Causality between appendiceal disease and cancers of the digestive system: A Mendelian randomization study

Abstract

The connection between appendiceal disease and cancers of the digestive tract has attracted widespread attention, but conclusion remain controversial. It is still unclear whether appendiceal disease has a definite causal effect on cancer. Our research aims to explore the causal relationship between appendiceal diseases and digestive system tumors. A two-sample Mendelian randomization (MR) analysis was conducted using genome-wide association study datasets to explore the causal impact of appendiceal disease on the risk of cancers. Five different MR methods were used to investigate causality. The stability, heterogeneity, and pleiotropy of MR were also assessed. The presence of appendiceal disease could reduce the incidence of rectal cancer (odds ratio = 0.588, 95% confidence interval: 0.392–0.881, P = .010, by inverse variance weighted method). There was no significant causal effect of appendiceal disease on the risk of other cancers. No horizontal pleiotropy was observed in the MR analysis, and leave-one-out analysis confirmed the stability of the results. The potential causal relationship between appendiceal disease and cancer risk was only observed with rectal cancer in populations of European ancestry. Future work will provide more robust evidence for the connection between appendiceal disease and cancers.

1. Introduction

The incidence and mortality rates of disorders related to cancer have rapidly increased during the last 10 years. Globally, there were 10 million cancer-related deaths and 19.3 million new cases of cancer in 2020. The management of public health issues is finding it more and more difficult to deal with malignant tumors.^[1] With 4,975,378 new cases and more than 3,524,932 fatalities yearly, gastrointestinal malignancies – including esophageal, gastric, liver, pancreatic, colon, and rectal cancers – are among the most frequent malignant tumors.^[1] According to reports, scientists project that by 2040, there will be 5.6 million deaths and 7.5 million new cases, respectively.^[2]

According to epidemiological research, Helicobacter pylori infection, obesity, smoking, excessive alcohol consumption, oxidative stress, and poor diet are the main causes of digestive system cancers.^[3] Population-level cancer burden can now be decreased with the help of early detection, screening, and prevention.^[4]

For a very long time, the human appendix was thought to be a vestigial organ – an entity that has lost its purpose as a result of evolution.^[5] There is evidence to support the idea that the appendix is crucial for preserving intestinal health. According to the notion, the appendix may act as a “safe house” or storage space that would rapidly replenish itself if the commensal microbiota were to be removed from the colon.^[5–7]

More and more studies are connecting the development and prevention of several diseases with the appendix.^[8–10] The relationship between gastrointestinal tumors and the appendix is still unknown, thus we need to develop a thorough analytical technique to evaluate the appendix’s involvement in gastrointestinal malignancies. The utilization of Mendelian randomization (MR) analysis is noteworthy,^[11,12] based on information from genome-wide association studies (GWAS), has been extensively employed in many public health fields, and may also aid in our investigation of causal linkages by establishing a genetic correlation between appendiceal features and cancer GWAS data. There are currently few data in the literature using MR techniques on the appendix’s possible role in the risk of gastrointestinal cancer. Here, we use a two-sample MR analysis to assess the causal relationship between illnesses of appendix characteristics and the risk of different gastrointestinal cancers. The study’s findings may offer more convincing proof in support of clinical judgment.

2. Materials and methods

2.1. Data source

Through appendix illnesses, this study assesses the appendix’s influence on digestive tract cancers. The GWAS database (https://gwas.mrcieu.ac.uk) of the Medical Research Council’s Integrative Epidemiology Unit is the source of the dataset used in this investigation for appendix illnesses.^[12] As a result, the GWAS data for appendix illnesses (ID: finn-b-K11_APPENDIX, ncase = 16,906) were obtained from the “MR-base” MR platform.^[12] Concurrently, GWAS information related to cancer was also taken from “MR-base.” In order to mitigate bias in the GWAS data analysis, we incorporated cancer types that satisfy the subsequent 3 standards: accessible through “MR-base”; the cancer GWAS dataset is current, having been updated within the last 5 years; and there are more than 1 million SNPs. Consequently, a total of 6 types of cancer, including esophageal cancer (ID: finn-b-C3_OESOPHAGUS, ncase = 232), gastric cancer (ID: finn-b-C3_STOMACH, ncase = 633), liver and intrahepatic bile ducts cancer (ID: finn-b-C3_LIVER_INTRAHEPATIC_BILE_DUCTS, ncase = 304), pancreatic cancer (ID: finn-b-C3_PANCREAS, ncase = 605), colon cancer (ID: finn-b-C3_COLON, ncase = 1803), and rectal cancer (ID: finn-b-C3_RECTUM, ncase = 1078), originated from the “MR-base” of unknown gene biobank. Table 1 summarizes the details of the GWAS datasets. Moreover, we defined the disease of the appendix as the exposure and various cancers as the outcomes to explore the causal effects between these 2 events.

