Association of MTHFR C677T and A1298C Polymorphisms with Glaucoma Risk: a Systematic Review Meta-Analysis based 42 Case-Control Studies.

Aim: Several epidemiological studies have been performed to explore the association of MTHFR polymorphisms with glaucoma risk. However, the results were inconsistent or even inconclusive. Hence, we performed a meta-analysis to evaluate the association of MTHFR C677T and A1298C polymorphisms with glaucoma risk. Methods: A comprehensive literature search on PubMed, Google Scholar, EMBASE, and CNKI databases was performed to find all eligible studies up to January 30, 2019. The pooled odds ratios (ORs) with 95% confidence intervals (CIs) were used to assess the strength of such association. Results: A total of 42 case-control studies including 33 studies for MTHFR C677T and nine studies for A1298C polymorphism were selected. Pooled results showed that there was no significant association between the MTHFR C677T polymorphism and glaucoma risk. Similarly, no associations were found in subgroup analysis based on ethnicity and glaucoma type. However, there was a significant association between the A1298C polymorphism and the increased risk of glaucoma under heterozygote model (OR=0.765, 95% CI=0.626-0.935, P=0.009). Moreover, the significant association between MTHFR A1298C polymorphism and glaucoma were found by ethnicity and primary open angle glaucoma (POAG). Conclusions: The present meta-analysis revealed that MTHFR A1298C polymorphism is significantly associated with the increased risk of glaucoma, but not MTHFR C677T polymorphism.


Introduction
Glaucoma is an optic neuropathy in which the optic nerve is damaged with typical loss of nerve fibers and increasing cupping of the optic disc, leading to progressive, irreversible loss of vision [1,2]. A leading cause of all blindness worldwide, secondary to cataracts, glaucoma is the main cause of irreversible vision loss [3]. It is estimated that more than 60 million people had Romanian Society of Ophthalmology © 2019 glaucoma in 2010, 8.4 million of whom are bilaterally blind as a result of this disease [4]. In general, glaucoma might be classified in three major categories: primary open angle glaucoma (POAG), primary congenital glaucoma (PCG) and primary angle-closure glaucoma (PACG) [5]. Glaucoma is a multifactorial disease involving both environmental and genetic factors [6,7]. During the past decade, molecular genetic studies of glaucoma have yielded some success. The importance of genetic factors in the etiology of glaucoma is supported by genome-wide association studies (GWASs) [8]. Recently, several candidate novel loci have been identified in a GWAS for POAG (e.g., ABCA1, AFAP1, GMDS, PMM2, TGFBR3, FNDC3B, ARHGEF12, GAS7, FOXC1, ATXN2, TXNRD2); PACG (e.g., EPDR1, CHAT, GLIS3, FERMT2, DPM2-FAM102); and exfoliation syndrome (XFS) glaucoma (CACNA1A) [8,9]. Furthermore, several epidemiological studies have reported a link between methylenetetrahydrofolate reductase (MTHFR) gene polymorphisms and glaucoma [10,11].
The MTHFR gene is located on chromosome 1p36.3 [12,13]. It is an important regulatory enzyme in the folate related one carbon metabolism, which is responsible for catalyzing 5, 10-methylenetetrahydrofolate to 5methyltetrahydrofolate [14,15]. In addition, MTHFR plays an important role by directing folate metabolites through the DNA methylation pathways [12]. An increased level of plasma homocysteine (Hcy) has been observed in patients with glaucoma [16]. The MTHFR gene is encoded by 11 exons and includes several SNP, some of which have functional relevance and result in high Hcy level. Many studies have shown an increased risk of glaucoma in patients with MTHFR C677T and A1298C polymorphism. However, results from these studies were inconsistent or inconclusive. It was suggested that this inconsistency might be related to the single studies with low statistical power, publication biases, and ethnicity differences. Thus, we have performed the current systematic review and meta-analysis to collecting and summarizing the evidence on the association of MTHFR C677T and A1298C polymorphisms with the risk of glaucoma.

