285 (0.83–1.15) 0.809 0.469 0.0% 0.24 0.371 (0.49–1.25) 0.304 0.001 73.4% 1.03 0.671 (0.71–1.14) 0.381 0.031 56.6% 0.88 0.941 Control source 1.00 1.15 1.02 HB 404/432 (0.49–2.02) 0.999 0.041 76.1% 0.00 (0.50–2.61) 0.741 0.108 61.3% 0.33 (0.48–2.17) 0.955 0.017 82.5% 0.06 0.93 0.66 0.81 PB 1205/853 (0.77–1.13) 0.467 0.980 0.0% 0.73 (0.42–1.04) 0.071 0.049 58.1% 1.81 (0.68–0.96) 0.017 0.591 0.0% 2.39 Case size 1.04 0.75 0.94 > 200 1212/626 (0.85–1.28) 0.687 0.185 40.7% 0.40 (0.34–1.65) 0.481 0.000 88.0% 0.71 (0.62–1.44) 0.786 0.004 81.8% 0.27 0.87 0.82 0.86 < 200 397/659 (0.66–1.15) 0.335 0.752 0.0% 0.96 (0.55–1.23) 0.340 0.141 45.0% 0.95 (0.66–1.11) 0.244 0.453 0.0% 1.16 Latitude 1.27 1.65 1.38 Mid-latitude 425/437 (0.93–1.73) 0.132 0.411 0.0% 1.51 (1.09–2.50) 0.017 0.789 0.0% 2.39 (1.05–1.82) 0.021 0.549 0.0% 2.31 High- 0.89 0.54 0.76 latitude 1085/848 (0.73–1.08) 0.220 0.873 0.0% 1.23 (0.41–0.70) 0.000 0.698 0.0% 4.62 (0.64–0.90) 0.002 0.983 0.0% 3.11 Histology 0.89 0.47 0.73 Seminomas 515/738 (0.69–1.16) 0.396 0.481 0.0% 0.85 0.182 (0.33–0.68) 0.000 0.314 13.6% 4.06 (0.57–0.92) 0.008 0.181 41.5% 2.66 Non- 0.94 0.52 0.78 seminomas 546/738 (0.73–1.22) 0.652 0.802 0.0% 0.45 0.870 (0.37–0.74) 0.000 0.285 20.3% 3.65 (0.62–0.98) 0.032 0.657 0.0% 2.14 Abbreviations: HB, hospital-based; PB, population-based. Note: P value of Q-test for heterogeneity test; P value of egger’s test for publication bias. h b I : 0–25, absence of heterogeneity; 25–50, modest heterogeneity; 50, high heterogeneity. Bold numbers indicate significant differences. susceptibility [18–20]. The development of TC is postulated that CAG repeat polymorphisms with > 25 and < 21 + > 25 to be due to endocrine disruption, particularly abnormalities repeats were associated with an increased risk of TC in in the action of gonadotropins and steroidal sex hormones the mid-latitude subgroup but were associated with [17]. Men with androgen insensitivity syndrome caused by a decreased risk of TC in the high-latitude subgroup, AR gene mutations have a higher risk of developing TC. indicating that latitude plays a key role in the effect of There is evidence of an inverse correlation between the CAG polymorphism on the risk of TC. In addition, long variability in AR CAG and GGN repeat numbers and the CAG repeats reduced AR activity and increased the risk of transactivation efficiency in AR [6, 9, 10, 15]. Irvine et al. TC in the mild mid-latitude environment. Previous studies [21] suggested that a longer CAG and GGN repeat region have indicated that men with CAG repeats > 25 have might reduce the transactivation activity in AR. lower androgen sensitivity [27, 28]. However, in the harsh Abnormalities in AR genes are also common in and cold, high-altitude environment, long CAG repeats other disorders, such as prostate cancer, hypospadias, may protect against TC. This is because the exposure to cryptorchidism, and infertility [22–25]. Many authors different environments or lifestyle-related factors may have have attempted to understand whether reduced androgen opposing effects on the male reproductive system [29]. sensitivity is caused by point mutations or by excessively In addition, we cannot exclude the possibility of the long CAG and GGN repeat segments, which might lead to moderate effect of CAG repeat polymorphisms on the risk the development of testicular agenesis and consequently of TC due to marginal associations. These polymorphisms increase susceptibility to TC [10, 26]. within or near the AR may drive malignant phenotypes. Giwercman et al. [13] and Rajpert-De Meyts et al. Therefore, large studies focusing on both gene-gene and [12] investigated the correlation between CAG and GGN gene-environment interactions are needed to explore the repeats and TC. No statistically significant differences mechanism of testicular carcinogenesis. in CAG or GGN repeat numbers were observed between However, this meta-analysis has some limitations. patients with TGCT and the control group. This was the First, some studies with small sample size may not have first study that demonstrated a correlation between AR CAG enough statistical power to determine the real association repeats, TGCT histology, and disease progression, albeit the and are thought to be more likely to report larger beneficial study size was limited [12, 13]. Grassetti et al. [10] observed effects compared with larger trials [30]. Second, our that there was a larger variability of CAG than GGN repeats results were only based on a Caucasian sample and in both patients and controls, especially among those with polymerase chain reaction (PCR) sequences, and a more rare alleles. When stratified, men with CAG repeats < 21 precise analysis would be conducted if more data were or > 24 were found to have a 50% and 76% higher risk of available. Third, clinical disorders are not the result of the TC, respectively, than those with CAG 21–24. Therefore, disruption of a single gene, and genetic disruptions are the risk of developing TC seems to be lower for men with a embedded within the entire genome and are affected by CAG repeat number between 21 and 24. environment exposure. In fact, other genes related to TC In the meta-analysis, our first finding was that long can also play a preeminent role in testis development. GGN repeats were associated with an increased risk of In conclusion, we found that long GGN repeats were TC, compared with repeats > 23; similarly, an increased associated with an increased risk of TC compared with risk was observed in studies with sample size > 200 and in a reference group. Furthermore, an association between the mid-latitude and seminoma subgroups. We speculated GGN repeats in AR and the risk of TC was found in that GGN > 23 was associated with lower AR activity studies with a sample size > 200 and in the mid-latitude compared with the more common genotype with GGN and seminoma subgroups. We found that CAG repeat ≤ 23, indicating that low androgen response could play a polymorphisms with > 25 and < 21 + > 25 repeats might role in disease progression, which is consistent with the confer a protective effect to the patients with TC in the PB, results of previous studies [10, 