Abstract
Multiple Sclerosis (MS) is an immune-related disease and the relationship between MS and cancer has raised attention. Previous studies of the relationship between MS and cancer have reached conflicting conclusions. In this study, the two-sample MR method is used to investigate whether MS has a causal correlation with cancers and offer scientific evidence for cancer prevention. Single nucleotide polymorphisms (SNPs) related to MS were obtained from the genome-wide association study (GWAS) based on International Multiple Sclerosis Genetics Consortium (IMSGC) and SNPs related to 15 types of cancers were obtained from the GWASs based on UK Biobank. Inverse variance weighted (IVW) method was mainly used to assess causal effects. Sensitivity analyses were conducted with Cochran’s Q-test, MR Egger intercept, leave-one-out test, and MR Steiger method. IVW analysis showed that MS was only associated with a marginal increased risk of cervical cancer (OR 1.0004, 95% CI 1.0002–1.0007, p = 0.0003). Sensitivity analyses showed that the results of MR analysis were robust and found no heterogeneity, no pleiotropy, and no reverse causation. In conclusion, this study finds no causal relationship between MS and 15 types of cancers except cervical cancer.
1. Introduction
Multiple Sclerosis (MS) is a chronic central nervous system demyelinating disease, originating from complex gene-environment interactions and characterized by the accumulation of demyelinating lesions in the white matter and the grey matter of the brain and spinal cord [1]. Self-reactive lymph cell, largely CD4+ T cells, plays significant roles in the MS autoimmune inflammatory chaos [2]. The onset of MS is associated with both genetic and environmental factors. Familial MS cases constitute 12.6% of all MS patients, and the incidence of MS in monozygotic twins (up to 30% [3]) is significantly higher than that in dizygotic twins [4], underscoring the genetic predispositon. Recent research has identified multiple genetic loci related to MS susceptibility [5]. Meanwhile, environmental factors such as viral infections [6], smoking [7], and sunlight [8] may also have a role in the onset of MS.
Cancer remains a predominant global health concern, ranking as the second leading cause of morbidity and mortality, imposing a substantial disease burden [9]. Recently, significant progress has been made in cancer treatment, but the harm of cancer to human health and quality of life still remains a challenge [10, 11]. There are evidences that immune-related diseases are closely related to cancer [12], and the relationship between MS and cancer has also raised attention. Existing research has shown that earlier cancer detection is associated with reduced mortality for certain cancers [13]. Thus, it is critical to investigate whether MS patients possess an increased risk for cancer. Previous studies of the relationship between MS and cancer have reached conflicting conclusions. Some studies have shown that MS may reduce the risk of cancer [14], such as lung cancer [15] and alimentary tract cancers [16], while other studies have shown that MS is a risk factor for cancer, both in overall cancers [17] and in certain cancers such as bladder cancer [18], breast cancer [19], brain tumors [20], etc. Interestingly, a cohort study has found that after starting disease-modifying therapies for the treatment of MS in the 1990s, the risk of cancer increases in patients with MS [21], suggesting a potential relationship between MS immune suppressive treatment and cancer risk.
Mendelian randomization (MR) study is a useful epidemiological method to evaluate the hypothetical correlation between exposure and outcome [22]. Single nucleotide polymorphisms (SNPs) are used as instrumental variables (IVs) to assume the correlation in MR study. Given the autonomous segregation and random assignment of alleles at meiosis, MR can effectively mitigate potential confounders and reverse causation, thus providing a more reliable causal inference [23]. Hence, MR is a powerful tool to investigate the correction between MS and cancer. In this study, the two-sample MR method is used to assess the potential causal relationship between MS and various cancers and offer scientific evidence for cancer prevention.
2. Methods
2.1 Study design and data sources
This study used the two-sample MR method to investigate the causal relationship between MS and cancer. The research flow chart was shown in Fig 1. The study is based on several hypotheses [24, 25]: first, genetic variation and MS are strongly correlated; second, genetic variation is independent of confounding factors; finally, genetic variants affect cancer risk only through MS and not through other ways.
Fig 1. An overview of the Mendelian randomization study.
The summary statistics of the GWAS used in this study were obtained from the Medical Research Council Integrative Epidemiology Unit OpenGWAS project (https://gwas.mrcieu.ac.uk/). Fifteen types of solid tumors (prostate cancer, breast cancer, bladder cancer, brain cancer, cervical cancer, laryngeal cancer, liver & bile duct cancer, lung cancer, malignant non-melanoma skin cancer, oesophageal cancer, oral cavity cancer, ovarian cancer, colorectal cancer, oropharyngeal cancer, melanoma skin cancer) were included as outcome traits, and these data were based on the UK Biobank consortium [26]. UK Biobank is a population-based prospective cohort which recruited more than 500000 UK volunteers aged 40 to 69 years and collected their genetic and health information [27]. The data of MS was based on International Multiple Sclerosis Genetics Consortium (IMSGC, https://imsgc.net/). IMSGC is an international collaborative organization dedicated to a comprehensively understanding of the genetic influences on etiology, pathogenesis, clinical course and treatment response of multiple sclerosis [28]. The summary of the data sources of exposure and outcome traits were shown in S1 Table. All data used in this study were obtained from publicly available databases; further ethical approval was not required.
