Tumor necrosis factor-α (TNF-α) -308G >a promoter polymorphism (rs1800629) promotes Asians in susceptibility to Plasmodium falciparum severe malaria: A meta-analysis

The multifactorial pathogenesis of severe malaria is partly attributed to host genes, such as those encoding cytokines involved in complex inflammatory reactions, namely tumor necrosis factor-alpha (TNF-α). However, the relationship between TNF-α -308G >A gene polymorphism (rs1800629) and the severity of Plasmodium falciparum (P. falciparum) malaria remains unclear, which prompts a meta-analysis to obtain more precise estimates. The present meta-analysis aimed to better understand this correlation and provide insight into its association in populations with different ethnicities. Literature search outcomes included eight eligible articles in which TNF-α -308G >A polymorphism was determined in uncomplicated malaria (UM) and severe malaria (SM) of P. falciparum as represented in the case and control groups. Pooled odds ratios (ORs) and 95% confidence intervals (95% CIs) were estimated in standard homozygous, recessive, dominant, and codominant genetic models. Subgroup analysis was based on ethnicity, i.e., Africans and Asians. The analyses included overall and the modified outcomes; the latter was obtained without the studies that deviated from the Hardy-Weinberg Equilibrium. The significant data also underwent sensitivity treatment but not publication bias tests because the number of studies was less than ten. Interaction tests were applied to differential outcomes between the subgroups. Overall and HWE-compliant analyses showed no significant association between the TNF-α -308G >A polymorphism and susceptibility to P. falciparum SM (ORs = 1.10–1.52, 95%CIs = 0.68–2.79; Pa = 0.24–0.68). Stratification by ethnicity revealed that two significant associations were found only in the Asians favoring SM for dominant (OR = 1.95, 95% CI = 1.06–3.61, Pa = 0.03) and codominant (OR = 1.83, 95% CI = 1.15–2.92, Pa = 0.01) under the random-effects model, but not among the African populations. The two significant Asian associations were improved with a test of interaction with P-value of0.02–0.03. The significant core outcomes were robust. Results of the meta-analysis suggest that TNF-α -308G >A polymorphism might affect the risk of P. falciparum SM, particularly in individuals of Asian descent. This supports ethnicity as one of the dependent factors of the TNF-α -308G >A association with the clinical severity of malaria. Further large and well-designed genetic studies are needed to confirm this conclusion.


Introduction
Malaria is one of the significant parasitic diseases.Host genetic and environmental components during malaria infection significantly contribute to morbidity and mortality in tropical and subtropical regions around the globe [1,2].Among the five species of plasmodial parasites in humans, Plasmodium falciparum infection is the main causative parasite of severe disease and malaria-related deaths [3,4].The spectrum of clinical severity of symptoms can range from asymptomatic, uncomplicated, to severe malaria [4,5].According to the recent World Health Organization Malaria Guidelines, uncomplicated malaria is defined as a patient who has a febrile illness and a positive parasitological test (microscopy or rapid diagnostic test) without features of severe malaria.In contrast, severe malaria involves the central nervous system (cerebral malaria), pulmonary system (respiratory failure), renal system (acute renal failure) and/or hematopoietic system (severe anaemia) [5,6].There is evidence showing clinical manifestations of malaria and fatality depends on the host's immune status, which is mainly orchestrated by inflammatory cytokine tumour necrosis factor-alpha (TNF-α) [7,8].In patients with malaria, the cytokine response was found to be responsible for elevated fever symptoms following the onset of blood infection and clinical feature [9].TNF-α levels has been found to be positively correlated with the severity of malaria [10][11][12].
TNF-α is produced by macrophages, neutrophils, activated lymphocytes, and natural killer cells [2].The major role of TNF-α in inflammatory responses is a yardstick of patients' reactivity towards malaria infection [13].Gene variation can regulate gene transcription, protein expression, and thus the biological function of these cytokines.The TNF-α gene lies within a region of ~7 kb in the center of the primary histocompatibility complex locus on the short arm of human chromosome 6 [14].Most extensive studies investigated the correlation between the PLOS NEGLECTED TROPICAL DISEASES TNF-α gene polymorphism in the -308 promoter region and disease status [15][16][17][18][19].A point substitution at position -308 (in relation to the TNFα transcription start site) of the TNF-α gene [20] defines the TNFα alleles *1 (G at -308) and *2 (A at -308).This polymorphism (rs1800629) appears to be important in TNF-α production [21][22][23].The circulating TNF-α level was found to be higher in TNF-α rs1800629A homozygous (AA) compared to TNFα rs1800629G homozygous (GG) individuals [24].Moreover, the activation potential of the TNF-α rs1800629A allele is more significant than TNF-α rs1800629G allele in producing TNFα [25].
Several case-control groups among primary studies have examined the association between rs1800629 polymorphism and disease severity.However, the results of these studies [26][27][28][29][30][31][32][33] are conflicting, probably because of small sample sizes and low statistical power.Although recent meta-analysis [34] reported a significant association between TNF-α -308G >A polymorphism and predisposition to severe P. falciparum malaria in the allele and dominant model, there were some obvious limitations in their meta-analytic procedure.First, the subjects in the control group consisted of heterogenous phenotypes, including asymptomatic malaria [35], uncomplicated malaria, and other illnesses [36].Second, additional five component studies were not included in their meta-analysis even genotype frequencies could be obtained.Furthermore, subgroup analysis was not performed in the different populations.Based on the above reasons, the present meta-analysis on the association of TNF-α -308G >A and severe malaria was performed to obtain a more precise estimate.Meta-analysis is helpful for detecting associations that may be obscured in studies of limited sample sizes, particularly in those evaluating rare allele frequency polymorphisms.The current study aimed to update the meta-analysis of the published studies to ascertain whether the polymorphisms of TNF-α -308G >A increases susceptibility to severe P. falciparum malaria and assess their possible association between ethnically diverse populations.

