Abstract

Purpose

Lynch syndrome is the most common hereditary cause of colorectal and endometrial cancers. Modifiable risk factors, including obesity, physical activity, alcohol intake, and smoking, are well-established in sporadic cancers but are less studied in Lynch syndrome.

Methods

Searches were conducted on MEDLINE, Embase, and Web of Science for cohort studies that investigated the association between modifiable risk factors and the risk of colorectal or endometrial cancer in people with Lynch syndrome. Adjusted hazard ratios (HRs) and 95% CIs for colorectal and endometrial cancers were pooled using a random effects model. The protocol was prospectively registered on PROSPERO (CRD 42022378462), and the meta-analysis was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses and Meta-Analysis of Observational Studies in Epidemiology reporting guidelines.

Results

A total of 770 citations were reviewed. Eighteen studies were identified for qualitative synthesis, with seven colorectal cancer (CRC) studies eligible for meta-analysis. Obesity (HR, 2.38 [95% CI, 1.52 to 3.73]) was associated with increased CRC risk. There was no increased CRC risk associated with smoking (HR, 1.04 [95% CI, 0.82 to 1.32]) or alcohol intake (HR, 1.32 [95% CI, 0.97 to 1.81]). Type 2 diabetes mellitus (T2DM) and some dietary factors might increase risk of CRC although more studies are needed. In a qualitative synthesis of three endometrial cancer cohort studies, female hormonal risk factors and T2DM may affect the risk of endometrial cancer, but obesity was not associated with an increased risk.

Conclusion

Lifestyle recommendations related to weight and physical activity may also be relevant to cancer prevention for individuals with Lynch syndrome. Further high-quality prospective cohort studies, in particular, including endometrial cancer as an end point, are needed to inform evidence-based cancer prevention strategies in this high-risk population.

Introduction

Lynch syndrome, formerly known as hereditary nonpolyposis colorectal cancer (CRC), is a cancer predisposition syndrome caused by inherited pathogenic variants in one of MLH1, MSH2, MSH6, PMS2, or epCAM genes.^1,2 It is the most common cause of inherited CRC and endometrial cancer, accounting for approximately 3% and 2%-5% of each malignancy, respectively.³ Individuals with Lynch syndrome are also at an increased risk of cancer of the ovary, stomach, small bowel, pancreas, and the genitourinary system, among others.^4–7

Studies in the general population have shown that obesity, diabetes mellitus, and insulin resistance increase the risk of both colorectal and endometrial cancers.^8–10 Although diabetes and other cardiometabolic risk factors are not classical modifiable risk factors, they are elements of the metabolic syndrome through which lifestyle affects cancer risk and can be ameliorated by behavioral changes.¹¹ For this reason, they have similar public health importance and are targets for intervention. Tobacco consumption and red and processed meat intake increase the risk of CRC, whereas physical activity and a diet rich in fiber, fruit, and vegetables may be protective.^12–15

Female hormonal factors, including age at menarche, age at menopause, parity, use of hormonal contraceptives, and hormonal replacement therapy, influence endometrial cancer risk.^15–17 The most recent recommendations from the World Cancer Research Fund Continuous Update Project Expert Report advise reducing excess body weight, increasing physical activity, and minimizing alcohol and tobacco consumption to reduce risk of colorectal and endometrial cancers.¹⁸ Reducing processed and red meat consumption is advised for CRC prevention.¹⁸

Although established in the general population, the association between these potentially modifiable lifestyle risk factors and the development of colorectal and endometrial cancers in Lynch syndrome is less well-characterized. A systematic review of energy-related lifestyle factors and risk of colorectal and endometrial cancers in Lynch syndrome was published previously.¹⁹ However, it was a qualitative evidence synthesis without meta-analysis, and it did not include studies of female hormonal or broader cardiometabolic risk factors. In addition, important studies have been published since and updated assessment of the data is warranted.