Table 1.

The list of genome-wide summary association studies included in the Mendelian randomization study.

Subtype	GWAS ID	Year	Population	Sex	ncases	ncontrol	No. SNPs	Consortium
Exposure
Diseases of appendix	finn-b-K11_APPENDIX	2021	European	F/M	16,906	201,886	16,380,466	NA
Cancer
Malignant neoplasm of esophagus	finn-b-C3_OESOPHAGUS	2021	European	F/M	232	218,560	16,380,466	NA
Malignant neoplasm of stomach	finn-b-C3_STOMACH	2021	European	F/M	633	218,159	16,380,466	NA
Malignant neoplasm of liver and intrahepatic bile ducts	finn-b-C3_LIVER_INTRAHEPATIC_BILE_DUCTS	2021	European	F/M	304	218,488	16,380,466	NA
Malignant neoplasm of pancreas	finn-b-C3_PANCREAS	2021	European	F/M	605	218,187	16,380,466	NA
Malignant neoplasm of colon	finn-b-C3_COLON	2021	European	F/M	1803	216,989	16,380,466	NA
Malignant neoplasm of rectum	finn-b-C3_RECTUM	2021	European	F/M	1078	217,714	16,380,466	NA

F/M = female and male, GWAS = genome-wide association studies, SNPs = single-nucleotide polymorphisms.

2.2. SNP selection and assumption

Each of the 3 MR analysis assumptions is met by our study (Fig. 1).^[13] Initially, there needs to be a good correlation between the SNPs chosen for MR analysis and the exposure. Only SNPs with P values below the genome-wide significance level (5 × 10⁻⁸) were included in the study to make sure this assumption is met. We eliminated instrumental variables with an F-value [formula: (R²/(1−R²))×((N−K−1)/K)] < 10 to guarantee a strong correlation between the exposure and the instrumental variables. Second, an independence test must be passed by the selected instrumental variables. In order to mitigate the effects of linkage disequilibrium and preserve the independence of the chosen instrumental variables, we fixed the genetic distance at 10,000 kb and the LD value at 0.001. The quality of our investigation was ensured by the selection of the aforementioned instrumental factors.

Figure 1.

Schematic representation of MR. MR can be used to test the hypothesis that diseases of appendix causes cancer risk, provided that 3 conditions are met adequately, governing that 5 SNPs is a valid instrument, in that 1st, it is associated with the intermediate phenotype of interest; 2nd, it has no association with the outcome except through the intermediate phenotype; and 3rd, it is not related to measured or unmeasured confounding factors. MR = Mendelian randomization, SNPs = single-nucleotide polymorphisms.

2.3. Two-sample MR analysis

There are currently numerous statistical techniques available for performing two-sample MR analysis. In this investigation, we employed 5 standard techniques to assess the causal link between appendix illnesses and the risk of digestive tract tumors: MR-Egger, weighted median, simple mode, weighted mode, and inverse variance weighted (IVW). IVW is a traditional analytical technique that integrates Wald ratio estimates from all pertinent instrumental variables. It is a fundamental analytical procedure. This approach is similar to using weighted linear regression to investigate the connection between outcomes and instrumental factors. Instrumental variable intercepts are limited to zero. An unbiased estimate can be obtained using IVW if there is no horizontal pleiotropy.

