Vitamin D receptor gene polymorphisms and multiple myeloma: a meta-analysis

Vitamin D acts through the vitamin D receptor (VDR), and vitamin D level decreases in multiple myeloma (MM) patients. Single nucleotide polymorphisms in VDR alter its functions to affect the vitamin D status. This raises the question of whether VDR gene polymorphisms are associated with MM risk, which has been investigated in case‒control studies, but the results have been inconsistent. This meta-analysis aimed to investigate the relationship between VDR gene polymorphisms and MM risk. The PubMed, Web of Science, Medline, Embase, Chinese National Knowledge Infrastructure (CNKI), Chinese Scientific Journal (VIP), Wanfang Databases (WANFANG) were searched from inception to June 1, 2023, without language restriction or publication preference. Pooled odds ratio (OR) and 95% confidence interval (CI) for each variable were calculated. Leave-one-out sensitivity analysis was performed to determine the source of heterogeneity. Publication bias was assessed using Begg’ and Egger’s tests, and the trim-and-fill method was used to compensate for publication bias. The correlation meta‐analysis was conducted using Comprehensive Meta‐Analysis 3.0 and STATA 12.0 software. All the included studies were based on Asian populations and involved four VDR gene polymorphisms, TaqI (rs731236), ApaI (rs7975232), BsmI (rs1544410) and FokI (rs2228570). The results showed that TaqI (C vs. T: OR = 1.487, 95% CI 1.052, 2.104, P = 0.025; CC + CT vs. TT: OR = 1.830, 95% CI 1.138, 2.944, P = 0.013), ApaI (T vs. G: OR = 1.292, 95% CI 1.101, 1.517, P = 0.002; TT vs. GG: OR = 1.600, 95% CI 1.106, 2.314, P = 0.013; TG vs. GG: OR 1.305, 95% CI 1.050, 1.622; P = 0.016; TT + TG vs. GG: OR = 1.353, 95% CI 1.103, 1.662, P = 0.004), BsmI (GG vs. AA: OR = 1.918, 95% CI 1.293, 2.844, P = 0.001; GA vs. AA: OR = 1.333, 95% CI 1.058, 1.679, P = 0.015; G vs. A: OR = 1.398, 95% CI 1.180, 1.657, P = 0.000; GG vs. AA + GA: OR = 1.686, 95% CI 1.174, 2.423, P = 0.005), and FokI (T vs. C: OR = 1.687, 95% CI 1.474, 1.931, P = 0.000; TT vs. CC: OR = 2.829, 95% CI 2.066, 3.872, P = 0.000; TC vs. CC: OR = 1.579, 95% CI 1.304, 1.913, P = 0.000, TT + TC vs. CC: OR = 1.771, 95% CI 1.477, 2.125, P = 0.000; TT vs. CC + TC: OR = 2.409, 95% CI 1.814, 3.200, P = 0.000) are associated with MM risk. VDR gene polymorphisms including ApaI, BsmI, TaqI, and FokI are associated with MM risk in Asian populations. Additional studies with large sample sizes and different ethnicities are needed. Supplementary Information The online version contains supplementary material available at 10.1007/s10238-024-01382-4.


