- Email: [email protected]
Association study between CYP24A1 gene polymorphisms and cancer risk
Pathology - Research and Practice xxx (xxxx) xxxx
Contents lists available at ScienceDirect
Pathology - Research and Practice journal homepage: www.elsevier.com/locate/prp
Association study between CYP24A1 gene polymorphisms and cancer risk Can Yia,1, Chao Huanga,1, Huan Wanga, Chen Wanga, Lijuan Donga, Xiuli Gub,c, Xianhong Fengd, Bifeng Chena,* a
Department of Biological Science and Technology, School of Chemistry, Chemical Engineering and Life Sciences, Wuhan University of Technology, Wuhan, China Center of Reproductive Medicine, Tongji Medical College, Huazhong University of Science and Technology, China Department of Reproductive Genetics, Wuhan Tongji Reproductive Medicine Hospital, Wuhan, China d Clinical Laboratory, Wuhan Xinzhou District People's Hospital, Wuhan, China b c
ARTICLE INFO
ABSTRACT
Keywords: CYP24A1 Single nucleotide polymorphism (SNP) Cancer risk Meta-analysis
CYP24A1, an essential gene in regulation of vitamin D, has been reported to play an important role in enhancing immune activity and inhibiting tumorigenesis. Previous studies proposed that rs2585428, rs4809960, rs6022999 and rs6068816 in CYP24A1 gene might be greatly associated with cancer risk. To validate the findings, we here investigated the associations of these four polymorphisms and colorectal cancer (CRC) risk in a central Chinese population (426 colon cancer patients, 361 rectal cancer patients and 800 healthy controls). The genotyping was conducted by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) and confirmed by sequencing. Our results revealed that the rs4809960 and rs6022999 were strongly associated with the CRC risk, especially with the colon cancer risk. Moreover, the analysis of haplotypes consisting of rs2585428(G > A), rs4809960(T > C), rs6022999(A > G) and rs6068816(C > T) indicated that haplotype ATGC significantly decreased the CRC risk, especially the colon cancer risk. Haplotype GCAT significantly increased the CRC risk, especially the rectal cancer risk. However, haplotype ACAC was only found to be associated with increased risk of CRC. To improve the statistical strength, an updated meta-analysis was further performed. The results showed that rs2585428 was associated with cancer risk in Caucasian population, rs4809960 was associated with breast cancer risk in Caucasian population, and rs6022999 was associated with cancer risk in Asian population. Collectively, the rs4809960 and rs6022999 may be the genetic biomarkers for prediction of colon cancer risk in Chinese population, the rs2585428 and rs6022999 may link to cancer susceptibility in Caucasian population and in Asian population respectly.
1. Introduction Vitamin D, an important fat-soluble vitamin and steroid prohormone, plays an essential role in human health for that it influences immune function, cell proliferation, differentiation and apoptosis. Vitamin D deficiency has been associated with numerous health outcomes, including bone disease, cancer, autoimmune disease and more [1,2]. In the process of vitamin D metabolism, the 25(OH)D (biomarker of vitamin D status) is first converted to 125(OH)2D3 by 1α-hydroxylase (encoded by CYP27B1 gene) in the kidney, then 125(OH)2D3 (active form of vitamin D) is released into the blood circulation. Lastly, the circulating 125(OH)2D3 will be degraded by 25-hydroxyvitamin D 24hydrolase (encoded by CYP24A1 gene), suggesting that CYP24A1 is the essential enzyme responsible for the degradation of vitamin D [3]. In
fact, some researchers have linked the abnormal expression of CYP24A1 gene to cancer risk, and elevated expression of CYP24A1 gene was observed in multiple types of cancer [4]. Collectively, CYP24A1 may be an oncogene and may contribute to tumor aggressiveness through degrading vitamin D. Recently, single nucleotide polymorphism (SNP), a well-defined molecular biomarker, has been extensively studied and applied to medical issues [5]. Interestingly, Four SNPs (rs2585428, rs4809960, rs6022999, and rs6068816) in CYP24A1 gene were found to be associated with cancer risks, including prostate cancer, breast cancer and pancreas cancer, indicating that the four CYP24A1 gene polymorphisms may serve as valuable biomarkers for cancer predisposition [6–14]. Colorectal carcinoma (CRC, colon cancer and rectal cancer) is an important contributor to cancer mortality and morbidity [15]. The
Corresponding author at: Department of Biological Science and Technology, School of Chemistry, Chemical Engineering and Life Sciences, Wuhan University of Technology, Wuhan, China. E-mail address: [email protected] (B. Chen). 1 Can Yi and Chao Huang contributed equally to this work. ⁎
https://doi.org/10.1016/j.prp.2019.152735 Received 17 August 2019; Received in revised form 23 October 2019; Accepted 10 November 2019 0344-0338/ © 2019 Elsevier GmbH. All rights reserved.
Please cite this article as: Can Yi, et al., Pathology - Research and Practice, https://doi.org/10.1016/j.prp.2019.152735
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 1 Characteristics of studied subjects. Group
Healthy controls (n = 800) Colon cancer patients (n = 426) Rectal cancer patients (n = 361) Colorectal cancer patients (n = 787)
Age, n (%)
Gender, n (%)
Smoking status, n (%)
Drinking status, n (%)
Never
Ever
591 298 261 559
237 135 122 257
≤60 years
> 60 years
Male
Female
Ever
434 245 210 455
366 181 151 332
558 309 246 555
242 117 115 232
209 128 100 228
(54.3) (57.5) (58.2) (57.8)
(45.7) (42.5) (41.8) (42.2)
(69.7) (72.5) (68.1) (70.5)
(30.3) (27.5) (31.9) (29.5)
(26.1) (30.0) (27.7) (29.0)
(73.9) (70.0) (72.3) (71.0)
P-value1
Never (29.6) (31.7) (33.8) (32.7)
563 291 239 530
(70.4) (68.3) (66.2) (67.3)
0.274 0.213 0.153
0.308 0.583 0.737
0.143 0.574 0.204
0.454 0.155 0.192
1
Two-sided χ2 test for the distributions of age (1st column), gender (2nd column), smoking status (3rd column) and drinking status (4th column) between colon/rectal/ colorectal cancer patients and healthy controls.
2.2. Genotyping assay
distinction between the colon cancer and the rectal cancer is largely anatomical. Both surgical and radiotherapeutic have been applied to CRC treatment, and achieved satisfactory results. However, due to the delay in diagnosis, most CRC patients get limited treatment and poor prognosis. Therefore, early and effective diagnostic methods will be particularly necessary [16]. Henrik et al. previously demonstrated that the CYP24A1 gene might be a potential biomarker for colorectal tumorigenesis [17], suggesting that there may be a potential association between CYP24A1 gene polymorphisms and CRC risk. To address this issue, we here examined the associations between the CYP24A1 gene polymorphisms (rs2585428, rs4809960, rs6022999, and rs6068816) and CRC risk in a central Chinese population. The meta-analysis is a statistical tool for combining the results from different studies on the same topic, thus increasing statistical strength and precision [18]. To resolve the discrepancies among the previous association studies, Zhu et al. have performed a meta-analysis to characterize the precise associations of rs2585428, rs4809960, rs6022999 and rs6068816 and cancer risk [19]. However, we found that the study of Oh et al. [11] was ignored by Zhu et al. Moreover, given the newly generated experiment data in present case–control study, we further performed a rigorous and updated meta-analysis (up to April of 2019) to determine the association of CYP24A1 gene polymorphism and cancer risk.
The peripheral blood samples (5 ml per participant) were collected into blood vacuum tubes containing EDTA and stored at 4 °C. Then, genomic DNA was extracted from blood samples using the TIANamp Blood DNA Kit (DP348; TianGen Biotech, Beijing, China), and stored at −20 °C before use. Next, the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique was conducted to genotype the four SNPs. The information of PCR-RFLP assay was described in supplementary Table S1. To ensure the reliability of genotyping results, the PCR-RFLP assay was repeated twice for all subjects, and 10 % randomly selected DNA samples were examined by DNA sequencing, all the results were 1.0 % concordant. The sequencing diagrams of the four CYP24A1 gene polymorphisms were shown in supplementary Figure S1. 2.3. Statistical analysis The Statistical Program for Social Sciences (SPSS, version 15.0, Chicago, IL, USA) was used to perform all the analyses. χ2 test was used to determine the statistical differences in age, gender, smoking status and drinking status. The numerical data of average ages for cases and controls were presented as means ± SD (standard deviation of the mean), and their differences were evaluated using the Student’s t-test. The genotypic frequencies of rs2585428, rs4809960, rs6022999 and rs6068816 in normal controls were tested for departure from HWE (Hardy-Weinberg equilibrium). Logistic regression analysis was used to estimate the associations between the four SNPs and cancer risk. Squared correlation coefficient r2 for each pair of polymorphic sites and haplotype analysis was calculated by using Haploview software available at http://www.broad.mit.edu/mpg/haploview. P values less than 0.05 were considered significant for all statistical analyses, and the Bonferroni correction for multiple testing was applied [20].
2. Material and methods 2.1. Studied subjects In this study, we recruited a total of 787 CRC patients (426 colon cancer patients and 361 rectal cancer) and 800 healthy controls. All the CRC patients were confirmed histopathologically and recruited from Hubei Cancer Hospital and Wuhan Xinzhou District People's Hospital between January 2015 and December 2016. Nowadays, more and more Chinese are inclined to have a physical examination every year. The healthy controls were selected from cancer-free individuals who visited Wuhan Xinzhou District People's Hospital for regular physical examinations between September 2014 and December 2016 or who volunteered to participate in the epidemiology survey during the same period. It was required that the normal controls passed all annual physical examinations in the latest three years. All participants were biologically unrelated Han Chinese living in central China (Hubei province). This study was approved by the Ethical Committees of Wuhan University of Technology (Approval No: WUT02720180705), and written informed consent for the genetics analysis was obtained from all participants or their guardians.
2.4. Meta-analysis We comprehensively searched the related literatures updated to April of 2019 on PubMed, EMBASE, ISI Web of Science, and CNKI and Wanfang databases without language restriction. The search terms used were as follows: "CYP24A1, rs2585428, and cancer/tumor/carcinoma", "CYP24A1, rs4809960, and cancer/tumor/carcinoma", "CYP24A1, rs6022999, and cancer/tumor/carcinoma", and "CYP24A1, rs6068816, and cancer/tumor/carcinoma". The included studies should meet the following criteria: (1) studies on humans, (2) investigation of the CYP24A1 gene polymorphisms (at least one of the four polymorphisms)
2
3
2
1
687 (80.6%) 165 (19.4%) 279 (65.5%) 129 (30.3%) 18 (4.2%)
553 (64.9%) 299 (35.1%) 181 (42.5%) 191 (44.8%) 54 (12.7%)
rs6022999 polymorphism A 1253 (79.6%) G 321 (20.4%) AA 504 (64.1%) AG 245 (31.1%) GG 38 (4.8%)
rs6068816 polymorphism C 1038 (65.9%) T 536 (34.1%) CC 342 (43.5%) CT 354 (45.0%) TT 91 (11.5%) 485 (67.2%) 237 (32.8%) 161 (44.6%) 163 (45.2%) 37 (10.2%)
566 (78.4%) 156 (21.6%) 225 (62.4%) 116 (32.1%) 20 (5.5%)
534 (74.0%) 188 (26.0%) 202 (56.0%) 130 (36.0%) 29 (8.0%)
352 (48.8%) 370 (51.2%) 85 (23.5%) 182 (50.4%) 94 (26.1%)
3. Rectal cancer patient
(50.9%) (49.1%) (27.3%) (47.4%) (25.3%)
1072 (67.0%) 528 (33.0%) 362 (45.3%) 348 (43.5%) 90 (11.2%)
1192 (74.5%) 408 (25.5%) 443 (55.4%) 306 (38.2%) 51 (6.4%)
1226 (76.6%) 374 (23.4%) 468 (58.5%) 290 (36.2%) 42 (5.3%)
815 785 218 379 203
4. Healthy controls
0.645
0.849
0.736
0.140
HWE1
C vs. T CC vs. TT CC vs. CT CT vs. TT CC vs. CT + TT CC + CT vs. TT
A vs. G AA vs. GG AA vs.AG AG vs. GG AA vs. AG + GG AA + AG vs. GG
T vs. C TT vs. CC CT vs. CC TT vs. CT TT vs. CT + CC TT + CT vs. CC
G vs. A GG vs. AA GG vs. AG AG vs. AA GG vs. AG + AA GG + AG vs. AA
Genetic Model
0.80 0.61 0.83 0.74 0.79 0.66
0.92 0.84 1.02 0.83 0.83 0.95 (0.69-0.94) (0.40-0.92) (0.67-1.02) (0.48-1.13) (0.65-0.97) (0.44-0.99)
(0.81-1.03) (0.66-1.07) (0.83-1.25) (0.67-1.02) (0.68-1.01) (0.78-1.16)
0.530, 0.684, 0.488, 0.971, 0.472, 0.845,
0.95 0.93 0.93 1.01 0.93 0.97
(0.82-1.11) (0.67-1.30) (0.75-1.15) (0.73-1.40) (0.76-1.13) (0.71-1.32)
0.001, 1.34 (1.13-1.58) 0.059, 1.53 (0.98-2.37) 0.001, 1.42 (1.15-1.76) 0.755, 1.08 (0.68-1.69) < 0.001, 1.44 (1.17-1.76) 0.182, 1.34 (0.87-2.07)
0.008, 0.020, 0.073, 0.161, 0.021, 0.044,
0.151, 0.157, 0.874, 0.077, 0.068, 0.626,
1 vs. 4
P value, OR(95% CI)2
Genotypic frequency of SNPs in normal controls were tested for departure from Hardy-Weinberg equilibrium (HWE) using the χ2 test. The OR (95% CI) and the corresponding P value were calculated by logistic regression analysis.