Study Identification and Selection
We have performed a comprehensive literature search using PubMed, Web of Science, Google Scholar, Cochrane Library, Embase, and Chinese Biomedical Literature database (CBM) databases to identify studies that evaluated the association between MTHFR C677T polymorphism and the risk of glaucoma up to October 2018, with the following keywords: "Methylenetetrahydrofolate reductase", "MTHFR", "MTHFR C677T", or "MTHFR A1298C" and "polymorphism", "mutation", or "variant" and "glaucoma" and "primary open-angle glaucoma" or "POAG" and "pseudoexfoliation glaucoma" or "PXFG", "pseudoexfoliation syndrome with glaucoma" or "PEXG" and "normal-tension glaucoma" or "NTG" and "primary angle-closure glaucoma" or "PACG" and "primary angle-closure glaucoma" or "PACG", "high-tension glaucoma" or "HTG", and "juvenileonset open-angle glaucoma" or "JOAG". We have retrieved any article matching the keywords and we evaluated it by reading the title and abstract. In addition, we have screened the references lists of the retrieved articles for original papers.

Inclusion and Exclusion Criteria
The following criteria were used for the study selection: 1) a case-control study evaluating the association of MTHFR C677T and A1298C polymorphisms with the risk of glaucoma and its types; 2) case-control or cohort studies; 3) sufficient data for estimating an odds ratio (OR) with 95% confidence interval (CI); 4) no overlapping data. In addition, if studies had the same or overlapping data, we have included only the largest study in the final analysis. The major excluding criteria for studies were the following: (1) not glaucoma research, (2) reviews, letters or case reports, (3) duplicate of previous publication, and (4) and those articles without definite information of genotypes.

Data Extraction
We have extracted information carefully from all the eligible studies independently by two investigators based on the above listed inclusion criteria. The following data were collected from each study: the first author's Romanian Society of Ophthalmology © 2019 name, the year of publication, ethnicity, country of origin, glaucoma type, genotyping method, source of control groups (population-based or hospital-based controls), total number of cases and controls, the frequencies of genotypes, minor allele frequencies (MAFs), and Hardy-Weinberg equilibrium (HWE) test in control subjects. Allele frequencies were calculated from the corresponding genotype distributions using an online website. Finally, the extracted data in terms of accuracy and any discrepancy between these two authors was resolved by reaching a consensus through discussion or the involvement of a third author who made the final decision through discussions.

