Introduction

Head and neck cancer (HNC) is a group of cancers that starts within the nose, mouth, throat, sinuses, larynx, or salivary glands (1, 2). According to the most recent GLOBOCAN estimates from 2020, HNC ranks as the seventh most prevalent cancer worldwide. It is responsible for approximately 890,000 new cases annually, representing roughly 4.5% of all cancer diagnoses globally (3). Additionally, HNC leads to approximately 450,000 deaths each year, accounting for approximately 4.6% of total cancer-related deaths worldwide (3). Men are more prone to HNC than women, irrespective of their alcohol consumption or tobacco use habits. This gender disparity in the incidence of HNC becomes particularly noticeable in individuals in their 60s. The lower regions of the upper aerodigestive tract, such as the larynx and hypopharynx, are the most commonly affected areas (4). Although tobacco and alcohol are the main risk factors for the development of HNC, a significant correlation has been observed between a subset of HNC and the human papillomavirus in epidemiological studies (5). A number of single nucleotide polymorphisms can associate with the risk of HNC reported in recent meta-analyses (6–10).

Glutathione S-transferase theta 1 (GSTT1) gene produces an enzyme that plays a key role in the neutralization of electrophilic compounds. These compounds include carcinogens, therapeutic drugs, environmental toxins, and by-products of oxidative stress (11). GSTT1 is located at 22q11.23 with a 480 bp fragment using specific primers (12, 13). and the GSTT1 null genotype, which results from a homozygous deletion of the GSTT1 gene, leads to a lack of enzyme activity (14).

However, the association between GSTT1 deletion polymorphism and HNC remains unclear due to inconsistent findings among studies. Some studies have reported a significant association (15–17), while others have found no such link (18–21). These discrepancies could be due to differences in study design, sample size, population characteristics, or methods of genotyping.

Several meta-analyses (22–36) reported the association of GSTT1 deletion polymorphism and HNC susceptibility. Three meta-analyses (24, 28, 32) were published after 2015 and all three meta-analyses just reported GSTT1 deletion polymorphism in oral cancer. The last meta-analysis (32) included 36 studies.

This systematic review aimed to provide a more definitive answer to this question by combining the results of 107 articles (50 studies reported oral cancer).

Materials and methods

Study design and registration

The meta-analysis was conducted by the protocols of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (37). The question posed in terms of PICO (population, intervention, comparison, and outcome) was (38): Does the deletion polymorphism of GSTT1 associate with the susceptibility to HNC in case-control studies? (Population (P): Patients with HNC. Intervention (I): GSTT1 deletion polymorphism. Comparison (C): Control subjects (non-HNC individuals). Outcome (O): Correlation with susceptibility to HNC in case-control studies). The study has not registered in any database.

Identification of articles

An exhaustive search was carried out by one author (M.S.) in the databases of PubMed/Medline, Web of Science, Scopus, and Cochrane Library from the beginning of each database until June 21, 2023, with no restrictions to identify pertinent articles. The titles/abstracts of the articles were evaluated by the same author (M.S), who also downloaded the full texts of the articles that satisfied the eligibility criteria. The search strategy encompassed: (“Glutathione S-transferase” OR “GSTT1” OR “GST”) AND (“head and neck” OR “oral” OR “OSCC” OR “tongue” OR “mouth” OR “HNSCC“ OR “nasopharyngeal” OR " oropharyngeal " OR “nasopharynx” OR “salivary gland” OR “hypopharyngeal” OR “pharyngeal” OR “pharynx” OR “oral cavity” OR “hypopharynx” OR “laryngeal” OR “larynx”) AND (“carcinoma*” OR “tumor*” OR “cancer*” OR “neoplasm*”) AND (“allele*” OR “variant*” OR “genotype*” OR “gene*” OR “polymorphism*”). To ensure no relevant study was overlooked, the reference lists of the articles were also scrutinized. The search and selection process was re-verified by another author (M.M.I.). In case of any disagreement between the two authors, a third author (M.S.) resolved it.

Inclusion and exclusion criteria

The inclusion criteria were as follows: (1) Studies of the case-control type that examined GSTT1 deletion polymorphism in patients with HNC and control subjects. (2) HNC was diagnosed clinically and pathologically. (3) Patients with HNC did not have any other systemic diseases and controls were either healthy or free from cancer. Conversely, the exclusion criteria included: review articles, meta-analyses, systematic reviews, articles that had incomplete data or lacked a control group, studies conducted on animals, conference papers, comment papers, duplicate studies, book chapters, studies that included controls with the disease, and studies that included cases under treatment.

Data summary

The information from the studies incorporated into the meta-analysis was independently gathered by two authors (M.S. and S.S.). Any disagreements were settled through discussion.