607 (71.2%) 245 (28.8%) 213 (50.0%) 181 (42.5%) 32 (7.5%)
rs4809960 polymorphism T 1141 (72.5%) C 433 (27.5%) TT 415 (52.7%) CT 311 (39.5%) CC 61 (7.8%)
(48.7%) (51.3%) (23.9%) (49.5%) (26.6%)
2. Colon cancer patients
415 437 102 211 113
1. Colorectal cancer
rs2585428 polymorphism G 767 (48.7%) A 807 (51.3%) GG 187 (23.8%) AG 393 (49.9%) AA 207 (26.3%)
Genotype
Table 2 Genotype and allele distributions of CYP24A1 polymorphisms and their associations with colorectal cancer risk.
0.296, 0.349, 0.467, 0.646, 0.354, 0.460,
0.001, 0.042, 0.002, 0.545, 0.001, 0.122,
0.004, 0.038, 0.012, 0.431, 0.004, 0.115,
0.293, 0.301, 0.999, 0.239, 0.210, 0.661,
2 vs. 4
0.91 0.83 0.91 0.92 0.89 0.87
1.43 1.79 1.49 1.19 1.53 1.54
0.76 0.60 0.73 0.82 0.71 0.68
0.92 0.84 1.00 0.84 0.84 0.94
(0.77-1.09) (0.57-1.22) (0.71-1.17) (0.63-1.34) (0.71-1.13) (0.61-1.25)
(1.16-1.75) (1.02-3.12) (1.16-1.93) (0.67-2.12) (1.20-1.95) (0.89-2.68)
(0.63-0.91) (0.37-0.97) (0.57-0.93) (0.50-1.35) (0.56-0.90) (0.42-1.10)
(0.78-1.08) (0.61-1.17) (0.75-1.33) (0.63-1.12) (0.64-1.10) (0.72-1.23)
0.934, 0.717, 0.699, 0.548, 0.836, 0.613,
0.043, 0.349, 0.032, 0.905, 0.027, 0.583,
0.166, 0.066, 0.779, 0.101, 0.417, 0.069,
0.330, 0.336, 0.813, 0.183, 0.184, 0.810,
3 vs. 4
1.01 1.08 0.95 1.14 0.97 1.11
1.24 1.30 1.34 0.97 1.33 1.16
0.87 0.63 0.96 0.65 0.90 0.63
0.92 0.84 1.04 0.81 0.82 0.97
(0.84-1.22) (0.71-1.66) (0.73-1.24) (0.74-1.74) (0.76-1.25) (0.74-1.66)
(1.01-1.53) (0.75-2.23) (1.03-1.75) (0.55-1.69) (1.03-1.72) (0.68-1.98)
(0.71-1.06) (0.38-1.03) (0.74-1.26) (0.39-1.09) (0.70-1.16) (0.39-1.04)
(0.77-1.09) (0.59-1.20) (0.77-1.40) (0.60-1.10) (0.62-1.10) (0.73-1.28)
C. Yi, et al.
Pathology - Research and Practice xxx (xxxx) xxxx
Pathology - Research and Practice xxx (xxxx) xxxx
The haplotype structure was rs2585428(G > A)/rs4809960(T > C)/rs6022999(A > G)/rs6068816(C > T). All haplotypes with frequency < 0.01 in both case and control groups would be ignored in analysis. The OR (95% CI) and the corresponding P value were calculated by logistic regression analysis.
0.99(0.70-1.41) 2.33(1.38-3.91) 0.97(0.81-1.15) 0.84(0.70-1.01) 0.956 0.001 0.720 0.070 < 0.0001 0.003 10.67 0.13 3.29 34.82
0.91(0.74-1.12) 0.88(0.71-1.09) 0.374 0.248 0.0002 0.79 1.34 30.1
0.95(0.75-1.20) 1.45(1.07-1.96) 0.72(0.59-0.88) 0.652 0.017 0.002 0.20 5.67 9.99
and cancer risk, (3) case-control study design, (4) sufficient data was accessible to estimate the OR and 95%CI, and (5) HWE equilibrium should be established in controls. The flowchart of the search strategy of present meta-analysis was shown in supplementary Figure S2. All analyses were performed using STATA 14.0 (Stata Corp, College Station, TX). The Cochran's Q-test and I2 were used to assess the heterogeneity of included studies. If Pheterogeneity ≥ 0.1, the fixed-effect model was applied to calculate the combined ORs [21], otherwise, random-effects model was used [22]. The significance of combined ORs was determined by the Z test. A P value < 0.05 was considered significantly, and the Bonferroni correction for multiple testing was also applied. 3. Results 3.1. Characteristics of studied subjects The distribution of participants’ variables was shown in Table 1, there were no significant differences in distribution of age, gender, smoking status, drinking status between cancer patients and healthy controls, suggesting that our case-control study was well matched. Moreover, the average ages of healthy controls, colon cancer patients, rectal cancer patients and colorectal cancer patients were 56.16 ± 10.85, 57.36 ± 10.47, 58.06 ± 10.65 and 57.87 ± 10.76 years old, respectly. No statistical differences were observed in average ages between patients (colon, rectal or colorectal cancer) and healthy controls. 3.2. The associations between CYP24A1 gene polymorphisms and CRC risk In this study, we performed comparisons for allele and genotype frequencies of the four CYP24A1 gene polymorphisms between cancer patients (colon cancer, rectal cancer and CRC) and healthy controls under six genetic models (allele model, carrier model, homozygote model, heterozygote model, recessive model, dominant model) (Table 2). The healthy controls in our study were in accordance with HWE for rs2585428, rs4809960, rs6022999 and rs6068816 polymorphisms and had a good representative. After Bonferroni correction (P < 0.0084, 0.05/6), we observed that the rs4809960 and rs6022999 polymorphisms were significantly associated with risk of CRC (especially colon cancer) but not rectal cancer, while rs2585428 and rs6068816 polymorphisms were not associated with risk of CRC, colon cancer or rectal cancer. For rs4809960 polymorphism, allele T was a significant protective allele of CRC (T vs. C, P = 0.008, OR = 0.80, 95%CI = 0.69-0.94) and colon cancer (T vs. C, P = 0.004, OR = 0.76, 95%CI = 0.63-0.91), and individuals with TT genotype had a lower risk for colon cancer compared with CT + CC (TT vs. CT + CC, P = 0.004, OR = 0.71, 95%CI = 0.56-0.90). For rs6022999 polymorphism, A allele had a higher risk for CRC (A vs. G, P = 0.001, OR = 1.34, 95%CI = 1.13–1.58) and colon cancer (A vs. G, P = 0.001, OR = 1.43, 95%CI = 1.16–1.75) than those carrying the G allele, and AA genotype conferred higher risk for CRC and colon cancer relative to AG and AG + GG genotypes (CRC: AA vs. AG, P = 0.001, OR = 1.42, 95%CI = 1.15–1.76; AA vs. AG + GG, P < 0.001, OR = 1.44, 95%CI = 1.17-1.76. Colon cancer: AA vs. AG, P = 0.002, OR = 1.49, 95%CI = 1.16–1.93; AA vs. AG + GG, P = 0.001, OR = 1.53, 95%CI = 1.20–1.95). 3.3. The associations between CYP24A1 haplotypes and CRC risk The patterns of linkage disequilibrium (LD) for the four CYP24A1 gene polymorphism were shown in supplementary Figure S3, which indicated low LD with each other. Since the haplotype analysis could
2
1
0.68(0.42-1.13) 2.67(1.49-4.80) 1.03(0.83-1.28) 0.78(0.62-0.99) 0.134 0.001 0.815 0.041 < 0.0001 2.25 11.7 0.06 4.19 32.3
0.77(0.56-1.06) 1.62(1.13-2.32) 0.85(0.66-1.09) 0.105 0.008 0.190 2.62 7.00 1.72
1.37(1.06-1.76) 0.014 6.01
1.36(1.07-1.72) 1.28(0.81-2.01) 1.06(0.81-1.40) 1.32(0.92-1.89) 0.61(0.48-0.79) 1.97(1.17-3.31) 1.24(0.84-1.85) 0.013 0.288 0.663 0.138 0.001 0.009 0.280 6.24 1.13 0.19 2.20 14.3 6.78 1.17 1.37(1.12-1.69) 0.002 9.31
0.120 0.030 0.098 0.047 0.162 0.018 0.041 0.013 0.204 0.192 0.925 0.163 0.015 0.081 0.077 0.147 0.013 0.029 0.035 0.218 0.165 0.943 0.160 0.026 0.095 0.068 0.126 0.026 0.041 0.030 0.204 0.172 0.948 ACAC ACAT ATAC ATAT ATGC ATGT GCAC GCAT GTAC GTAT Global result
0.163 0.015 0.081 0.077 0.147 0.013 0.029 0.035 0.218 0.165 0.943
χ2
3. Rectal cancer 2. Rectal cancer 1. Colorectal cancer Haplotype1
Table 3 Association between CYP24A1 haplotypes with colorectal cancer risk.
4. Healthy controls
1 vs. 4
2 vs. 4
P χ2 OR(95% CI)2 P2
3 vs. 4
P χ2 OR(95% CI)
OR(95% CI)
C. Yi, et al.
4
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 4 Meta-analysis of association between rs2585428 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs2585428 and cancer risk G vs. A 23.68 0.003 GG vs. AA 17.26 0.027 AG vs. AA 5.14 0.742 GG vs. AG 15.82 0.045 GG vs. AG + AA 20.10 0.010 GG + AG vs. AA 9.26 0.321 rs2585428 and cancer risk in Asian population G vs. A 8.90 0.012 GG vs. AA 8.12 0.017 AG vs. AA 1.41 0.495 GG vs. AG 5.31 0.070 GG vs. AG + AA 7.94 0.019 GG + AG vs. AA 4.17 0.124 rs2585428 and cancer risk based on Illumina method G vs. A 1.58 0.208 GG vs. AA 1.63 0.201 AG vs. AA 0.03 0.854 GG vs. AG 11.88 0.001 GG vs. AG + AA 1.56 0.211 GG + AG vs. AA 0.30 0.584 rs2585428 and cancer risk based on PCR-RFLP method G vs. A 2.79 0.426 GG vs. AA 2.63 0.452 AG vs. AA 0.69 0.877 GG vs. AG 2.96 0.397 GG vs. AG + AA 3.40 0.333 GG + AG vs. AA 1.14 0.767 rs2585428 and prostate cancer risk G vs. A 4.84 0.184 GG vs. AA 4.52 0.211 AG vs. AA 2.27 0.518 GG vs. AG 1.63 0.653 GG vs. AG + AA 2.91 0.406 GG + AG vs. AA 3.76 0.289
I
2
Hypothesis test
Studies (n)
Z
P
66.2% 53.7% 0.0% 49.4% 60.2% 13.6%
1.11(1.10-1.24) 1.13(0.94-1.36) 1.01(0.91-1.12) 1.09(0.99-1.20) 1.12(0.96-1.31) 1.04(0.94-1.15)
1.92 1.31 0.18 1.78 1.43 0.80
0.055 0.191 0.858 0.076 0.153 0.426
9 9 9 9 9 9
77.5% 75.4% 0.0% 62.3% 74.8% 75.2%
1.04(0.82-1.33) 1.07(0.67-1.72) 1.06(0.88-1.29) 0.96(0.70-1.32) 1.01(0.70-1.44) 1.07(0.81-1.42)
0.32 0.30 0.61 0.24 0.01 0.48
0.746 0.766 0.543 0.807 0.992 0.635
3 3 3 3 3 3
36.8% 38.7% 0.0% 91.6% 36.0% 0.0%
1.20(1.10-1.31) 1.32(1.14-1.53) 1.09(0.96-1.23) 1.13(0.29-4.45) 1.88(1.23-2.86) 1.17(1.12-1.23)
3.93 3.69 1.34 0.17 2.92 2.60
< 0.001 < 0.001 0.182 0.866 0.004 < 0.001
2 2 2 2 2 2
0.0% 0.0% 0.0% 0.0% 11.9% 0.0%
0.97(0.89-1.06) 0.94(0.78-1.13) 1.02(0.86-1.19) 0.94(0.80-1.09) 0.94(0.81-1.09) 0.99(0.85-1.15)
0.65 0.67 0.18 0.84 0.85 0.18
0.516 0.500 0.856 0.398 0.393 0.855
4 4 4 4 4 4
38.0% 33.6% 0.0% 0.0% 0.0% 20.1%
1.12(1.02-1.24) 1.25(1.02-1.53) 1.13(0.93-1.36) 1.20(0.95-1.32) 1.16(0.10-1.35) 1.17(0.98-1.40)
2.25 2.18 1.23 1.37 1.90 1.75
0.024 0.029 0.218 0.172 0.057 0.080
4 4 4 4 4 4
enhance the statistical power in the mapping of human complex trait loci, the analysis of haplotypes consisting of rs2585428(G > A), rs4809960(T > C), rs6022999(A > G) and rs6068816(C > T) was performed to estimate the association between CYP24A1 gene and cancer risk in this study. In Table 3, haplotype ATGC was significant associated with a decreased risk for CRC (P = 0.002, OR = 0.72, 95%CI = 0.59-0.88) and colon cancer (P = 0.001, OR = 0.61, 95%CI = 0.48-0.79), haplotype GCAT was significant associated with an increased risk for CRC (P = 0.001, OR = 2.33, 95%CI = 1.38–3.91) and rectal cancer (P = 0.001, OR = 2.67, 95%CI = 1.49–4.80), and haplotype ACAC was significant associated with an increased risk for CRC (P = 0.002, OR = 1.37, 95%CI = 1.12–1.69).