A better understanding of these risk factors in the Lynch syndrome population will directly contribute to the implementation of adjunctive cancer prevention strategies. To this end, we conducted a systematic review and meta-analysis of additional modifiable and nonmodifiable risk factors for colorectal and endometrial cancers in individuals with Lynch syndrome. Through this, we aim to inform evidence-based lifestyle recommendations in this high-risk cohort.

Discussion

Determining whether potentially modifiable lifestyle and environmental risk factors for sporadic malignancies are also applicable in germline cancer predisposition syndromes is an important clinical question. As the lifetime risk of cancer is many times higher, there may be a greater absolute benefit to positive health behaviors compared with the general population.⁴⁴ This systematic review and meta-analysis evaluated the association between lifestyle risk factors and colorectal and endometrial cancers in people with Lynch syndrome. In the case of CRC, obesity and a lack of physical activity were found to be associated with an increased risk, while there was insufficient evidence to conclude that alcohol or tobacco intake had a consistent association. As Lynch syndrome–associated malignancies are highly immunogenic, the immunosuppressive effect of the aforementioned risk factors^45,46 may plausibly amplify their already carcinogenic actions. Limited evidence suggests that certain dietary habits might be protective, but more robust studies are needed to inform public health recommendations. In endometrial cancer, the data are particularly limited, and it is challenging to draw conclusions. The limited available evidence suggests that T2DM and the same female hormonal risk factors that affect sporadic endometrial cancer also hold in those with inherited mismatch repair deficiency. Interestingly, obesity may be less of a risk factor in Lynch syndrome–associated endometrial cancer. This finding contrasts with the strong and consistent relationship between obesity and sporadic endometrial cancer, which is believed to be mediated through hormonal, insulinemic, and proinflammatory mechanisms of carcinogenesis.⁴⁷ Lynch syndrome–associated endometrial cancer is biologically distinct, in terms of both tumor histology and molecular subtype.^48–52 A hypothesis for this lack of association is related to the fact that Lynch syndrome–associated endometrial cancer occurs in the setting of a constantly replenishing endometrium, with a high level of replication. In the error-prone mismatch repair phenotype in Lynch syndrome, the high level of cellular turnover may increase the generation of oncogenic mutations.⁵¹ In comparison, in women with functioning mismatch repair system, errors occurring during this frequent turnover are repaired. It is possible that the high estrogenic drive associated with obesity¹¹ is a greater driver of malignancy in that context.

Our results are generally consistent with another systematic review that examined energy-related lifestyle factors and the risk of colorectal and endometrial cancers in patients with Lynch syndrome.¹⁹ The review’s conclusion that health recommendations for the general public are generally relevant to Lynch syndrome, with a few caveats, is consistent with this article. Our results are also in agreement with a recent meta-analysis of obesity and the risk of CRC in patients with Lynch syndrome.⁵³ The authors found a twofold increased risk of CRC in obese men compared with nonobese men, but this was not statistically significant in women. A recent case-control study examined the association between obesity and the risk of CRC in the general population across five Jass molecular subtypes.⁵⁴ Interestingly, subtype 5, most commonly associated with Lynch syndrome,⁵⁵ was the only subtype in which obesity did not increase odds of CRC (odds ratio, 1.04 [95% CI, 0.9 to 1.2]). This discordance in results between this study and the rest of the literature may be due to assuming the presence of Lynch syndrome by tumor molecular subtype rather than a confirmed diagnosis, the use of a case-control methodology, and the use of cohorts from different populations. Further high-quality cohort studies including those with germline-proven Lynch syndrome are needed to determine the exact extent to which BMI determines the risk of CRC.