Under the premise that the instrument strength is independent of the direct effects,^[14] the MR-Egger method can principally exhibit the dose relationship between instrumental variables and outcomes, while also accounting for some forms of pleiotropy. The use of the weighted median method can reduce the type I error rate, which also allows for the possibility that some specific genetic variants are invalid. Even if some instrumental variables do not satisfy the requirements for causal inference of MR technology, the weighted mode method remains valid when the vast majority of tools with the same causal estimate are effective. If the results of these methods are inconsistent, IVW takes precedence as the primary finding.

2.4. Bonferroni correction

To account for the multiple comparisons across the 6 cancer outcomes in this study, we applied a Bonferroni correction after testing. This method adjusts the significance threshold by dividing the conventional alpha level (α = 0.05) by the number of tests performed, thereby reducing the likelihood of false-positive results and ensuring that the reported associations are statistically robust.

2.5. Heterogeneity, pleiotropy, and sensitive analyses

Additionally, the MR-Egger approach is used to assess pleiotropy, which is a genetic variable with several independent phenotypic effects that can influence causal effects. Heterogeneity tests are performed using the MR-Egger and IVW methods to confirm the compliance of each SNP, and the Cochran Q statistic is computed to identify heterogeneity among genetic variations.^[15] The remaining genetic variants are subjected to MR analysis while each genetic variant is progressively excluded in a leave-one-out analysis. The causal association is regarded as reliable and stable if the findings of the IVW analysis and the leave-one-out analysis agree. Causal effects are deemed nominal effects when the P value is between .05 and the corrected value.

2.6. Statistical analysis

The two-tailed P value <.05 was considered statistically significant. All of the statistical analyses were performed by R (version 4.3.2; The University of Auckland, Auckland, New Zealand, https://www.rproject.org/) via the “TwoSampleMR” package (the package could be downloaded from the website: https://github.com/MRCIEU/TwoSampleMR).

3. Results

3.1. The selected SNPs in MR analysis

In a European population, a total of 16,380,466 SNPs were identified as being associated with appendiceal diseases (including appendicitis and appendiceal tumors, but excluding appendectomy status). After meeting the genome-wide significance threshold (P < 5 × 10⁻⁸), 5 instrumental variables were ultimately included in the two-sample MR analysis. The strength of these instruments was supported by an F-statistic of 64.39. These 5 SNPs in the exposure dataset were then used to estimate the effects of appendiceal diseases on various cancer outcomes (Table S1, Supplemental Digital content, https://links.lww.com/MD/R325).

3.2. Primary two-sample MR analysis

IVW analysis suggests a potential causal relationship between genetically predicted appendiceal disease and the incidence of rectal cancer, with the presence of appendiceal disorders associated with a reduced risk of rectal cancer (odds ratio [OR] = 0.588, 95% confidence interval [CI]: 0.392–0.881, P = .010, after Bonferroni correction; P = .05, by the IVW method). Conversely, rectal cancer does not appear to causally influence the development of appendiceal disease (OR = 0.999, 95% CI: 0.977–1.02, P = .934). Appendiceal disease also does not significantly increase the risk of other cancers, such as esophageal cancer. On the other hand, malignant neoplasms of the liver and intrahepatic bile ducts were more common among individuals with appendiceal disease; however, the association did not reach statistical significance (OR = 1.967, 95% CI: 0.926–4.176, P = .078, by the IVW method) (Fig. 2). The weighted median model analysis indicates that there is a causal relationship between genetically predicted disorders of the appendix and rectal cancer, with the presence of appendix diseases linked to a decreased risk of rectal cancer (OR = 0.592, 95% CI: 0.368–0.951, P = .030, by weighted median method) (Tables 2 and S2, Supplemental Digital content, https://links.lww.com/MD/R325).

Figure 2.

A forest plot of causal relationship between genetic prediction of appendiceal disease and risk of malignant tumors, by IVW method, indicates a causal relationship between appendiceal disease and rectal cancer (P < .05). CI = confidence interval, IVW = inverse variance weighted, MR = Mendelian randomization, OR = odds ratio.

Table 2.

Two-sample Mendelian randomization analysis results of diseases of appendix and pan-cancer via inverse variance weighted method.