for rs2585428, rs4809960, rs6022999 and rs6068816 in each control group. For the four CYP24A1 gene polymorphisms, their associations with cancer risk were evaluated in total population, as well as in stratified analysis according to ethnicity (Asian and Caucasian), genotyping method (PCR-RFLP and Illumina) and cancer type (prostate, breast and colon cancer). Compared with Zhu et al’s meta-analysis, another two association studies in Asian population [11,14] and present case-control study in Chinese (Asian) population have been added in ours. Therefore, the analysis results of CYP24A1 gene polymorphisms and overall cancer risk in Caucasian population, as well as CYP24A1 gene polymorphisms and breast cancer risk were not updated and not shown in present study. Due to the insufficient numbers of included studies (n < 10), the assessment of publication bias was not performed through Begg’s funnel plot and Egger’s linear regression method in this metaanalysis [23]. A sum of 9 publications for rs2585428 that met the inclusion criteria were finally included. As shown in Table 4, we found that rs2585428 was not associated with cancer risk in total population or Asian population. In the cancer type-stratified analysis, it was observed
3.4. Meta-analysis of CYP24A1 gene polymorphisms and cancer risk As shown in supplementary Table S2, the NOS score of all articles are not < 6, indicating that each included literature was a high-quality study. The characteristics of included studies for this meta-analysis were demonstrated in Supplementary Table S3. HWE was established
5
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 5 Meta-analysis of association between rs4809960 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs4809960 and cancer risk in total population T vs. C 21.16 0.004 TT vs. CC 10.24 0.193 TT vs. CT 24.01 0.001 CT vs. CC 7.95 0.329 TT vs. CT + CC 24.17 0.001 TT + CT vs. CC 8.45 0.312 rs4809960 and cancer risk in Asian population T vs. C 3.40 0.182 TT vs. CC 1.65 0.438 TT vs. CT 3.47 0.176 CT vs. CC 1.01 0.604 TT vs. CT + CC 3.80 0.150 TT + CT vs. CC 1.26 0.533 rs4809960 and cancer risk based on Illumina method T vs. C 0.28 0.597 TT vs. CC 0.00 0.998 TT vs. CT 0.55 0.457 CT vs. CC 0.10 0.748 TT vs. CT + CC 0.46 0.498 TT + CT vs. CC 0.01 0.909 rs4809960 and cancer risk based on PCR-RFLP method T vs. C 15.02 0.002 TT vs. CC 4.87 0.181 TT vs. CT 14.96 0.002 CT vs. CC 0.95 0.813 TT vs. CT + CC 16.51 0.001 TT + CT vs. CC 2.66 0.447 rs4809960 and prostate cancer risk T vs. C 3.49 0.323 TT vs. CC 0.49 0.922 TT vs. CT 10.21 0.017 CT vs. CC 4.20 0.241 TT vs. CT + CC 7.40 0.060 TT + CT vs. CC 1.48 0.688
2
I
Hypothesis test
Studies (n)
Z
P
66.9% 31.7% 70.8% 11.9% 71.0% 17.2%
1.02 (0.90-1.16) 0.98(0.82-1.16) 1.06(0.89-1.26) 0.91(0.76-1.08) 1.04(0.88-1.23) 0.95(0.80-1.13)
0.31 0.29 0.66 1.06 0.51 0.59
0.754 0.775 0.512 0.287 0.608 0.554
8 8 8 8 8 8
41.2% 0.0% 42.4% 0.0% 47.3% 0.0%
0.84(0.74-0.96) 0.66(0.48-0.92) 0.86(0.73-1.02) 0.77(0.55-1.08) 0.83(0.71-0.98) 0.71(0.51-0.97)
2.67 2.44 1.73 1.51 2.26 2.12
0.009 0.015 0.083 0.131 0.024 0.034
3 3 3 3 3 3
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
1.16(1.01-1.34) 1.18(0.81-1.71) 1.23(1.03-1.46) 0.96(0.65-1.41) 1.22(1.03-1.44) 1.09(0.75-1.28)
2.12 0.84 2.28 0.23 2.32 0.47
0.034 0.403 0.022 0.816 0.020 0.636
2 2 2 2 2 2
80.0% 38.4% 79.9% 0.0% 81.8% 0.0%
0.98(0.75-1.28) 0.79(0.60-1.03) 1.07(0.76-1.50) 0.78(0.59-1.03) 1.02(0.73-1.44) 0.79(0.60-1.03)
0.17 1.73 0.36 1.77 0.13 1.76
0.864 0.084 0.717 0.077 0.895 0.078
4 4 4 4 4 4
13.9% 0.0% 70.6% 28.6% 59.5% 0.0%
1.10(0.97-1.24) 1.23(0.90-1.67) 1.13(0.82-1.55) 1.13(0.82-1.56) 1.13(0.87-1.46) 1.19(0.88-1.62)
1.53 1.28 0.74 0.75 0.90 1.12
0.127 0.201 0.459 0.454 0.370 0.261
4 4 4 4 4 4
that rs2585428 was not associated with the risk of prostate cancer or breast cancer. However, the stratified analysis by genotyping method suggested that rs2585428 significantly associated with cancer risk based on Illumina method but not PCR-RFLP method, and rs2585428 was inclined to development cancer under three models (G vs. A, GG vs. AA, GG vs. AG + AA and GG + AG vs. AA). A sum of 8 publications for rs4809960 that met the inclusion criteria were finally retrieved. In Table 5, no significant association was found for rs4809960 and cancer risk in total population or Asian population. When performing stratified analysis by genotyping method and cancer type, it was also observed that rs4809960 was not associated with cancer risk or prostate cancer risk. We finally included 9 related publications for rs6080550 in this meta-analysis. As shown in Table 6, rs6022999 was suggested to significantly associate with cancer risk in Asian population, but not in total population. For Asians, A allele had a higher cancer risk than G allele (A vs. G), and AA genotype conferred higher cancer risk relative to AG and/or GG genotypes (AA vs. AG, AA vs. GG, and AA vs. AG + GG). Moreover, no association was found for rs6080550 with cancer in the stratified analysis according to genotyping method or cancer type. A sum of 8 publications for rs6068816 that met the inclusion criteria were
finally retrieved. In Table 7, rs6068816 was not associated with cancer risk in total population or Asian population, and was also not associated with cancer risk in the stratified analysis. Moreover, the sensitivity analysis identified that the removal of any single case-control study did not influence the stability of the results (Fig. 1). 4. Discussion CYP24A1 gene has been proposed as a potential diagnostic and prognostic indicator in cancer for its critical role in the metabolism process of Vitamin D. To be an important molecular biomarker, SNP has been extensively studied and applied to medical issues. Currently, the relations between CYP24A1 gene polymorphisms and cancer risk have attracted considerable attention, and revealed several possible susceptibility loci (e.g. rs2585428, rs4809960, rs6022999, and rs6068816). To support the findings, we here explored the associations between these four CYP24A1 polymorphisms and CRC risk in a central Chinese population. Interestingly, the association of rs6022999 and colon cancer risk has been examined by Dong et al., and produced a negative result [13]. In contrast, our results demonstrated that
6
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 6 Meta-analysis of association between rs6022999 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs6022999 and cancer risk A vs. G 55.66 < 0.001 AA vs. GG 17.38 0.026 AG vs. GG 5.97 0.651 AA vs. AG 18.34 0.019 AA + AG vs.GG 11.63 0.169 AA vs. AG + GG 24.14 0.002 rs6022999 and cancer risk in Asian population A vs. G 1.32 0.517 AA vs. GG 1.35 0.508 AG vs. GG 0.74 0.692 AA vs. AG 0.4 0.818 AA + AG vs.GG 1.16 0.559 AA vs. AG + GG 0.81 0.667 rs6022999 and cancer risk based on Illumina method A vs. G 2.27 0.132 AA vs. GG 1.02 0.313 AG vs. GG 0.18 0.673 AA vs. AG 1.53 0.217 AA + AG vs.GG 0.63 0.427 AA vs. AG + GG 2.08 0.149 rs6022999 and cancer risk based on PCR-RFLP method A vs. G 42.51 < 0.001 AA vs. GG 6.64 0.084 AG vs. GG 1.67 0.644 AA vs. AG 6.35 0.096 AA + AG vs.GG 4.01 0.260 AA vs. AG + GG 8.66 0.034 rs6022999 and prostate cancer risk A vs. G 30.83 < 0.001 AA vs. GG 6.54 0.088 AG vs. GG 3.01 0.390 AA vs. AG 5.19 0.158 AA + AG vs.GG 4.34 0.227 AA vs. AG + GG 6.56 0.087 rs6022999 and colon cancer risk A vs. G 8.80 0.003 AA vs. GG 4.81 0.028 AG vs. GG 1.26 0.262 AA vs. AG 4.78 0.029 AA + AG vs.GG 3.42 0.064 AA vs. AG + GG 7.18 0.007
I
2
Hypothesis test
Studies (n)
Z
P
85.6% 54.0% 0.0% 56.4% 31.2% 66.9%
1.08(0.91-1.27) 1.15(0.90-1.45) 0.95(0.82-1.10) 1.14 (1.01-1.29) 1.01(0.87-1.16) 1.15(1.00-1.32)
0.86 1.12 0.73 2.12 0.06 2.03
0.388 0.262 0.466 0.034 0.953 0.042
9 9 9 9 9 9
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
1.36 (1.19-1.56) 1.62(1.14-2.31) 1.14(0.79-1.64) 1.43 (1.21-1.69) 1.42 (1.00-2.01) 1.46(1.24-1.71)
4.55 2.66 0.69 4.20 1.97 4.61
< 0.001 0.008 0.488 < 0.001 0.048 < 0.001
3 3 3 3 3 3
55.9% 2.0% 0.0% 34.5% 0.0% 52.0%
1.27(0.98-1.64) 1.53(1.01-2.31) 1.27(0.83-1.93) 1.27(0.83-1.93) 1.19(1.00-1.42) 1.29(0.96-1.74)
1.81 2.02 1.11 1.98 1.73 1.70
0.070 0.043 0.267 0.047 0.083 0.089
2 2 2 2 2 2
92.9% 54.8% 0.0% 52.8% 25.3% 65.4%
1.09(0.70-1.68) 1.30(0.75-2.23) 0.93(0.70-1.23) 1.30(1.04-1.63) 1.09(0.84-1.42) 1.31(1.01-1.68)
0.37 1.14 0.51 2.30 0.63 2.06
0.708 0.255 0.608 0.021 0.526 0.039
4 4 4 4 4 4
90.3% 54.2% 0.3% 42.2% 30.9% 54.3%
1.05(0.70-1.57) 1.29(0.80-2.07) 0.10 (0.75-1.33) 1.10 (0.95-1.28) 1.08(0.82-1.42) 1.18(0.92-1.50)
0.22 1.05 0.01 1.25 0.54 1.31
0.822 0.294 0.993 0.212 0.591 0.189
4 4 4 4 4 4
88.6% 79.2% 20.7% 79.1% 70.7% 86.1%
1.18 (0.84-1.67) 1.21(0.62-2.36) 0.89(0.70-1.14) 1.25(0.91-1.71) 1.10(0.63-1.91) 1.25(0.86-1.81)
0.96 0.55 0.94 1.37 0.33 1.16
0.335 0.582 0.347 0.172 0.740 0.245
2 2 2 2 2 2
rs6022999 was significantly associated with colon cancer risk, as well as CRC risk. Similarly, a significant association between rs4809960 and risk of colon and CRC was also observed. However, the situation was different for rs2585428 and rs6068816. Consisted with previous studies, the four CYP24A1 gene polymorphisms were not simultaneously associated with risk of the same cancer in each study, and each CYP24A1 gene polymorphism showed inconsistent results across different studies. The possible reasons for these discrepancies might be as follows. First, the limited sample size may lead to inadequate statistical strength, which should be solved in well-designed study with large sample size. Second, the action of CYP24A1 may be varied in different cancer types, which contributes to the discrepancies of CYP24A1 gene polymorphisms to affect individuals' cancer susceptibility. Third, different environments, lifestyles, and genetic backgrounds among different ethnic populations