2.2 Selection of instrumental variables
SNPs meeting the following criteria were selected as instrumental variables: (1) SNPs should be significantly correlated with multiple sclerosis (p < 1×10−6). (2) SNPs should be independent with each other to avoid the impact of linkage disequilibrium (LD), which means R2 < 0.001 and genetic distance kb > 10000. (3) SNPs should not be correlated with any confounders and outcomes. Phenoscanner database (http://www.phenoscanner.medschl.cam.ac.uk/) was used to identify associations with confounders and outcomes. We calculated F-statistic to evaluate the strength of the instruments and F-statistic > 10 indicated sufficient strength [29]. The calculation is: F = 2 × MAF × (1−MAF) β2 (N−2) / [1−2 × MAF × (1−MAF) β2], MAF is the minor allele frequency, β is the estimated corresponding effect and N is the exposure GWAS sample size.
2.3 Mendelian randomization analysis
Mendelian randomization analysis was performed by the TwoSampleMR package with R version 4.2.2. We mainly used inverse variance weighted (IVW) method to assess causal effects, but also referred to the results of four other methods (MR Egger, weighted median, simple mode, weighted mode). Results were expressed as odds ratios (OR) and 95% confidence intervals (CI). Bonferroni correction was performed on the p values of multiple testing. The significance threshold was set at 3.33×10−3 (0.05/15). In sensitivity analysis, we used Cochran’s Q-test to assess heterogeneity, MR Egger intercept for pleiotropy [30], leave-one-out test to assess robustness of results, and MR Steiger method to conduct the directionality test [31].
3. Results
We screened out 107 SNPs from the GWAS data of the exposure factors, all of which had a strong correlation with MS (p < 1×10−6), and the 107 SNPs were independent of each other without LD (S2 Table). Upon a review of the literature related to MS and cancer, we identified several potential confounders. Evidence suggests that smoking is a risk factor for MS [7], with risks increasing with cumulative smoking exposure [32]. Prolonged exposure to tobacco can accelerate the progression of MS [33]. Smoking is also a recognized cancer risk factor [34]. Epstein-Barr virus infection significantly elevates the risk for MS [6] and is also a risk factor for several cancers [35], such as nasopharyngeal carcinoma [36], gastric cancer [37], and lymphomas [38]. We searched the Phenoscanner database for phenotypes associated with the screened SNPs and found 5 confounding factors that might be associated with MS and cancer. The 102 SNPs were normalized to the outcome data to obtain the final IVs. All IVs had F-statistics > 10, indicating that they were not weak instruments.
The results of the causal analysis of MS and 15 types of cancers were shown in the Table 1. The determination of causality mainly referred to the results of IVW analysis. In general, the MR results showed no causal relationship between MS and almost all cancers. IVW analysis showed that MS was only associated with higher risk of cervical cancer (OR 1.0004, 95% CI 1.0002–1.0007, p = 0.0003). After multiple testing correction, the p value of cervical cancer was still less than 3.33×10−3 (0.05/15), indicating that MS was a risk factor for cervical cancer. Scatter plots of the MR results were shown in the Fig 2.
Table 1. Causal analysis of MS and 15 types of cancers.
| Outcome | Method | SNP | OR (95% CI) | p value |
|---|---|---|---|---|
| Prostate cancer | MR Egger | 92 | 1.0010(0.9990, 1.0029) | 0.3357 |
| Weighted median | 92 | 1.0016(0.9999, 1.0032) | 0.0652 | |
| Inverse variance weighted | 92 | 0.9998(0.9986, 1.0011) | 0.8037 | |
| Simple mode | 92 | 1.0009(0.9971, 1.0048) | 0.6310 | |
| Weighted mode | 92 | 1.0013(0.9995, 1.0032) | 0.1652 | |
| Breast cancer | MR Egger | 92 | 0.9996(0.9975, 1.0017) | 0.7057 |
| Weighted median | 92 | 0.9991(0.9972, 1.0010) | 0.3348 | |
| Inverse variance weighted | 92 | 1.0004(0.9991, 1.0017) | 0.5484 | |
| Simple mode | 92 | 0.9986(0.9945, 1.0027) | 0.5125 | |
| Weighted mode | 92 | 0.9989(0.9968, 1.0011) | 0.3426 | |
| Bladder cancer | MR Egger | 92 | 1.0000(0.9997, 1.0004) | 0.8530 |