Introduction
Multiple myeloma (MM) is a malignancy of clonal plasma cells that manifests clinically as bone pain, pathologic fractures, anemia, and renal insufficiency [1].An increasing number of studies have reported that patients newly diagnosed with MM have low levels of vitamin D [2], and vitamin D deficiency is correlated with myeloma activity [3], the occurrence of peripheral neuropathy (PN) [4,5], and poorer outcomes [6].Evidence from laboratory has shown that analogs of vitamin D inhibit the growth of the myeloma cell line NCI-H929 via cell cycle arrest and apoptosis [7][8][9][10][11], and vitamin D supplements have potential for improving cancer prognosis and outcome in deficient individuals [12].
Vitamin D functions through its vitamin D receptor (VDR), and genetic variations of VDR could influence an individual's vitamin D status [13,14].Polymorphisms caused by single nucleotides at the genome level are described as single nucleotide polymorphisms (SNPs), the most common form of variations among individuals.The most reported VDR gene polymorphisms are the BsmI (rs1544410), FokI (rs2228570), TaqI (rs731236) and ApaI (rs7975232) [15].Studies have explored the association between these VDR gene polymorphisms and cancer development, and revealed that BsmI is associated with overall survival in patients with cancer, that ApaI is associated with progression-free survival in patients with cancer, and that FokI is associated with overall survival in patients with lung cancer [16].In this study, we hypothesized that VDR gene polymorphisms could affect the binding of the VDR and have the potential to influence the risk of MM.MM is a multifactorial disease involving genetics [17], and factors such as the lifestyle and environmental exposures might change the actual effect of SNPs.Additionally, there is genetic heterogeneity in different ethnic populations [18].These are uncertainties and conflicts that underlie the hypotheticals.
In MM, VDR gene polymorphisms have been investigated in several case-control studies [19][20][21][22][23][24], and a meta-analysis indicated that the heterozygote and homozygote models of FokI and the homozygote model of ApaI are associated with an increased risk of MM [18].However, conflicting conclusions have been reported in case-control studies, and the associations of BsmI and TaqI with MM were not explored in previous meta-analysis.Further, significant heterogeneity was observed in the existing meta-analysis.In this study, we focused on the following clinical question: Is there an association between VDR gene polymorphisms and the risk of MM?This evidence will indicate the existence of VDR gene polymorphisms that may increase genetic susceptibility to MM in healthy individuals to promote earlier screening and diagnosis, and will provide evidence for developing therapies against VDR gene polymorphisms for MM patients.Therefore, a meta-analysis was conducted with the aim of examining and summarizing the evidence on the association between VDR gene polymorphisms and MM risk.

Ethics
This study did not require patient recruitment or personal data collection, and there was no need for ethics committee approval.

Protocol registration
This meta-analysis was conducted following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 checklist [25] and Cochrane Handbook criteria, and the protocol was registered with the International Platform of Registered Systematic Review and Meta-analysis Protocols (INPLASY) with registration number INPLASY202330076.The PRISMA 2020 checklist is provided in Supplementary file 1.

Eligibility criteria and study selection
The research question was formulated using Population, Intervention, Comparison, Outcomes, and Study (PICOS) guidelines [26].In the PICOS form, the studies were described as follows: P, patients diagnosed with MM; I, VDR gene polymorphisms; C, healthy control groups; O, effect of VDR gene polymorphisms associated with MM risk; S, case-control study.The inclusion criteria were as follows: (i) case-control study in humans in which the association between VDR polymorphisms and MM risk was investigated; (ii) the case group included patients who meet the diagnostic criteria for MM, and the control group included healthy individuals; (iii) the number of individuals with each genotype in the case and control groups was sufficient to calculate odds ratio (OR) and 95% confidence interval (95% CI); and (iv) the genotype distributions of the control group followed the Hardy-Weinberg equilibrium (HWE).The exclusion criteria were as follows: (i) not a case-control study reporting the association between VDR gene polymorphisms and MM risk; (ii) studies containing data repeated from another published study; (iii) studies containing incomplete data or data that could not be analyzed; (iv) studies with missing or apparently incorrect data; and (v) reviews, case reports, or basic experimental studies.

Literature search and extraction strategy
The PubMed, Web of Science, Medline, Embase, Chinese National Knowledge Infrastructure (CNKI), Chinese Scientific Journal Database (VIP), and Wanfang Database (WANFANG) were searched from their respective inception to June 1, 2023.The retrieval keywords were "multiple myeloma", "vitamin D receptor", and "polymorphism".The retrieval strategy is provided in Supplementary file 2. Two reviewers independently screened and extracted the data.If there were any discrepancies, a third individual reviewed the articles.The data extracted included the following: first author, publication data, sample source, ethnicity, genotyping method, sample size, and genotype characterization for the cases and controls.

Quality evaluation of the included studies
The quality of the studies included in this meta-analysis was evaluated using the Newcastle-Ottawa Scale (NOS) [27].The full score was 9, and high quality was defined as a study with ≥ 7.