Materials and methods
Case-control definition Patients with severe P. falciparum malaria were included as the study populations (cases).Intervention or exposure was not applied in this case.The control group included patients who manifested uncomplicated malaria's clinical symptoms as described by the World Health Organization [6].

Publication search and selection of studies
This meta-analysis was performed by following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [37].We searched the electronic MEDLINE database using PubMed, Scopus, Google Scholar, and MedNar (deep web search engine) for association studies as of April 1, 2023.The search terms applied were "Tumor necrosis factoralpha", "TNF-α -308", "TNF-α-308G/A", "polymorphism", "malaria", and "rs1800629", with no restriction of language.The cited references of full-text and review articles were manually scanned to increase additional studies.Inclusion criteria were as follows: (i) original case-control design that examined the associations between TNF-α-308G >A polymorphism and risk of uncomplicated or severe P. falciparum malaria; and (ii) provision of genotype frequency data that allowed calculation of odds ratios (ORs) and 95% confidence intervals (CIs).Exclusion criteria were (i) review or systematic review articles; (ii) studies that did not cover the polymorphisms TNF-α -308G >A or disease in question; (iii) non-human subjects; (iv) studies without controls or studies whose genotype or allele frequencies were unusable or absent; (v) not uncomplicated malaria control type; (vi) not P. falciparum malaria; and (vii) articles whose genotype data originated from the same group of the population (duplicated genotype data).

Data extraction and study quality
Two investigators (PK and NP) independently extracted data.Disagreements were adjudicated by a third investigator (PT) and arrived at a consensus.Key information from the selected articles included: the first author's last name, year of publication, country of origin, ethnicity, age of case and control, genotyping method, and malaria type of case and control.Quantitative features from the articles were also extracted.These included numbers of cases and controls, the sum of which generated the sample size, statistical power, genotype frequencies (GG, GA, and AA), and those of the minor allele and the HWE.Using the application in https://genecalc.pl/hardy-weinberg-page., the HWE was assessed, and the P-value of the controls from Pearson's goodness-of-fit χ 2 -square test was reported.The potential impact of HWE-noncompliant studies was examined to ascertain genotyping quality.The Clark-Baudouin scale was used to assess the methodological quality of the included studies [38].Criteria for the evaluation included P-values, statistical power, multiple-comparison correction, comparative sample sizes between cases and controls, genotyping methods, and the HWE.On this scale, the scores of < 5, 5-6, and � 7 were considered low, moderate and high quality, respectively.