| Weighted median | 92 | 1.0000(0.9996, 1.0003) | 0.8011 | |
| Inverse variance weighted | 92 | 1.0001(0.9999, 1.0003) | 0.3744 | |
| Simple mode | 92 | 1.0001(0.9994, 1.0008) | 0.7807 | |
| Weighted mode | 92 | 1.0000(0.9996, 1.0003) | 0.8985 | |
| brain cancer | MR Egger | 90 | 0.9999(0.9997, 1.0001) | 0.4464 |
| Weighted median | 90 | 1.0000(0.9998, 1.0002) | 0.8707 | |
| Inverse variance weighted | 90 | 1.0000(0.9998, 1.0001) | 0.8739 | |
| Simple mode | 90 | 1.0001(0.9996, 1.0006) | 0.6176 | |
| Weighted mode | 90 | 1.0000(0.9998, 1.0002) | 0.9492 | |
| cervical cancer | MR Egger | 90 | 1.0004(1.0001, 1.0008) | 0.0141 |
| Weighted median | 90 | 1.0004(1.0001, 1.0008) | 0.0202 | |
| Inverse variance weighted | 90 | 1.0004(1.0002, 1.0007) | 0.0003 | |
| Simple mode | 90 | 1.0002(0.9994, 1.0010) | 0.6203 | |
| Weighted mode | 90 | 1.0004(1.0001, 1.0007) | 0.0202 | |
| Laryngeal cancer | MR Egger | 87 | 1.0000(0.9999, 1.0002) | 0.5282 |
| Weighted median | 87 | 1.0000(0.9999, 1.0002) | 0.6518 | |
| Inverse variance weighted | 87 | 1.0000(0.9999, 1.0001) | 0.8215 | |
| Simple mode | 87 | 1.0000(0.9997, 1.0004) | 0.9377 | |
| Weighted mode | 87 | 1.0000(0.9999, 1.0001) | 0.7312 | |
| Liver & bile duct cancer | MR Egger | 89 | 1.0000(0.9999, 1.0002) | 0.7995 |
| Weighted median | 89 | 1.0001(0.9999, 1.0002) | 0.3577 | |
| Inverse variance weighted | 89 | 1.0000(0.9999, 1.0001) | 0.3864 | |
| Simple mode | 89 | 0.9997(0.9993, 1.0001) | 0.2104 | |
| Weighted mode | 89 | 1.0001(0.9999, 1.0002) | 0.3412 | |
| Lung cancer | MR Egger | 92 | 1.0004(0.9999, 1.0010) | 0.1352 |
| Weighted median | 92 | 1.0003(0.9997, 1.0009) | 0.2956 | |
| Inverse variance weighted | 92 | 1.0003(1.0000, 1.0007) | 0.0498 | |
| Simple mode | 92 | 0.9994(0.9980, 1.0007) | 0.3533 | |
| Weighted mode | 92 | 1.0002(0.9997, 1.0008) | 0.4329 | |
| Malignant non-melanoma skin cancer | MR Egger | 92 | 1.0014(0.9992, 1.0035) | 0.2245 |
| Weighted median | 92 | 1.0008(0.9993, 1.0024) | 0.3020 | |
| Inverse variance weighted | 92 | 1.0003(0.9990, 1.0016) | 0.6314 | |
| Simple mode | 92 | 1.0027(0.9994, 1.0060) | 0.1081 | |
| Weighted mode | 92 | 1.0012(0.9997, 1.0027) | 0.1099 | |
| Oesophageal cancer | MR Egger | 90 | 1.0002(1.0000, 1.0004) | 0.0879 |
| Weighted median | 90 | 1.0001(0.9999, 1.0003) | 0.2945 | |
| Inverse variance weighted | 90 | 1.0000(0.9999, 1.0002) | 0.5105 | |
| Simple mode | 90 | 1.0003(0.9998, 1.0008) | 0.2556 | |
| Weighted mode | 90 | 1.0001(0.9999, 1.0003) | 0.2578 | |
| Oral cavity cancer | MR Egger | 90 | 0.9998(0.9997, 1.0000) | 0.0270 |
| Weighted median | 90 | 0.9998(0.9997, 1.0000) | 0.0373 | |
| Inverse variance weighted | 90 | 1.0000(0.9999, 1.0001) | 0.6826 | |
| Simple mode | 90 | 1.0000(0.9996, 1.0004) | 0.9636 | |
| Weighted mode | 90 | 0.9999(0.9997, 1.0000) | 0.0792 | |
| Ovarian cancer | MR Egger | 92 | 1.0001(0.9994, 1.0008) | 0.8158 |
| Weighted median | 92 | 1.0002(0.9996, 1.0008) | 0.5159 | |
| Inverse variance weighted | 92 | 1.0000(0.9995, 1.0004) | 0.9004 | |
| Simple mode | 92 | 0.9995(0.9983, 1.0007) | 0.4159 | |
| Weighted mode | 92 | 1.0001(0.9995, 1.0007) | 0.8308 | |
| Colorectal cancer | MR Egger | 92 | 1.0008(1.0000, 1.0016) | 0.0547 |
| Weighted median | 92 | 1.0003(0.9995, 1.0011) | 0.4108 | |
| Inverse variance weighted | 92 | 1.0003(0.9998, 1.0008) | 0.2711 | |
| Simple mode | 92 | 1.0002(0.9987, 1.0018) | 0.7564 | |
| Weighted mode | 92 | 1.0006(0.9998, 1.0013) | 0.1485 | |
| Oropharyngeal cancer | MR Egger | 90 | 1.0002(1.0000, 1.0003) | 0.0712 |
| Weighted median | 90 | 1.0002(1.0000, 1.0003) | 0.0510 | |
| Inverse variance weighted | 90 | 1.0001(1.0000, 1.0002) | 0.0772 | |
| Simple mode | 90 | 1.0001(0.9996, 1.0005) | 0.7052 | |
| Weighted mode | 90 | 1.0002(1.0000, 1.0003) | 0.0461 | |
| Melanoma skin cancer | MR Egger | 92 | 1.0005(0.9999, 1.0011) | 0.1248 |
| Weighted median | 92 | 1.0003(0.9997, 1.0009) | 0.3979 | |
| Inverse variance weighted | 92 | 1.0000(0.9996, 1.0004) | 0.9544 | |
| Simple mode | 92 | 0.9991(0.9978, 1.0005) | 0.2228 | |
| Weighted mode | 92 | 1.0007(1.0001, 1.0013) | 0.0216 |
Fig 2.
