Statins and the Risk of Gastric Cancer: A Systematic Review and Meta-Analysis

Previous epidemiological studies have reported that the use of statins is associated with a decreased risk of gastric cancer, although the beneficial effects of statins on the reduction of gastric cancer remain unclear. Therefore, we conducted a systematic review and meta-analysis to investigate the association between the use of statins and the risk of gastric cancer. Electronic databases such as PubMed, EMBASE, Scopus, and Web of Science were searched between 1 January 2000 and 31 August 2022. Two authors used predefined selection criteria to independently screen all titles, abstracts, and potential full texts. Observational studies (cohort and case-control) or randomized control trials that assessed the association between statins and gastric cancer were included in the primary and secondary analyses. The pooled effect sizes were calculated using the random-effects model. The Meta-analysis of Observational Studies in Epidemiology (MOOSE) reporting guidelines were followed to conduct this study. The total sample size across the 20 included studies was 11,870,553. The use of statins was associated with a reduced risk of gastric cancer (RRadjusted: 0.72; 95%CI: 0.64–0.81, p < 0.001). However, the effect size of statin use on the risk of gastric cancer was lower in Asian studies compared to Western studies (RRAsian: 0.62; 95%CI: 0.53–0.73 vs. RRwestern: 0.88; 95%CI: 0.79–0.99). These findings suggest that the use of statins is associated with a reduced risk of gastric cancer. This reverse association was even stronger among Asian people than Western individuals.


Introduction
Gastric cancer is a common public health problem associated with a substantial healthcare burden [1]. Gastric cancer ranks fourth in terms of incidence, and is the fifth leading cause of cancer mortality worldwide, with an estimated approximately 1.1 million new cases and 770,000 deaths in 2020 [2,3]. Since a considerable proportion of gastric cancer patients are diagnosed at advanced stages, the five-year survival rate is only 32% [4]. Even though the prevalence and mortality of gastric cancer have been declining in some parts of the world, meaningful prevention strategies are needed to treat it early and reduce the overall healthcare burden. Previous studies have reported that several modifiable (e.g., tobacco smoking, alcohol consumption, obesity, gastroesophageal reflux, and Helicobacter pylori (H. pylori) infection) and non-modifiable (e.g., age, gender, and ethnicity) risk factors are associated with gastric cancer [5][6][7][8].
Statins are the most commonly prescribed medications and are considered to be effective for lowering cholesterol and protecting against cardiovascular diseases [9,10]. Previous epidemiological studies have highlighted the beneficial effects of statins against

Methods
This study was based on two main questions. Question 1 was defined as follows: Are patients treated with statins associated with a reduced risk of gastric cancer? Question 2 was defined as follows: Do subgroup analyses support the effect size of the association between statin use and gastric cancer risk? The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement and the guidelines for Meta-analysis Of Observational Studies in Epidemiology (MOOSE) were properly followed to conduct this study [18,19] (Supplementary Table S1).

Databases and Search Strategy
We conducted a comprehensive search on the most popular databases for scientific literatures such as PubMed, EMBASE, Scopus, and Web of Science to obtain all Englishlanguage studies published between 1 January 2000 and 31 August 2022. We considered studies that evaluated the association between statin use and the risk of gastric cancer. The following search terms were used: "statin/s" or "hydroxymethylglutaryl-CoA reductase inhibitors" or "lipid-lowering drugs" or "atorvastatin" or "fluvastatin" or "lovastatin" or "pitavastatin" or "pravastatin" or "rosuvastatin" or "simvastatin" and "gastric cancer" or "gastric carcinoma" or "gastric neoplasm" or "stomach cancer" or "stomach carcinoma" or "stomach adenocarcinoma" or "stomach neoplasm" or "cancer" (Supplementary Table S2). Additional searches were conducted in the bibliographies of relevant articles to obtain missing articles.

Inclusion and Exclusion Criteria
All observational studies (e.g., cohort or case-control studies) and randomized control trials (RCTs) that met the following criteria were included: (1) studies that carried out at least one analysis assessing the effect of statins use on gastric cancer, (2) patients with an established diagnosis of gastric cancer by accepted clinical and/or histologic criteria, (3) studies published in English, and (4) studies that provided sufficient information to calculate the pooled effect size. We excluded studies if they were reviews, letters, case reports, or editorials.