Quality evaluation

One author (M.S) performed the quality scoring using the Newcastle-Ottawa Scale (NOS) tool (39). This tool evaluates a study based on three broad perspectives: the selection of the study groups (4 scores), the comparability of the groups (2 scores), and the ascertainment of either the exposure or outcome of interest (3 scores) for case-control studies. The maximum possible score is nine, and a score of ≥ 7 is considered to be of high quality. Another author (N.K.) re-checked the scores. Disagreement between the authors was resolved by a short discussion.

Statistical analyses

The Review Manager 5.3 (RevMan 5.3) software was used to calculate the effect sizes, which were displayed as the odds ratio (OR) along with a 95% confidence interval (CI) for the prevalence of the null genotype of GSTT1 polymorphism in HNC patients and controls. The significance of the pooled OR was determined using the Z-test, with a two-sided p-value less than 0.05 deemed significant. A random-effects model (40) was employed if Pheterogeneity was < 0.10 (I2 > 50%), indicating significant heterogeneity. If the heterogeneity was not significant, a fixed-effect model (41) was applied.

A subgroup analysis was conducted to ascertain whether the combined effect sizes in these subgroups differed significantly from one another. Furthermore, a meta-regression analysis using a random-effects model was carried out to illustrate a linear correlation between auxiliary variables in the study and the effect size.

The extent of publication bias was assessed using the funnel plot and Egger’s regression test. The possibility of publication bias was evaluated using Begg’s funnel plot and Begg’s test, and the level of asymmetry was tested with Egger’s test. The p-values from both Egger’s and Begg’s tests were obtained, and a 2-sided p-value less than 0.10 indicated the existence of publication bias. In terms of sensitivity analysis, both “one-study-removed” (This is done to determine if any single study has a disproportionate impact on the overall estimate.) and “cumulative” (This is done to assess the impact of each additional study on the overall estimate.) analyses were employed to assess the stability and consistency of the pooled SMDs. Both the publication bias and sensitivity analyses were performed using the Comprehensive Meta-Analysis version 3.0 (CMA 3.0) software.

The Radial plot, also known as the Galbraith plot, was designed using the NCSS 2021 version 21.0.2 software. This plot displays the z-statistic (obtained by dividing by the standard error) on the vertical axis and the weight measurement on the horizontal axis (42). A p-value less than 0.05, indicates statistically significant heterogeneity.

To mitigate the risk of false-positive or negative conclusions from meta-analyses (43), a TSA was conducted using TSA software (version 0.9.5.10 beta) from the Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Copenhagen, Denmark (44). TSA allows for the testing of a futility threshold to establish a result of no effect before reaching the necessary information size. The required information size (RIS) was computed with an alpha risk of 5%, a beta risk of 20%, and a two-sided boundary type. Heterogeneity (D2) was evaluated for the prevalence of the null genotype of GSTT1 polymorphism in HNC patients and controls. If the Z-curve reached the RIS line or traced the boundary line or futility area, it suggested that the studies included a sufficient number of cases and that the conclusions were reliable. If not, it indicated that the information available was insufficient and additional data was required.

Results

Study selection

A total of 1966 records were initially retrieved from four databases, along with 8 records from other electronic sources (Fig. 1). After the removal of duplicates, 1052 records remained and were screened. Of these, 887 records were deemed irrelevant and subsequently removed. This left 165 full-text articles that met the eligibility criteria. However, 58 of these were excluded for various reasons. Ultimately, 107 full-text articles were included in the analysis.

Fig. 1
figure 1

Flowchart of study selection for systematic review and meta-analysis

Study’s characteristics

Table 1 presents a comprehensive list of 107 articles (15–21, 45–142) including 109 studies conducted on the null genotype of GSTT1 polymorphism in HNC patients and controls. The studies span multiple countries and ethnicities, with a variety of cancer types and control sources. Each study includes the number of cases and controls, with some studies matching controls based on age and sex. The quality score of each study is also provided, offering insight into the reliability of the data.

Table 1 Characteristics of the studies

Meta-analysis

A forest plot analysis using a random-effects model was conducted to examine the association between GSTT1 polymorphism and the risk of HNC, as depicted in Fig. 2. The combined analysis revealed that the pooled OR was 1.28, with a 95% CI ranging from 1.14 to 1.44. This result was statistically significant with a p-value less than 0.0001. However, there was substantial heterogeneity among the studies, as indicated by an I2 value of 82%. The result suggests that there is a significant association between GSTT1 polymorphism and the risk of HNC, with the null genotype of GSTT1 associated with a 28% increased risk of HNC. However, due to the high heterogeneity (I2 = 83%), the results should be interpreted with caution as the studies included in the analysis may have varied in aspects such as study and population characteristics.

