Clinical Gastroenterology and Hepatology 2015;13:11–26

PERSPECTIVES IN CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Sessile Serrated Adenomas: An Evidence-Based Guide to Management Seth D. Crockett,* Dale C. Snover,‡ Dennis J. Ahnen,§ and John A. Baron*

*Division of Gastroenterology and Hepatology, University of North Carolina School of Medicine, Chapel Hill, North Carolina; ‡Department of Pathology, Fairview Southdale Hospital, Edina, Minnesota; and §Division of Gastroenterology, Department of Veterans Affairs Eastern Colorado Health Care System and University of Colorado School of Medicine, Denver, Colorado

The concept of serrated colorectal neoplasia and a serrated adenoma (TSA), in addition to the more serrated pathway to colorectal cancer (CRC) is relatively innocuous hyperplastic polyp (HP).5 SSAs are the pro- new and continuing to evolve, but it has become highly totypical precursor lesions of the “serrated pathway” to relevant to gastroenterologists, pathologist, and oncolo- colorectal cancer (CRC), now understood to have the gists alike. Sessile serrated adenomas (SSA) are now potential to develop into sporadic CRCs via a series of thought to be the major precursor lesion of serrated molecular alterations including BRAF mutation and CpG pathway cancers, which represent up to one-third of all island methylation, with epigenetic inactivation of the sporadic CRC cases. However, despite their increasingly mismatch repair gene MLH1 resulting in microsatellite recognized importance, relatively little is known about the 5–9 epidemiology and natural history of SSAs, and the molec- instability (MSI) (Figure 1). SSAs are probably the ular and epigenetic aspects are incompletely understood. precursor to CpG island methylator phenotype (CIMP) 10 Endoscopists must be aware of the unique features of SSAs high microsatellite stable (MSS) carcinomas as well. so that the practice of colonoscopic screening for CRC can SSAs are a furtive foe for both the gastroenterologist include optimized detection, removal, and appropriate and pathologist. They tend to be proximally located, surveillance of SSAs and other serrated precursor lesions. have subtle endoscopic features, and are susceptible to In this review, we discuss the history, epidemiology, and being both overlooked and incompletely resected by pathologic aspects of SSAs, as well as a recommended endoscopists. To complicate matters further, there re- management approach and a discussion of uncertainties mains considerable interpathologist variability in the and opportunities for future research. diagnosis of serrated polyp subtypes. These features and their potential to progress may explain (at least in part) Keywords: Sessile Serrated Adenoma; Sessile Serrated Polyp; why colonoscopy offers less protection from right-sided Serrated Neoplasia; Colorectal Cancer Screening; Colonoscopy; CRC.11–13 In addition, interval CRCs (ie, those occurring Endoscopic Detection; Polypectomy. after a previous colonoscopy) frequently are located in the proximal colon and have molecular features of serrated pathway cancers such as CIMP and MSI, sug- olorectal cancer (CRC) kills more than 600,000 gesting that SSAs contribute to this problem as well people annually worldwide, and is the second – C (Figure 2).13 16 Serrated pathway cancers are thought to leading cause of cancer death in the United States.1,2 Until be responsible for 20% to 35% of CRC cases,4,17 which recently, adenomatous polyps (now referred to as con- makes this CRC subtype a substantial public health ventional adenomas) were considered the precursor problem, with an incidence potentially greater than that lesions of all cases of sporadic colon cancer, which was of pancreatic, gastric, or esophageal cancer.2 thought to occur via a series of well-defined molecular Given their potential public health importance, this genetic steps along the chromosomal instability pathway.3 review focuses on SSAs, with particular attention to Recent research has indicated that this paradigm does not pertain to all sporadic CRC.4 Previously, all hyperplastic colorectal polyps were regarded as rela- Abbreviations used in this paper: CI, confidence interval; CIMP, CpG tively harmless lesions without malignant potential. It is island methylator phenotype; CRC, colorectal cancer; CT, computed tomography; FIT, fecal immunochemical testing; HP, hyperplastic polyp; now recognized that these polyps are in fact a hetero- MSI, microsatellite instability; MSS, microsatellite stable; MVHP, micro- geneous group, now called serrated polyps, characterized vesicular hyperplastic polyp; OR, odds ratio; PSP, proximal serrated polyp; SPS, serrated polyposis syndrome; SSA, sessile serrated adenoma; SSAD, primarily by a saw-toothed appearance of colonic crypts. sessile serrated adenomas with dysplasia; TSA, traditional serrated ade- This group includes premalignant lesions such as the noma; WHO, World Health Organization. sessile serrated adenoma (SSA) (also known as the © 2015 by the AGA Institute sessile serrated polyp or lesion) and the traditional 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2013.10.035 12 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1