Data distribution and statistical power
The distribution of the data was assessed with the Shapiro-Wilks test [39] using SPSS 20.0 (IBM Corp., Armonk, NY, USA).Gaussian (normal) distribution (P > 0.05) warranted descriptive and inferential expressions of the mean ± standard deviation (SD) as well as the parametric approach.Otherwise, the median with interquartile range (IQR) and non-parametric tests were used.The statistical power was assessed using the statistical software G*Power [40], assuming an OR of 1.5 with a genotypic risk level of α = 0.05 (two-sided), where power was considered adequate at � 80%.With complete linkage disequilibrium for all the TNF-α polymorphisms, only the rs1800629 (-308G>A) polymorphism was focused, which best represented all included articles.

Meta-analysis
The risk for severe P. falciparum malaria (using raw data for frequencies) was estimated for each study, and comparing the effects on the same baseline.Pooled ORs and 95% Cis were calculated with a significance threshold of P � 0.05 (two-tailed).TNF-α associations with OR were estimated for each study.The presence of zero genotype values warranted applying the Laplace correction, which involves adding a pseudo-count of one to all data set values [41].This method could be used in the current study because MAF of all included studies are approximately less than 2%, indicating a low prevalence of mutation probability.Pooled ORs with 95% CIs were calculated for the following genetic models: (a) homozygous: AA and GG genotypes compared with GG; (b) recessive: AA versus GA + GG; (c) dominant: AA + GA versus GG; and (d) codominant: A versus G. Heterogeneity of the study was estimated using χ 2based Q test, where significance was set at P � 0.10 [42] and quantified with the measure of variability (I 2 ) statistic [43].Either the presence of heterogeneity or not warranted using random and fixed-effects models of analyses, respectively [44,45].
Subgroup analysis was performed with respect to ethnicity, including Africans and Asians.The probability of differential risk associations between comparisons warranted testing for the presence of interactions [46].Sensitivity analysis, which involves omitting one study at a time and recalculating the pooled OR, was used to test for the robustness of the summary effects.Tests for publication bias were used to evaluate the influence of small-study specific effects [47] and were applied on comparisons that met two conditions: (i) associatively significant (P a < 0.05), and (ii) � 10 studies [48].

Characteristics of the included studies
The flow chart of the study selection was outlined in Fig 1 according to PRISMA guidelines.The initial search of four databases and four search strings yielded 13,087 citations, which were further reduced to 115 after reviewing titles and abstracts and removing duplicates (S1 Table ).After full-text evaluation, 107 articles were excluded because of not conforming to the inclusion criteria.A manual search of the cited reference list yielded no additional articles.Table 1 lists the eight case-control studies included in the meta-analysis [26][27][28][29][30][31][32][33].Of them, five [26,27,29,32,33] were not included (new) in a recent meta-analysis [34].The year range of the articles was 1994-2019 and comprised 1,448 severe P. falciparum malaria cases and 1,371 uncomplicated malaria control.Patients in three studies were Africans (843 cases/718 controls), and five studies were Asians (605 cases/653 controls).The age of the subjects was heterogeneous (children to adults).The mean and SD values of the normally distributed Clark-Baudouin scores (Shapiro-Wilks: P = 0.139) were 6.38 ± 1.69, with most (62.5%) of the articles scoring � 7.These values indicate the high methodological quality of the included studies.
Table 2 shows the quantitative traits of the included studies.Sample sizes ranged from 67 to 819 and the statistical powers ranged from 12.4 to 80.1%, in which only one study [26] had power of more than 80%.The mean and SD of the minor allele frequency in Africans (0.13 ± 0.02) were not significantly different (Unpaired t-test, P = 0.73) from those in Asians (0.11 ± 0.04).The control group in one article (12.5%) [33] showed a significant deviation from HWE.The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines checklist provides detailed description of this meta-analysis (S3 Table ).

Tests of interaction
Interaction tests were calculated on the significant ethnic subgroup outcomes.Table 4 shows that of 2 comparisons subjected to these tests, the two non-significant African effects (dominant: OR 0.72, P a = 0.30; codominant: OR 0.73, P a = 0.34) compared with that of the two significant Asian effects (dominant: OR 1.95, P a = 0.03; codominant: OR 1.83, P a = 0.01) resulted in significant interaction (dominant: P interaction = 0.023, codominant: P interaction =

Sensitivity analysis and publication bias
Sensitivity analysis was performed using a modified protocol that confined this treatment to the significant findings.Pooled effects that retained (P < 0.05) or lost (P > 0.05) significance were considered robust and not robust, respectively.The results indicate the robustness of two significant outcomes (Table 3).Since there was a limited number of studies (< 10), a test for publication could not be conducted in this meta-analysis.