Fig. 2
figure 2

Forest plot analysis of the association between GSTT1 polymorphism and the risk of head and neck cancer

Subgroup analysis

Table 2 presents a subgroup analysis of the association between GSTT1 polymorphism and the risk of HNC. The subgroups are divided based on ethnicity, cancer type, sample size, control source, and quality score. In terms of ethnicity, the pooled OR was highest in Asian ethnicities (OR = 1.31), followed by Mixed (OR = 1.28), and Caucasians (OR = 1.17). For cancer type, nasopharyngeal cancer had the highest OR (1.84), followed by oral cancer with an OR of 1.20, and laryngeal cancer with an OR of 1.17. When considering sample size, studies with less than 200 samples had a higher OR (1.59) compared to those with 200 or more samples (OR = 1.23). The control source did not significantly affect the OR, with both population-based and hospital-based controls showing similar ORs of 1.29 and 1.24 respectively. The studies with a quality score of 7 or more had a higher OR (1.37) compared to those with a score of less than 7 (OR = 1.05). When considering age, the OR of 1.41 suggests a higher risk, but the high heterogeneity and p-value of 0.31 indicate that this result is not statistically significant. The analysis based on sex shows a significant association, with an OR of 1.36 and a p-value of 0.006. However, when both age and sex are considered, while the OR of 1.42 is significant, the high heterogeneity suggests caution in interpreting these results. Finally, in the group where neither age nor sex was considered, no significant association was found. These findings highlight the complexity of the relationship between the GSTT1 null genotype and HNC risk, and how it can be influenced by factors such as age and sex. It’s important to note that all these results should be interpreted with caution due to the high heterogeneity observed in most subgroups (I2 > 50%).

Table 2 Subgroup analysis

Meta-regression

Table 3 presents a meta-regression analysis of the variables: publication year, sample size, and quality score. For the publication year, the coefficient is ˗ 0.0003 with a p-value of 0.1213. For the sample size, the coefficient is -0.0002 with a p-value of 0.1965. For the quality score, the coefficient is 0.1283 with a p-value of 0.0147. In this case, only the quality score shows statistical significance as its p-value is less than 0.05. The results indicate that quality score increased, the effect size significantly increased.

Table 3 Meta-regression analysis

Sensitivity analysis

The sensitivity analysis, which included both one-study-removed and cumulative analyses, showed that the results were robust and reliable. In this case, the fact that the results did not change significantly in either analysis indicates that no single study unduly influenced the results and that the results were consistent across all studies. This adds to the validity and reliability of your findings.

Publication bias.

Figure 3 shows the funnel plot of the association between GSTT1 polymorphism and the risk of HNC. The p-values for both Egger’s test (0.895) and Begg’s test (0.108) are greater than 0.10. This suggests that there is no evidence of publication bias in the meta-analysis. Therefore, it can be interpreted that your results are likely not influenced by publication bias.

Fig. 3
figure 3

Funnel plot of the association between GSTT1 polymorphism and the risk of head and neck cancer

Heterogeneity analysis

Figure 4 identifies the radial plot of the association between GSTT1 polymorphism and the risk of HNC. The p-value of less than 0.0001 suggests that there is significant heterogeneity among the studies included in the meta-analysis. This means that there are substantial differences in the results of these studies that cannot be attributed to chance alone. The presence of outlier data in some studies could be contributing to this heterogeneity. It’s important to investigate these outliers further to understand their source and consider their impact on the overall results of the meta-analysis. Therefore, while your analysis shows a significant association between GSTT1 polymorphism and the risk of HNC, the high heterogeneity suggests that caution should be taken when interpreting these results. Further research may be needed to explore the sources of this heterogeneity.

Fig. 4
figure 4

Radial plot of the association between GSTT1 polymorphism and the risk of head and neck cancer

TSA

Figure 5 shows the TSA of the association between GSTT1 polymorphism and the risk of HNC (D2 = 85%, the incidence in the intervention arm (IIA) = 28.24%; the incidence in the control arm (ICA) = 25.76%). IIA is higher than ICA. This indicates that the occurrence of the GSTT1 null genotype under study is more frequent in the HNC group compared to the control group. The D2 value represents the diversity (or heterogeneity) of the study results. A high D2 value suggests a high degree of variability in the study results, which could be due to differences in study characteristics. The Z-curve crossing the boundary for harm suggests that the GSTT1 polymorphism being studied may have harmful effects. However, since the number of patients in the study (43,555) is less than the RIS (65,384), the study does not have sufficient statistical power. This means that the results should be interpreted with caution as they may be prone to random errors. In other words, while the current data suggests potential harm, it does not conclusively prove it due to insufficient information size. Therefore, more research or larger studies may be needed to conclusively determine whether the GSTT1 polymorphism is harmful.