Among the strengths of this study is the substantial amount of recently published studies, assessing a large number of individuals, with long-term follow-up. It is the most comprehensive evidence synthesis of nongenetic risk factors and their association with cancer in Lynch syndrome to date, providing up-to-date results that may inform public health recommendations. Nevertheless, our study had several limitations. First, most cohort studies were retrospective in methodology, which are subject to recall bias, especially as many risk factors were self-reported. Second, few studies assessed each MMR mutation individually and, in many cases, the mutations were pooled. This meant how different risk factors interact with the different Lynch syndrome genotypes was not explored. One study³⁶ included 17% of participants diagnosed with Lynch syndrome by the Amsterdam criteria, which are neither sensitive nor specific.^56,57 We included this cohort as this small subset is unlikely to have affected the strong association between obesity and CRC found in the study. There were few studies assessing certain risk factors and very few studies overall on endometrial cancer. Because of this, just two studies were eligible for meta-analysis of obesity in CRC. This limits the precision of the pooled estimate of the risk of CRC and assessment of between-study heterogeneity.⁵⁸ There was just one high-quality study investigating diet and the risk of CRC.²⁴ Although it found an association between calcium and multivitamin intake and cancer, the fact that these risk factors are not conclusively associated with sporadic CRC¹⁸ precludes any definitive conclusion in this genetic subpopulation. A major limitation of our meta-analysis is that, for most risk factors examined, extreme values were dichotomized (eg, BMI >30 v <25 kg/m²) in the meta-analysis. This was performed to ensure that comparable end points were being pooled but have reduced the precision of our pooled estimate. This dichotomization may also account for our finding that alcohol and tobacco intake is not associated with Lynch syndrome–associated CRC, in contrast to the general population. A dose-response meta-analysis can be performed to overcome this issue.⁵⁹ The relative lack of dose-response data in the included cohort studies means that this is not feasible for this analysis. However, if more granular data related to modifiable risk factors and cancer risk in Lynch syndrome are available in the future, this would be a worthwhile approach. Finally, our study was limited in that it just included studies looking at the risk of colorectal and endometrial cancers. Other Lynch syndrome–associated cancers were usually not accounted for in most cohort studies, or if they were, they were grouped as non-CRCs,³⁶ preventing an effective evidence synthesis.

Our findings highlight several areas that warrant further research. There is a paucity of research investigating risk factors for endometrial cancer among people with Lynch syndrome, with our literature search identifying just three studies. Given that sporadic endometrial cancer is among the cancers with the greatest lifestyle risk,⁶⁰ and there is no established benefit of endometrial surveillance,⁶¹ determining whether these risk factors apply to the Lynch syndrome population would have a substantial public health benefit. Future research should also examine how gene-environment and environment-environment interactions can affect risk. The risk of cancer among people with Lynch syndrome depends on the genotype, with MLH1 and MSH2 generally being the most penetrant.^62–65 The differential risk between genotypes is also likely to affect the magnitude of a given lifestyle risk factor. Two studies included in this review found that MLH1 mutation carriers had the greatest increase in the risk of CRC because of obesity,^34,36 and this was also seen in a recent meta-analysis.⁵³ However, most other studies did not stratify the risk of CRC by specific MMR gene variant, so more research is needed to explore gene-specific recommendations for cancer risk reduction. Interactions between multiple nongenetic risk factors are also an emerging research focus. A cohort study on the basis of the CAPP2 trial of aspirin for CRC prevention in Lynch syndrome found that, although obesity was associated with an increased risk of CRC, this was abrogated in those taking aspirin. This suggests that the protective association of aspirin may counteract the carcinogenic mechanism of obesity, and those with obesity may be more likely to benefit from aspirin use.³⁶ This example of interaction between two different nongenetic risk factors could be a future method of precision cancer prevention, on the basis of not just one’s genotype but also the lifestyle risk profile.

In conclusion, this systematic review and meta-analysis demonstrates that public health lifestyle recommendations, with some caveats, apply to the Lynch syndrome population. Obesity and lack of physical activity may be associated with an increased risk of CRC among individuals with Lynch syndrome. Female hormonal risk factors and type 2 diabetes might confer an increased risk of endometrial cancer. These findings can inform adjunctive cancer prevention strategies. Further high-quality prospective cohort studies, particularly in endometrial cancer, are needed to inform more robust, evidence-based lifestyle recommendations for people living with a diagnosis of Lynch syndrome.