may also contribute to the differences of CYP24A1 gene polymorphisms on cancer risk. Nowadays, the successful completion of the HapMap project suggested that haplotype analysis would enhance the statistical power in the mapping of human complex trait loci, with the potential of reducing the sample size of association studies [24,25]. Indeed, none of the four CYP24A1 gene polymorphisms was identified to be associate with rectal cancer in the single-locus analysis, but haplotype GCAT (rs2585428rs4809960-rs6022999-rs6068816) was shown to increase the rectal cancer risk. Moreover, haplotype ATGC was associated with a decreased risk of CRC and colon cancer, and haplotypes (ACAC and GCAT) were significant associated an increased CRC risk. Of note, rs4809960 T/C allele was the protective/risk allele for colon and CRC, while rs6022999 G/A allele was the protective/risk allele of colon and CRC. These findings suggested that rs4809960(T > C)-rs6022999(A
7
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 7 Meta-analysis of associations between rs6068816 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs6068816 and cancer risk C vs. T 17.25 0.016 CC vs. TT 13.62 0.058 CC vs. CT 10.07 0.185 CT vs. TT 5.76 0.568 CC vs. CT + TT 14.48 0.043 CC + CT vs. TT 9.97 0.190 rs6068816 and cancer risk in Asian population C vs. T 10.17 0.017 CC vs. TT 7.94 0.047 CC vs. CT 6.40 0.094 CT vs. TT 2.00 0.572 CC vs. CT + TT 9.11 0.028 CC + CT vs. TT 4.65 0.199 rs6068816 and cancer risk by Illumina method C vs. T 2.67 0.102 CC vs. TT 2.49 0.114 CC vs. CT 0.98 0.322 CT vs. TT 1.27 0.260 CC vs. CT + TT 1.98 0.160 CC + CT vs. TT 1.87 0.172 rs6068816 and cancer risk by PCR-RFLP method C vs. T 0.74 0.691 CC vs. TT 0.86 0.650 CC vs. CT 0.08 0.963 CT vs. TT 0.61 0.737 CC vs. CT + TT 0.28 0.869 CC + CT vs. TT 0.83 0.660 rs6068816 and prostate cancer risk C vs. T 0.73 0.693 CC vs. TT 0.29 0.866 CC vs. CT 0.46 0.795 CT vs. TT 0.14 0.930 CC vs. CT + TT 0.79 0.675 CC + CT vs. TT 0.11 0.946
2
I
Hypothesis test
Studies`(n)
Z
P
59.4% 48.6% 30.5% 0.0% 51.7% 29.8%
1.00(0.88-1.13) 1.05(0.85-1.28) 0.99(0.90-1.09) 1.06(0.87-1.31) 0.99(0.86-1.13) 1.05(0.86-1.27)
0.07 0.42 0.29 0.59 0.18 0.45
0.944 0.674 0.774 0.557 0.861 0.655
8 8 8 8 8 8
70.5% 62.2% 53.2% 0.0% 67.1% 35.5%
0.99(0.82-1.20) 0.99(0.67-1.45) 0.10(0.80-1.24) 1.00(0.80-1.26) 0.99(0.77-1.27) 0.99(0.80-1.23)
0.10 0.07 0.03 0.02 0.07 0.09
0.923 0.945 0.973 0.984 0.947 0.927
4 4 4 4 4 4
62.5% 59.9% 0.0% 21.3% 49.4% 46.4%
0.91(0.69-1.21) 0.86(0.37-1.96) 0.94(0.77-1.15) 0.92(0.57-1.48) 0.94(0.77-1.14) 0.84(0.53-1.32)
0.66 0.37 0.58 0.36 0.64 0.75
0.508 0.712 0.560 0.721 0.521 0.453
2 2 2 2 2 2
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
0.94(0.84-1.06) 0.93(0.71-1.23) 0.92(0.79-1.07) 1.00(0.76-1.32) 0.92(0.80-1.07) 0.96(0.74-1.25)
0.99 0.52 1.06 0.02 1.11 0.27
0.322 0.605 0.287 0.983 0.269 0.786
3 3 3 3 3 3
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
0.84(0.72-0.97) 0.66(0.41-1.04) 0.85(0.71-1.01) 0.81(0.51-1.29) 0.83(0.70-0.99) 0.72(0.46-1.15)
2.32 1.78 1.82 0.89 2.09 1.48
0.020 0.075 0.069 0.373 0.036 0.140
3 3 3 3 3 3
> G) might be a susceptible marker for CRC (colon cancer and rectal cancer) in central Chinese population. To solve the discrepancies and improve the statistical strength, we further performed an updated meta-analysis to explore the real impact of CYP24A1 gene polymorphisms on cancer risk. However, none of the four SNPs were shown to be associated with cancer risk in total population, while rs2585428 and rs6022999 were significantly associated with cancer risk in Caucasian population and in Asian population respectly. Interestingly, the pooled results reinforced our current findings for that rs6022999 but not rs2585428 was significantly associated with the risk of CRC (especially colon cancer) in Chinese (Asian) population. Moreover, both the case-control study and meta-analysis showed no association for rs6068816 and cancer risk in Chinese (Asian) population. However, future studies with larger sample size in different ethnic populations are needed to confirm our present findings, and the molecule mechanism of rs2585428 and rs6022999 involving cancer susceptibility should be elucidated in future functional studies. In conclusion, our results indicated that the rs4809960 and rs6022999 may be strongly associated with risk of colon cancer in Chinese population. The updated meta-analysis showed rs2585428 and
rs6022999 may link to cancer susceptibility in Caucasian population and in Asian population respectly, and rs6068816 may not be associated with cancer risk. However, cohort expansion and further mechanistic studies on the role of these CYP24A1 gene polymorphisms that influence carcinogenesis are necessary in future. Declaration of Competing Interest The authors declare that they have no conflict of interest. Acknowledgements This work was supported by grants from the National Natural Science Foundation of China (81502427), the Natural Science Foundation of Hubei Province (2019CFB756), and the Fundamental Research Funds for the Central Universities (WUT: 2018IB023, 2019IB005 and 2019-HS-B1-13).
8
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Fig. 1. Sensitivity analysis of the meta‐analysis. (A) rs2585428 (G vs. A), (B) rs4809960 (T vs. C), (C) rs6022999 (A vs. G) and (D) rs6068816 (C vs. T).
Appendix A. Supplementary data
[9] L.N. Anderson, M. Cotterchio, J.A. Knight, A. Borgida, S. Gallinger, S.P. Cleary, Genetic variants in vitamin d pathway genes and risk of pancreas cancer; results from a population-based case-control study in ontario, Canada, PLoS One 8 (2013) e66768. [10] L.L. Reimers, K.D. Crew, P.T. Bradshaw, R.M. Santella, S.E. Steck, I. Sirosh, M.B. Terry, D.L. Hershman, E. Shane, S. Cremers, E. Dworakowski, S.L. Teitelbaum, A.I. Neugut, M.D. Gammon, Vitamin D-related gene polymorphisms, plasma 25hydroxyvitamin D, and breast cancer risk, Cancer Causes Control 26 (2015) 187–203. [11] J.J. Oh, S.S. Byun, S.E. Lee, S.K. Hong, C.W. Jeong, W.S. Choi, D. Kim, H.J. Kim, S.C. Myung, Genetic variants in the CYP24A1 gene are associated with prostate cancer risk and aggressiveness in a Korean study population, Prostate Cancer Prostatic Dis. 17 (2014) 149–156. [12] T.V. Clendenen, W. Ge, K.L. Koenig, T. Axelsson, M. Liu, Y. Afanasyeva, A. Andersson, A.A. Arslan, Y. Chen, G. Hallmans, P. Lenner, T. Kirchhoff, E. Lundin, R.E. Shore, M. Sund, A. Zeleniuch-Jacquotte, Genetic polymorphisms in vitamin d metabolism and signaling genes and risk of breast Cancer: a nested case-control study, PLoS One 10 (2015) e0140478. [13] L.M. Dong, C.M. Ulrich, L. Hsu, D.J. Duggan, D.S. Benitez, E. White, M.L. Slattery, F.M. Farin, K.W. Makar, C.S. Carlson, B.J. Caan, J.D. Potter, U. Peters, Vitamin D related genes, CYP24A1 and CYP27B1, and colon cancer risk, Cancer Epidemiol. Biomarkers Prev. 18 (2009) 2540–2548. [14] R. Qu, X. Li, X. Quan, L. Xia, X. Fang, H. Li, B. Zhou, Polymorphism in CYP24A1 is associated with lung Cancer risk: a case-control study in chinese female nonsmokers, DNA Cell Biol. 38 (2019) 243–249. [15] F. Bray, J. Ferlay, I. Soerjomataram, R.L. Siegel, L.A. Torre, A. Jemal, Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA Cancer J. Clin. 68 (2018) 394–424. [16] J. Gu, N. Chen, Current status of rectal cancer treatment in China, Colorectal Dis. 15 (2013) 1345–1350. [17] H.C. Horvath, P. Lakatos, J.P. Kosa, K. Bacsi, K. Borka, G. Bises, T. Nittke, P.A. Hershberger, G. Speer, E. Kallay, The candidate oncogene CYP24A1: a potential biomarker for colorectal tumorigenesis, J. Histochem. Cytochem. 58 (2010) 277–285.
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.prp.2019.152735. References [1] S. Kimball, H. Fuleihan Gel, R. Vieth, Vitamin D: a growing perspective, Crit. Rev. Clin. Lab. Sci. 45 (2008) 339–414. [2] J.C. Fleet, M. DeSmet, R. Johnson, Y. Li, Vitamin D and cancer: a review of molecular mechanisms, Biochem. J. 441 (2012) 61–76. [3] G. Jones, D.E. Prosser, M. Kaufmann, 25-Hydroxyvitamin D-24-hydroxylase (CYP24A1): its important role in the degradation of vitamin D, Arch. Biochem. Biophys. 523 (2012) 9–18. [4] A.N. King, D.G. Beer, P.J. Christensen, R.U. Simpson, N. Ramnath, The vitamin D/ CYP24A1 story in cancer, Anticancer Agents Med. Chem. 10 (2010) 213–224. [5] P.E. Thomas, R. Klinger, L.I. Furlong, M. Hofmann-Apitius, C.M. Friedrich, Challenges in the association of human single nucleotide polymorphism mentions with unique database identifiers, BMC Bioinformatics 12 (Suppl 4) (2011) S4. [6] C.N. Holick, J.L. Stanford, E.M. Kwon, E.A. Ostrander, S. Nejentsev, U. Peters, Comprehensive association analysis of the vitamin D pathway genes, VDR, CYP27B1, and CYP24A1, in prostate cancer, Cancer Epidemiol. Biomarkers Prev. 16 (2007) 1990–1999. [7] S.K. Holt, E.M. Kwon, U. Peters, E.A. Ostrander, J.L. Stanford, Vitamin D pathway gene variants and prostate cancer risk, Cancer Epidemiol. Biomarkers Prev. 18 (2009) 1929–1933. [8] S. Yao, G. Zirpoli, D.H. Bovbjerg, L. Jandorf, C.C. Hong, H. Zhao, L.E. Sucheston, L. Tang, M. Roberts, G. Ciupak, W. Davis, H. Hwang, C.S. Johnson, D.L. Trump, S.E. McCann, F. Ademuyiwa, K.S. Pawlish, E.V. Bandera, C.B. Ambrosone, Variants in the vitamin D pathway, serum levels of vitamin D, and estrogen receptor negative breast cancer among African-American women: a case-control study, Breast Cancer Res. 14 (2012) R58.
9
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al. [18] M. MR, F. J, Meta-analysis of genetic association studies, Trends Genet. 20 (2004) 439–444. [19] M. Zhu, S. Qiu, X. Zhang, Y. Wang, T.D.M. Souraka, X. Wen, C. Liang, J. Tu, The associations between CYP24A1 polymorphisms and cancer susceptibility: a metaanalysis and trial sequential analysis, Pathol. Res. Pract. 214 (2018) 53–63. [20] P. Sedgwick, Multiple significance tests: the Bonferroni correction, BMJ 344 (2012). [21] N. Mantel, W. Haenszel, Statistical aspects of the analysis of data from retrospective studies of disease, J. Natl. Cancer Inst. 22 (1959) 719–748.
[22] R. DerSimonian, N. Laird, Meta-analysis in clinical trials, Control. Clin. Trials 7 (1986) 177–188. [23] Y. Hayashino, Y. Noguchi, T. Fukui, Systematic evaluation and comparison of statistical tests for publication bias, J. Epidemiol. 15 (2005) 235–243. [24] C. International HapMap, A haplotype map of the human genome, Nature 437 (2005) 1299–1320. [25] A. Montpetit, F. Chagnon, [The Haplotype Map of the human genome: a revolution in the genetics of complex diseases], Med. Sci. 22 (2006) 1061–1067.