Statistical Analysis
Pooled odds ratios (ORs) and corresponding 95% confidence intervals (CIs) were calculated to assess the association of MTHFR C677T and A1298C polymorphisms with the risk of glaucoma. The significance of the pooled OR was determined by the Z-test. The pooled ORs were performed under five genetic models, i.e., allele (B vs. A), homozygote (BB vs. AA), heterozygote (BA vs. AA), dominant (BB+BA vs. AA), and recessive (BB vs. BA+AA), which a "A" denotes a major allele; "B" denotes a minor allele.
Heterogeneity (between-study inconsistency) was assessed by the Cochran Χ 2based Q test (Heterogeneity was considered statistically significant if P<0.10) and the I 2 statistics. An I 2 value of 0% represents no heterogeneity, with values of 25%, 50%, 75%, or more represent low, moderate, high, and extreme heterogeneity, respectively. A fixed effect model (Mantel-Haenszel method) was used to calculate pooled OR when there was no heterogeneity among the studies. Otherwise, the fixed-effects model (Mantel-Haenszel approach) was used. We have calculated the Hardy-Weinberg equilibriums (HWEs) with goodnessof-fit tests (i.e., chi-square or Fisher's exact tests). In addition, one-way sensitivity analyses were carried out by consecutively omitting one study at a time to assess power of the metaanalysis [15]. In addition, sensitivity analysis was also performed, excluding studies whose allele frequencies in controls exhibited a significant deviation from the Hardy-Weinberg equilibrium (HWE), given that the deviation may denote bias. Deviation of HWE may reflect methodological problems such as genotyping errors, population stratification or selection bias. Visual inspection of the asymmetry of funnel plots was carried out to assess potential publication bias. Begg's funnel plot, a scatter plot of effect against a measure of study size was used as a visual aid to detect bias or systematic heterogeneity. Publication bias was assessed by Egger's test (p<0.05 was considered statistically significant). If publication bias existed, the Duval and Tweedie non-parametric "trim and fill" method was used to adjust for it. A metaregression analysis was carried out to identify the major sources of between-studies variation in the results, using the log of the ORs from each study as dependent variables, and ethnicity and source of controls as the possible sources of heterogeneity. All the statistical calculations were performed using Comprehensive Meta-Analysis (CMA) software version 2.0 (Biostat, USA). Two-sided P-values < 0.05 were considered statistically significant.
The main characteristics of studies included in the current meta-analysis are presented in Tables 1 and 2. Among these studies, six types of glaucoma, including primary open angle glaucoma (POAG), pseudoexfoliation glaucoma (PXFG) or pseudoexfoliation syndrome with glaucoma (PEXG), normal-tension glaucoma (NTG), primary angle-closure glaucoma (PACG), hightension glaucoma (HTG), and juvenile-onset open-angle glaucoma (JOAG) were involved. Among the selected studies, 23 case-control studies were conducted in the Asians, 18 studies were conducted in the Caucasians, and one study was conducted in the Latinos. Genotyping methods used in the studies include PCR-RFLP, Real-time PCR, TaqMan, and sequencing. The genotype frequencies in the control group for three publications did not fit well in the Hardy-Weinberg equilibrium (P>0.05). Table 3 listed the main results of the metaanalysis of MTHFR C677T polymorphism and glaucoma risk. After the 33 case-control studies were pooled into meta-analysis, no evidence of a significant association between MTHFR C677T polymorphism and glaucoma risk was observed under all genetic models (T vs. C: OR = 1.120. 95% CI 0.994-1.262, P = 0.062, Fig. 2A   In the subgroup analysis by glaucoma type, no significant associations with POAG, PACG, PXFG, and NTG subgroups were observed. Moreover, no significant association was found in a subgroup analysis by ethnicity among Asian and Caucasian populations ( Table 3). The studies were further stratified based on genotyping technique, source of control subjects and HWE. In the PCR-RFLP group, significantly increased association between MTHFR C677T polymorphism and glaucoma risk were found in the recessive model (TT vs. TC+CC: OR = 1.438. 95% CI 1.056-1.958, P = 0.021). The population based subgroup analysis also revealed that the presence of the MTHFR C677T, which was related to a higher risk of glaucoma under the heterozygote model (TT vs. TC: OR = 1.350, 95% CI 1.012-1.802, P = 0.041). Subgroup analysis of studies in accordance with HWE showed that there was a significant association between MTHFR C677T polymorphism and the increased risk of glaucoma under the allele model (OR = 1.156, 95% CI 1.020-1.309, p = 0.023) (data not shown). Table 4 listed the main results of the metaanalysis of MTHFR A1298C polymorphism and glaucoma risk. When all the eligible studies were pooled into the meta-analysis of MTHFR A1298C polymorphism, significantly increased risk of glaucoma was observed in the heterozygote model (CA vs. AA: OR = 0.765, 95% CI 0.626-0.935, p = 0.009, Fig. 2B).

Heterogeneity Test and Sensitivity Analyses
There was a significant heterogeneity among these studies for MTHFR C677T polymorphism under allele model comparison (T vs. C: Ph = ≤ 0.001), homozygote model comparison (TT vs. CC: Ph = 0.005) and dominant model comparison (TT + CT vs. CC: Ph = 0.001). Then, we assessed the source of heterogeneity by meta-regression analysis. However, we found that ethnicity, glaucoma types, genotyping methods, source of controls and HWE did not contribute to substantial heterogeneity among the meta-analysis ( Table 2). Sensitivity analyses were conducted to determine whether modification of the inclusion criteria of the current meta-analysis affected the findings. Although the sample size for cases and controls in all eligible studies ranged from 18 to 243, the pooled ORs were not qualitatively altered by omitting the study of small sample. Three studies were not in HWE; however, the overall association was unchanged after the exclusion of these studies, which indicated that the results from this meta-analysis were statistically robust. Moreover, the heterogeneity test showed that there was no significant between-study heterogeneity in terms of the MTHFR A1298C polymorphism in the overall comparisons and subgroup analyses ( Table 3).