Figure 1. Proposed serrated pathway to MSI CRC, with characteristic crypt alterations, histologic and endoscopic appear- ance, and epigenetic and genetic alterations. Dotted line between MVHP and SSA indicates a possible but unproven pro- gression step. Regarding crypt alterations, the thickened line indicates a proliferative zone, which is expanded in HPs, and migrates upward in SSAs. Arrows indicate the direction of maturation. See text for details. MSI-H, high microsatellite instability. Modified with permission from Snover et al.17 clinical aspects relevant to gastroenterologists. A number intramucosal carcinoma, establishing a risk of neoplastic of reviews have been published on the pathologic aspects transformation, and thus this report served to codify the of these lesions,6–8,17 and these aspects are covered only premalignant potential of some serrated lesions. In 1996, briefly here. Torlakovic and Snover23 analyzed polyps from persons with HPs (now known as serrated polyposis, defined later), and identified a previously unrecognized serrated History polyp characterized by large size and proximal location, and histologically by architectural distortion resulting from abnormalities in proliferation (Figure 1). This is The understanding of serrated polyps has been fi evolving over the past 30 years. Originally, all such lesions generally considered the rst description of the SSA, although this nomenclature was not generally acknowl- were classified as HPs, and considered benign, non- edged until a 2003 report that described TSAs and SSAs as neoplastic colorectal polyps, in contrast to adenomatous separate entities.24 Subsequent reports have helped to polyps (ie, conventional adenomas). Occasional reports in refine the pathologic and molecular features of SSAs.8,25,26 the 1970s and 1980s challenged the convention that HPs lacked malignant potential,18–21 but it was not until the 1990s when the term “serrated adenoma” was coined by Taxonomy and Nomenclature Longacre and Fenoglio-Presier22 to describe a polyp with crypt serration resembling that of HPs, but also with The World Health Organization (WHO) now recog- cytologic dysplasia not characteristic of HPs. A number of nizes 3 categories of serrated polyps, all characterized by these serrated adenomas (which would be classified as serration of the glandular epithelium: HPs, SSAs, and TSAs using contemporary terminology) had foci of TSAs (Supplementary Figure 1).27 January 2015 Sessile Serrated Adenomas 13

Figure 2. SSAs likely contribute to the problem of interval (or postcolono- scopy) cancers for 3 chief reasons: (1) SSAs can be difficult to detect during colonoscopy and are com- monly missed; (2) it is thought that SSAs have the potential for rapid progres- sion owing to the develop- ment of MSI, especially after dysplasia develops; and (3) incomplete poly- pectomy is common with SSAs. Hollow arrows represent potential targets for combating this problem.

Hyperplastic Polyps the proximal colon, a trait it shares with SSAs.33 For these reasons, it has been proposed that MVHPs may progress HPs are typically small (5 mm), and roughly 70% to SSAs, but this has not been proven.34 are located in the distal colon and rectum.28,29 Endo- fl scopically, or whitish color, sessile or at morphology, Sessile Serrated Adenomas and have a type II asteroid, stellate, or papillary Kudo pit pattern when examined with chromoendoscopy or fi 30 SSAs are signi cantly less common than HPs. narrow-band imaging. Histologically, HPs have crypt Depending on the series, SSAs represent up to one fifth serration involving the upper segment of colonic crypts, of all serrated polyps.32,35–37 There remains consider- with normal proliferation and no nuclear atypia or 26 able controversy as to the best name for this lesion dysplasia. HPs are the most common type of serrated (Table 1). Some investigators prefer the term sessile polyp, representing 29% to 40% of all polyps, and 80% 24,31,32 serrated polyp to differentiate it from conventional ad- to 95% of all serrated polyps. There are several enomas, whereas other investigators (particularly in subtypes of HPs including the most common micro- Europe) prefer the term sessile serrated lesion because – vesicular hyperplastic polyp (MVHP), the goblet cell rich SSAs can be nonpolypoid and flat. Although SSA (and HP, and the mucin-poor HP (Supplementary Figure 1). TSA) contain the word “adenoma,” it is important to Although HPs as a group do not directly progress to recognize that these lesions are quite distinct from malignancy, some MVHPs do contain a BRAFV600E muta- 33 conventional adenomas with respect to histology tion, which disrupts normal apoptosis of epithelial cells. and molecular biology. Conventional adenomas by MVHPs also have a higher incidence of CIMP, especially in definition contain dysplasia, but SSAs typically are 14 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1