10
Contents lists available at ScienceDirect
Pathology - Research and Practice journal homepage: www.elsevier.com/locate/prp
Association study between CYP24A1 gene polymorphisms and cancer risk Can Yia,1, Chao Huanga,1, Huan Wanga, Chen Wanga, Lijuan Donga, Xiuli Gub,c, Xianhong Fengd, Bifeng Chena,* a
Department of Biological Science and Technology, School of Chemistry, Chemical Engineering and Life Sciences, Wuhan University of Technology, Wuhan, China Center of Reproductive Medicine, Tongji Medical College, Huazhong University of Science and Technology, China Department of Reproductive Genetics, Wuhan Tongji Reproductive Medicine Hospital, Wuhan, China d Clinical Laboratory, Wuhan Xinzhou District People's Hospital, Wuhan, China b c
ARTICLE INFO
ABSTRACT
Keywords: CYP24A1 Single nucleotide polymorphism (SNP) Cancer risk Meta-analysis
CYP24A1, an essential gene in regulation of vitamin D, has been reported to play an important role in enhancing immune activity and inhibiting tumorigenesis. Previous studies proposed that rs2585428, rs4809960, rs6022999 and rs6068816 in CYP24A1 gene might be greatly associated with cancer risk. To validate the findings, we here investigated the associations of these four polymorphisms and colorectal cancer (CRC) risk in a central Chinese population (426 colon cancer patients, 361 rectal cancer patients and 800 healthy controls). The genotyping was conducted by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) and confirmed by sequencing. Our results revealed that the rs4809960 and rs6022999 were strongly associated with the CRC risk, especially with the colon cancer risk. Moreover, the analysis of haplotypes consisting of rs2585428(G > A), rs4809960(T > C), rs6022999(A > G) and rs6068816(C > T) indicated that haplotype ATGC significantly decreased the CRC risk, especially the colon cancer risk. Haplotype GCAT significantly increased the CRC risk, especially the rectal cancer risk. However, haplotype ACAC was only found to be associated with increased risk of CRC. To improve the statistical strength, an updated meta-analysis was further performed. The results showed that rs2585428 was associated with cancer risk in Caucasian population, rs4809960 was associated with breast cancer risk in Caucasian population, and rs6022999 was associated with cancer risk in Asian population. Collectively, the rs4809960 and rs6022999 may be the genetic biomarkers for prediction of colon cancer risk in Chinese population, the rs2585428 and rs6022999 may link to cancer susceptibility in Caucasian population and in Asian population respectly.
1. Introduction Vitamin D, an important fat-soluble vitamin and steroid prohormone, plays an essential role in human health for that it influences immune function, cell proliferation, differentiation and apoptosis. Vitamin D deficiency has been associated with numerous health outcomes, including bone disease, cancer, autoimmune disease and more [1,2]. In the process of vitamin D metabolism, the 25(OH)D (biomarker of vitamin D status) is first converted to 125(OH)2D3 by 1α-hydroxylase (encoded by CYP27B1 gene) in the kidney, then 125(OH)2D3 (active form of vitamin D) is released into the blood circulation. Lastly, the circulating 125(OH)2D3 will be degraded by 25-hydroxyvitamin D 24hydrolase (encoded by CYP24A1 gene), suggesting that CYP24A1 is the essential enzyme responsible for the degradation of vitamin D [3]. In
fact, some researchers have linked the abnormal expression of CYP24A1 gene to cancer risk, and elevated expression of CYP24A1 gene was observed in multiple types of cancer [4]. Collectively, CYP24A1 may be an oncogene and may contribute to tumor aggressiveness through degrading vitamin D. Recently, single nucleotide polymorphism (SNP), a well-defined molecular biomarker, has been extensively studied and applied to medical issues [5]. Interestingly, Four SNPs (rs2585428, rs4809960, rs6022999, and rs6068816) in CYP24A1 gene were found to be associated with cancer risks, including prostate cancer, breast cancer and pancreas cancer, indicating that the four CYP24A1 gene polymorphisms may serve as valuable biomarkers for cancer predisposition [6–14]. Colorectal carcinoma (CRC, colon cancer and rectal cancer) is an important contributor to cancer mortality and morbidity [15]. The
Corresponding author at: Department of Biological Science and Technology, School of Chemistry, Chemical Engineering and Life Sciences, Wuhan University of Technology, Wuhan, China. E-mail address: [email protected] (B. Chen). 1 Can Yi and Chao Huang contributed equally to this work. ⁎
https://doi.org/10.1016/j.prp.2019.152735 Received 17 August 2019; Received in revised form 23 October 2019; Accepted 10 November 2019 0344-0338/ © 2019 Elsevier GmbH. All rights reserved.
Please cite this article as: Can Yi, et al., Pathology - Research and Practice, https://doi.org/10.1016/j.prp.2019.152735
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 1 Characteristics of studied subjects. Group
Healthy controls (n = 800) Colon cancer patients (n = 426) Rectal cancer patients (n = 361) Colorectal cancer patients (n = 787)
Age, n (%)
Gender, n (%)
Smoking status, n (%)
Drinking status, n (%)
Never
Ever
591 298 261 559
237 135 122 257
≤60 years
> 60 years
Male
Female
Ever
434 245 210 455
366 181 151 332
558 309 246 555
242 117 115 232
209 128 100 228
(54.3) (57.5) (58.2) (57.8)
(45.7) (42.5) (41.8) (42.2)
(69.7) (72.5) (68.1) (70.5)
(30.3) (27.5) (31.9) (29.5)
(26.1) (30.0) (27.7) (29.0)
(73.9) (70.0) (72.3) (71.0)
P-value1
Never (29.6) (31.7) (33.8) (32.7)
563 291 239 530
(70.4) (68.3) (66.2) (67.3)
0.274 0.213 0.153
0.308 0.583 0.737
0.143 0.574 0.204
0.454 0.155 0.192
1
Two-sided χ2 test for the distributions of age (1st column), gender (2nd column), smoking status (3rd column) and drinking status (4th column) between colon/rectal/ colorectal cancer patients and healthy controls.
2.2. Genotyping assay
distinction between the colon cancer and the rectal cancer is largely anatomical. Both surgical and radiotherapeutic have been applied to CRC treatment, and achieved satisfactory results. However, due to the delay in diagnosis, most CRC patients get limited treatment and poor prognosis. Therefore, early and effective diagnostic methods will be particularly necessary [16]. Henrik et al. previously demonstrated that the CYP24A1 gene might be a potential biomarker for colorectal tumorigenesis [17], suggesting that there may be a potential association between CYP24A1 gene polymorphisms and CRC risk. To address this issue, we here examined the associations between the CYP24A1 gene polymorphisms (rs2585428, rs4809960, rs6022999, and rs6068816) and CRC risk in a central Chinese population. The meta-analysis is a statistical tool for combining the results from different studies on the same topic, thus increasing statistical strength and precision [18]. To resolve the discrepancies among the previous association studies, Zhu et al. have performed a meta-analysis to characterize the precise associations of rs2585428, rs4809960, rs6022999 and rs6068816 and cancer risk [19]. However, we found that the study of Oh et al. [11] was ignored by Zhu et al. Moreover, given the newly generated experiment data in present case–control study, we further performed a rigorous and updated meta-analysis (up to April of 2019) to determine the association of CYP24A1 gene polymorphism and cancer risk.
The peripheral blood samples (5 ml per participant) were collected into blood vacuum tubes containing EDTA and stored at 4 °C. Then, genomic DNA was extracted from blood samples using the TIANamp Blood DNA Kit (DP348; TianGen Biotech, Beijing, China), and stored at −20 °C before use. Next, the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique was conducted to genotype the four SNPs. The information of PCR-RFLP assay was described in supplementary Table S1. To ensure the reliability of genotyping results, the PCR-RFLP assay was repeated twice for all subjects, and 10 % randomly selected DNA samples were examined by DNA sequencing, all the results were 1.0 % concordant. The sequencing diagrams of the four CYP24A1 gene polymorphisms were shown in supplementary Figure S1. 2.3. Statistical analysis The Statistical Program for Social Sciences (SPSS, version 15.0, Chicago, IL, USA) was used to perform all the analyses. χ2 test was used to determine the statistical differences in age, gender, smoking status and drinking status. The numerical data of average ages for cases and controls were presented as means ± SD (standard deviation of the mean), and their differences were evaluated using the Student’s t-test. The genotypic frequencies of rs2585428, rs4809960, rs6022999 and rs6068816 in normal controls were tested for departure from HWE (Hardy-Weinberg equilibrium). Logistic regression analysis was used to estimate the associations between the four SNPs and cancer risk. Squared correlation coefficient r2 for each pair of polymorphic sites and haplotype analysis was calculated by using Haploview software available at http://www.broad.mit.edu/mpg/haploview. P values less than 0.05 were considered significant for all statistical analyses, and the Bonferroni correction for multiple testing was applied [20].
2. Material and methods 2.1. Studied subjects In this study, we recruited a total of 787 CRC patients (426 colon cancer patients and 361 rectal cancer) and 800 healthy controls. All the CRC patients were confirmed histopathologically and recruited from Hubei Cancer Hospital and Wuhan Xinzhou District People's Hospital between January 2015 and December 2016. Nowadays, more and more Chinese are inclined to have a physical examination every year. The healthy controls were selected from cancer-free individuals who visited Wuhan Xinzhou District People's Hospital for regular physical examinations between September 2014 and December 2016 or who volunteered to participate in the epidemiology survey during the same period. It was required that the normal controls passed all annual physical examinations in the latest three years. All participants were biologically unrelated Han Chinese living in central China (Hubei province). This study was approved by the Ethical Committees of Wuhan University of Technology (Approval No: WUT02720180705), and written informed consent for the genetics analysis was obtained from all participants or their guardians.
2.4. Meta-analysis We comprehensively searched the related literatures updated to April of 2019 on PubMed, EMBASE, ISI Web of Science, and CNKI and Wanfang databases without language restriction. The search terms used were as follows: "CYP24A1, rs2585428, and cancer/tumor/carcinoma", "CYP24A1, rs4809960, and cancer/tumor/carcinoma", "CYP24A1, rs6022999, and cancer/tumor/carcinoma", and "CYP24A1, rs6068816, and cancer/tumor/carcinoma". The included studies should meet the following criteria: (1) studies on humans, (2) investigation of the CYP24A1 gene polymorphisms (at least one of the four polymorphisms)
2
3
2
1
687 (80.6%) 165 (19.4%) 279 (65.5%) 129 (30.3%) 18 (4.2%)
553 (64.9%) 299 (35.1%) 181 (42.5%) 191 (44.8%) 54 (12.7%)
rs6022999 polymorphism A 1253 (79.6%) G 321 (20.4%) AA 504 (64.1%) AG 245 (31.1%) GG 38 (4.8%)
rs6068816 polymorphism C 1038 (65.9%) T 536 (34.1%) CC 342 (43.5%) CT 354 (45.0%) TT 91 (11.5%) 485 (67.2%) 237 (32.8%) 161 (44.6%) 163 (45.2%) 37 (10.2%)
566 (78.4%) 156 (21.6%) 225 (62.4%) 116 (32.1%) 20 (5.5%)
534 (74.0%) 188 (26.0%) 202 (56.0%) 130 (36.0%) 29 (8.0%)
352 (48.8%) 370 (51.2%) 85 (23.5%) 182 (50.4%) 94 (26.1%)
3. Rectal cancer patient
(50.9%) (49.1%) (27.3%) (47.4%) (25.3%)
1072 (67.0%) 528 (33.0%) 362 (45.3%) 348 (43.5%) 90 (11.2%)
1192 (74.5%) 408 (25.5%) 443 (55.4%) 306 (38.2%) 51 (6.4%)
1226 (76.6%) 374 (23.4%) 468 (58.5%) 290 (36.2%) 42 (5.3%)
815 785 218 379 203
4. Healthy controls
0.645
0.849
0.736
0.140
HWE1
C vs. T CC vs. TT CC vs. CT CT vs. TT CC vs. CT + TT CC + CT vs. TT
A vs. G AA vs. GG AA vs.AG AG vs. GG AA vs. AG + GG AA + AG vs. GG
T vs. C TT vs. CC CT vs. CC TT vs. CT TT vs. CT + CC TT + CT vs. CC
G vs. A GG vs. AA GG vs. AG AG vs. AA GG vs. AG + AA GG + AG vs. AA
Genetic Model
0.80 0.61 0.83 0.74 0.79 0.66
0.92 0.84 1.02 0.83 0.83 0.95 (0.69-0.94) (0.40-0.92) (0.67-1.02) (0.48-1.13) (0.65-0.97) (0.44-0.99)
(0.81-1.03) (0.66-1.07) (0.83-1.25) (0.67-1.02) (0.68-1.01) (0.78-1.16)
0.530, 0.684, 0.488, 0.971, 0.472, 0.845,
0.95 0.93 0.93 1.01 0.93 0.97
(0.82-1.11) (0.67-1.30) (0.75-1.15) (0.73-1.40) (0.76-1.13) (0.71-1.32)
0.001, 1.34 (1.13-1.58) 0.059, 1.53 (0.98-2.37) 0.001, 1.42 (1.15-1.76) 0.755, 1.08 (0.68-1.69) < 0.001, 1.44 (1.17-1.76) 0.182, 1.34 (0.87-2.07)
0.008, 0.020, 0.073, 0.161, 0.021, 0.044,
0.151, 0.157, 0.874, 0.077, 0.068, 0.626,
1 vs. 4
P value, OR(95% CI)2
Genotypic frequency of SNPs in normal controls were tested for departure from Hardy-Weinberg equilibrium (HWE) using the χ2 test. The OR (95% CI) and the corresponding P value were calculated by logistic regression analysis.