Discussion
Examination of the association between -308G >A (rs1800629) and the risk of severe P. falciparum malaria relied on the literature's usable data, which paved the way to use the arsenal of meta-analytical processes, including modifier (HWE-compliant), subgrouping, sensitivity, and test of interaction.This is the second meta-analysis to address the association between TNF-α -308G >A polymorphisms and the risk of severe P. falciparum malaria and the first metaanalysis to unravel this association in different populations.Contrary to the previous metaanalysis [34], the study failed to reveal an association between TNF-α -308G >A polymorphism and susceptibility to severe P. falciparum malaria in the overall analysis under all genetic models.Moreover, the analyses (excluding the HWE-deviated study) further validated the non-significant association.Interestingly, after stratification by ethnicity, significant associations were found in dominant and codominant carriers in Asians favoring SM, which is the principle finding of the current meta-analysis.Also, both important outcomes showed statistically significant subgroup differences by interaction test as well as the stability of the outcomes.

TNF-α -308G >A and malaria
The diversity of clinical manifestations of malaria reflects a diverse pathophysiological mechanism depending on factors such as the degree of infection and host immunity [49].TNF-α is a crucial proinflammatory cytokine in the pathogenesis and control of P. falciparum infection https://doi.org/10.1371/journal.pntd.0011735.g002[30].Physiologically, the inappropriate or excessive production of TNF-α can contribute to severity and mortality [50,51].The underlying mechanism can be involved in toxicity and inflammatory processes such as promoting extravasation of neutrophils, lymphocytes, and monocyte to endothelial cells, affecting immune responses by controlling T cell activation.These result in inducing the synthesis of numerous pro-inflammatory cytokines and apoptosis of different cell types, increasing malaria severity as well as other complications [50,51].The rare allele (A) for TNF-α (-308G >A) polymorphism produces a higher level of TNF-αmessenger ribonucleic acid and thus, a high TNF-α phenotype, compared to that of major allele G [23].Therefore, the blood TNF-α level based on the genetic background may partly contributable to the intensity of the host immune response.Previous meta-analysis indicated that patients with UM had higher TNF-α levels than those with asymptomatic malaria [52].However, there was no difference in TNF-α levels between fatal and non-fatal cases, suggesting that TNF-α levels are not associated with mortality.In turn, the diverse genetic background likely impacts the progression of clinical severity.Moreover, the severity of malaria symptoms may involve the contribution of multi-genes encoding vital host proteins.Previous investigations reported interactions between the TNF-α and other genes [53,54], suggesting that multiple functional polymorphism's effects seem more vital than a single polymorphism's effects.Of note, half of the included articles in the meta-analysis addressed haplotype analysis [27,28,30,32].
The subgroup analysis of the current study showed that Asians significantly favor SM.Meanwhile, Africans favor UM, although there was non-significance in statistics.The finding of different associations is somewhat surprising.A possible explanation for TNF-α -308G >A playing a different role in different ethnic could be a result of genetic and clinical heterogeneity between the different populations.Whole-genome linkage studies on severe P. falciparum malaria have shown genetic heterogeneity [55].Although the frequency of TNFα -308G >A in Asian populations such as Thailand, Myanmar, Sri Lanka, India, and Vietnam was found to be low (0.06-2.00%), the association between TNFα -308G >A polymorphism and malaria clinical outcomes remains evidently clear.The homozygous AA of TNFα at -308 was associated with severe malaria in Sri Lanka [32], but no association was found in Myanmar patients [33].As well, it was of interest that different linkage disequilibrium (LD) patterns may contribute to the discrepancy.The polymorphism may be in LD with a nearby causal variant in one ethnic group but not in another.Because Africa is the origin of modern humans, there is a high level of genetic diversity and weak LD in African compared to non-African populations [55].However, the chance of type I error from multiple testing could not be ruled out.It is unlikely that the discrepancy arises from the number of articles.Fewer studies in Africans (n = 3) generated more total patients in case and control than Asians (n = 5).Further investigation could be performed to strengthen the current results.