Scatter plots of the MR results: (a) prostate cancer, (b) breast cancer, (c) bladder cancer, (d) brain cancer, (e) cervical cancer, (f) laryngeal cancer, (g) liver & bile duct cancer, (h) lung cancer, (i) malignant non-melanoma skin cancer, (j) oesophageal cancer, (k) oral cavity cancer, (l) ovarian cancer, (m) colorectal cancer, (n) oropharyngeal cancer, (o) melanoma skin cancer.
Sensitivity analyses were conducted to verify the reliability of MR results. The Cochran’s Q-test found no heterogeneity in cervical cancer (Table 2), and the MR Egger intercept method found no pleiotropy in cervical cancer (Table 3). The MR Steiger directionality test found no reverse causation (Table 4), and the leave-one-out test showed that the results of MR analysis were robust (Fig 3).
Table 2. Cochran’s Q-test.
| Outcome | Method | Q | Q_df | Q_pval |
|---|---|---|---|---|
| Prostate cancer | MR Egger | 111.2732 | 90 | 0.0637 |
| Inverse variance weighted | 113.7652 | 91 | 0.0534 | |
| Breast cancer | MR Egger | 116.6085 | 90 | 0.0311 |
| Inverse variance weighted | 117.7485 | 91 | 0.0311 | |
| Bladder cancer | MR Egger | 106.7965 | 90 | 0.1092 |
| Inverse variance weighted | 107.0348 | 91 | 0.1202 | |
| brain cancer | MR Egger | 115.0544 | 88 | 0.0280 |
| Inverse variance weighted | 116.1095 | 89 | 0.0284 | |
| cervical cancer | MR Egger | 98.0164 | 88 | 0.2183 |
| Inverse variance weighted | 98.0173 | 89 | 0.2407 | |
| Laryngeal cancer | MR Egger | 55.6793 | 85 | 0.9942 |
| Inverse variance weighted | 56.8075 | 86 | 0.9936 | |
| Liver & bile duct cancer | MR Egger | 92.0023 | 87 | 0.3363 |
| Inverse variance weighted | 92.2195 | 88 | 0.3582 | |
| Lung cancer | MR Egger | 119.6247 | 90 | 0.0201 |
| Inverse variance weighted | 119.8344 | 91 | 0.0231 | |
| Malignant non-melanoma skin cancer | MR Egger | 246.0540 | 90 | 0.0000 |
| Inverse variance weighted | 249.8197 | 91 | 0.0000 | |
| Oesophageal cancer | MR Egger | 88.0616 | 88 | 0.4781 |
| Inverse variance weighted | 91.0891 | 89 | 0.4187 | |
| Oral cavity cancer | MR Egger | 91.4637 | 88 | 0.3791 |
| Inverse variance weighted | 98.9387 | 89 | 0.2211 | |
| Ovarian cancer | MR Egger | 128.8383 | 90 | 0.0046 |
| Inverse variance weighted | 129.0560 | 91 | 0.0054 | |
| Colorectal cancer | MR Egger | 124.5669 | 90 | 0.0093 |
| Inverse variance weighted | 128.1216 | 91 | 0.0063 | |
| Oropharyngeal cancer | MR Egger | 85.1771 | 88 | 0.5654 |
| Inverse variance weighted | 85.8976 | 89 | 0.5734 | |
| Melanoma skin cancer | MR Egger | 114.4810 | 90 | 0.0418 |
| Inverse variance weighted | 119.1332 | 91 | 0.0255 |
Table 3. MR Egger intercept.