Publication Bias
We have used both Begg's funnel plot and Egger's test to access the small study effects of articles in literature. The shape of the funnel plots did not reveal an obvious asymmetry. Then, the Egger's test was used to provide statistical evidence of funnel plot symmetry. Egger's test found evidence for the publication bias between MTHFR C677T polymorphism and glaucoma risk under the allele model (T vs. C: PBegg = 0.052, PEgger = 0.031, Fig. 3), homozygote model (TT vs. CC: PBegg = 0.010, PEgger = 0.008) and the recessive model (TT vs. CT + CC: PBegg = 0.022, PEgger = 0.005). This finding might be a limitation for this meta-analysis because studies with null findings, especially those with small sample size, are less likely to be published. The Duval and Tweedie non-parametric "trim and fill" method was used to adjust for publication bias. Meta-analysis with and without "trim and fill" did not draw a different conclusion, indicating that our results were statistically robust. Moreover, no significant publication bias for MTHFR A1298C polymorphism was found by Egger's test in the overall or subgroup analyses.

Discussion
To the best of our knowledge, this is the first and most comprehensive meta-analysis assessing the associations of MTHFR C677T and A1298C polymorphisms with risk of different types of glaucoma. A total of 33 case-control studies in 19 publications (3,504 cases and 2,525 controls) and nine case-control studies in six publications (1,073 cases and 775 controls) have investigated the associations of MTHFR C677T and A1298C polymorphisms with glaucoma risk, respectively. Our meta-analysis showed that MTHFR C677T polymorphism was not associated with glaucoma risk. Similar results were observed in the subgroup analyses based on ethnicity and types of glaucoma (POAG, PACG, PEXG, and NTG). However, we have found that the MTHFR A1298C may be associated with an increased glaucoma risk overall and by ethnicity. Moreover, in a subgroup analysis of glaucoma types, MTHFR A1298C polymorphism was significantly associated with an increased risk of POAG, but not with PXFG and NTG subgroups.
Interestingly, stratified analysis according to genotyping technique revealed a significantly increased risk of glaucoma in participants with the C677T polymorphism in those studies involving PCR-RFLP under recessive genetic model (TT vs. TC+CC: OR = 1.438, 95% CI 1.056-1.958, P = 0.021). With the recent advent of sophisticated high-throughput genotyping technologies such as semi nested PCR, the TaqMan allelic discrimination test, or real-time PCR, we may witness a significant progress in the association studies in the future [32]. High sensitivity of real-time PCR makes the technique applicable to very small samples [33]. However, this trend is possible because studies involving Caucasians mainly utilized Real-Time PCR. While, in studies involving Asians, PCR-RFLP was the main genotyping technique. We proposed that the sensitivity and specificity of genotyping techniques are further explored to seek out optimal approaches that could minimize the genotyping errors. Therefore, this result should be carefully interpreted and confirmed by conducting a further analysis of additional published studies. Moreover, the population based subgroup analysis also revealed that the presence of the MTHFR C677T was related to a between MTHFR C677T polymorphism and POAG in allelic genetic model and additive genetic model for population-based subgroup, which indicated that the T allele or TT genotype might increase the risk of POAG [34].
Pathogenesis of POAG is a complex process. It is known that genetic factors play an important role in POAG susceptibility [35]. However, most of the molecular mechanisms leading to POAG development are still unknown [36]. It seems that approximately 5% of POAG is currently attributed to a single-gene or Mendelian forms of glaucoma. Gene mutations in various loci have been identified by genetic studies and a genetic basis for glaucoma pathogenesis has been established [18,37]. Although many epidemiological studies have been conducted to assess the roles of MTHFR C677T polymorphism and POAG risk in different populations, results have been inconclusive. Recently, in a case-control study of 144 POAG cases and 280 controls in Saudi Arabia, Al-Sharani et al. indicated that the allele T and genotype CT of MTHFR C677T polymorphism confer risk of POAG, while allele C and CC genotype had a different role [30]. However, four studies did not find an association between MTHFR C677T polymorphism and POAG risk in Iranian, Mexican, Indian and Greek populations [25,[27][28][29]. In 2012, Xu et al. have conducted the first meta-analysis including ten studies with 1,406 cases and 1,216 controls on MTHFR C677T polymorphism [38]. They found no impact of MTHFR C677T polymorphism on POAG susceptibility in the pooled analysis. Since then, a series of better-designed case-control studies on the association between MTHFR C677T polymorphism and POAG were performed. In the current meta-analysis, 16 eligible studies with 2,179 cases and 2,069 controls were identified and analyzed. The present meta-analysis suggested that there was no significant association between MTHFR C677T and POAG risk in the overall comparisons. Consistent with our study, a previous meta-analysis was undergone in 2015, which included 13 studies with 1,970 POAG patients and 1,712 control subjects, suggesting that the MTHFR C677T was Romanian Society of Ophthalmology © 2019 not associated with increased genetic susceptibility to POAG [39]. However, we found out they wrongly included one study evaluated about the MTHFR C677T polymorphism and PACG risk in their meta-analysis. Our literature search was more thorough, containing four more articles, which increased the total number of cases and controls, thus, providing a greater power to our conclusions. Moreover, we used one more genetic model, the allele genetic model, to gain a more comprehensive and accurate understanding of the MTHFR C677T polymorphism association.
Assessing heterogeneity in the metaanalysis of genetic associations is critical for model selection and interpretation of the results. On the other hand, heterogeneity and publication bias might influence the results of the metaanalysis. It is well known that different factors, such as population stratification, source of controls, population size, deviation from Hardy-Weinberg equilibrium, and other covariates could be the source of heterogeneity. In the current meta-analysis, moderate between-study heterogeneity was detected across studies under allele, heterozygote and dominant genetic models for MTHFR C677T polymorphism and thus we selected the random-effects model to summarize the ORs. Therefore, we performed a meta-regression analysis to find the source of between-study heterogeneity. The results showed that ethnicity, glaucoma types, genotyping methods, source of controls and HWE status did not contribute to substantial between-study heterogeneity in the current meta-analysis.
It was obvious that some limitations of this meta-analysis should be considered. First, the sample size reported in literature is still relatively small and might not provide sufficient power to estimate the association between the null MTHFR A1298C polymorphism and the glaucoma risk. Second, the language of the publications was limited to English. Third, the current meta-analysis was based predominantly on Asian and Caucasian research. No study from other parts of the world was found, such as the Africans. This suggested a partial result that is only relevant to the Asian and Caucasian subgroups. Forth, the existence of between-study heterogeneity in some comparisons might compromise the reliability of conclusion. Finally, glaucoma is a multifactorial disease that results from complex interactions between various genetic and environmental factors. Due to the unavailability of other detailed information, our results were based on single-factor estimates without adjustments for other risk factors. Further evaluation of glaucoma risk should pay more attention to the potential interactions among gene-gene, gene-environment, and even different polymorphism of the MTHFR gene and other loci. Despite these limitations, our metaanalysis had some clear advantages. Our metaanalysis contained the largest sample size to date to assess the association between the MTHFR C677T and A1298C polymorphisms and glaucoma risk.
In summary, the current meta-analysis indicated that MTHFR C677T might not be associated with the glaucoma risk, and yet the MTHFR A1298C polymorphism may be a risk factor for glaucoma. In the future, large sample studies should be warranted to investigate the association of MTHFR C677T and A1298C polymorphisms with glaucoma, and to examine the potential gene-gene and gene-environment interactions.

Funding
No specific funding was obtained to support the conduct of this study.