Novelties of the present meta-analysis
Meta-analyses are confronted with a host of epidemiological heterogeneities (different ethnic backgrounds of sample populations) and methodological issues, e.g., diagnostic criteria, inclusion/exclusion criteria, lack of uniformity in how outcomes are measured, poorly defined phenotypes as well as population stratification, which may contribute to difficulties in detecting significant allelic association with phenotypes [56].
Our study delineates the role of TNF-α -308G >A polymorphism and the severity of P. falciparum malaria, showing there was no statistically significant association for all genetic models.Instead, the significant association was only observed in the racial subgroup.A recent meta-analysis [34] found a significant association between the -308G >A polymorphism and predisposition to the P. falciparum clinical severity in the allele comparison and dominant model.There are several possible reasons for such different results.First, the current metaanalysis included five more published studies that were not included in the previous one.In this study, 8 eligible studies with 1,448 cases and 1,371 controls to yield summary statistics, however, Sarangi et al. (2023) analyzed data with 982 cases and 773 controls.Second, the malaria phenotype in the control group was not uniform in previous meta-analyses [34].The control subjects from two studies [35,36] were asymptomatic malaria and other illness, while the remaining were UM.Third, the present analysis included a subgroup analysis of ethnically diverse populations.

Strengths and limitations
Pointing out the limitations and strengths contextualizes the interpretation of the present meta-analysis results.There are still some limitations that could be addressed in this study.First, the majority (7/8: 87.5%) of published studies were underpowered (< 80%) to prove authentic associations.Second, the number of patients and studies was not sufficiently large, particularly for subgroup analyses.Third, the present analysis included only English articles.It may have articles in other languages or in local journals.Fourth, the effects of gene-gene and gene-environment interactions were not addressed due to inadequate data.Fifth, the influence of TNF-α -308G >A did not preclude effects from other polymorphisms in proximity to TNFα given the complete LD between them.Sixth, publication bias was not analyzed (number of studies < 10), which may have distorted the results.Seventh, the fact that Laplace correction was applied to the genotype frequencies may have distorted our results, possibly highly sensitive modification of a SNP that already has a low MAF.However, this method improves algorithms' exploring power and is not likely affected the results when the variance loss at maximum value of 0.5.Eighth, great disparity of minor allele A among the Asians studies may affect the outcome of overall and Asian subgroup.On the other hand, our main results showed that two significant associations were retained in the domianant and codominant model after excluding the highest MAF study (S2 Table ).Despite these limitations, the strength of present meta-analyses include (i) the most control group in the selected literature (7/8, 87.5%) were distributed in compliance with HWE, indicating the high genotyping quality of the component studies; (ii) the type of malaria parasite in all patients was homogenous, i.e., P. falciparum; (iii) all of the included articles were population-based which facilitates extrapolation of our findings to the general populations in Africans and Asians; (iv) most of the quality of the case-control studies included in the current meta-analysis (7/8, 87.5%) had moderate to high Clark-Baudouin scores indicating good methodological quality; and (v) modifier analysis was performed, which excludes non-HWE compliant study.The outcome of the non-association of TNF-α -308G >A polymorphism and SM remains.

Fig 2
illustrates the forest plots of the association between TNF-α -308G >A gene polymorphism and the severity of falciparum malaria showing the differences in outcome between Africans and Asians in the dominant and codominant models.

Table 1 . Characteristics of the included studies in the TNF-α -308G >A polymorphism and its associations with severe P. falciparum malaria.
[34]α: tumor necrosis factor-alpha; [R]: reference number; G: guanine nucleobase; A: alanine nucleobase; NM: no mentioned PCR: polymerase chain reaction; ASO: hybridized with allele-specific oligonucleotides; RFLP: restriction fragment length polymorphism; SSOP: sequence-specific oligonucleotide probes Age is presented with mean ± standard deviation or an average value UM: uncomplicated malaria (defined as febrile illness with positive parasitological test and without any of symptom of severe malaria) SM: severe malaria (see definition in reference no. 6) * not included in the previous meta-nalysis[34].https://doi.org/10.1371/journal.pntd.0011735.t001