| Outcome | Egger_intercept | SE | pval |
|---|---|---|---|
| Prostate cancer | -2.25E-04 | 1.59E-04 | 0.1591 |
| Breast cancer | 1.61E-04 | 1.72E-04 | 0.3507 |
| Bladder cancer | 1.31E-05 | 2.93E-05 | 0.6552 |
| brain cancer | 1.71E-05 | 1.90E-05 | 0.3714 |
| cervical cancer | -9.20E-07 | 3.16E-05 | 0.9768 |
| Laryngeal cancer | -1.21E-05 | 1.14E-05 | 0.2912 |
| Liver & bile duct cancer | 5.91E-06 | 1.31E-05 | 0.6516 |
| Lung cancer | -1.77E-05 | 4.46E-05 | 0.6921 |
| Malignant non-melanoma skin cancer | -2.06E-04 | 1.75E-04 | 0.2436 |
| Oesophageal cancer | -3.19E-05 | 1.84E-05 | 0.0855 |
| Oral cavity cancer | 3.49E-05 | 1.30E-05 | 0.0087 |
| Ovarian cancer | -2.30E-05 | 5.89E-05 | 0.6974 |
| Colorectal cancer | -1.05E-04 | 6.55E-05 | 0.1125 |
| Oropharyngeal cancer | -1.27E-05 | 1.50E-05 | 0.3983 |
| Melanoma skin cancer | -9.85E-05 | 5.15E-05 | 0.0590 |
Table 4. MR Steiger directionality test.
| Outcome | Snp_r2 (exposure) | Snp_r2 (outcome) | Correct causal direction | Steiger pval |
|---|---|---|---|---|
| Prostate cancer | 0.0453 | 0.0008 | TRUE | 0 |
| Breast cancer | 0.0453 | 0.0006 | TRUE | 0 |
| Bladder cancer | 0.0453 | 0.0003 | TRUE | 0 |
| brain cancer | 0.0443 | 0.0003 | TRUE | 0 |
| cervical cancer | 0.0443 | 0.0006 | TRUE | 0 |
| Laryngeal cancer | 0.0429 | 0.0002 | TRUE | 0 |
| Liver & bile duct cancer | 0.0434 | 0.0003 | TRUE | 0 |
| Lung cancer | 0.0453 | 0.0004 | TRUE | 0 |
| Malignant non-melanoma skin cancer | 0.0453 | 0.0008 | TRUE | 0 |
| Oesophageal cancer | 0.0443 | 0.0003 | TRUE | 0 |
| Oral cavity cancer | 0.0443 | 0.0003 | TRUE | 0 |
| Ovarian cancer | 0.0453 | 0.0007 | TRUE | 0 |
| Colorectal cancer | 0.0453 | 0.0004 | TRUE | 0 |
| Oropharyngeal cancer | 0.0443 | 0.0003 | TRUE | 0 |
| Melanoma skin cancer | 0.0453 | 0.0003 | TRUE | 0 |
Fig 3.
Leave-one-out test results: (a) prostate cancer, (b) breast cancer, (c) bladder cancer, (d) brain cancer, (e) cervical cancer, (f) laryngeal cancer, (g) liver & bile duct cancer, (h) lung cancer, (i) malignant non-melanoma skin cancer, (j) oesophageal cancer, (k) oral cavity cancer, (l) ovarian cancer, (m) colorectal cancer, (n) oropharyngeal cancer, (o) melanoma skin cancer.
4. Discussions
In this study, we evaluated the causal relationship between MS and 15 types of cancers. Our findings indicate that genetic susceptibility to MS is associated with an increased risk of cervical cancer and no other types of cancer. Sensitivity analyses confirmed that these outcomes were robust.
MR employs genetic variations as IVs to assess causality between the exposure and the outcome. Since alleles are randomly assigned to offspring, MR can effectively avoid reverse causation. A previous study used MR to analyze the relationship between MS and lung cancer [39], and found that MS increased the risk of overall lung cancer but there were no significant causal relationships between MS and lung adenocarcinoma or squamous cell cancer. By contrast, no causal association between MS and lung cancer risk was found in our study. The difference was possibly due to heterogeneity in the GWAS population included in the two studies. Another study explored the association between MS and breast cancer based on MR analysis and the result does not support the correlation between them [40], which is consistent with our results.
We found that MS was associated with a marginal increased risk of cervical cancer through MR and remained significant after multiple testing correction. A previous cohort study identified that cervical benign cellular changes, considered as a premalignant condition, was more frequently observed in MS patients treated with cytotoxic immunosuppressive agents [41]. A clinical trial on the safety and tolerability of cladribine tablets in multiple sclerosis reported pre-malignant cervical carcinoma in situ [42]. Additionally, there are several case reports of patients developing cervical dysplasia during natalizumab therapy [43], and in some instances, rapidly progressing to squamous cell carcinoma of the cervix [44]. Researchers suspect that the occurrence of premalignant conditions and cervical cancer might be associated with the drugs used in MS treatment. The pathogenesis of MS involves disrupted immune function; hence, most current therapies, including corticosteroids and disease-modifying therapies (DMTs), have immunosuppressive effects. Some agents with immunosuppressive effect (i.e. S1P receptor modulators) might reduce the number of lymphocytes that are needed to identify and eliminate malignant cells [45]. The immunosuppression may also hinder the body’s ability to clear HPV [46], a well-known risk factor for cervical cancer. Besides, medications for treating MS could promote extramedullary hematopoiesis and myeloid-derived suppressor cells accumulation, thus it might have a protumorigenic potential [47]. However, it’s crucial to note that, according to our analysis of publicly available GWAS data, MS is associated with only a slight increase (0.04%) in cervical cancer risk. Due to the limitations of data sources and sample size, obtaining a more precise odds ratio is challenging and necessitates further epidemiological studies. Although the identified increase in risk is minimal, the growing body of evidence suggesting a potential rise in cervical cancer risk due to MS medications underscores the need for vigilant post-treatment monitoring, including routine cervical HPV testing, histological examination, and HPV vaccination for MS patients.