607 (71.2%) 245 (28.8%) 213 (50.0%) 181 (42.5%) 32 (7.5%)
rs4809960 polymorphism T 1141 (72.5%) C 433 (27.5%) TT 415 (52.7%) CT 311 (39.5%) CC 61 (7.8%)
(48.7%) (51.3%) (23.9%) (49.5%) (26.6%)
2. Colon cancer patients
415 437 102 211 113
1. Colorectal cancer
rs2585428 polymorphism G 767 (48.7%) A 807 (51.3%) GG 187 (23.8%) AG 393 (49.9%) AA 207 (26.3%)
Genotype
Table 2 Genotype and allele distributions of CYP24A1 polymorphisms and their associations with colorectal cancer risk.
0.296, 0.349, 0.467, 0.646, 0.354, 0.460,
0.001, 0.042, 0.002, 0.545, 0.001, 0.122,
0.004, 0.038, 0.012, 0.431, 0.004, 0.115,
0.293, 0.301, 0.999, 0.239, 0.210, 0.661,
2 vs. 4
0.91 0.83 0.91 0.92 0.89 0.87
1.43 1.79 1.49 1.19 1.53 1.54
0.76 0.60 0.73 0.82 0.71 0.68
0.92 0.84 1.00 0.84 0.84 0.94
(0.77-1.09) (0.57-1.22) (0.71-1.17) (0.63-1.34) (0.71-1.13) (0.61-1.25)
(1.16-1.75) (1.02-3.12) (1.16-1.93) (0.67-2.12) (1.20-1.95) (0.89-2.68)
(0.63-0.91) (0.37-0.97) (0.57-0.93) (0.50-1.35) (0.56-0.90) (0.42-1.10)
(0.78-1.08) (0.61-1.17) (0.75-1.33) (0.63-1.12) (0.64-1.10) (0.72-1.23)
0.934, 0.717, 0.699, 0.548, 0.836, 0.613,
0.043, 0.349, 0.032, 0.905, 0.027, 0.583,
0.166, 0.066, 0.779, 0.101, 0.417, 0.069,
0.330, 0.336, 0.813, 0.183, 0.184, 0.810,
3 vs. 4
1.01 1.08 0.95 1.14 0.97 1.11
1.24 1.30 1.34 0.97 1.33 1.16
0.87 0.63 0.96 0.65 0.90 0.63
0.92 0.84 1.04 0.81 0.82 0.97
(0.84-1.22) (0.71-1.66) (0.73-1.24) (0.74-1.74) (0.76-1.25) (0.74-1.66)
(1.01-1.53) (0.75-2.23) (1.03-1.75) (0.55-1.69) (1.03-1.72) (0.68-1.98)
(0.71-1.06) (0.38-1.03) (0.74-1.26) (0.39-1.09) (0.70-1.16) (0.39-1.04)
(0.77-1.09) (0.59-1.20) (0.77-1.40) (0.60-1.10) (0.62-1.10) (0.73-1.28)
C. Yi, et al.
Pathology - Research and Practice xxx (xxxx) xxxx
Pathology - Research and Practice xxx (xxxx) xxxx
The haplotype structure was rs2585428(G > A)/rs4809960(T > C)/rs6022999(A > G)/rs6068816(C > T). All haplotypes with frequency < 0.01 in both case and control groups would be ignored in analysis. The OR (95% CI) and the corresponding P value were calculated by logistic regression analysis.
0.99(0.70-1.41) 2.33(1.38-3.91) 0.97(0.81-1.15) 0.84(0.70-1.01) 0.956 0.001 0.720 0.070 < 0.0001 0.003 10.67 0.13 3.29 34.82
0.91(0.74-1.12) 0.88(0.71-1.09) 0.374 0.248 0.0002 0.79 1.34 30.1
0.95(0.75-1.20) 1.45(1.07-1.96) 0.72(0.59-0.88) 0.652 0.017 0.002 0.20 5.67 9.99
and cancer risk, (3) case-control study design, (4) sufficient data was accessible to estimate the OR and 95%CI, and (5) HWE equilibrium should be established in controls. The flowchart of the search strategy of present meta-analysis was shown in supplementary Figure S2. All analyses were performed using STATA 14.0 (Stata Corp, College Station, TX). The Cochran's Q-test and I2 were used to assess the heterogeneity of included studies. If Pheterogeneity ≥ 0.1, the fixed-effect model was applied to calculate the combined ORs [21], otherwise, random-effects model was used [22]. The significance of combined ORs was determined by the Z test. A P value < 0.05 was considered significantly, and the Bonferroni correction for multiple testing was also applied. 3. Results 3.1. Characteristics of studied subjects The distribution of participants’ variables was shown in Table 1, there were no significant differences in distribution of age, gender, smoking status, drinking status between cancer patients and healthy controls, suggesting that our case-control study was well matched. Moreover, the average ages of healthy controls, colon cancer patients, rectal cancer patients and colorectal cancer patients were 56.16 ± 10.85, 57.36 ± 10.47, 58.06 ± 10.65 and 57.87 ± 10.76 years old, respectly. No statistical differences were observed in average ages between patients (colon, rectal or colorectal cancer) and healthy controls. 3.2. The associations between CYP24A1 gene polymorphisms and CRC risk In this study, we performed comparisons for allele and genotype frequencies of the four CYP24A1 gene polymorphisms between cancer patients (colon cancer, rectal cancer and CRC) and healthy controls under six genetic models (allele model, carrier model, homozygote model, heterozygote model, recessive model, dominant model) (Table 2). The healthy controls in our study were in accordance with HWE for rs2585428, rs4809960, rs6022999 and rs6068816 polymorphisms and had a good representative. After Bonferroni correction (P < 0.0084, 0.05/6), we observed that the rs4809960 and rs6022999 polymorphisms were significantly associated with risk of CRC (especially colon cancer) but not rectal cancer, while rs2585428 and rs6068816 polymorphisms were not associated with risk of CRC, colon cancer or rectal cancer. For rs4809960 polymorphism, allele T was a significant protective allele of CRC (T vs. C, P = 0.008, OR = 0.80, 95%CI = 0.69-0.94) and colon cancer (T vs. C, P = 0.004, OR = 0.76, 95%CI = 0.63-0.91), and individuals with TT genotype had a lower risk for colon cancer compared with CT + CC (TT vs. CT + CC, P = 0.004, OR = 0.71, 95%CI = 0.56-0.90). For rs6022999 polymorphism, A allele had a higher risk for CRC (A vs. G, P = 0.001, OR = 1.34, 95%CI = 1.13–1.58) and colon cancer (A vs. G, P = 0.001, OR = 1.43, 95%CI = 1.16–1.75) than those carrying the G allele, and AA genotype conferred higher risk for CRC and colon cancer relative to AG and AG + GG genotypes (CRC: AA vs. AG, P = 0.001, OR = 1.42, 95%CI = 1.15–1.76; AA vs. AG + GG, P < 0.001, OR = 1.44, 95%CI = 1.17-1.76. Colon cancer: AA vs. AG, P = 0.002, OR = 1.49, 95%CI = 1.16–1.93; AA vs. AG + GG, P = 0.001, OR = 1.53, 95%CI = 1.20–1.95). 3.3. The associations between CYP24A1 haplotypes and CRC risk The patterns of linkage disequilibrium (LD) for the four CYP24A1 gene polymorphism were shown in supplementary Figure S3, which indicated low LD with each other. Since the haplotype analysis could
2
1
0.68(0.42-1.13) 2.67(1.49-4.80) 1.03(0.83-1.28) 0.78(0.62-0.99) 0.134 0.001 0.815 0.041 < 0.0001 2.25 11.7 0.06 4.19 32.3
0.77(0.56-1.06) 1.62(1.13-2.32) 0.85(0.66-1.09) 0.105 0.008 0.190 2.62 7.00 1.72
1.37(1.06-1.76) 0.014 6.01
1.36(1.07-1.72) 1.28(0.81-2.01) 1.06(0.81-1.40) 1.32(0.92-1.89) 0.61(0.48-0.79) 1.97(1.17-3.31) 1.24(0.84-1.85) 0.013 0.288 0.663 0.138 0.001 0.009 0.280 6.24 1.13 0.19 2.20 14.3 6.78 1.17 1.37(1.12-1.69) 0.002 9.31
0.120 0.030 0.098 0.047 0.162 0.018 0.041 0.013 0.204 0.192 0.925 0.163 0.015 0.081 0.077 0.147 0.013 0.029 0.035 0.218 0.165 0.943 0.160 0.026 0.095 0.068 0.126 0.026 0.041 0.030 0.204 0.172 0.948 ACAC ACAT ATAC ATAT ATGC ATGT GCAC GCAT GTAC GTAT Global result
0.163 0.015 0.081 0.077 0.147 0.013 0.029 0.035 0.218 0.165 0.943
χ2
3. Rectal cancer 2. Rectal cancer 1. Colorectal cancer Haplotype1
Table 3 Association between CYP24A1 haplotypes with colorectal cancer risk.
4. Healthy controls
1 vs. 4
2 vs. 4
P χ2 OR(95% CI)2 P2
3 vs. 4
P χ2 OR(95% CI)
OR(95% CI)
C. Yi, et al.
4
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 4 Meta-analysis of association between rs2585428 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs2585428 and cancer risk G vs. A 23.68 0.003 GG vs. AA 17.26 0.027 AG vs. AA 5.14 0.742 GG vs. AG 15.82 0.045 GG vs. AG + AA 20.10 0.010 GG + AG vs. AA 9.26 0.321 rs2585428 and cancer risk in Asian population G vs. A 8.90 0.012 GG vs. AA 8.12 0.017 AG vs. AA 1.41 0.495 GG vs. AG 5.31 0.070 GG vs. AG + AA 7.94 0.019 GG + AG vs. AA 4.17 0.124 rs2585428 and cancer risk based on Illumina method G vs. A 1.58 0.208 GG vs. AA 1.63 0.201 AG vs. AA 0.03 0.854 GG vs. AG 11.88 0.001 GG vs. AG + AA 1.56 0.211 GG + AG vs. AA 0.30 0.584 rs2585428 and cancer risk based on PCR-RFLP method G vs. A 2.79 0.426 GG vs. AA 2.63 0.452 AG vs. AA 0.69 0.877 GG vs. AG 2.96 0.397 GG vs. AG + AA 3.40 0.333 GG + AG vs. AA 1.14 0.767 rs2585428 and prostate cancer risk G vs. A 4.84 0.184 GG vs. AA 4.52 0.211 AG vs. AA 2.27 0.518 GG vs. AG 1.63 0.653 GG vs. AG + AA 2.91 0.406 GG + AG vs. AA 3.76 0.289
I
2
Hypothesis test
Studies (n)
Z
P
66.2% 53.7% 0.0% 49.4% 60.2% 13.6%
1.11(1.10-1.24) 1.13(0.94-1.36) 1.01(0.91-1.12) 1.09(0.99-1.20) 1.12(0.96-1.31) 1.04(0.94-1.15)
1.92 1.31 0.18 1.78 1.43 0.80
0.055 0.191 0.858 0.076 0.153 0.426
9 9 9 9 9 9
77.5% 75.4% 0.0% 62.3% 74.8% 75.2%
1.04(0.82-1.33) 1.07(0.67-1.72) 1.06(0.88-1.29) 0.96(0.70-1.32) 1.01(0.70-1.44) 1.07(0.81-1.42)
0.32 0.30 0.61 0.24 0.01 0.48
0.746 0.766 0.543 0.807 0.992 0.635
3 3 3 3 3 3
36.8% 38.7% 0.0% 91.6% 36.0% 0.0%
1.20(1.10-1.31) 1.32(1.14-1.53) 1.09(0.96-1.23) 1.13(0.29-4.45) 1.88(1.23-2.86) 1.17(1.12-1.23)
3.93 3.69 1.34 0.17 2.92 2.60
< 0.001 < 0.001 0.182 0.866 0.004 < 0.001
2 2 2 2 2 2
0.0% 0.0% 0.0% 0.0% 11.9% 0.0%
0.97(0.89-1.06) 0.94(0.78-1.13) 1.02(0.86-1.19) 0.94(0.80-1.09) 0.94(0.81-1.09) 0.99(0.85-1.15)
0.65 0.67 0.18 0.84 0.85 0.18
0.516 0.500 0.856 0.398 0.393 0.855
4 4 4 4 4 4
38.0% 33.6% 0.0% 0.0% 0.0% 20.1%
1.12(1.02-1.24) 1.25(1.02-1.53) 1.13(0.93-1.36) 1.20(0.95-1.32) 1.16(0.10-1.35) 1.17(0.98-1.40)
2.25 2.18 1.23 1.37 1.90 1.75
0.024 0.029 0.218 0.172 0.057 0.080
4 4 4 4 4 4
enhance the statistical power in the mapping of human complex trait loci, the analysis of haplotypes consisting of rs2585428(G > A), rs4809960(T > C), rs6022999(A > G) and rs6068816(C > T) was performed to estimate the association between CYP24A1 gene and cancer risk in this study. In Table 3, haplotype ATGC was significant associated with a decreased risk for CRC (P = 0.002, OR = 0.72, 95%CI = 0.59-0.88) and colon cancer (P = 0.001, OR = 0.61, 95%CI = 0.48-0.79), haplotype GCAT was significant associated with an increased risk for CRC (P = 0.001, OR = 2.33, 95%CI = 1.38–3.91) and rectal cancer (P = 0.001, OR = 2.67, 95%CI = 1.49–4.80), and haplotype ACAC was significant associated with an increased risk for CRC (P = 0.002, OR = 1.37, 95%CI = 1.12–1.69).