17]. high-latitude, seminoma, and non-seminoma subgroups. Overall, the present meta-analysis reports for the However, it CAG repeat polymorphisms with > 25 and first time the association between AR CAG and GGN < 21 + > 25 repeats in the mid-latitude subgroup were repeat polymorphisms and the risk of TC. No significant associated with an increased risk of TC. association was observed between CAG repeat and TC in the models evaluated in the overall analysis, and the MATERIALS AND METHODS groups were heterogeneous. We then evaluated the source of heterogeneity in these groups. Furthermore, in the Literature selection subgroup analysis of latitude, case size, control source, and histology, a significant association was found between Data from single reports were extracted (Figure 1). CAG repeats and TC in the PB, mid-latitude, high-latitude, We searched PubMed and Web of Science until July 2015 seminoma, and non-seminoma subgroups. to identify publications on the association between TC and Interestingly, we observed no heterogeneity after CAG and/or GGN trinucleotide repeat lengths in AR. We stratifying according to latitude and histology. We found focused on the studies performed in humans and on those www.impactjournals.com/oncotarget 13762 Oncotarget that utilized the following key words: testicular cancer or ACKNOWLEDGMENTS AND FUNDING TC, androgen receptor or AR, combined with CAG and/ or GGN. We appreciate the assistance from all the members The inclusion criteria were as follows: (1) studies of our laboratory. This study was supported by the Medical that evaluated the association between AR CAG or Health and Scientific Research Foundation of Nanjing GGC/GGN repeat polymorphisms and the risk of TC; Military Command (No. 2014MS098), the Key Foundation of (2) studies with a case-control design; (3) studies that Jiangsu Science and Technology Bureau (No. BM2013058), provided sufficient information on CAG or GGC/GGN and the Foundation of Nanjing General Hospital of Nanjing repeat distributions between patients and controls; Military Command, PLA. (No. 2014044). (4) studies for which the full text was available. Two reviewers assessed the full text of eligible studies from CONFLICTS OF INTEREST the above databases. Additional studies were identified by manually searching references of original and review The authors declare that no conflicts of interest are articles on this topic. associated with this publication. Data extraction and verification REFERENCES Information on the enrolled studies is shown in 1. Bosl GJ, Motzer RJ. Testicular germ-cell cancer. N Engl J Tables 1 and 2, including: (I) the first author’s name; Med. 1997; 337:242–254. (II) year of publication; (III) country or region of 2. Huyghe E, Matsuda T, Thonneau P. Increasing incidence of origin, ethnicity, and method; (IV) number of cases and testicular cancer worldwide: a review. J Urol. 2003; 170:5–11 controls. Two authors (WJ Jiang and SM Liu) of the study 3. Chia VM, Quraishi SM, Devesa SS, Purdue MP, Cook MB, extracted the information independently and screened the McGlynn KA. International trends in the incidence of citations that met the inclusion criteria. The discrepancies testicular cancer, 1973–2002. Cancer Epidemiol Biomarkers were adjudicated via discussions until a consensus was Prev. 2010; 19:1151–1159. reached. 4. Manecksha RP, Fitzpatrick JM. Epidemiology of testicular cancer. BJU Int. 2009; 104:1329–1333. Data and statistical analysis 5. Garner MJ, Turner MC, Ghadirian P, Krewski D. Epidemiology of Data were divided into three categories: CAG testicular cancer: an overview. Int J Cancer .2005; 116:331–339. repeat length of 21–25, which was used as the reference 6. Kristiansen W, Aschim EL, Andersen JM, Witczak O, group, CAG length < 21, and CAG length > 25. The GGN Fosså SD, Haugen TB. Variations in testosterone pathway repeat length was divided into 2 categories: GGN repeat genes and susceptibility to testicular cancer in Norwegian length ≤ 23, which was used as the reference group, and men. Int J Androl. 2012; 35:819–827. GGN length > 23. To obtain specific data from these 7. Gu M, Dong X, Zhang X, Niu W. The CAG repeat categories, we thoroughly analyzed and carefully obtained polymorphism of androgen receptor gene and prostate information from all the studies that met the inclusion cancer: a meta-analysis. Mol Biol Rep. 2012; 39:2615–2624. criteria. 8. Giwercman YL, Xu C, Arver S, Pousette A, Reneland R. We predicted the contribution of AR CAG and No association between the androgen receptor gene CAG GGN repeat polymorphisms to the risk of TC using the repeat and impaired sperm production in Swedish men. Clin Stata software version 11.0. The OR with 95% (CI) was Genet. 1998; 54:435–436. calculated to measure the strength of the associations 9. Västermark Å, Giwercman YL, Hagströmer O, Rajpert De- [31]. A test of heterogeneity of the pooled results was Meyts E, Eberhard J, Ståhl O, Cedermark GC, Rastkhani H, performed using Cochran’s Q test and Higgins I statistic Daugaard G, Arver S, Giwercman A. Polymorphic variation [32]. I > 50% is considered as a measure of significant in the androgen receptor gene: association with risk of heterogeneity. If the result of the Q test was P > 0.10, testicular germ cell cancer and metastatic disease. Eur J the OR was analyzed using the fixed-effects model Cancer. 2011; 47:413–419. (Mantel–Haenszel method). Otherwise, a random- 10. Grassetti D, Giannandrea F, Paoli D, Masciandaro P, effects model (DerSimonian–Laird method) was used in Figura V, Carlini T, Rizzo F, Lombardo F, Lenzi A, cases of significant heterogeneity. Sensitivity analysis Gandini L. Androgen receptor polymorphisms and testicular was performed to estimate the stability of the results by cancer risk. Andrology. 2015; 3:27–33. removing each study from the analysis, one at a time. Publication bias was evaluated using Begg’s funnel plot. 11. Nenonen HA, Giwercman A, Hallengren E, Giwercman YL. In addition to the visual inspection of the funnel plot, a Non-linear association between androgen receptor CAG value of P < 0.05 was considered to indicate the presence repeat length and risk of male subfertility - a meta-analysis. of significant publication bias [33–37]. Int J Androl. 