Compared with conventional epidemiological studies, our study’s strength lies in its application of MR to discern the causal relationship between MS and cancer. By using genetic variations as the IVs, we minimized the impacts of confounders and reverse causation. In this study, there is small sample overlap, ensuring the independence of exposure and outcomes. Different methods are used for sensitivity analyses to find out potential pleiotropy and heterogeneity, making MR results more robust. Compared to previous MR studies on MS and cancer, this study covers a wider range of cancer types.
However, our study also has some limitations. First, the populations included were exclusively of European ancestry, limiting its generalizability to other ethnicities. Second, although we excluded potential confounding factors and the influence of gene pleiotropy on the results as much as possible, residual effects might exist. Third, drug use in MS patients is not recorded and a deeper analysis of the effects of drugs is not possible. Fourth, the increased risk of cervical cancer is marginal and more studies are required to confirm this finding. Lastly, while MR analysis offers a preliminary insight into the MS-cancer causality, the intricate biological mechanisms underpinning this relationship remain elusive and warrant further exploration in laboratory settings.
5. Conclusion
In conclusion, by using MR method, this study reveals that MS is only causally associated with a marginal increased risk of cervical cancer and shows no association with other types of cancer. The specific biological mechanism remains to be studied.
Supporting information
(DOCX)
(DOCX)
Abbreviations
- MS
multiple sclerosis
- MR
Mendelian randomization
- SNP
single nucleotide polymorphism
- IV
instrumental variable
- LD
linkage disequilibrium
- GWAS
genome-wide association study
- IVW
inverse variance weight
- OR
odds ratio
- CI
confidence interval
- DMT
disease-modifying therapy
Data Availability
The data presented in this study are available in article and Supplementary Material. Publicly available GWAS data on cancers and multiple sclerosis were obtained from the Medical Research Council Integrative Epidemiology Unit OpenGWAS project (https://gwas.mrcieu.ac.uk/, accessed on 20 December 2023).
Funding Statement
This study was supported by the Guangzhou Municipal Science and Technology Bureau (202206080012 to B.Z.) and China postdoctoral science foundation (2022M723682 to T.F.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Filippi M., et al., Multiple sclerosis. Nat Rev Dis Primers, 2018. 4(1): p. 43. doi: 10.1038/s41572-018-0041-4 [DOI] [PubMed] [Google Scholar]
- 2.Leray E., et al., Evidence for a two-stage disability progression in multiple sclerosis. Brain, 2010. 133(Pt 7): p. 1900–13. doi: 10.1093/brain/awq076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sadovnick A.D., et al., A population-based study of multiple sclerosis in twins: update. Annals of neurology, 1993. 33(3): p. 281–5. doi: 10.1002/ana.410330309 [DOI] [PubMed] [Google Scholar]
- 4.Harirchian M.H., et al., Worldwide prevalence of familial multiple sclerosis: A systematic review and meta-analysis. Multiple Sclerosis and Related Disorders, 2018. 20: p. 43–47. doi: 10.1016/j.msard.2017.12.015 [DOI] [PubMed] [Google Scholar]
- 5.Moutsianas L., et al., Class II HLA interactions modulate genetic risk for multiple sclerosis. Nature Genetics, 2015. 47(10): p. 1107-+. doi: 10.1038/ng.3395 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bjornevik K., et al., Longitudinal analysis reveals high prevalence of Epstein-Barr virus associated with multiple sclerosis. Science, 2022. 375(6578): p. 296-+. doi: 10.1126/science.abj8222 [DOI] [PubMed] [Google Scholar]
- 7.Handel A.E., et al., Smoking and Multiple Sclerosis: An Updated Meta-Analysis. Plos One, 2011. 6(1). doi: 10.1371/journal.pone.0016149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kampman M.T., Wilsgaard T., and Mellgren S.I., Outdoor activities and diet in childhood and adolescence relate to MS risk above the Arctic Circle. Journal of Neurology, 2007. 254(4): p. 471–477. doi: 10.1007/s00415-006-0395-5 [DOI] [PubMed] [Google Scholar]
- 9.Siegel R.L., Miller K.D., and Jemal A., Cancer statistics, 2020. CA Cancer J Clin, 2020. 70(1): p. 7–30. doi: 10.3322/caac.21590 [DOI] [PubMed] [Google Scholar]
