Age and CRC Risk in the Serrated Pathway

Serrated polyps are precursors in an epigenetic pathway that may account for a significant proportion of colorectal cancer (CRC) and may be detected in >20% of screening examinations.¹ Serrated polyps can be subclassified into hyperplastic polyps (HPs), sessile serrated adenomas/polyps (SSA/Ps), and traditional serrated adenomas (TSAs). While most HPs predict low malignant potential, all TSAs harbor dysplasia and SSA/Ps can acquire dysplasia and thus can progress to CRC. Given their collective risk for cancer, endoscopists are greatly interested in understanding their clinical natural history.

The biology of the serrated pathway involves both point mutation of an oncogene, BRAF, and the epigenetic (DNA sequence intact) silencing of other genes. SSA/Ps may progress to SSA/P with dysplasia (SSA/P w/D), a more proximate step to serrated CRC. This transition may occur through hypermethylation of CpG dinucleotide islands, in the promoter region (CIMP) of the MHL1 gene, leading to microsatellite instability and subsequent BRAF-mutated CRC with microsatellite instability.² Alternatively, other gene-silencing sequences, including methylation of the CDKN2A gene with loss of the p-16 protein, may culminate in a BRAF MSS cancer with a lesser clinical prognosis.²

Because of the relative infrequency of serrated polyps such as SSA/P, their natural history and timeline for malignant transformation is not as evident as that for conventional adenomas.^3,4 Data from adults with serrated polyps, undergoing endoscopic surveillance, suggest that the 5-year CRC risk is not elevated, over polyp-free individuals.^5,6 In addition, over time, the risk is not for developing metachronous conventional high-risk adenomas but for large (≥1 cm) serrated polyps, surrogates of SSA/Ps. Data from a 10-year follow-up Norwegian study showed that these large serrated polyps increase long-term CRC risk, similar to the risk associated with advanced adenomas,⁷ and should be regarded as important intermediate lesions in the serrated pathway.

For clinicians, it is imperative to understand the cancer risk associated with detected polyps. Studies that have compared the average ages of individuals with SSA/Ps and SSA/P w/D to those with a serrated pathway carcinoma have estimated the serrated polyp dwell time to be about 15 or more years.^8,9 Conversely, as the age of adults with SSA/P w/D and those with cancer were similar, it was concluded that the speed of transition from SSA to SSA/P w/D and CRC might increase with age. Overall, data from various studies suggest that, although serrated polyps may take a stepwise route to CRC, as compared with conventional high-risk adenomas, their overall transition time is comparable.

Given the observed rise in CRC in young adults,¹⁰ an important clinical question is whether these serrated polyps pose a threat for CRC in adults younger than 50 years of age, similar to conventional adenomas. In the current issue of this journal, Bettington et al¹¹ address this by examining the prevalence of SSA/Ps and CRCs in 2 populations, whose ages varied from 20 to 80 years and older. The CRCs were tested for KRAS and BRAF mutations, the latter being a strong predictor of CRCs in the serrated pathway. The authors observe that, while the prevalence of SSA/Ps was similar across all decades, the prevalence for BRAF-mutated CRCs increased after the 40 to 49 years age group. Specifically, the prevalence for SSA/Ps was similar for those under and over 50 years. In contrast, only 2.5% (1/40) of all CRCs in adults <50 years were positive for BRAF mutations, whereas 24.3% (98/404) were positive for CRCs in adults over 50 years of age (P=0.003; Fisher exact). In addition, the overall prevalence for BRAF-mutated CRCs dramatically increased after 60 years of age. These data suggest that, while SSA/Ps can occur in younger adults, serrated pathway CRCs are more likely to present in individuals over 50 years, and possibly even as late as 60 years of age. It appears, therefore, that the transition from SSA/P takes several years, not progressing to CRC until at least after 50 years, and further increasing in the ensuing decades.

