The “scope” of colorectal cancer screening in Lynch syndrome: is there an optimal interval?

Lynch syndrome (LS) is one of the most common hereditary colorectal cancer (CRC) predisposition syndromes and is present in 1 out of 279 individuals in the general population (1) and in approximately 3% of unselected cases of CRC (2). LS is caused by the autosomal dominant inheritance of a pathologic variant in 1 of 4 mismatch repair (MMR) genes (MLH1, MSH2, MSH6, PMS2) or deletions in the EPCAM gene (which lead to epigenetic silencing of MSH2). LS carriers are at increased risk of CRC as well as endometrial, ovarian, stomach, small intestine, pancreaticobiliary, urinary tract, and skin cancers, among an increasingly long list of other rarer cancer types (3). It is now well established that rather than being a singular entity, each individual LS gene phenotype is different with respect to cancer type and risk (4). MLH1 and MSH2 are highly penetrant, with a cumulative incidence of CRC by age 75 years in MLH1 of 48.3% women and 57.1% men and in MSH2 of 46.6% women and 51.4% men (4). Cumulative incidence of CRC for MSH6 carriers is less than that for MLH1 and MSH2, at 20.3% and 18.2% for women and men, respectively. PMS2 is the most prevalent LS gene (1) but the least penetrant, with a cumulative CRC incidence of 10.4% in both sexes by age 75 years (4), and one large study suggested that PMS2 carriers are at increased risk of only CRC and endometrial cancer but not other LS-associated cancers (5). Across all genes, colonoscopy screening reduces the risk of CRC and improves overall survival among LS carriers (6,7) and remains the mainstay of current risk reduction recommendations. What remain less clear are the optimal age at which to start screening and the best interval between colonoscopy screenings for LS carriers, such that many international professional society guidelines vary with respect to these recommendations (Table 1) (8-12).

Table 1.

Recent professional society guidelines for Lynch syndrome colonoscopy screening, stratified by gene^a

Society guidelines	MLH1	MSH2/EPCAM	MSH6	PMS2
NCCN V2 (12)	Start at age 20-25 y, continue 1-2 y	Start at age 20-25 y, continue 1-2 y	Start at age 30-35 y, continue 1-3 y	Start at age 30-35 y, continue 1-3 y
ESGE (11)	Start at age 25 y, continue every 2 y	Start at age 25 y, continue every 2 y	Start at age 35 y, continue every 2 y	Start at age 35 y, continue every 2 y
ESMO (10)	Start at age 25 y, continue every 1-2 y	Start at age 25 y, continue every 1-2 y	Start at age 35 y, continue every 1-2 y	Start at age 35 y, continue every 1-2 y
BSG/ACPGBI/UKCGG (8)	Start at age 25 y, continue every 2 y	Start at age 25 y, continue every 2 y	Start at age 35 y, continue every 2 y	Start at age 35 y, continue every 2 y
EHTG/ESCP (9)	Start at age 25 y, continue every 2-3 y	Start at age 25 y, continue every 2-3 y	Start at age 35 y, continue every 2-3 y	Start at age 35 y, continue every 5 y

^aACPGBI = Association of Coloproctology of Great Britain and Ireland; BSG = British Society of Gastroenterology; EHTG = European Hereditary Tumor Group; ESCP = European Society of Coloproctology; ESGE = European Society of Gastrointestinal Endoscopy; ESMO = European Society for Medical Oncology; NCCN = National Comprehensive Cancer Network; UKCGG = UK Cancer Genetics Group.

In this issue of the Journal, Aronson et al. (13) attempt to evaluate how colonoscopy screening intervals impact CRC risk by analyzing the colonoscopy patterns and adenoma and CRC history of 429 LS carriers in the Familial Gastrointestinal Cancer Registry (Zane Cohen Centre, Sinai Health System, Toronto, Canada). The authors used an instrumental variable approach where the relationship between screening intervals and CRC is predicated first on the number of adenomas detected in the screening interval between 2 consecutive colonoscopies (categorized as $\le$1 year, 1-2 years, 2-3 years, or >3 years), with each individual patient contributing multiple intervals during a median follow-up of 9.2 years. They then analyzed the predicted number of adenomas in each interval with the time to CRC and found that any new adenoma detected during screening decreased the 10-year CRC risk by 11.3%. The 20-year cumulative CRC risk for women was 10.0%, 9.2%, 13.4%, and 16.4% for screening intervals of ≤1 year, 1-2 years, 2-3 years, or more than 3 years, respectively; for men, rates were higher at 36.2%, 35.6%, 43.4%, and 47.7%. When stratified by specific gene, the authors found a 20-year cumulative risk reduction for MLH1 carriers of 28% women and 14% men for a 1- to 2-year interval vs 2- to 3-year interval; for MSH2 carriers, of 29% women and 17% men; and for MSH6 carriers of 18% men (unable to calculate women), respectively (13). These results support a 1- to 2-year screening interval for adenoma detection and reduced CRC risk.

This article (13) adds to the body of literature exploring the optimal interval for colonoscopy screening among LS carriers. One of the first studies to investigate this important question prospectively followed LS carriers getting colonoscopies every 3 years compared with those getting no screening and found that the 3-year screening interval decreased CRC risk and improved overall survival, establishing routine colonoscopy screening as the standard of care (6). Since then, researchers have attempted to define the ideal screening interval, with results varying depending on the study and approach (14-16). For example, a large European study compared the incidence of CRC among LS carriers from 3 countries (each of which had different national screening guidelines for colonoscopies: annual, 1-2 years, and 2-3 years) and did not find a statistically significant difference in CRC incidence or stage of diagnosis between the different intervals (14). This study by Aronson et al. (13) presents a new approach for evaluating screening intervals based on an instrumental variable analysis that accounts for variation of screening intervals both within and between individual LS carriers.