- 10.Mun E.J., et al., Tumor-Treating Fields: A Fourth Modality in Cancer Treatment. Clin Cancer Res, 2018. 24(2): p. 266–275. doi: 10.1158/1078-0432.CCR-17-1117 [DOI] [PubMed] [Google Scholar]
- 11.Ilbawi A.M. and Anderson B.O., Cancer in global health: how do prevention and early detection strategies relate? Sci Transl Med, 2015. 7(278): p. 278cm1. doi: 10.1126/scitranslmed.3008853 [DOI] [PubMed] [Google Scholar]
- 12.Zhou Z., et al., The five major autoimmune diseases increase the risk of cancer: epidemiological data from a large-scale cohort study in China. Cancer Commun (Lond), 2022. 42(5): p. 435–446. doi: 10.1002/cac2.12283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yang J.H. and Graves J.S., Cancer Screening in Patients With Multiple Sclerosis: Are We Doing Enough? Neurology, 2022. 98(18): p. 737–738. doi: 10.1212/WNL.0000000000200366 [DOI] [PubMed] [Google Scholar]
- 14.Ghajarzadeh M., Mohammadi A., and Sahraian M.A., Risk of cancer in multiple sclerosis (MS): A systematic review and meta-analysis. Autoimmun Rev, 2020. 19(10): p. 102650. doi: 10.1016/j.autrev.2020.102650 [DOI] [PubMed] [Google Scholar]
- 15.Handel A.E., Joseph A., and Ramagopalan S.V., Multiple sclerosis and lung cancer: an unexpected inverse association. QJM, 2010. 103(8): p. 625–6. doi: 10.1093/qjmed/hcq071 [DOI] [PubMed] [Google Scholar]
- 16.Landgren A.M., et al., Autoimmune disease and subsequent risk of developing alimentary tract cancers among 4.5 million US male veterans. Cancer, 2011. 117(6): p. 1163–71. doi: 10.1002/cncr.25524 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bosco-Levy P., et al., Incidence and risk of cancer among multiple sclerosis patients: A matched population-based cohort study. Eur J Neurol, 2022. 29(4): p. 1091–1099. doi: 10.1111/ene.15226 [DOI] [PubMed] [Google Scholar]
- 18.Marrie R.A., et al., Cancer Incidence and Mortality Rates in Multiple Sclerosis: A Matched Cohort Study. Neurology, 2021. 96(4): p. e501–e512. doi: 10.1212/WNL.0000000000011219 [DOI] [PubMed] [Google Scholar]
- 19.Nielsen N.M., et al., Cancer risk among patients with multiple sclerosis: a population-based register study. Int J Cancer, 2006. 118(4): p. 979–84. doi: 10.1002/ijc.21437 [DOI] [PubMed] [Google Scholar]
- 20.Bahmanyar S., et al., Cancer risk among patients with multiple sclerosis and their parents. Neurology, 2009. 72(13): p. 1170–7. doi: 10.1212/01.wnl.0000345366.10455.62 [DOI] [PubMed] [Google Scholar]
- 21.Grytten N., et al., Incidence of cancer in multiple sclerosis before and after the treatment era- a registry- based cohort study. Mult Scler Relat Disord, 2021. 55: p. 103209. doi: 10.1016/j.msard.2021.103209 [DOI] [PubMed] [Google Scholar]
- 22.Swanson S.A., et al., Nature as a Trialist?: Deconstructing the Analogy Between Mendelian Randomization and Randomized Trials. Epidemiology, 2017. 28(5): p. 653–659. doi: 10.1097/EDE.0000000000000699 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Davies N.M., et al., Within family Mendelian randomization studies. Hum Mol Genet, 2019. 28(R2): p. R170–R179. doi: 10.1093/hmg/ddz204 [DOI] [PubMed] [Google Scholar]
- 24.Bowden J., Davey Smith G., and Burgess S., Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol, 2015. 44(2): p. 512–25. doi: 10.1093/ije/dyv080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Burgess S., Small D.S., and Thompson S.G., A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res, 2017. 26(5): p. 2333–2355. doi: 10.1177/0962280215597579 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mullard A., The UK Biobank at 20. Nat Rev Drug Discov, 2022. 21(9): p. 628–629. doi: 10.1038/d41573-022-00137-8 [DOI] [PubMed] [Google Scholar]
- 27.Allen N.E., et al., UK biobank data: come and get it. Sci Transl Med, 2014. 6(224): p. 224ed4. doi: 10.1126/scitranslmed.3008601 [DOI] [PubMed] [Google Scholar]