Data Extraction
Two authors (CCW and MMI) independently extracted information from all selected studies using piloted data extraction sheets. Extracted data included (1) demographics: author name, publication year, country, (2) population: age, gender, percentage of male, number of statin users, number of gastric cancer patients, (3) methods: study design, inclusion and exclusion criteria, and (4) results: effect sizes (hazard ratio, odds ratio). Any disagreement during the study screening process was resolved by discussing with the third author.

Study Quality Assessment
The risk of bias of RCTs was assessed using the Cochrane Collaboration tool [20], which is comprised of the following domains: (a) sequence generation, (b) allocation concealment, (c) blinding of participants, (d) incomplete outcome data, (e) selective reporting, and (f) other risk of bias. Moreover, the Newcastle-Ottawa Scale was used to assess the methodological quality of the observational studies [21]. The NOS scale is recommended by the Cochrane Handbook for Systematic Reviews of Interventions, and it is divided into three categories such as (a) study selection, (b) comparability, and (c) the ascertainment of exposure (for case-control studies) or the outcome of interest (for cohort studies).

Statistical Analysis
We used ORs and HRs to calculate the overall pooled effect sizes of statin use on the risk of gastric cancer. The pooled RRs were calculated using random-effects models. We calculated heterogeneity across studies using the Q-statistic and quantified using the inconsistency I 2 . The I 2 values were classified into four groups: of 0~29%, 30~49%, 50~74%, and 75~100%, representing very low, low, medium, and high inconsistency, respectively [22][23][24]. We also conducted subgroup analyses to assess the potential impact of study design, region, quality of observational studies (NOS ≤ 7 vs. NOS > 7), duration, and statin types. In the sensitivity analysis, the impact of each study on the summary statistics was evaluated by excluding one study at a time from the meta-analysis. A forest plot was drawn to visually represent the effect size of each study and the pooled analyses. Finally, funnel plots and the Egger regression test of funnel plot asymmetry were used to calculate overall publication bias. The statistical analyses were performed using comprehensive meta-analysis software (CMA). A p < 0.05 was considered statistically significant. Figure 1 shows the PRISMA diagram of the selection of studies. The search identified 6976 studies. However, 3700 studies were excluded for duplications and 3247 studies were excluded after screening the titles and abstracts. In total, 29 studies were selected for full-text screening based on prespecified selection criteria. Finally, 20 studies were selected for the meta-analysis [17,[25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43].

Study Selection and Characteristics
These 20 studies consisted of 9 cohort, 8 case-control, and 3 RCTs studies, which comprised of 11,870,553 participants. Twelve studies were conducted in Asia and eight studies were conducted in Western countries. A total of 17 out of 20 studies used administrative databases to identify statin users and gastric cancer patients. More than half of observational studies were of good quality, reflected by a Newcastle-Ottawa score of at least 8. Table 1 presents the characteristics of the included studies, intervention, outcomes, included and excluded criteria, and study quality. These 20 studies consisted of 9 cohort, 8 case-control, and 3 RCTs studies, which comprised of 11,870,553 participants. Twelve studies were conducted in Asia and eight studies were conducted in Western countries. A total of 17 out of 20 studies used administrative databases to identify statin users and gastric cancer patients. More than half of observational studies were of good quality, reflected by a Newcastle-Ottawa score of at least 8. Table 1 presents the characteristics of the included studies, intervention, outcomes, included and excluded criteria, and study quality.

Statin Use and the Risk of Gastric Cancer
Twenty studies (three RCTs and 17 observational studies) reporting the effect of statins on gastric cancer were included in the meta-analysis. When all adjusted effect sizes were pooled, the summary RR in statin users compared with statin nonusers from the randomeffects models was 0.72 (95%CI: 0.64-0.81). There was significant heterogeneity across studies (I 2 = 92.63, Q = 272.69, τ 2 = 0.05) in the random-effects model. Figure 2 shows the forest plot of the association between statin use and the risk of gastric cancer.
Twenty studies (three RCTs and 17 observational studies) reporting the effect of statins on gastric cancer were included in the meta-analysis. When all adjusted effect sizes were pooled, the summary RR in statin users compared with statin nonusers from the random-effects models was 0.72 (95%CI: 0.64-0.81). There was significant heterogeneity across studies (I 2 = 92.63, Q = 272.69, τ 2 = 0.05) in the random-effects model. Figure 2 shows the forest plot of the association between statin use and the risk of gastric cancer.