Statistical analysis
The data for the meta-analysis and sensitivity analysis were analyzed using Comprehensive Meta-Analysis (CMA, version 3.0).The trim-and-fill method was used to test publication bias using STATA 12.0 (STATA Corporation, Texas, USA).According to whether the heterogeneity was low (I 2 < 50), or high (I 2 ≥ 50%), we used fixed-or randomeffects models, respectively.The OR was used as a summary statistic for dichotomous variables.The 95% CI was calculated for all mean values.P values > 0.05 were considered not significant.
To explore the source of heterogeneity among the studies included in the review, a leave-one-out sensitivity analysis in which a meta-analysis was conducted on each subset of the studies obtained by leaving out exactly one study, was performed.Publication bias was tested using Begg's and Egger's tests.If P < 0.05, publication bias existed, and the trim-and-fill method was used to identify and correct for publication bias.

Characteristics of the included studies
A total of 38 references were identified in the initial examination.After layer-by-layer screening, six case-control studies were included in this analysis.Four case-control studies with 342 cases and 336 controls were included for TaqI (rs731236), four case-control studies with 689 cases and 1222 controls were included for ApaI (rs7975232), four case-control studies with 689 cases and 1147 controls included for BsmI (rs1544410), and six case-control studies with 877 cases and 1414 controls were included for FokI (rs2228570).The flow of the literature screening is detailed in Fig. 1.The included studies involved individuals of Asian ethnicity, four studies included individuals of Chinese Han ethnicity, one study included individuals of Indian ethnicity, and one study included individuals of Kashmiri ethnicity.PCR was the main genotyping method employed.The results of the quality assessment of the literature showed that the scores of the included studies ranged from ranged from 7 to 8 points, indicating that the studies were rated good or better.The basic characteristics of the included studies and the results of the quality evaluation are shown in Table 1.All genotype frequencies of the control group were consistent with the HWE.The genotype data are provided in Supplementary file 3.

Association between BsmI (rs1544410) and MM risk
Six hundred eighty-nine patients with MM and 1147 healthy control subjects were included.A summary of the heterogeneity results, meta-analysis results, and publication bias assessment is shown in Table 6.The results of Begg's and Egger's tests showed no publication bias.Heterogeneity tests revealed significant heterogeneity in the allelic model (G vs. A, I 2 = 75.272%,P = 0.007) and recessive model (GG vs. AA + GA, I 2 = 74.332%,P = 0.009).Primary meta-analysis results showed that the allelic model (OR 1.131, 95% CI 0.632, 2.023; P = 0.679; Fig. 4a) and recessive model (OR 1.158, 95% CI 0.506, 2.653; P = 0.728; Fig. 4e) were independent of MM risk.But heterogeneity was high, so  7.