for rs2585428, rs4809960, rs6022999 and rs6068816 in each control group. For the four CYP24A1 gene polymorphisms, their associations with cancer risk were evaluated in total population, as well as in stratified analysis according to ethnicity (Asian and Caucasian), genotyping method (PCR-RFLP and Illumina) and cancer type (prostate, breast and colon cancer). Compared with Zhu et al’s meta-analysis, another two association studies in Asian population [11,14] and present case-control study in Chinese (Asian) population have been added in ours. Therefore, the analysis results of CYP24A1 gene polymorphisms and overall cancer risk in Caucasian population, as well as CYP24A1 gene polymorphisms and breast cancer risk were not updated and not shown in present study. Due to the insufficient numbers of included studies (n < 10), the assessment of publication bias was not performed through Begg’s funnel plot and Egger’s linear regression method in this metaanalysis [23]. A sum of 9 publications for rs2585428 that met the inclusion criteria were finally included. As shown in Table 4, we found that rs2585428 was not associated with cancer risk in total population or Asian population. In the cancer type-stratified analysis, it was observed
3.4. Meta-analysis of CYP24A1 gene polymorphisms and cancer risk As shown in supplementary Table S2, the NOS score of all articles are not < 6, indicating that each included literature was a high-quality study. The characteristics of included studies for this meta-analysis were demonstrated in Supplementary Table S3. HWE was established
5
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 5 Meta-analysis of association between rs4809960 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs4809960 and cancer risk in total population T vs. C 21.16 0.004 TT vs. CC 10.24 0.193 TT vs. CT 24.01 0.001 CT vs. CC 7.95 0.329 TT vs. CT + CC 24.17 0.001 TT + CT vs. CC 8.45 0.312 rs4809960 and cancer risk in Asian population T vs. C 3.40 0.182 TT vs. CC 1.65 0.438 TT vs. CT 3.47 0.176 CT vs. CC 1.01 0.604 TT vs. CT + CC 3.80 0.150 TT + CT vs. CC 1.26 0.533 rs4809960 and cancer risk based on Illumina method T vs. C 0.28 0.597 TT vs. CC 0.00 0.998 TT vs. CT 0.55 0.457 CT vs. CC 0.10 0.748 TT vs. CT + CC 0.46 0.498 TT + CT vs. CC 0.01 0.909 rs4809960 and cancer risk based on PCR-RFLP method T vs. C 15.02 0.002 TT vs. CC 4.87 0.181 TT vs. CT 14.96 0.002 CT vs. CC 0.95 0.813 TT vs. CT + CC 16.51 0.001 TT + CT vs. CC 2.66 0.447 rs4809960 and prostate cancer risk T vs. C 3.49 0.323 TT vs. CC 0.49 0.922 TT vs. CT 10.21 0.017 CT vs. CC 4.20 0.241 TT vs. CT + CC 7.40 0.060 TT + CT vs. CC 1.48 0.688
2
I
Hypothesis test
Studies (n)
Z
P
66.9% 31.7% 70.8% 11.9% 71.0% 17.2%
1.02 (0.90-1.16) 0.98(0.82-1.16) 1.06(0.89-1.26) 0.91(0.76-1.08) 1.04(0.88-1.23) 0.95(0.80-1.13)
0.31 0.29 0.66 1.06 0.51 0.59
0.754 0.775 0.512 0.287 0.608 0.554
8 8 8 8 8 8
41.2% 0.0% 42.4% 0.0% 47.3% 0.0%
0.84(0.74-0.96) 0.66(0.48-0.92) 0.86(0.73-1.02) 0.77(0.55-1.08) 0.83(0.71-0.98) 0.71(0.51-0.97)
2.67 2.44 1.73 1.51 2.26 2.12
0.009 0.015 0.083 0.131 0.024 0.034
3 3 3 3 3 3
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
1.16(1.01-1.34) 1.18(0.81-1.71) 1.23(1.03-1.46) 0.96(0.65-1.41) 1.22(1.03-1.44) 1.09(0.75-1.28)
2.12 0.84 2.28 0.23 2.32 0.47
0.034 0.403 0.022 0.816 0.020 0.636
2 2 2 2 2 2
80.0% 38.4% 79.9% 0.0% 81.8% 0.0%
0.98(0.75-1.28) 0.79(0.60-1.03) 1.07(0.76-1.50) 0.78(0.59-1.03) 1.02(0.73-1.44) 0.79(0.60-1.03)
0.17 1.73 0.36 1.77 0.13 1.76
0.864 0.084 0.717 0.077 0.895 0.078
4 4 4 4 4 4
13.9% 0.0% 70.6% 28.6% 59.5% 0.0%
1.10(0.97-1.24) 1.23(0.90-1.67) 1.13(0.82-1.55) 1.13(0.82-1.56) 1.13(0.87-1.46) 1.19(0.88-1.62)
1.53 1.28 0.74 0.75 0.90 1.12
0.127 0.201 0.459 0.454 0.370 0.261
4 4 4 4 4 4
that rs2585428 was not associated with the risk of prostate cancer or breast cancer. However, the stratified analysis by genotyping method suggested that rs2585428 significantly associated with cancer risk based on Illumina method but not PCR-RFLP method, and rs2585428 was inclined to development cancer under three models (G vs. A, GG vs. AA, GG vs. AG + AA and GG + AG vs. AA). A sum of 8 publications for rs4809960 that met the inclusion criteria were finally retrieved. In Table 5, no significant association was found for rs4809960 and cancer risk in total population or Asian population. When performing stratified analysis by genotyping method and cancer type, it was also observed that rs4809960 was not associated with cancer risk or prostate cancer risk. We finally included 9 related publications for rs6080550 in this meta-analysis. As shown in Table 6, rs6022999 was suggested to significantly associate with cancer risk in Asian population, but not in total population. For Asians, A allele had a higher cancer risk than G allele (A vs. G), and AA genotype conferred higher cancer risk relative to AG and/or GG genotypes (AA vs. AG, AA vs. GG, and AA vs. AG + GG). Moreover, no association was found for rs6080550 with cancer in the stratified analysis according to genotyping method or cancer type. A sum of 8 publications for rs6068816 that met the inclusion criteria were
finally retrieved. In Table 7, rs6068816 was not associated with cancer risk in total population or Asian population, and was also not associated with cancer risk in the stratified analysis. Moreover, the sensitivity analysis identified that the removal of any single case-control study did not influence the stability of the results (Fig. 1). 4. Discussion CYP24A1 gene has been proposed as a potential diagnostic and prognostic indicator in cancer for its critical role in the metabolism process of Vitamin D. To be an important molecular biomarker, SNP has been extensively studied and applied to medical issues. Currently, the relations between CYP24A1 gene polymorphisms and cancer risk have attracted considerable attention, and revealed several possible susceptibility loci (e.g. rs2585428, rs4809960, rs6022999, and rs6068816). To support the findings, we here explored the associations between these four CYP24A1 polymorphisms and CRC risk in a central Chinese population. Interestingly, the association of rs6022999 and colon cancer risk has been examined by Dong et al., and produced a negative result [13]. In contrast, our results demonstrated that
6
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 6 Meta-analysis of association between rs6022999 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs6022999 and cancer risk A vs. G 55.66 < 0.001 AA vs. GG 17.38 0.026 AG vs. GG 5.97 0.651 AA vs. AG 18.34 0.019 AA + AG vs.GG 11.63 0.169 AA vs. AG + GG 24.14 0.002 rs6022999 and cancer risk in Asian population A vs. G 1.32 0.517 AA vs. GG 1.35 0.508 AG vs. GG 0.74 0.692 AA vs. AG 0.4 0.818 AA + AG vs.GG 1.16 0.559 AA vs. AG + GG 0.81 0.667 rs6022999 and cancer risk based on Illumina method A vs. G 2.27 0.132 AA vs. GG 1.02 0.313 AG vs. GG 0.18 0.673 AA vs. AG 1.53 0.217 AA + AG vs.GG 0.63 0.427 AA vs. AG + GG 2.08 0.149 rs6022999 and cancer risk based on PCR-RFLP method A vs. G 42.51 < 0.001 AA vs. GG 6.64 0.084 AG vs. GG 1.67 0.644 AA vs. AG 6.35 0.096 AA + AG vs.GG 4.01 0.260 AA vs. AG + GG 8.66 0.034 rs6022999 and prostate cancer risk A vs. G 30.83 < 0.001 AA vs. GG 6.54 0.088 AG vs. GG 3.01 0.390 AA vs. AG 5.19 0.158 AA + AG vs.GG 4.34 0.227 AA vs. AG + GG 6.56 0.087 rs6022999 and colon cancer risk A vs. G 8.80 0.003 AA vs. GG 4.81 0.028 AG vs. GG 1.26 0.262 AA vs. AG 4.78 0.029 AA + AG vs.GG 3.42 0.064 AA vs. AG + GG 7.18 0.007
I
2
Hypothesis test
Studies (n)
Z
P
85.6% 54.0% 0.0% 56.4% 31.2% 66.9%
1.08(0.91-1.27) 1.15(0.90-1.45) 0.95(0.82-1.10) 1.14 (1.01-1.29) 1.01(0.87-1.16) 1.15(1.00-1.32)
0.86 1.12 0.73 2.12 0.06 2.03
0.388 0.262 0.466 0.034 0.953 0.042
9 9 9 9 9 9
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
1.36 (1.19-1.56) 1.62(1.14-2.31) 1.14(0.79-1.64) 1.43 (1.21-1.69) 1.42 (1.00-2.01) 1.46(1.24-1.71)
4.55 2.66 0.69 4.20 1.97 4.61
< 0.001 0.008 0.488 < 0.001 0.048 < 0.001
3 3 3 3 3 3
55.9% 2.0% 0.0% 34.5% 0.0% 52.0%
1.27(0.98-1.64) 1.53(1.01-2.31) 1.27(0.83-1.93) 1.27(0.83-1.93) 1.19(1.00-1.42) 1.29(0.96-1.74)
1.81 2.02 1.11 1.98 1.73 1.70
0.070 0.043 0.267 0.047 0.083 0.089
2 2 2 2 2 2
92.9% 54.8% 0.0% 52.8% 25.3% 65.4%
1.09(0.70-1.68) 1.30(0.75-2.23) 0.93(0.70-1.23) 1.30(1.04-1.63) 1.09(0.84-1.42) 1.31(1.01-1.68)
0.37 1.14 0.51 2.30 0.63 2.06
0.708 0.255 0.608 0.021 0.526 0.039
4 4 4 4 4 4
90.3% 54.2% 0.3% 42.2% 30.9% 54.3%
1.05(0.70-1.57) 1.29(0.80-2.07) 0.10 (0.75-1.33) 1.10 (0.95-1.28) 1.08(0.82-1.42) 1.18(0.92-1.50)
0.22 1.05 0.01 1.25 0.54 1.31
0.822 0.294 0.993 0.212 0.591 0.189
4 4 4 4 4 4
88.6% 79.2% 20.7% 79.1% 70.7% 86.1%
1.18 (0.84-1.67) 1.21(0.62-2.36) 0.89(0.70-1.14) 1.25(0.91-1.71) 1.10(0.63-1.91) 1.25(0.86-1.81)
0.96 0.55 0.94 1.37 0.33 1.16
0.335 0.582 0.347 0.172 0.740 0.245
2 2 2 2 2 2
rs6022999 was significantly associated with colon cancer risk, as well as CRC risk. Similarly, a significant association between rs4809960 and risk of colon and CRC was also observed. However, the situation was different for rs2585428 and rs6068816. Consisted with previous studies, the four CYP24A1 gene polymorphisms were not simultaneously associated with risk of the same cancer in each study, and each CYP24A1 gene polymorphism showed inconsistent results across different studies. The possible reasons for these discrepancies might be as follows. First, the limited sample size may lead to inadequate statistical strength, which should be solved in well-designed study with large sample size. Second, the action of CYP24A1 may be varied in different cancer types, which contributes to the discrepancies of CYP24A1 gene polymorphisms to affect individuals' cancer susceptibility. Third, different environments, lifestyles, and genetic backgrounds among different ethnic populations