Sensitivity and Subgroup Analyses
We also conducted a sensitivity analysis to assess changing the overall effect sizes of the use of statin on gastric cancer and the presence of heterogeneity by omitting one study from the main analysis (Supplementary table S3). The overall effect size and heterogeneity did not change after sensitivity analyses.

Sensitivity and Subgroup Analyses
We also conducted a sensitivity analysis to assess changing the overall effect sizes of the use of statin on gastric cancer and the presence of heterogeneity by omitting one study from the main analysis (Supplementary Table S3). The overall effect size and heterogeneity did not change after sensitivity analyses.

Dose-Response Association between Statins and Gastric Cancer
Kim et al. [26] evaluated the risk of gastric acid among statin users who received <1460 and ≥1460 cumulative defined daily doses (cDDDs). After propensity score matching, the effect size for the association between gastric cancer risk and statin users with <1460 cDDDs and ≥1460 cDDDs were 0.98 (95%CI: 0.91-1.06) and 0.22 (95%CI: 0.19-0.26), respectively. Cho et al. [27] classified the cumulative duration of statin use into four categories such as <182. 5

Duration-Response Association between Statins and Gastric Cancer
Cheung et al. [29] invested the duration-response relationship and categorized statin users into two groups:

Duration-Response Association between Statins and Gastric Cancer
Cheung et al. [29] invested the duration-response relationship and categorized statin users into two groups: (i) <5 years and (ii) ≥5 years. Lower risk of gastric cancer was observed among patients who used statins longer (0.46 [95%CI:0.25-0.86] for <5 years of use and 0.43 [95%CI:0.29-0.66] for ≥5 years of use). Le et al. [31] also evaluated the impact of duration of statin use on gastric cancer risk. Diabetes patients who received statins less than 1 year before the index date had an effect size of 0.45; however, the risk of gastric cancer was even lower among patients with statin use of more than 2 years (0.154 [95%CI 0.09-0.26]). Figure 3 shows the funnel plot for publication bias. Out of 20 data points, six lay outside the triangle, and only one lay on the left side of the triangle altitude. The Egger regression test of the funnel asymmetry showed no publication bias (p = 0.09).

Discussion
In this updated systematic review and meta-analysis, we evaluated the association between statin use and the risk of gastric cancer, which is currently unclear. Our findings are similar to previous evidence [15,[44][45][46][47] that showed statin use is associated with a significantly reduced risk of gastric cancer. The strength of this association varied among study designs, and no association was observed in the RCTs. In addition, a more protective effect was observed in Asian people compared to Western individuals.
The possible biological mechanisms as to how statin use reduces the risk of gastric cancer is still unclear. However, several biological pathways may help to understand the process of their association. Statin reduces the production of cholesterol, dolichol, and coenzyme Q10 by inhibiting HMG-CoA reductase in the mevalonate pathway [48,49]. Previous studies highlighted that statin increases apoptosis, suppress angiogenesis, and changes the tumor microenvironment downstream of the mevalonate pathway [50][51][52] Moreover, suppression of the mevalonate pathway due to statin use can reduce radiosensitization or chemosensitization [53,54]. Statins interrupt the production of primary geranylgeranyl pyrophosphate (GGPP) and farnesylpyrophophosphate (FPP) and delay