2011; 34:327–332. www.impactjournals.com/oncotarget 13763 Oncotarget 12. Rajpert-De Meyts E, Leffers H, Daugaard G, Andersen CB, cryptorchidism risk among Hispanic white boys. J Clin Petersen PM, Hinrichsen J, Pedersen LG, Skakkebaek NE. Endocrinol Metab. 2012; 97:E393–E399. Analysis of the polymorphic CAG repeat length in the 25. Lazaros L, Xita N, Takenaka A, Sofikitis N, Makrydimas G, androgen receptor gene in patients with testicular germ cell Stefos T, Kosmas I, Zikopoulos K, Hatzi E, Georgiou I. Semen cancer. Int J Cancer. 2002; 102:201–204. quality is influenced by androgen receptor and aromatase gene 13. Giwercman A, Lundin KB, Eberhard J, Ståhl O, Cwikiel M, synergism. Hum Reprod. 2012; 27:3385–3392. Cavallin-Ståhl E, Giwercman YL. Linkage between 26. Giwercman A, Rylander L, Hagmar L, Giwercman YL. androgen receptor gene CAG trinucleotide repeat length Ethnic differences in occurrence of TDS-genetics and/or and testicular germ cell cancer histological type and clinical environment? Int J Androl. 2006; 29:291–297. stage. Eur J Cancer. 2004; 40:2152–2158. 27. Lundin KB, Giwercman YL, Rylander L, Hagmar L, 14. Eberhard J, Ståhl O, Giwercman Y, Cwikiel M, Cavallin- Giwercman A. Androgen receptor gene GGN repeat length Ståhl E, Lundin KB, Flodgren P, Giwercman A. Impact of and reproductive characteristics in young Swedish men. Eur therapy and androgen receptor polymorphism on sperm J Endocrinol. 2006; 155:347–354. concentration in men treated for testicular germ cell cancer: 28. Buchanan G, Yang M, Cheong A, Harris JM, Irvine a longitudinal study. Hum Reprod. 2004; 19:1418–1425. RA, Lambert PF, Moore NL, Raynor M, Neufing PJ, 15. Aschim EL, Oldenburg J, Kristiansen W, Giwercman A, Coetzee GA. Structural and functional consequences of Witczak O, Fosså SD, Haugen TB. Genetic variations glutamine tract variation in the androgen receptor. Hum Mol associated with the effect of testicular cancer treatment on Genet. 2004; 13:1677–1692. gonadal hormones. Hum Reprod. 2014; 29:2844–2851. 29. Davis-Dao CA, Siegmund KD, Vandenberg DJ, Skinner EC, 16. Davis-Dao CA, Tuazon ED, Sokol RZ, Cortessis VK. Coetzee GA, Thomas DC, Pike MC, Cortessis VK. Male infertility and variation in CAG repeat length in the Heterogenous effect of androgen receptor CAG tract androgen receptor gene: a meta-analysis. J Clin Endocrinol length on testicular germ cell tumor risk: shorter repeats Metab. 2007; 92:4319–4326. associated with seminoma but not other histologic types. 17. Garolla A, Ferlin A, Vinanzi C, Roverato A, Sotti G, Carcinogenesis. 2011; 32:1238–1243. Artibani W, Foresta C. Molecular analysis of the androgen 30. Zhang Z, Xu X, Ni H. Small studies may overestimate receptor gene in testicular cancer. Endocr Relat Cancer. the effect sizes in critical care meta-analyses: a meta- 2005; 12:645–655. epidemiological study. Crit Care. 2013; 17:R2. 18. Krausz C, Looijenga LH. Genetic aspects of testicular germ 31. Gu XB, Tian T, Tian XJ, Zhang XJ. Prognostic significance cell tumors. Cell Cycle. 2008; 7:3519–3524. of neutrophil-to-lymphocyte ratio in non-small cell lung 19. Giannandrea F, Gandini L, Paoli D, Turci R, Figà- cancer: a meta-analysis. Sci Rep. 2015; 5:12493. Talamanca I. Pesticide exposure and serum organochlorine 32. He J, Liao XY, Zhu JH, Xue WQ, Shen G-P, Huang SY, residuals among testicular cancer patients and healthy Chen W, Jia WH. Association of MTHFR C677T and A1298C controls. J Environ Sci Health B. 2011; 46:780–787. polymorphisms with non-Hodgkin lymphoma susceptibility: 20. Giannandrea F, Paoli D, Figà-Talamanca I, Lombardo F, evidence from a meta-analysis. Sci Rep. 2014; 4:6159 Lenzi A, Gandini L. Effect of endogenous and exogenous 33. Qin T, Du M, Du H, Shu Y, Wang M, Zhu L. Folic acid hormones on testicular cancer: the epidemiological supplements and colorectal cancer risk: meta-analysis of evidence. Int J Dev Biol. 2013; 57:255–263. randomized controlled trials. Sci Rep. 2015; 5:12044. 21. Irvine RA, Ma H, Yu MC, Ross RK, Stallcup MR, 34. Ntais C, Polycarpou A, Ioannidis JP. Association of Coetzee GA. Inhibition of p160-mediated coactivation with GSTM1, GSTT1, and GSTP1 gene polymorphisms with the increasing androgen receptor polyglutamine length. Hum risk of prostate cancer: a meta-analysis. Cancer Epidemiol Mol Genet. 2000; 9:267–274. Biomarkers Prev. 2005; 14:176–181. 22. Giwercman A, Richiardi L, Kaijser M, Ekbom A, Akre O. 35. Yang Y, Yin X, Sheng L, Xu S, Dong L, Liu L. Perioperative Reduced risk of prostate cancer in men who are childless as chemotherapy more of a benefit for overall survival than compared to those who have fathered a child: a population adjuvant chemotherapy for operable gastric cancer: an based case-control study. Int J Cancer. 2005; 115:994–997. updated Meta-analysis. Sci Rep. 2015; 5:12850. 23. Vottero A, Minari R, Viani I, Tassi F, Bonatti F, Neri TM, 36. Jiang W, Sun H, Zhang J, Zhou Q, Wu Q, Li T, Zhang C, Bertolini L, Bernasconi S, Ghizzoni L. Evidence for epigenetic Li W, Zhang M, Xia X. Polymorphisms in Protamine 1 abnormalities of the androgen receptor gene in foreskin from and Protamine 2 predict the risk of male infertility: a meta- children with hypospadias. J Clin Endocrinol Metab. 2011; analysis. Sci Rep. 2015; 5:15300. 96:E1953–E1962. 37. He J, Wang MY, Qiu LX, Zhu ML, Shi TY, Zhou XY, 24. Davis-Dao C, Koh CJ, Hardy BE, Chang A, Kim SS, De Sun MH, Yang YJ, Wang JC, Jin L, Yu HP, Wei QY. Genetic Filippo R, Hwang A, Pike MC, Carroll JD, Coetzee GA, variations of mTORC1 genes and risk of gastric cancer Vandenberg D, Siegmund K, Cortessis VK. Shorter in an Eastern Chinese population. Mol Carcinog. 2013; androgen receptor CAG repeat lengths associated with 52:E70–E79. www.impactjournals.com/oncotarget 13764 Oncotarget http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Oncotarget Pubmed Central

Predictive value of GGN and CAG repeat polymorphisms of androgen receptors in testicular cancer: a meta-analysis

Loading next page...
 
/lp/pubmed-central/predictive-value-of-ggn-and-cag-repeat-polymorphisms-of-androgen-0ZvdkrSh0c

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
Copyright
Copyright: © 2016 Jiang et al.