Importantly, this article’s (13) instrumental variable methodology relies on the adenoma–carcinoma paradigm that was first described in the 1990s (17). However, 53.4% of LS carriers did not have any adenomas detected, although these individuals still had an elevated 20-year cumulative risk of CRC of 25.7% and 57.2% in womens and men, respectively (13). It has been hypothesized that some CRCs may develop within MMR-deficient crypt foci (ie, histologically normal intestinal crypts with absent MMR protein expression), which in turn acquire somatic mutations leading to invasive cancer while bypassing adenomatous precursors (18,19). The fact that cancer may develop via an adenoma-independent pathway may also support shorter interval colonoscopies, as more frequent direct visualization of the colon could potentially lead to earlier detection of cancer. However, an analysis from the Prospective Lynch Syndrome Database found that CRC stage did not correlate with interval from last colonoscopy, suggesting the potential for overdiagnosis with more frequent colonoscopies; the Prospective Lynch Syndrome Database authors hypothesized that some of these cancers may not be clinically relevant, with host immune response potentially leading to regression of early tumors (20). Notably, in the Aronson et al. (13) study, more than half of all LS carriers and 32% of those diagnosed with CRC had no adenomas detected during colonoscopy screening. There was no difference in colonoscopy quality or clinical expertise for these adenoma-free cases, suggesting that this was not likely due to missed adenomas. It is therefore difficult to reconcile these cases with the underlying analytic premise that the primary mediator between colonoscopy screening interval and CRC risk is via the detection of adenomas. Further investigation is required to better define the biology and optimal management of adenoma-independent carcinogenesis, as well as the potential interactions of host immune response on early LS cancers.

It is now well understood that there are LS gene-specific variations in CRC cancer risk (4), and as such, screening recommendations must reflect these differences. The results from Aronson et al. (13) support colonoscopy intervals of 1-2 years for MLH1 and MSH2 carriers, who comprised more than 70% of the cohort, but there were a limited number of MSH6 and PMS2 carriers (87 and 26 individuals, respectively). Although most professional society guidelines generally agree with an every 1- to 2-year interval for MLH1 and MSH2 carriers (8,10-12), the optimal management of MSH6 and especially PMS2 carriers is more controversial, with colonoscopies recommended anywhere from every 1 to every 5 years (8-12). This subset of LS carriers may also be able to begin screening at a later age than their MLH1 and MSH2 counterparts, because of the later age of CRC onset. Although the results from Aronson et al. (13) contribute further evidence to the literature to support a 1- to 2-year interval for MLH1 and MSH2 carriers, definitive conclusions cannot be drawn about the optimal screening interval for MSH6 and PMS2 carriers because of the small number of patients with these mutations; additional studies with larger sample size are warranted.

Many important questions remain unanswered. Aronson et al. (13) reported sex differences in CRC risk, consistent with other studies (4). In this article, men were at higher risk compared with women across all screening intervals, and the risk reduction for a 1- to 2-year interval compared with more than 2 years was higher in women (13). The authors suggest that the enhanced risk reduction may be explained by a higher number of adenomas detected in women, but other studies of non-LS carriers have actually found higher rates of adenomas (in addition to CRC) among individuals of male sex (21). The reason for sex-specific differences in prevalence of adenomas remains uncertain and cannot be explained solely by differences in risk factors (22). Similarly, potential risk-modifying factors, including aspirin use (23) and resistant starch (24) intake that have been associated with a reduction in CRC in LS carriers, were not available for this analysis and may impact screening and CRC development. The authors were also unable to model familial relationships because most families have only 1 or 2 relatives (13), and we know that family history of cancer influences risk considerations across LS-associated cancers (25). As we attempt to better personalize surveillance and risk-reducing strategies for LS carriers, understanding the interplay of clinical and lifestyle factors, along with family history, is critical for a precision screening approach.

Overall, the study by Aronson et al. (13) supports the recommendation for colonoscopy surveillance every 1-2 years, especially for MLH1 and MSH2 carriers. We hope that in the near future, novel interventions and technological advances for enhanced early detection and prevention, including noninvasive circulating tumor DNA testing (26) and vaccination strategies (27), will aid in CRC diagnosis and risk reduction. At present, however, colonoscopy screening remains critical to LS surveillance, CRC risk reduction, and improved outcomes for all LS carriers.

Data availability

No new data were generated or analyzed for this editorial.

Author contributions

Leah Biller, MD (Conceptualization; Writing – original draft; Writing – review & editing) and Kimmie Ng, MD, MPH (Conceptualization; Supervision; Writing – review & editing).

Funding

No funding was used for this editorial.

Conflicts of interest

LHB has no disclosures. KN reports research funding to institution: NIH/NCI, Colorectal Cancer Alliance, Cancer Research UK, Pharmavite, Evergrande Group, Janssen, Revolution Medicines; advisory board or consulting for Bayer, SeaGen, BiomX, Bicara Therapeutics, GlaxoSmithKline, Pfizer, X-Biotix Therapeutics, Redesign Health.