may also contribute to the differences of CYP24A1 gene polymorphisms on cancer risk. Nowadays, the successful completion of the HapMap project suggested that haplotype analysis would enhance the statistical power in the mapping of human complex trait loci, with the potential of reducing the sample size of association studies [24,25]. Indeed, none of the four CYP24A1 gene polymorphisms was identified to be associate with rectal cancer in the single-locus analysis, but haplotype GCAT (rs2585428rs4809960-rs6022999-rs6068816) was shown to increase the rectal cancer risk. Moreover, haplotype ATGC was associated with a decreased risk of CRC and colon cancer, and haplotypes (ACAC and GCAT) were significant associated an increased CRC risk. Of note, rs4809960 T/C allele was the protective/risk allele for colon and CRC, while rs6022999 G/A allele was the protective/risk allele of colon and CRC. These findings suggested that rs4809960(T > C)-rs6022999(A
7
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Table 7 Meta-analysis of associations between rs6068816 polymorphism and cancer risk. Genetic Model
Heterogeneity test Q
Summary OR (95%CI) P
rs6068816 and cancer risk C vs. T 17.25 0.016 CC vs. TT 13.62 0.058 CC vs. CT 10.07 0.185 CT vs. TT 5.76 0.568 CC vs. CT + TT 14.48 0.043 CC + CT vs. TT 9.97 0.190 rs6068816 and cancer risk in Asian population C vs. T 10.17 0.017 CC vs. TT 7.94 0.047 CC vs. CT 6.40 0.094 CT vs. TT 2.00 0.572 CC vs. CT + TT 9.11 0.028 CC + CT vs. TT 4.65 0.199 rs6068816 and cancer risk by Illumina method C vs. T 2.67 0.102 CC vs. TT 2.49 0.114 CC vs. CT 0.98 0.322 CT vs. TT 1.27 0.260 CC vs. CT + TT 1.98 0.160 CC + CT vs. TT 1.87 0.172 rs6068816 and cancer risk by PCR-RFLP method C vs. T 0.74 0.691 CC vs. TT 0.86 0.650 CC vs. CT 0.08 0.963 CT vs. TT 0.61 0.737 CC vs. CT + TT 0.28 0.869 CC + CT vs. TT 0.83 0.660 rs6068816 and prostate cancer risk C vs. T 0.73 0.693 CC vs. TT 0.29 0.866 CC vs. CT 0.46 0.795 CT vs. TT 0.14 0.930 CC vs. CT + TT 0.79 0.675 CC + CT vs. TT 0.11 0.946
2
I
Hypothesis test
Studies`(n)
Z
P
59.4% 48.6% 30.5% 0.0% 51.7% 29.8%
1.00(0.88-1.13) 1.05(0.85-1.28) 0.99(0.90-1.09) 1.06(0.87-1.31) 0.99(0.86-1.13) 1.05(0.86-1.27)
0.07 0.42 0.29 0.59 0.18 0.45
0.944 0.674 0.774 0.557 0.861 0.655
8 8 8 8 8 8
70.5% 62.2% 53.2% 0.0% 67.1% 35.5%
0.99(0.82-1.20) 0.99(0.67-1.45) 0.10(0.80-1.24) 1.00(0.80-1.26) 0.99(0.77-1.27) 0.99(0.80-1.23)
0.10 0.07 0.03 0.02 0.07 0.09
0.923 0.945 0.973 0.984 0.947 0.927
4 4 4 4 4 4
62.5% 59.9% 0.0% 21.3% 49.4% 46.4%
0.91(0.69-1.21) 0.86(0.37-1.96) 0.94(0.77-1.15) 0.92(0.57-1.48) 0.94(0.77-1.14) 0.84(0.53-1.32)
0.66 0.37 0.58 0.36 0.64 0.75
0.508 0.712 0.560 0.721 0.521 0.453
2 2 2 2 2 2
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
0.94(0.84-1.06) 0.93(0.71-1.23) 0.92(0.79-1.07) 1.00(0.76-1.32) 0.92(0.80-1.07) 0.96(0.74-1.25)
0.99 0.52 1.06 0.02 1.11 0.27
0.322 0.605 0.287 0.983 0.269 0.786
3 3 3 3 3 3
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
0.84(0.72-0.97) 0.66(0.41-1.04) 0.85(0.71-1.01) 0.81(0.51-1.29) 0.83(0.70-0.99) 0.72(0.46-1.15)
2.32 1.78 1.82 0.89 2.09 1.48
0.020 0.075 0.069 0.373 0.036 0.140
3 3 3 3 3 3
> G) might be a susceptible marker for CRC (colon cancer and rectal cancer) in central Chinese population. To solve the discrepancies and improve the statistical strength, we further performed an updated meta-analysis to explore the real impact of CYP24A1 gene polymorphisms on cancer risk. However, none of the four SNPs were shown to be associated with cancer risk in total population, while rs2585428 and rs6022999 were significantly associated with cancer risk in Caucasian population and in Asian population respectly. Interestingly, the pooled results reinforced our current findings for that rs6022999 but not rs2585428 was significantly associated with the risk of CRC (especially colon cancer) in Chinese (Asian) population. Moreover, both the case-control study and meta-analysis showed no association for rs6068816 and cancer risk in Chinese (Asian) population. However, future studies with larger sample size in different ethnic populations are needed to confirm our present findings, and the molecule mechanism of rs2585428 and rs6022999 involving cancer susceptibility should be elucidated in future functional studies. In conclusion, our results indicated that the rs4809960 and rs6022999 may be strongly associated with risk of colon cancer in Chinese population. The updated meta-analysis showed rs2585428 and
rs6022999 may link to cancer susceptibility in Caucasian population and in Asian population respectly, and rs6068816 may not be associated with cancer risk. However, cohort expansion and further mechanistic studies on the role of these CYP24A1 gene polymorphisms that influence carcinogenesis are necessary in future. Declaration of Competing Interest The authors declare that they have no conflict of interest. Acknowledgements This work was supported by grants from the National Natural Science Foundation of China (81502427), the Natural Science Foundation of Hubei Province (2019CFB756), and the Fundamental Research Funds for the Central Universities (WUT: 2018IB023, 2019IB005 and 2019-HS-B1-13).
8
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al.
Fig. 1. Sensitivity analysis of the meta‐analysis. (A) rs2585428 (G vs. A), (B) rs4809960 (T vs. C), (C) rs6022999 (A vs. G) and (D) rs6068816 (C vs. T).
Appendix A. Supplementary data
[9] L.N. Anderson, M. Cotterchio, J.A. Knight, A. Borgida, S. Gallinger, S.P. Cleary, Genetic variants in vitamin d pathway genes and risk of pancreas cancer; results from a population-based case-control study in ontario, Canada, PLoS One 8 (2013) e66768. [10] L.L. Reimers, K.D. Crew, P.T. Bradshaw, R.M. Santella, S.E. Steck, I. Sirosh, M.B. Terry, D.L. Hershman, E. Shane, S. Cremers, E. Dworakowski, S.L. Teitelbaum, A.I. Neugut, M.D. Gammon, Vitamin D-related gene polymorphisms, plasma 25hydroxyvitamin D, and breast cancer risk, Cancer Causes Control 26 (2015) 187–203. [11] J.J. Oh, S.S. Byun, S.E. Lee, S.K. Hong, C.W. Jeong, W.S. Choi, D. Kim, H.J. Kim, S.C. Myung, Genetic variants in the CYP24A1 gene are associated with prostate cancer risk and aggressiveness in a Korean study population, Prostate Cancer Prostatic Dis. 17 (2014) 149–156. [12] T.V. Clendenen, W. Ge, K.L. Koenig, T. Axelsson, M. Liu, Y. Afanasyeva, A. Andersson, A.A. Arslan, Y. Chen, G. Hallmans, P. Lenner, T. Kirchhoff, E. Lundin, R.E. Shore, M. Sund, A. Zeleniuch-Jacquotte, Genetic polymorphisms in vitamin d metabolism and signaling genes and risk of breast Cancer: a nested case-control study, PLoS One 10 (2015) e0140478. [13] L.M. Dong, C.M. Ulrich, L. Hsu, D.J. Duggan, D.S. Benitez, E. White, M.L. Slattery, F.M. Farin, K.W. Makar, C.S. Carlson, B.J. Caan, J.D. Potter, U. Peters, Vitamin D related genes, CYP24A1 and CYP27B1, and colon cancer risk, Cancer Epidemiol. Biomarkers Prev. 18 (2009) 2540–2548. [14] R. Qu, X. Li, X. Quan, L. Xia, X. Fang, H. Li, B. Zhou, Polymorphism in CYP24A1 is associated with lung Cancer risk: a case-control study in chinese female nonsmokers, DNA Cell Biol. 38 (2019) 243–249. [15] F. Bray, J. Ferlay, I. Soerjomataram, R.L. Siegel, L.A. Torre, A. Jemal, Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA Cancer J. Clin. 68 (2018) 394–424. [16] J. Gu, N. Chen, Current status of rectal cancer treatment in China, Colorectal Dis. 15 (2013) 1345–1350. [17] H.C. Horvath, P. Lakatos, J.P. Kosa, K. Bacsi, K. Borka, G. Bises, T. Nittke, P.A. Hershberger, G. Speer, E. Kallay, The candidate oncogene CYP24A1: a potential biomarker for colorectal tumorigenesis, J. Histochem. Cytochem. 58 (2010) 277–285.
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.prp.2019.152735. References [1] S. Kimball, H. Fuleihan Gel, R. Vieth, Vitamin D: a growing perspective, Crit. Rev. Clin. Lab. Sci. 45 (2008) 339–414. [2] J.C. Fleet, M. DeSmet, R. Johnson, Y. Li, Vitamin D and cancer: a review of molecular mechanisms, Biochem. J. 441 (2012) 61–76. [3] G. Jones, D.E. Prosser, M. Kaufmann, 25-Hydroxyvitamin D-24-hydroxylase (CYP24A1): its important role in the degradation of vitamin D, Arch. Biochem. Biophys. 523 (2012) 9–18. [4] A.N. King, D.G. Beer, P.J. Christensen, R.U. Simpson, N. Ramnath, The vitamin D/ CYP24A1 story in cancer, Anticancer Agents Med. Chem. 10 (2010) 213–224. [5] P.E. Thomas, R. Klinger, L.I. Furlong, M. Hofmann-Apitius, C.M. Friedrich, Challenges in the association of human single nucleotide polymorphism mentions with unique database identifiers, BMC Bioinformatics 12 (Suppl 4) (2011) S4. [6] C.N. Holick, J.L. Stanford, E.M. Kwon, E.A. Ostrander, S. Nejentsev, U. Peters, Comprehensive association analysis of the vitamin D pathway genes, VDR, CYP27B1, and CYP24A1, in prostate cancer, Cancer Epidemiol. Biomarkers Prev. 16 (2007) 1990–1999. [7] S.K. Holt, E.M. Kwon, U. Peters, E.A. Ostrander, J.L. Stanford, Vitamin D pathway gene variants and prostate cancer risk, Cancer Epidemiol. Biomarkers Prev. 18 (2009) 1929–1933. [8] S. Yao, G. Zirpoli, D.H. Bovbjerg, L. Jandorf, C.C. Hong, H. Zhao, L.E. Sucheston, L. Tang, M. Roberts, G. Ciupak, W. Davis, H. Hwang, C.S. Johnson, D.L. Trump, S.E. McCann, F. Ademuyiwa, K.S. Pawlish, E.V. Bandera, C.B. Ambrosone, Variants in the vitamin D pathway, serum levels of vitamin D, and estrogen receptor negative breast cancer among African-American women: a case-control study, Breast Cancer Res. 14 (2012) R58.
9
Pathology - Research and Practice xxx (xxxx) xxxx
C. Yi, et al. [18] M. MR, F. J, Meta-analysis of genetic association studies, Trends Genet. 20 (2004) 439–444. [19] M. Zhu, S. Qiu, X. Zhang, Y. Wang, T.D.M. Souraka, X. Wen, C. Liang, J. Tu, The associations between CYP24A1 polymorphisms and cancer susceptibility: a metaanalysis and trial sequential analysis, Pathol. Res. Pract. 214 (2018) 53–63. [20] P. Sedgwick, Multiple significance tests: the Bonferroni correction, BMJ 344 (2012). [21] N. Mantel, W. Haenszel, Statistical aspects of the analysis of data from retrospective studies of disease, J. Natl. Cancer Inst. 22 (1959) 719–748.
[22] R. DerSimonian, N. Laird, Meta-analysis in clinical trials, Control. Clin. Trials 7 (1986) 177–188. [23] Y. Hayashino, Y. Noguchi, T. Fukui, Systematic evaluation and comparison of statistical tests for publication bias, J. Epidemiol. 15 (2005) 235–243. [24] C. International HapMap, A haplotype map of the human genome, Nature 437 (2005) 1299–1320. [25] A. Montpetit, F. Chagnon, [The Haplotype Map of the human genome: a revolution in the genetics of complex diseases], Med. Sci. 22 (2006) 1061–1067.
10