Outcome	No. of SNPs	β	SEβ	OR (95% CI)	P value
Esophagus	5	0.279	0.439	1.322 (0.559–3.128)	.526
Stomach	5	0.173	0.394	1.188 (0.549–2.570)	.661
INTRAHEPATIC_BILE_DUCTS	5	0.676	0.384	1.967 (0.926–4.176)	.078
Pancreas	5	−0.0435	0.347	0.957 (0.485–1.889)	.900
Colon	5	0.189	0.160	1.208 (0.883–1.654)	.237
Rectum	5	−0.532	0.207	0.588 (0.392–0.881)	.010

CI = confidence interval, MR = Mendelian randomization, OR = odds ratio, SE = standard error, SNPs = single-nucleotide polymorphisms.

Bold values indicate statistical significance (P < .05).

3.3. Heterogeneity, pleiotropy, and sensitive analyses

Heterogeneity was assessed using Cochran’s Q test. The MR-Egger method identified heterogeneity in the effect of appendiceal diseases on gastric cancer (P = .041), while no heterogeneity was detected for the other diseases. To account for this issue, we adopted a random-effects IVW model as the primary analytical framework. Compared with the fixed-effects model, the random-effects approach incorporates between-instrument variability, allowing it to better accommodate heterogeneity in instrument effects and thereby provide more conservative and generalizable causal estimates. This modeling choice ensures that our conclusions remain robust even in the presence of heterogeneity (Table S3, Supplemental Digital content, https://links.lww.com/MD/R325).

Furthermore, no horizontal pleiotropy was seen in the exposure group (P > .05), suggesting that plausible confounding pathways had no effect on the results, which makes them robust and credible (Table S4, Supplemental Digital content, https://links.lww.com/MD/R325).

To evaluate the stability of the results for appendix illnesses, a leave-one-out analysis was conducted for each cancer outcome. The results of the leave-one-out analysis also align with those of the initial IVW analysis (Fig. 3). Moreover, funnel plots were created to display the distribution balance of the single SNP effects (Fig. 4).

Figure 3.

Leave-one-out analysis for appendiceal disease on cancers of the digestive. The vertical axis is SNP. The horizontal axis is the beta value; red and black dots represent beta value estimates. Solid lines represent the 95% CIs of the beta values. CI = confidence interval, MR = Mendelian randomization, SNP = single-nucleotide polymorphism.

Figure 4.

Funnel plots for appendiceal disease on cancers of the digestive. MR = Mendelian randomization, SE = standard error.

4. Discussion

This is, to the best of our knowledge, the 1st study to evaluate the causal association between appendiceal diseases and malignant digestive system cancers using a two-sample MR approach. We aimed to investigate the relationship between appendiceal disorders and 6 types of digestive system malignancies, including esophageal cancer, gastric cancer, malignant neoplasm of the liver and intrahepatic bile ducts, pancreatic cancer, colon cancer, and rectal cancer. Our results indicate that appendiceal diseases may reduce the risk of rectal cancer (OR = 0.588, by the IVW method; OR = 0.592, by the weighted median method). Conversely, rectal cancer does not appear to affect the risk of appendiceal diseases (OR = 0.999, P = .934, by the IVW method). No clear causal association was observed between appendiceal diseases and other digestive system malignancies.

The majority of research so far has established a connection between malignant digestive system cancers and the appendix. Researchers 1st considered the appendix as a protective device in 1964 after discovering, through retrospective analysis, that patients who died from colon cancer had a higher proportion of appendectomies than the control group.^[16]

A retrospective study conducted in 1983 also discovered that a greater percentage of men who had colon cancer had previously had appendectomies, suggesting that the appendix may protect men from colon cancer.^[17]

An increased risk of colon cancer was linked to appendectomy, according to an animal study. Additionally, they discovered that appendectomy patients with ulcerative colitis had a higher chance of colon cancer following the treatment.^[18]

According to a study conducted on an Asian population, colorectal cancer incidence rose by 14% following appendectomy as compared to the control group (P < .05). This finding supports our hypothesis that the appendix, even in the event of disease, provides protection against rectal cancer.^[19]

Simultaneously, a Swedish population cohort research found that individuals who underwent appendix surgery had a marginally higher risk of esophageal cancer than the whole population.^[20] An elevated relative risk of stomach cancer following appendectomy was noted in another Swedish cohort study involving children as opposed to the general population.^[21] Our findings do not demonstrate a direct causative link between the incidence of gastric and esophageal malignancies and appendiceal disease; however, this could be attributed to variations in the population and research design.