ISSN
1949-2553
eISSN
1949-2553
DOI
10.18632/oncotarget.7337
Publisher site
See Article on Publisher Site

Abstract

www.impactjournals.com/oncotarget/ Oncotarget, Vol. 7, No. 12 Predictive value of GGN and CAG repeat polymorphisms of androgen receptors in testicular cancer: a meta-analysis 1 1 1 1 1 1 Weijun Jiang , Jing Zhang , Qing Zhou , Shuaimei Liu , Mengxia Ni , Peiran Zhu , 1 1 1 1 Qiuyue Wu , Weiwei Li , Mingchao Zhang , Xinyi Xia Department of Reproduction and Genetics, Institute of Laboratory Medicine, Jinling Hospital, Nanjing University School of Medicine, Nanjing 210002, P.R. China Correspondence to: Xin yi Xia, e-mail: xiaxynju@163.com Keywords: androgen receptor, GGN repeat, CAG repeat, testicular cancer, polymorphism Received: October 22, 2015 Accepted: January 29, 2016 Published: February 12, 2016 ABSTRACT The risk of testicular cancer (TC) is markedly increased in subjects with androgen insensitivity, and previous studies have proposed that GGN and CAG repeats in androgen receptors (AR) could be related to the risk of TC. To evaluate the association between the length of GGN and CAG repeats in AR and TC, a meta-analysis involving 3255 TC cases and 2804 controls was performed. The results suggested that long GGN repeats are associated with an increased risk of TC compared with those < 23 [odds ratio (OR) = 1.22, 95% confidence interval (CI) = 1.05–1.41]; similarly, a subgroup analysis revealed that this association occurred in studies with case sizes > 200, and in the mid-latitude, and seminoma subgroups. The subgroup analysis based on populations, high-latitude, and seminomas/non-seminomas suggested that AR CAG repeat polymorphisms with > 25 and < 21 + > 25 repeats might confer a protective effect to the patients with TC (in the high-latitude subgroup analysis, for > 25 vs. 21–25: OR = 0.54, 95% CI = 0.41–0.70). In contrast, an increased risk of TC was observed for AR CAG repeat polymorphisms with > 25 and < 21 + > 25 repeats in the mid-latitude subgroup (for > 25 vs. 21–25: OR = 1.65, 95% CI = 1.09–2.50). In addition, no associations between the remaining subgroups and male infertility were observed. In short, this meta-analysis suggested that AR GGN and CAG repeat polymorphisms may be involved in the etiology of TC. encoded by a (CAG) CAA stretch, as well as a INTRODUCTION (GGT) GGG(GGT) (GGC) repeat, and these repeats 3 2 n are designated CAG and GGN repeats, respectively [9]. Testicular cancer (TC) is an malignancy, accounting The extreme variability of the number of these repeats for 1%–2% of all tumors among men worldwide [1], determines the different lengths of the polyglutamine and and affects primarily young men in the age group 15–44 polyglycine segments in the N-terminal transactivation years. The incidence of TC is increasing worldwide and domain of the AR [10]. In men, the number of CAG has steeply increased in the past 40 years in almost all repeats can vary from 8 to 37, with an average of 20–22, Western countries [2–4]. Clinical studies reported that depending on the ethnic origin. Africans and Asians have 95% of all TCs are testicular germ cell tumors (TGCT), a lower number of repeats than Caucasians and a reduced with an approximately equal division between seminomas risk of TGCT [10]. Changes in the length of the CAG and non-seminomas, and epidemiological studies have polymorphic trinucleotide repeat in the AR gene may suggested that environmental factors, including endocrine lead to the altered transactivation of the AR gene and disrupting agents, which act as either weak estrogen have been implicated to play a role in the pathogenesis of agonists or androgen antagonists, are primarily responsible several forms of endocrine cancer and certain reproductive for the increased incidence of TC [5, 6]. disorders [11]. Subjects with reproductive disorders that The AR gene, located on Xq11-12, has 8 exons are associated with a relative deficiency in androgen and 7 introns, and in exon 1, this gene contains function have an increased risk of TC [12, 13]. In the two important polymorphic trinucleotide repeats of past decade, some studies have attempted to evaluate the polyglutamine and polyglycine tracts [7, 8], which are www.impactjournals.com/oncotarget 13754 Oncotarget association between CAG and GGN repeat number and the 95% CI = 1.02– 1.41, P = 0.028; for > 23 vs. ≤ 23: OR = risk of TC [6, 9, 10, 12– 15]; however, the results appear 1.24, 95% CI = 1.00–1.54, P = 0.050). contradictory because of differences in the sources of the study participants and inconsistencies in the inclusion Association between CAG repeat length and the criteria in case and control subjects among the studies risk of TC [9, 16]. To the best of our knowledge, to date, no meta- analysis has analyzed the results of all the studies that The association between CAG repeat in AR and evaluated this association. Therefore, this meta-analysis the risk of TC is summarized in Figures 3 and 4 and was conducted to investigate the association between in Table 4. In brief, the overall analysis indicated no CAG and GGN repeat polymorphisms and the risk of significant association between CAG repeats and the risk TC, as well as the genetic heterogeneity across different of TC for the models evaluated; however, in the subgroup control sources and study designs. Herein, seven reports analysis based on latitude, sample size, control source, involving 3255 TC cases and 2804 controls were identified and histology, significant associations were found between according to the inclusion criteria for the pooled analysis. CAG repeats and TC in the population-based (PB) subgroup (for < 21 + > 25 vs. 21–25: OR = 0.81, 95% CI = 0.68–0.96, P = 0.017), mid-latitude subgroup (for > 25 vs. RESULTS 21–25: OR = 1.65, 95% CI = 1.09–2.50, P = 0.017; for < 21 + > 25 vs. 21–25: OR = 1.38, 95% CI = 1.05–1.82, Study characteristics P = 0.021), high-latitude subgroup (for > 25 vs. 21–25: OR = 0.54, 95% CI = 0.41–0.70, P = 0.000; for < 21 + > 25 Because not all of the studies evaluated provided vs. 21–25: OR = 0.76, 95% CI= 0.64–0.90, P = 0.002), specific distributions of AR CAG or GGN repeat counts, seminoma subgroup (for > 25 vs. 21–25: OR = 0.47, 95% we used a CAG repeat length of 21–25 as reference to CI = 0.33–0.68, P = 0.000; for < 21 + > 25 vs. 21–25: evaluate dichotomous comparisons (< 21 CAG repeats OR = 0.73, 95% CI = 0.57–0.92, P = 0.008), and non- vs. the reference, > 25 CAG repeats vs. the reference, and seminoma subgroup (for > 25 vs. 21–25: OR = 0.52, 95% < 21 + > 25 CAG repeats vs. the reference). Similarly, the CI = 0.37–0.74, P = 0.000; for < 21 + > 25 vs. 21–25: OR GGN genotype of ≤ 23 repeats was used as reference to = 0.78, 95% CI = 0.62–0.98, P = 0.032). assess