Additional findings in the study may provide added insight to the progression of SSA/P to CRC. The observed prevalence of adults with both synchronous SSA/Ps and conventional adenomas increased dramatically with age. It has been previously reported that SSA/Ps may be present synchronously with conventional high-risk adenomas.¹² Such dual pathway patients are, in fact, more likely to develop metachronous high-risk adenomas, inferring that the concurrent presence of serrated polyps increases the overall colon cancer risk.⁵

There were some interesting findings regarding SSA/P w/D that may also provide insight into the progression of SSA/Ps. All of the SSA/Ps w/D in this study occurred in people older than 65 years (65, 80, and 86 y), suggesting that the transition from SSA/P to SSA/P w/D occurs with advanced age. Biologically, chronologic aging has been associated with increasing DNA methylation, the presumed primary mechanism for gene silencing and progression of SSA/Ps.^13,14 Other factors such as smoking,¹⁵ which also can be strongly associated with CIMP,¹⁶ may also require a long exposure time in order to promote development of SSA/P w/D. Thus, increased long-term exposure to factors like smoking may explain the increased CRC risk associated with aging. It should also be recognized that the number of SSA/Ps w/D in this sample was small (n=3). This is a finding that will resonate with many endoscopists whose experience with SSA w/D is limited because of their scarcity, likely reflecting the rapid progression of these lesions to CRC, as suggested by the studies cited above.^8,9

Despite an increase in BRAF CRCs at an older age, the sizes of the SSA/Ps in this study remained constant, <1 cm for all age groups except 80+ years. These data suggest that SSA/P progression to CRC or SSA/P w/D may not be initially predicated on the SSA/P size. Furthermore, as these lesions were <1 cm, they may be more likely to be missed during colonoscopy. This further reinforces the precept that endoscopists need to ensure adequate bowel preparation¹⁷ and consider a longer withdrawal time to maximize SSA detection.¹⁸

Finally, the authors note that the one person less than 50 years of age with a BRAF-mutated CRC had a diagnosis of serrated polyposis syndrome (SPS). Previously known as hyperplastic polyposis syndrome, the criteria for SPS include any of the following: (1) 5 or more serrated polyps proximal to sigmoid, with 2 or more that are 10 mm or larger; (2) 1 or more serrated polyps proximal to the sigmoid and a first degree relative with SPS; (3) having 20 or more serrated polyps of any size and location.¹⁹ The authors correctly point out that individuals with SPS represent a different phenotype and may be at risk at an earlier age than individuals with “nonsyndromic” serrated polyps. Given that the serrated polyps can be sequential and cumulative, it is important that clinicians recognize this syndrome and provide the proper follow-up, which can include yearly colonoscopies to control the serrated polyp burden.²⁰

There are some limitations that need to be mentioned. As there were 2 different populations from which the samples were culled, there may be significant differences in specific risk factors associated with SSA/Ps, such as smoking.¹⁵ Another limitation may be the difficulty in diagnosing SSA/Ps and especially distinguishing them from HPs.²¹ The authors used an experienced central pathologist, minimizing variation of interpretation across sites. To address the concern that some SSA/Ps may have been mistakenly diagnosed as HPs, the authors present data showing that the prevalence of HPs, regardless of size and location, was similar across younger and older groups.

In summary, Bettington and colleagues provide more data regarding the natural history of serrated polyps (see Table 1 for summary). Specifically, serrated CRCs do not appear to pose a threat to younger adults, despite the presence of SSA/Ps. While there are some limitations to their analyses, it is reassuring that, despite a consistent presence of SSA/Ps in one group, risk for BRAF-mutated CRC did not increase until after 50 years. However, many issues remain, including whether potential molecular markers can help to predict progression of SSA/Ps to CRC. Challenges remain as to how endoscopists can accurately identify all SSA/Ps in adults presenting for colonoscopies, especially those with dysplasia or those with a high likelihood of progressing to SSA w/D and CRC. For example, recent reporting of changes in colon microbiome diversity, particularly the correlation with Fusobacterium nucleatum,²² may yet provide clues as to who is at risk for CRC arising in the serrated pathway.

TABLE 1: Study Findings and Impact of Findings on Clinical Practice