- 28.Consortium I.M.S.G., Multiple sclerosis genomic map implicates peripheral immune cells and microglia in susceptibility. Science, 2019. 365(6460). doi: 10.1126/science.aav7188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Pongdee T., et al., White blood cells and chronic rhinosinusitis: a Mendelian randomization study. Allergy Asthma Clin Immunol, 2022. 18(1): p. 98. doi: 10.1186/s13223-022-00739-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rees J.M.B., Wood A.M., and Burgess S., Extending the MR-Egger method for multivariable Mendelian randomization to correct for both measured and unmeasured pleiotropy. Stat Med, 2017. 36(29): p. 4705–4718. doi: 10.1002/sim.7492 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hemani G., Tilling K., and Davey Smith G., Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet, 2017. 13(11): p. e1007081. doi: 10.1371/journal.pgen.1007081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hedstrom A.K., et al., Tobacco smoking, but not Swedish snuff use, increases the risk of multiple sclerosis. Neurology, 2009. 73(9): p. 696–701. doi: 10.1212/WNL.0b013e3181b59c40 [DOI] [PubMed] [Google Scholar]
- 33.Ramanujam R., et al., Effect of Smoking Cessation on Multiple Sclerosis Prognosis. Jama Neurology, 2015. 72(10): p. 1117–1123. doi: 10.1001/jamaneurol.2015.1788 [DOI] [PubMed] [Google Scholar]
- 34.Caliri A.W., Tommasi S., and Besaratinia A., Relationships among smoking, oxidative stress, inflammation, macromolecular damage, and cancer. Mutation Research-Reviews in Mutation Research, 2021. 787. doi: 10.1016/j.mrrev.2021.108365 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Farrell P.J., Epstein- Barr Virus and Cancer, in Annual Review of Pathology: Mechanisms of Disease, Vol 14, Abbas A.K., Aster J.C., and Feany M.B., Editors. 2019. p. 29–53. [DOI] [PubMed] [Google Scholar]
- 36.Tsao S.W., Tsang C.M., and Lo K.W., Epstein-Barr virus infection and nasopharyngeal carcinoma. Philosophical Transactions of the Royal Society B-Biological Sciences, 2017. 372(1732). doi: 10.1098/rstb.2016.0270 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Saito M. and Kono K., Landscape of EBV-positive gastric cancer. Gastric Cancer, 2021. 24(5): p. 983–989. doi: 10.1007/s10120-021-01215-3 [DOI] [PubMed] [Google Scholar]
- 38.Vockerodt M., et al., The Epstein-Barr virus and the pathogenesis of lymphoma. Journal of Pathology, 2015. 235(2): p. 312–322. doi: 10.1002/path.4459 [DOI] [PubMed] [Google Scholar]
- 39.Ge F., et al., Lung cancer risk in patients with multiple sclerosis: a Mendelian randomization analysis. Mult Scler Relat Disord, 2021. 51: p. 102927. [DOI] [PubMed] [Google Scholar]
- 40.Fang T., et al., Multiple sclerosis and breast cancer risk: a meta-analysis of observational and Mendelian randomization studies. Front Neuroinform, 2023. 17: p. 1154916. doi: 10.3389/fninf.2023.1154916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Doosti R., et al., Evaluation of the risk of cervical cancer in patients with Multiple Sclerosis treated with cytotoxic agents: A cohort study. Iran J Neurol, 2018. 17(2): p. 64–70. [PMC free article] [PubMed] [Google Scholar]
- 42.Cook S., et al., Safety and tolerability of cladribine tablets in multiple sclerosis: the CLARITY (CLAdRIbine Tablets treating multiple sclerosis orallY) study. Mult Scler, 2011. 17(5): p. 578–93. doi: 10.1177/1352458510391344 [DOI] [PubMed] [Google Scholar]
- 43.Durrieu G., et al., Cervical dysplasia in a patient with multiple sclerosis treated with natalizumab. Fundam Clin Pharmacol, 2019. 33(1): p. 125–126. doi: 10.1111/fcp.12394 [DOI] [PubMed] [Google Scholar]
- 44.Wan K.M. and Oehler M.K., Rapid Progression of Low-Grade Cervical Dysplasia into Invasive Cancer during Natalizumab Treatment for Relapsing Remitting Multiple Sclerosis. Case Rep Oncol, 2019. 12(1): p. 59–62. doi: 10.1159/000496198 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Killestein J., et al., Five cases of malignant melanoma during fingolimod treatment in Dutch patients with MS. Neurology, 2017. 89(9): p. 970–972. doi: 10.1212/WNL.0000000000004293 [DOI] [PubMed] [Google Scholar]
- 46.Triplett J., et al., Warts and all: Fingolimod and unusual HPV-associated lesions. Mult Scler, 2019. 25(11): p. 1547–1550. doi: 10.1177/1352458518807088 [DOI] [PubMed] [Google Scholar]
- 47.Li Y., et al., The protumorigenic potential of FTY720 by promoting extramedullary hematopoiesis and MDSC accumulation. Oncogene, 2017. 36(26): p. 3760–3771. doi: 10.1038/onc.2017.2 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX)
(DOCX)
Data Availability Statement
The data presented in this study are available in article and Supplementary Material. Publicly available GWAS data on cancers and multiple sclerosis were obtained from the Medical Research Council Integrative Epidemiology Unit OpenGWAS project (https://gwas.mrcieu.ac.uk/, accessed on 20 December 2023).