the association between > 23 repeats and the risk of TC. Through literature search and selection based on the Publication bias and small-study effects inclusion criteria, 7 articles published between 2002 and 2015 were identified after reviewing potentially relevant To assess the publication bias of the studies, Begg’s articles (Figure 1). The characteristics of the enrolled funnel plot and Egger’s test were performed. The shapes studies are summarized in Tables 1 and 2. of the funnel plots revealed no evidence of obvious For the GGN repeats, seven studies involving 1636 asymmetry. Egger’s test was used to provide statistical cases (range of 74–635, average of 272.67 ± 180.5) and evidence of funnel plot symmetry (Supplementary 1519 controls (range of 115–576, average of 304 ± 154) Figure S1) and indicated no evidence of publication bias were included in the meta-analysis. or small-study effects across the studies (for GGN repeats For the CAG repeats, 6 studies involving 1609 > 23 vs. ≤ 23, P = 0.840; for CAG repeat < 21 vs. 21–25, cases (range of 83–635, average of 230 ± 185) and 1285 P = 0.371; for CAG repeats > 25 vs. 21–25, P = 0.671; for controls (range of 110–322, average of 214 ± 82.1) met the CAG repeats < 21 + > 25 vs. 21–25, P = 0.941). inclusion criteria and were selected for the meta-analysis. Sensitivity analysis Association between GGN repeat length and the To confirm the results of the current study, the I risk of TC statistics calculated for the overall analysis of the CAG The association between GGN repeats and the risk repeats > 25 vs. 21–25 and CAG repeats < 21 + > 25 vs. of TC is summarized in Figure 2 and Table 3. The overall 21–25 were 73.4% and 56.6%, respectively, indicating analysis indicated a significant association between GGN that more than 50% of the abnormal CAG repeats may repeats and TC [odds ratio (OR) = 1.22, 95% confidence be due to between-study heterogeneity. We then evaluated interval (CI) = 1.05–1.41, P = 0.010]. To clarify the the source of heterogeneity in these comparisons by potential effect of latitude, sample size, and histological sample size, latitude, and histology stratifications, and differences, a subgroup analysis of study populations we observed no heterogeneity in latitude and histology was also conducted, and a significant association stratifications (Figure 4). Sensitivity analyses were was found between GGN repeats and TC in studies conducted to determine whether modifications in the with a sample size > 200 and in the mid-latitude and inclusion criteria of the meta-analysis affected the results. seminoma subgroups (for > 23 vs. ≤ 23: OR = 1.23, 95% Our results indicated that the studies by Grassetti et al. CI = 1.00– 1.51, P = 0.050; for > 23 vs. ≤ 23: OR = 1.20, [10] and Garolla et al. [17] caused this heterogeneity. www.impactjournals.com/oncotarget 13755 Oncotarget seminoma, and non-seminoma subgroups. However, AR DISCUSSION CAG repeat polymorphism with > 25 and < 21 + > 25 repeats in the mid-latitude subgroup were associated with The present meta-analysis, including 3255 TC an increased risk of TC. cases and 2804 controls from seven case-control studies, TC is a very common disease and its incidence has explored the association between GGN and CAG repeat increased worldwide in recent decades. TGCT makes up polymorphisms in AR and the risk of TC. Our results 95% of all TCs and is the most common solid tumor in indicated that long GGN repeats were associated with men aged 15–39 years [2, 10]. Although there has been an increased risk of TC, compared with repeats < 23; enormous progress in the clinical treatment of TC and similarly, sample size > 200 and the mid-latitude and preservation of fertility through sperm banks in recent seminoma subgroups were associated with an increased years, the main causes of the disease remain unclear. risk of TC. In contrast, AR CAG repeat polymorphism with However, important risk factors include work, lifestyle, > 25 and < 21 + > 25 repeats may confer a protective effect diet, familial history, environmental conditions, and genetic to the TC patients in the analysis of the PB, high-latitude, Figure 1: Flow diagram of the study selection process. www.impactjournals.com/oncotarget 13756 Oncotarget www.impactjournals.com/oncotarget 13757 Oncotarget Table 1: Main characteristics of the studies on the GGN repeats included in the meta-analysis Case Control Control Case Cases/ Author (year) Country Ethnicity Method Latitude Group source size Controls ≤ 23 > 23 ≤ 23 > 23 Grassetti D (2015) Italy Caucasian Sequence HB Mid-latitude > 200 302/322 Total 182 120 206 116 166 Seminomas 97 69 136 Non-seminomas 85 51 Kristiansen W (2012) Norway Caucasian Sequence PB High-latitude > 200 635/292 Total 391 244 196 96 315 Seminomas 191 124 320 Non-seminomas 200 120 Vastermark A (2011a) Sweden Caucasian Sequence PB High-latitude > 200 267/214 Total 175 92 141 73 Vastermark A (2011b) Denmark Caucasian Sequence PB High-latitude < 200 74/214 Total 43 31 141 73 158 Seminomas 101 57 178 Non-seminomas 114 64 Biggs ML (2008) USA Mixed Sequence PB Mid-latitude > 200 235/576 Total 133 102 359 217 Garolla A (2005) Italy Caucasian Sequence PB Mid-latitude < 200 123/115 Total 69 54 71 44 Abbreviations: HB, hospital-based; PB, population-based. www.impactjournals.com/oncotarget 13758 Oncotarget Table 2: Main characteristics of the studies on the CAG repeats included in the meta-analysis Case Control Author Control Case Cases/ Country Ethnicity Method Latitude Group (year) source size controls < 21 21–25 > 25 < 21 21–25 > 25 Grassetti D Italy Caucasian Sequence HB > 200 302/322 Total 83 171 48 74 211 37 (2015) Mid-latitude Kristiansen Norway Caucasian Sequence PB High-latitude > 200 635/304 Total 189 374 72 91 174 72 W (2012) 316 Seminomas 92 187 37 319 Non-seminomas 97 187 35 Vastermark Sweden Caucasian Sequence PB High-latitude > 200 275/214 Total 89 161 25 72 112 30 A (2011a) Vastermark Denmark Caucasian Sequence PB High-latitude < 200 89/214 Total 31 52 6 72 112 30 A (2011b) 172 Seminomas 59 97 16 186 Non-seminomas 60 111 15 Garolla A Italy Caucasian Sequence PB Mid-latitude < 200 123/115 Total 38 67 18 37 68 10 (2005) Giwercman Sweden Caucasian Sequence PB High-latitude < 200 83/220 Total 28 41 14 78 99 43 A (2004) 27 Seminomas 8 19 0 41 Non-seminomas 15 18 8 Rajpert-De Meyts E Denmark Caucasian Sequence HB High-latitude < 200 102/110 Total 32 61 9 43 55 12 (2002) Abbreviations: HB, hospital-based; PB, population-based. Table 3: Main results for the GGN repeats included in the meta-analysis Cases/Controls OR (95% CI) P P I Z P h b Overall 1636/1519 1.22 (1.05–1.41) 0.010 0.929 0.0% 2.59 0.840 Case size > 200 1439/1190 1.20 (1.02–1.41) 0.028 0.788 0.0% 2.19 < 200 197/329 1.32 (0.91–1.92) 0.143 0.798 0.0% 1.46 Latitude Mid-latitude 660/1013 1.23 (1.00–1.51) 0.050 0.934 0.0% 1.96 High-latitude 976/506 1.20 (0.97–1.49) 0.089 0.549 0.0% 1.7 Histology Seminomas 639/828 1.24 (1.00–1.54) 0.050 0.776 0.0% 1.96 Non-seminomas 634/828 1.14 (0.91–1.42) 0.249 0.845 0.0% 1.15 Note: P value of the Q-test for the heterogeneity test; P value of Egger’s test for publication bias. h b I : 0–25, absence of heterogeneity; 25–50, modest heterogeneity; 50, high heterogeneity. Bold numbers indicate significant differences. Figure 2: Forest plot of the GGN repeat polymorphism and the risk of testicular cancer. (A) Overall analysis. (B) Case size subgroup. (C) Latitude subgroup. (D) Histology subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). www.impactjournals.com/oncotarget 13759 Oncotarget Figure 3: Forest plot of the long CAG repeat polymorphism and the risk of TC in the subgroup analysis. (A) Latitude subgroup. (B) Histology subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). Figure 4: Forest plot of the abnormal CAG repeat polymorphism and the risk of testicular cancer in the subgroup analysis. (A). Control source subgroup. (B). Latitude subgroup. (C). Histology subgroup. (D). Case size subgroup. Studies are plotted according to the last name of the first author, followed by the publication year in parentheses. Each square represents the OR point estimate and its size is proportional to the weight of the study. The diamond (and broken line) represents the overall summary estimate and its width indicates the confidence interval. The unbroken vertical line is at the null value (OR = 1.0). www.impactjournals.com/oncotarget 13760 Oncotarget www.impactjournals.com/oncotarget 13761 Oncotarget Table 4: Main results for the CAG repeats included in the meta-analysis < 21 vs. 21–25 > 25 vs. 21–25 < 21 + > 25 vs. 21–25 Cases/Controls OR (95% OR (95% OR (95% 2 2 2 P P I Z P P P I Z P P P I Z P h b h b h b CI) CI) CI) 0.98 0.78 0.90 Overall 1609/1285 (0.83–1.15) 0.809 0.469 0.0% 0.24 0.371 (0.49–1.25) 0.304 0.001 73.4% 1.03 0.671 (0.71–1.14) 0.381 0.031 56.6% 0.88 0.941 Control source 1.00 1.15 1.02 HB 404/432 (0.49–2.02) 0.999 0.041 76.1% 0.00 (0.50–2.61) 0.741 0.108 61.3% 0.33 (0.48–2.17) 0.955 0.017 82.5% 0.06 0.93 0.66 0.81 PB 1205/853 (0.77–1.13) 0.467 0.980 0.0% 0.73 (0.42–1.04) 0.071 0.049 58.1% 1.81 (0.68–0.96) 0.017 0.591 0.0% 2.39 Case size 1.04 0.75 0.94 > 200 1212/626 (0.85–1.28) 0.687 0.185 40.7% 0.40 (0.34–1.65) 0.481 0.000 88.0% 0.71 (0.62–1.44) 0.786 0.004 81.8% 0.27 0.87 0.82 0.86 < 200 397/659 (0.66–1.15) 0.335 0.752 0.0% 0.96 (0.55–1.23) 0.340 0.141 45.0% 0.95 (0.66–1.11) 0.244 0.453 0.0% 1.16 Latitude 1.27 1.65 1.38 Mid-latitude 425/437 (0.93–1.73) 0.132 0.411 0.0% 1.51 (1.09–2.50) 0.017 0.789 0.0% 2.39 (1.05–1.82) 0.021 0.549 0.0% 2.31 High- 0.89 0.54 0.76 latitude 1085/848 (0.73–1.08) 0.220 0.873 0.0% 1.23 (0.41–0.70) 0.000 0.698 0.0% 4.62 (0.64–0.90) 0.002 0.983 0.0% 3.11 Histology 0.89 0.47 0.73 Seminomas 515/738 (0.69–1.16) 0.396 0.481 0.0% 0.85 0.182 (0.33–0.68) 0.000 0.314 13.6% 4.06 (0.57–0.92) 0.008 0.181 41.5% 2.66 Non- 0.94 0.52 0.78 seminomas 546/738 (0.73–1.22) 0.652 0.802 0.0% 0.45 0.870 (0.37–0.74) 0.000 0.285 20.3% 3.65 (0.62–0.98) 0.032 0.657 0.0% 2.14 Abbreviations: HB, hospital-based; PB, population-based. Note: P value of Q-test for heterogeneity test; P value of egger’s test for publication bias. h b I : 0–25, absence of heterogeneity; 25–50, modest heterogeneity; 50, high heterogeneity. Bold numbers indicate significant differences. susceptibility [18–20]. The development of TC is postulated that CAG repeat polymorphisms with > 25 and < 21 + > 25 to be due to endocrine disruption, particularly abnormalities repeats were associated with an increased risk of TC in in the action of gonadotropins and steroidal sex hormones the mid-latitude subgroup but were associated with [17]. Men with androgen insensitivity syndrome caused by a decreased risk of TC in the high-latitude subgroup, AR gene mutations have a higher risk of developing TC. indicating that latitude plays a key role in the effect of There is evidence of an inverse correlation between the CAG polymorphism on the risk of TC. In addition, long variability in AR CAG and GGN repeat numbers and the CAG repeats reduced AR activity and increased the risk of transactivation efficiency in AR [6, 9, 10, 15]. Irvine et al. TC in the mild mid-latitude environment. Previous studies [21] suggested that a longer CAG and GGN repeat region have indicated that men with CAG repeats > 25 have might reduce the transactivation activity in AR. lower androgen sensitivity [27, 28]. However, in the harsh Abnormalities in AR genes are also common in and cold, high-altitude environment, long CAG repeats other disorders, such as prostate cancer, hypospadias, may protect against TC. This is because the exposure to cryptorchidism, and infertility [22–25]. Many authors different environments or lifestyle-related factors may have have attempted to understand whether reduced androgen opposing effects on the male reproductive system [29]. sensitivity is caused by point mutations or by excessively In addition, we cannot exclude the possibility of the long CAG and GGN repeat segments, which might lead to moderate effect of CAG repeat polymorphisms on the risk the development of testicular agenesis and consequently of TC due to marginal associations. These polymorphisms increase susceptibility to TC [10, 26]. within or near the AR may drive malignant phenotypes. Giwercman et al. [13] and Rajpert-De Meyts et al. Therefore, large studies focusing on both gene-gene and [12] investigated the correlation between CAG and GGN gene-environment interactions are needed to explore the repeats and TC. No statistically significant differences mechanism of testicular carcinogenesis. in CAG or GGN repeat numbers were observed between However, this meta-analysis has some limitations. patients with TGCT and the control group. This was the First, some studies with small sample size may not have first study that demonstrated a correlation between AR CAG enough statistical power to determine the real association repeats, TGCT histology, and disease progression, albeit the and are thought to be more likely to report larger beneficial study size was limited [12, 13]. Grassetti et al. [10] observed effects compared with larger trials [30]. Second, our that there was a larger variability of CAG than GGN repeats results were only based on a Caucasian sample and in both patients and controls, especially among those with polymerase chain reaction (PCR) sequences, and a more rare alleles. When stratified, men with CAG repeats < 21 precise analysis would be conducted if more data were or > 24 were found