In our study, the potential mechanism by which appendiceal disease may specifically influence rectal cancer, rather than colon cancer, remains unclear. We speculate that this could be related to the appendix’s role as a reservoir for gut microbiota. Although the exact pathways are still unknown, our findings provide preliminary evidence supporting a potential biological basis for this selective association.

Traditional observational studies have several limitations.^[22] First, residual confounding cannot be ruled out. For example, appendiceal diseases and upper gastrointestinal cancers may share certain unknown inflammation- or immunity-related risk factors, meaning that associations observed in observational studies could be driven by these shared etiologies rather than true causal relationships. Second, the possibility of reverse causation cannot be excluded – for instance, atypical intestinal lesions might be misdiagnosed as appendiceal disease. Moreover, previous observational studies vary in population characteristics and study design, and often suffer from limited statistical power, which may lead to unstable association estimates. Therefore, we employed a two-sample MR approach to investigate the causal relationship between appendiceal diseases and digestive tract malignancies. By using genetic instrumental variables, this method better accounts for these sources of bias and provides more reliable evidence for assessing causality between the two. There are some connections between malignant digestive system tumors and the appendix as a component of the digestive system, according to current studies.^[16–21] When treating appendiceal disease in clinical practice, appendectomy is frequently the preferred course of action. Nonetheless, from a genetic perspective, an ill appendix also appears to offer defense against rectum malignant tumors in European people. Consequently, individuals with appendix illnesses who are also at a high risk for other rectal cancers should carefully decide whether to undergo an appendectomy.

The large sample size and single population distribution lessened the bias of population stratification, ensuring the accurate causal relationship between diseases of the appendix and the risk of incidence of malignant rectal tumors. We used heterogeneity and sensitivity analyses in our study to improve the stability of the results.

However, this study also has several limitations. First, because all participants were of European ancestry, additional data and analyses are needed to determine whether our findings can be generalized to other populations. Second, because of the lack of access to individual-level source data, we were unable to assess potential sample overlap and, therefore, cannot fully exclude bias arising from this issue. Third, although we employed multiple strategies to minimize confounding, the potential influence of horizontal pleiotropy cannot be completely ruled out. In addition, we were able to identify only 5 SNPs that reached genome-wide significance; relaxing the threshold would introduce additional instruments but may reduce the reliability of the results. Furthermore, given the broad clinical definition of appendiceal diseases, our conclusion requires further validation in future studies.

Given these constraints, we eagerly await additional studies to validate the association between appendix disorders and digestive system malignancies. Patients with appendix disorders must have a wide range of parameters carefully evaluated in order to determine whether surgical removal is necessary. Additionally, after treating appendix diseases, patients must closely monitor their digestive systems for malignancies.

5. Conclusion

In summary, this extensive pan-cancer MR analysis suggests a potential causal association between appendiceal diseases and rectal cancer in European populations. The impact of appendiceal diseases on other malignant tumors in different populations remains controversial. Therefore, researchers propose that, in Europeans, appendiceal diseases may have a potential causal relationship with rectal cancer and could possibly delay its onset to some extent. Consequently, the diagnosis and management of patients with appendiceal diseases may require additional considerations. Moreover, enhanced rectal cancer screening might be warranted following appendiceal disease treatment to prevent potential tumor development. We also emphasize the need for further studies to clarify the relationship between the two.

Acknowledgments

We thank all the investigators of the FinnGen biobank and the Medical Research Council’s Integrative Epidemiology Unit for exposing the data publicly.

Author contributions

Conceptualization: Feizhuan Lin, Kangze Wu.

Data curation: Li Bao, Baochun Lu, Zhihan Jiang.

Methodology: Feizhuan Lin.

Software: Kangze Wu.

Validation: Li Bao.

Writing – original draft: Feizhuan Lin.

Writing – review & editing: Kangze Wu.