Fig. 5
figure 5

A trial sequential analysis of the association between GSTT1 polymorphism and the risk of head and neck cancer

Discussion

A meta-analysis found that people with a certain gene variant (GSTT1 null) had a higher risk of HNC, especially nasopharyngeal cancer. However, the studies included in the meta-analysis were very different from each other and had some limitations. Subgroup analysis revealed differences in ORs based on factors such as ethnicity, cancer type, sample size, control source, and quality score. The quality score was found to significantly impact the effect size in the meta-regression analysis. Despite these findings, the high heterogeneity and the smaller sample size compared to the RIS suggest that these results should be interpreted cautiously.

The relationship between GSTT1 and HNC can be influenced by several factors. A study presented the results of the analysis of the joint effects or interaction between tobacco use and GSTT1 null genotype (111). Another study (89) discovered that the GSTT1 null genotype was linked to a higher risk among individuals who had a longer history of smoking. In addition, interaction between GSTT1 polymorphism with other genes such as GSTs and CYPs can affect the risk of HNC (63, 129, 143, 144). The studies recommended also that occupational hazards can affect the association between GST polymorphisms and HNC risk (145–147). In this meta-analysis, we couldn’t check these factors due to a lack of sufficient data that future studies can check them in HNC.

GST is a family of enzymes that play an important role in detoxification by catalyzing the conjugation of many hydrophobic and electrophilic compounds with reduced glutathione (148, 149). They have been linked to the development of various cancers (150–153), but the specific role they play in HNC may require further research. The protection provided by the GSTT1 enzyme is viewed as more comprehensive, given that the gene is not only expressed in the liver but also erythrocytes. This results in a systemic action of the enzyme (154).

Research on the GSTT1 null genotype indicates that in the United States, the absence of GSTT1 is less prevalent compared to the GSTM1 deletion genotype. Among individuals of European descent, it’s found that 15–31% lack a functional GSTT1 enzyme. For African Americans, the frequencies range from 22 to 29%. Meanwhile, individuals of Hispanic origin exhibit GSTT1 deletions at a rate of 10–12% (155–158). In terms of ethnicity, Asians are more susceptible to HNC associated with GSTT1 null genotype, compared to their European and American counterparts (107). The present meta-analysis reported that GSTT1 polymorphism has an association with the risk of HNC in Asians and mixed ethnicities, not Caucasians. Therefore, the prevalence of GSTT1 null genotype may differ by geographic region (30).

In diabetic patients, the GSTM1 null genotype was found to be significantly more prevalent in the 24–36 year age group compared to other age groups (159). The present meta-analysis reported the relationship between the GSTT1 null genotype and HNC risk, and how it can be influenced by factors such as age and sex. Therefore, outliers based on radial plot, lack of sufficient cases based on TSA, variation in age range and sex percentage can be main factors for a high heterogeneity in this meta-analysis.

The present systematic review and meta-analysis included four limitations: (1) there was substantial heterogeneity among the studies included in the meta-analysis. (2) Due to the high heterogeneity, the results should be interpreted with caution. (3) the study may not have sufficient statistical power to detect small effect sizes or rare events. (4) the study was based on published data rather than individual patient data, which may limit the ability to control for potential confounding factors. The present systematic review and meta-analysis included four strengths: (1) the study included a comprehensive analysis of 107 full-text articles, providing a broad overview of the existing literature on the association between GSTT1 polymorphism and the risk of HNC. (2) Subgroup analysis allowed for a more nuanced understanding of how the factors might influence the association between GSTT1 polymorphism and HNC risk. (3) Meta-regression analysis provided insights into how these variables might impact the effect size. (4) the study found no evidence of publication bias, suggesting that the results are not skewed by the selective publication of studies with positive results.

Conclusions

This comprehensive meta-analysis revealed a significant association between the GSTT1 null genotype and an increased risk of HNC, with variations based on factors such as ethnicity, cancer type, sample size, control source, and quality score. Despite the robustness of the results, there was high heterogeneity among studies and limited statistical power due to a smaller number of cases. From a clinical perspective, these findings underscore the potential of the GSTT1 null genotype as a genetic marker for HNC susceptibility, which could have significant implications for early detection and prevention strategies. However, further research is needed to confirm these findings and elucidate the underlying mechanisms. This study sets the stage for future research in this area, highlighting the importance of considering factors such as ethnicity, cancer type, sample size, control source, and quality score in understanding the complex relationship between GSTT1 null genotype and HNC risk.