to have a 50% and 76% higher risk of available. Third, clinical disorders are not the result of the TC, respectively, than those with CAG 21–24. Therefore, disruption of a single gene, and genetic disruptions are the risk of developing TC seems to be lower for men with a embedded within the entire genome and are affected by CAG repeat number between 21 and 24. environment exposure. In fact, other genes related to TC In the meta-analysis, our first finding was that long can also play a preeminent role in testis development. GGN repeats were associated with an increased risk of In conclusion, we found that long GGN repeats were TC, compared with repeats > 23; similarly, an increased associated with an increased risk of TC compared with risk was observed in studies with sample size > 200 and in a reference group. Furthermore, an association between the mid-latitude and seminoma subgroups. We speculated GGN repeats in AR and the risk of TC was found in that GGN > 23 was associated with lower AR activity studies with a sample size > 200 and in the mid-latitude compared with the more common genotype with GGN and seminoma subgroups. We found that CAG repeat ≤ 23, indicating that low androgen response could play a polymorphisms with > 25 and < 21 + > 25 repeats might role in disease progression, which is consistent with the confer a protective effect to the patients with TC in the PB, results of previous studies [10, 17]. high-latitude, seminoma, and non-seminoma subgroups. Overall, the present meta-analysis reports for the However, it CAG repeat polymorphisms with > 25 and first time the association between AR CAG and GGN < 21 + > 25 repeats in the mid-latitude subgroup were repeat polymorphisms and the risk of TC. No significant associated with an increased risk of TC. association was observed between CAG repeat and TC in the models evaluated in the overall analysis, and the MATERIALS AND METHODS groups were heterogeneous. We then evaluated the source of heterogeneity in these groups. Furthermore, in the Literature selection subgroup analysis of latitude, case size, control source, and histology, a significant association was found between Data from single reports were extracted (Figure 1). CAG repeats and TC in the PB, mid-latitude, high-latitude, We searched PubMed and Web of Science until July 2015 seminoma, and non-seminoma subgroups. to identify publications on the association between TC and Interestingly, we observed no heterogeneity after CAG and/or GGN trinucleotide repeat lengths in AR. We stratifying according to latitude and histology. We found focused on the studies performed in humans and on those www.impactjournals.com/oncotarget 13762 Oncotarget that utilized the following key words: testicular cancer or ACKNOWLEDGMENTS AND FUNDING TC, androgen receptor or AR, combined with CAG and/ or GGN. We appreciate the assistance from all the members The inclusion criteria were as follows: (1) studies of our laboratory. This study was supported by the Medical that evaluated the association between AR CAG or Health and Scientific Research Foundation of Nanjing GGC/GGN repeat polymorphisms and the risk of TC; Military Command (No. 2014MS098), the Key Foundation of (2) studies with a case-control design; (3) studies that Jiangsu Science and Technology Bureau (No. BM2013058), provided sufficient information on CAG or GGC/GGN and the Foundation of Nanjing General Hospital of Nanjing repeat distributions between patients and controls; Military Command, PLA. (No. 2014044). (4) studies for which the full text was available. Two reviewers assessed the full text of eligible studies from CONFLICTS OF INTEREST the above databases. Additional studies were identified by manually searching references of original and review The authors declare that no conflicts of interest are articles on this topic. associated with this publication. Data extraction and verification REFERENCES Information on the enrolled studies is shown in 1. Bosl GJ, Motzer RJ. Testicular germ-cell cancer. N Engl J Tables 1 and 2, including: (I) the first author’s name; Med. 1997; 337:242–254. (II) year of publication; (III) country or region of 2. Huyghe E, Matsuda T, Thonneau P. Increasing incidence of origin, ethnicity, and method; (IV) number of cases and testicular cancer worldwide: a review. J Urol. 2003; 170:5–11 controls. Two authors (WJ Jiang and SM Liu) of the study 3. Chia VM, Quraishi SM, Devesa SS, Purdue MP, Cook MB, extracted the information independently and screened the McGlynn KA. International trends in the incidence of citations that met the inclusion criteria. The discrepancies testicular cancer, 1973–2002. Cancer Epidemiol Biomarkers were adjudicated via discussions until a consensus was Prev. 2010; 19:1151–1159. reached. 4. Manecksha RP, Fitzpatrick JM. Epidemiology of testicular cancer. BJU Int. 2009; 104:1329–1333. Data and statistical analysis 5. Garner MJ, Turner MC, Ghadirian P, Krewski D. Epidemiology of Data were divided into three categories: CAG testicular cancer: an overview. Int J Cancer .2005; 116:331–339. repeat length of 21–25, which was used as the reference 6. Kristiansen W, Aschim EL, Andersen JM, Witczak O, group, CAG length < 21, and CAG length > 25. The GGN Fosså SD, Haugen TB. Variations in testosterone pathway repeat length was divided into 2 categories: GGN repeat genes and susceptibility to testicular cancer in Norwegian length ≤ 23, which was used as the reference group, and men. Int J Androl. 2012; 35:819–827. GGN length > 23. To obtain specific data from these 7. Gu M, Dong X, Zhang X, Niu W. The CAG repeat categories, we thoroughly analyzed and carefully obtained polymorphism of androgen receptor gene and prostate information from all the studies that met the inclusion cancer: a meta-analysis. Mol Biol Rep. 2012; 39:2615–2624. criteria. 8. Giwercman YL, Xu C, Arver S, Pousette A, Reneland R. We predicted the contribution of AR CAG and No association between the androgen receptor gene CAG GGN repeat polymorphisms to the risk of TC using the repeat and impaired sperm production in Swedish men. Clin Stata software version 11.0. The OR with 95% (CI) was Genet. 1998; 54:435–436. calculated to measure the strength of the associations 9. Västermark Å, Giwercman YL, Hagströmer O, Rajpert De- [31]. A test of heterogeneity of the pooled results was Meyts E, Eberhard J, Ståhl O, Cedermark GC, Rastkhani H, performed using Cochran’s Q test and Higgins I statistic Daugaard G, Arver S, Giwercman A. Polymorphic variation [32]. I > 50% is considered as a measure of significant in the androgen receptor gene: association with risk of heterogeneity. If the result of the Q test was P > 0.10, testicular germ cell cancer and metastatic disease. Eur J the OR was analyzed using the fixed-effects model Cancer. 