Discussion
Low level of vitamin D is related to increased risk of several cancers [28].Vitamin D regulates the process of oncogenesis through immunomodulation, antioxidant defense, and DNA damage repair, which affects cancel cell proliferation and apoptosis [29].Previous studies have investigated the relationship between VDR gene polymorphisms and MM [19][20][21][22][23][24].However, the results of different studies are controversial.We therefore performed this meta-analysis to comprehensively evaluate these inconclusive findings.TaqI (rs731236) is located at codon 352 in exon 9 of VDR gene, which creates a TaqI restriction site and related to transcriptional activity, mRNA stability, and the serum level of 1,25(OH) 2 D 3 [30,31].Previous studies have reported that MM patients have a significantly greater frequency of the C allele at the TaqI than healthy controls [19,23], but other studies have reached inconsistent conclusions [21,22].We included 342 patients and 336 controls from four studies.The results from this meta-analysis showed that TaqI is associated with MM in the dominant model and heterozygous model but is independent of MM risk in the allelic model, homozygote model, and recessive model.ApaI (rs7975232) is positioned in intron 8 near the 3′ end of VDR gene and has the potential to affect alternative splicing of the VDR mRNA [32].Four studies did not support the association between ApaI and MM risk [20,21,23,24], but one study reached the opposite conclusion [22].In this meta-analysis, 689 patients and 1222 controls were included.There was heterogeneity in the primary analysis, and the heterogeneity disappeared after the elimination of the study from Syed Shafia [24].Syed Shafia's study including individuals of Kashmiri ethnicity, while the subjects of the remaining four study subjects were of the Chinese Han ethnicity, which suggests that ethnicity might be the source of the heterogeneity.The meta-analysis results showed that   ApaI is correlated with MM risk in the allelic model, dominant model, and homozygote model.Previous studies reported that BsmI (rs1544410) is located at the 3' end of the noncoding region of VDR gene, and enhances the stability and transcriptional activity of VDR gene [30].Previous studies reported that BsmI is significantly associated with MM risk [21][22][23], but one study reported that there is no association [20].This meta-analysis included 689 cases and 1147 healthy controls.Removing the data from Ni Zhai's study [21] reduced the heterogeneity in the allelic model and dominant model, and the meta-analysis indicated a highly significant association between BsmI and the risk of MM in the allelic model, recessive model, homozygote model and heterozygous model.FokI (rs2228570) is located in exon 2 of VDR gene.The association between FokI polymorphism and increased MM risk was described in four studies [19,20,22,24], whereas opposite results were reported in the other two studies [21,23].In this meta-analysis, we pooled data of 877 MM patients and 1414 healthy controls from six studies.The meta-analysis revealed that FokI is associated with MM risk in the allelic model, dominant model, recessive model, homozygote model, and heterozygous model.Collectively, individuals with the T genotype had a significantly greater risk of MM than those with the C genotype.Previous studies have reported that the C to T conversion causes shortening of the generated proteins by three amino acids and the loss of the FokI-recognition site, causing the VDR to be less effective as a transcriptional activator and resulting in a lower vitamin D status, thus, decreasing the anti-cancer effects of vitamin D [33].
In this study, a meta-analysis to evaluate the polymorphisms of VDR gene (FokI, BsmI and ApaI) and MM risk was performed for the first time, and TaqI, FokI, ApaI, and BsmI polymorphisms were found to be associated with MM risk.However, SNPs have geographical and ethnic differences; the results of this meta-analysis may be difficult to extrapolate to non-Asian populations because the included studies examining the association between VDR polymorphisms and MM were limited to Asian participants.The association between VDR gene polymorphisms and MM has not been widely studied, resulting in the relatively small number of included studies.Further highquality studies with multiple ethnicities and larger sample sizes are needed.Although a meta-analysis of VDR gene polymorphisms and MM risk provides valuable information, we must be cautious in how interpret these associations and continue to conduct in-depth research to reveal the potential underlying biological mechanism involved.

Fig. 1
Fig. 1 Flowchart of the literature-screening process

Fig. 2
Fig. 2 Forest plot of the association between TaqI(rs731236) and multiple myeloma for the allelic model (a), homozygote model (b), heterozygous model (c), dominant model (d), and recessive model (e).The adjusted funnel plot of the association between TaqI

Fig. 3
Fig. 3 Forest plot for the association between ApaI (rs7975232) and multiple myeloma for the allelic model (a), homozygote model (b), heterozygous model (c), dominant model (d), and recessive model (e)

Fig. 4
Fig. 4 Forest plots of the association of BsmI (rs1544410) and multiple myeloma for the allelic model (a), homozygote model (b), heterozygous model (c), dominant model (d), and recessive model (e)

Fig. 5
Fig. 5 Forest plots for the association between FokI (rs2228570) and multiple myeloma risk for the allelic model (a), homozygote model (b), heterozygous model (c), dominant model (d), and recessive model (e)

Table 1
Basic characteristics of the included studies NOS Newcastle-Ottawa Scale, PCR polymerase chain reaction, PCR-RFLP polymerase chain reaction-restriction fragment length polymorphism

Table 3
Trim-and-fill analysis of the eligible studies for the metaanalysis of the association between multiple myeloma risk and TaqI (rs731236)

Table 8
Meta-analysis of the association between multiple myeloma risk and FokI (rs2228570)