Discussion
In this updated systematic review and meta-analysis, we evaluated the association between statin use and the risk of gastric cancer, which is currently unclear. Our findings are similar to previous evidence [15,[44][45][46][47] that showed statin use is associated with a significantly reduced risk of gastric cancer. The strength of this association varied among study designs, and no association was observed in the RCTs. In addition, a more protective effect was observed in Asian people compared to Western individuals.
The possible biological mechanisms as to how statin use reduces the risk of gastric cancer is still unclear. However, several biological pathways may help to understand the process of their association. Statin reduces the production of cholesterol, dolichol, and coen-zyme Q10 by inhibiting HMG-CoA reductase in the mevalonate pathway [48,49]. Previous studies highlighted that statin increases apoptosis, suppress angiogenesis, and changes the tumor microenvironment downstream of the mevalonate pathway [50][51][52]. Moreover, suppression of the mevalonate pathway due to statin use can reduce radiosensitization or chemosensitization [53,54]. Statins interrupt the production of primary geranylgeranyl pyrophosphate (GGPP) and farnesylpyrophophosphate (FPP) and delay the growth of malignant cells, eventually leading to apoptosis [55]. Statins alter the activation of the proteasome pathway, inhibiting the breakdown of both p21 and p27 [56]. Finally, statins allow p21 and p27 molecules to utilize their growth-inhibitory effects and try to slow down gastric cancer cell mitosis [57].
The effect size of the relationship between statin use and the risk of gastric cancer was different in Asian and Western population. Statin use significantly showed a reduction of gastric cancer in Asian populations, but no association was observed in Western populations. Previous studies assessed statin responses between Asians and Westerners (European Americans), showing a difference in pharmacokinetic and pharmacodynamic effects [58,59]. Studies highlighted that statin responses were significantly different even after adjustment for potential confounding factors such as age, comorbidities, and/or socioeconomic status. Body size differences between Asians and Westerners may also contribute to pharmacokinetic variation (<10%) [60,61]. Moreover, genetic variation among Asians can influence statin pharmacokinetics and pharmacodynamics [62].
The findings of our study also showed that the reduction of gastric cancer among statin users was higher in the case-control studies compared to cohort and RCTs. The inherent risk of bias among case-control studies is consistently high due to potential confounding factors. In the cohort studies, maximum efforts are given to reduce the possible biases, although it does demonstrate a causal relationship. In the clinical-decision making, RCTs are considered the gold standard of study design, revealing the causal relationship between a drug and gastric cancer risk. On the other hand, more chemo-preventive effects of statins were observed in the low-quality studies, which may overestimate their true effect. Low-quality studies often contain a lack of random patient allocation and a lack of adjusting for numerous covariates; therefore, it is not possible to eliminate the possibility of residual confounding factors. Included studies also demonstrated that statins are even more protective against gastric cancer if patients take a higher dose over a longer duration of time [26][27][28]. Kwon et al. [17] showed that a significant reduction in gastric cancer was observed among patients taking hydrophilic statins. Our subgroup analyses showed that pravastatin (hydrophilic) was associated with an insignificant reduction of gastric cancer, whereas rosuvastatin showed an increased risk of gastric cancer. More studies with controlled confounding factors and larger follow-up periods are needed to provide supporting evidence that demonstrates the effectiveness of statins against gastric cancer.
There are several strengths in this study. First, this is the most updated systematic review and meta-analysis so far that evaluated the relationship between statin use and gastric cancer risk. We included both RCTs and observational studies to summarize overall effect size that ensures the best available evidence through a meta-analytic approach. Second, this meta-analysis included a large sample size and the quality of included studies were high. Therefore, the findings of this study were more reliable. Third, we could identify statin as an independent risk factor for gastric cancer since we adjusted with potential confounding factors to estimate the pooled summary size.
This study also has several limitations. First, most studies were observational studies with a highly heterogeneous population with different patient characteristics. The heterogeneity among the studies was high, although this can be explained by study design, region, and study quality. Second, there was limited information regarding the types of statins, indication, follow-up, and duration of statins; therefore, we were unable to pool effect size based on dosage, duration, and follow-up. Third, several potential covariates/confounders related to gastric cancer risk have not been adjusted in the included study. Future studies should adjust all possible confounders to pool the effect sizes and to show a stronger association between statin use and gastric cancer risk.

Conclusions
The findings of this updated meta-analysis suggest that statin use is associated with a reduced risk of gastric cancer. Although the pooled effect size of observational studies showed a possible role of statin therapy in the prevention of gastric cancer, the pooled effect size of the clinical trials and the risk of the bias of observational studies do not encourage physicians to consider statin use to achieve a reduction in gastric cancer. More RCTs with larger populations, longer follow-up, and standardized methods are required to consider statin use as a strategy for reducing gastric cancer.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11237180/s1, Table S1: PRISMA checklist; Table S2: Search strategy; Table S3: Sensitivity analyses for the association between statin use and the risk of gastric cancer. Reference [63] has been cited in Supplementary Materials.