2011; 47:413–419. (Mantel–Haenszel method). Otherwise, a random- 10. Grassetti D, Giannandrea F, Paoli D, Masciandaro P, effects model (DerSimonian–Laird method) was used in Figura V, Carlini T, Rizzo F, Lombardo F, Lenzi A, cases of significant heterogeneity. Sensitivity analysis Gandini L. Androgen receptor polymorphisms and testicular was performed to estimate the stability of the results by cancer risk. Andrology. 2015; 3:27–33. removing each study from the analysis, one at a time. Publication bias was evaluated using Begg’s funnel plot. 11. Nenonen HA, Giwercman A, Hallengren E, Giwercman YL. In addition to the visual inspection of the funnel plot, a Non-linear association between androgen receptor CAG value of P < 0.05 was considered to indicate the presence repeat length and risk of male subfertility - a meta-analysis. of significant publication bias [33–37]. Int J Androl. 2011; 34:327–332. www.impactjournals.com/oncotarget 13763 Oncotarget 12. Rajpert-De Meyts E, Leffers H, Daugaard G, Andersen CB, cryptorchidism risk among Hispanic white boys. J Clin Petersen PM, Hinrichsen J, Pedersen LG, Skakkebaek NE. Endocrinol Metab. 2012; 97:E393–E399. Analysis of the polymorphic CAG repeat length in the 25. Lazaros L, Xita N, Takenaka A, Sofikitis N, Makrydimas G, androgen receptor gene in patients with testicular germ cell Stefos T, Kosmas I, Zikopoulos K, Hatzi E, Georgiou I. Semen cancer. Int J Cancer. 2002; 102:201–204. quality is influenced by androgen receptor and aromatase gene 13. Giwercman A, Lundin KB, Eberhard J, Ståhl O, Cwikiel M, synergism. Hum Reprod. 2012; 27:3385–3392. Cavallin-Ståhl E, Giwercman YL. Linkage between 26. Giwercman A, Rylander L, Hagmar L, Giwercman YL. androgen receptor gene CAG trinucleotide repeat length Ethnic differences in occurrence of TDS-genetics and/or and testicular germ cell cancer histological type and clinical environment? Int J Androl. 2006; 29:291–297. stage. Eur J Cancer. 2004; 40:2152–2158. 27. Lundin KB, Giwercman YL, Rylander L, Hagmar L, 14. Eberhard J, Ståhl O, Giwercman Y, Cwikiel M, Cavallin- Giwercman A. Androgen receptor gene GGN repeat length Ståhl E, Lundin KB, Flodgren P, Giwercman A. Impact of and reproductive characteristics in young Swedish men. Eur therapy and androgen receptor polymorphism on sperm J Endocrinol. 2006; 155:347–354. concentration in men treated for testicular germ cell cancer: 28. Buchanan G, Yang M, Cheong A, Harris JM, Irvine a longitudinal study. Hum Reprod. 2004; 19:1418–1425. RA, Lambert PF, Moore NL, Raynor M, Neufing PJ, 15. Aschim EL, Oldenburg J, Kristiansen W, Giwercman A, Coetzee GA. Structural and functional consequences of Witczak O, Fosså SD, Haugen TB. Genetic variations glutamine tract variation in the androgen receptor. Hum Mol associated with the effect of testicular cancer treatment on Genet. 2004; 13:1677–1692. gonadal hormones. Hum Reprod. 2014; 29:2844–2851. 29. Davis-Dao CA, Siegmund KD, Vandenberg DJ, Skinner EC, 16. Davis-Dao CA, Tuazon ED, Sokol RZ, Cortessis VK. Coetzee GA, Thomas DC, Pike MC, Cortessis VK. Male infertility and variation in CAG repeat length in the Heterogenous effect of androgen receptor CAG tract androgen receptor gene: a meta-analysis. J Clin Endocrinol length on testicular germ cell tumor risk: shorter repeats Metab. 2007; 92:4319–4326. associated with seminoma but not other histologic types. 17. Garolla A, Ferlin A, Vinanzi C, Roverato A, Sotti G, Carcinogenesis. 2011; 32:1238–1243. Artibani W, Foresta C. Molecular analysis of the androgen 30. Zhang Z, Xu X, Ni H. Small studies may overestimate receptor gene in testicular cancer. Endocr Relat Cancer. the effect sizes in critical care meta-analyses: a meta- 2005; 12:645–655. epidemiological study. Crit Care. 2013; 17:R2. 18. Krausz C, Looijenga LH. Genetic aspects of testicular germ 31. Gu XB, Tian T, Tian XJ, Zhang XJ. Prognostic significance cell tumors. Cell Cycle. 2008; 7:3519–3524. of neutrophil-to-lymphocyte ratio in non-small cell lung 19. Giannandrea F, Gandini L, Paoli D, Turci R, Figà- cancer: a meta-analysis. Sci Rep. 2015; 5:12493. Talamanca I. Pesticide exposure and serum organochlorine 32. He J, Liao XY, Zhu JH, Xue WQ, Shen G-P, Huang SY, residuals among testicular cancer patients and healthy Chen W, Jia WH. Association of MTHFR C677T and A1298C controls. J Environ Sci Health B. 2011; 46:780–787. polymorphisms with non-Hodgkin lymphoma susceptibility: 20. Giannandrea F, Paoli D, Figà-Talamanca I, Lombardo F, evidence from a meta-analysis. Sci Rep. 2014; 4:6159 Lenzi A, Gandini L. Effect of endogenous and exogenous 33. Qin T, Du M, Du H, Shu Y, Wang M, Zhu L. Folic acid hormones on testicular cancer: the epidemiological supplements and colorectal cancer risk: meta-analysis of evidence. Int J Dev Biol. 2013; 57:255–263. randomized controlled trials. Sci Rep. 2015; 5:12044. 21. Irvine RA, Ma H, Yu MC, Ross RK, Stallcup MR, 34. Ntais C, Polycarpou A, Ioannidis JP. Association of Coetzee GA. Inhibition of p160-mediated coactivation with GSTM1, GSTT1, and GSTP1 gene polymorphisms with the increasing androgen receptor polyglutamine length. Hum risk of prostate cancer: a meta-analysis. Cancer Epidemiol Mol Genet. 2000; 9:267–274. Biomarkers Prev. 2005; 14:176–181. 22. Giwercman A, Richiardi L, Kaijser M, Ekbom A, Akre O. 35. Yang Y, Yin X, Sheng L, Xu S, Dong L, Liu L. Perioperative Reduced risk of prostate cancer in men who are childless as chemotherapy more of a benefit for overall survival than compared to those who have fathered a child: a population adjuvant chemotherapy for operable gastric cancer: an based case-control study. Int J Cancer. 2005; 115:994–997. updated Meta-analysis. Sci Rep. 2015; 5:12850. 23. Vottero A, Minari R, Viani I, Tassi F, Bonatti F, Neri TM, 36. Jiang W, Sun H, Zhang J, Zhou Q, Wu Q, Li T, Zhang C, Bertolini L, Bernasconi S, Ghizzoni L. Evidence for epigenetic Li W, Zhang M, Xia X. Polymorphisms in Protamine 1 abnormalities of the androgen receptor gene in foreskin from and Protamine 2 predict the risk of male infertility: a meta- children with hypospadias. J Clin Endocrinol Metab. 2011; analysis. Sci Rep. 2015; 5:15300. 96:E1953–E1962. 37. He J, Wang MY, Qiu LX, Zhu ML, Shi TY, Zhou XY, 24. Davis-Dao C, Koh CJ, Hardy BE, Chang A, Kim SS, De Sun MH, Yang YJ, Wang JC, Jin L, Yu HP, Wei QY. Genetic Filippo R, Hwang A, Pike MC, Carroll JD, Coetzee GA, variations of mTORC1 genes and risk of gastric cancer Vandenberg D, Siegmund K, Cortessis VK. Shorter in an Eastern Chinese population. Mol Carcinog. 2013; androgen receptor CAG repeat lengths associated with 52:E70–E79. www.impactjournals.com/oncotarget 13764 Oncotarget

Journal

OncotargetPubmed Central

Published: Feb 12, 2016

References