Table 1. Nomenclature Issues of SSA

Proposed term Pros Cons

SSA Term adenoma connotes premalignant Creates confusion with conventional adenomas for providers potential and patients Term differentiates SSAs from HPs Some argue that the term adenoma should apply only to dysplastic lesions Sessile serrated polyp Avoids confusion with conventional adenomas Some SSAs are flat and nonpolypoid by Paris classification Does not imply dysplasia is present HPs also are polyps that are sessile and serrated Fails to convey the premalignant potential of these lesions May cause problems with insurance coverage for repeat endoscopy Sessile serrated lesion Avoids confusion with conventional adenomas The term lesion does not necessarily imply neoplasia Does not imply dysplasia is present HPs also are lesions that are sessile and serrated Term is inclusive of nonpolypoid (eg, flat) Fails to convey the premalignant potential of these lesions lesions May cause problems with insurance coverage for repeat Preferred term of some European groups endoscopy SSA/P Hybrid term of SSA and sessile serrated Cumbersome polyp indicates that both terms apply Cons of both terms apply to same lesion Most inclusive term Serrated adenoma Congruent with terminology of other Nonspecific term could include both SSAs and TSAs premalignant colon polyps (eg, tubular Creates confusion with conventional adenomas for providers adenomas, villous adenomas) and patients Some argue that the term adenoma should apply only to dysplastic lesions Use of this term is discouraged Serrated polyp with abnormal Indicates that lesions have abnormal Nonspecific term could include both SSAs and TSAs proliferation proliferation even in absence of true Cumbersome dysplasia Not frequently used Giant HP Historical importance (first term used Does not differentiate from HP apart from size by Jass)124 Does not connote premalignant potential Indicates large size Some SSAs are small Differentiates from diminutive HPs Use of this term is discouraged Variant HP Differentiates from typical HPs Suggests SSAs are a subset of HPs Does not connote premalignant potential Use of this term is discouraged

SSA/P, sessile serrated adenoma or polyp.

nondysplastic.24 The characteristic histology, molecular MSS cancers.6 Furthermore, because TSAs have an endo- features, epidemiology, and endoscopic appearance of scopic appearance similar to conventional adenomas and SSAs will be described further later. typically are located distally,39,40 they do not present the same challenges as SSAs with respect to endoscopic detection, and are unlikely to be overlooked. TSAs are also Traditional Serrated Adenomas less likely to be misclassified pathologically as HPs.41 TSAs are the least common serrated lesions, repre- senting only 1% of these polyps.31 TSAs tend to be left Serrated Polyposis Syndrome sided, protuberant, and/or pedunculated, and a signifi- cant proportion harbor conventional dysplasia.22 Histo- Serrated polyposis syndrome (SPS) (formerly known logically, TSAs have villiform features and cells with as hyperplastic polyposis) is a condition of multiplicity of elongated nuclei and eosinophilic cytoplasm along with serrated polyps. The serrated polyps found in patients ectopic crypts.8,26 Given their rarity, these polyps and with SPS include mainly SSAs, but HPs also often are their behavior are not as well characterized as HPs and present, in addition to occasional TSAs.42 Conventional SSAs. Although TSAs also have malignant potential,22 adenomas also can co-exist with serrated lesions in SPS their pathway to cancer involves different steps than and the presence of adenomas may portend an SSAs, and is not as well understood. In contrast to SSAs, increased CRC risk in syndromic patients.42 The mean TSAs more commonly are KRAS-mutated (vs BRAF- age at presentation is approximately 55 years, and mutated),38 may show epigenetic inactivation of males and females are affected equally.43 Although other DNA repair genes such as O6-methylguanine-DNA some cases may be heritable, the exact genetic defect methyl-transferase (MGMT), and likely give rise to associated with SPS has not yet been identified, and the January 2015 Sessile Serrated Adenomas 15 pattern of inheritance is uncertain.43,44 Therefore, the although this is difficult to prove definitively. However, diagnosis currently is based on meeting somewhat the clinical consequences of specific epigenetic alterations arbitrary clinical criteria. According to the WHO, per- found in SSAs are largely unstudied. sonsmaybediagnosedwithSPSiftheyhavethe following findings on a single colonoscopic screening Epidemiology examination: (1) at least 5 serrated polyps proximal to the sigmoid colon, 2 of which are greater than 10 mm in The understanding of the epidemiology of SSAs is in diameter; (2) any number of serrated polyps occurring evolution. Most studies of the epidemiology of serrated proximal to the sigmoid colon in an individual who has a polyps predate the contemporary understanding of first-degree relative with serrated polyposis; or (3) serrated neoplasia. The early investigations of HPs are more than 20 serrated polyps of any size distributed really studies of what we now know to be serrated polyps throughout the colon.27 Patients with SPS generally are in general, which may have included some SSAs. Never- considered to be at increased risk of CRC. Case series of theless, more recent studies have begun to focus on risk patients with SPS have reported that between 25% and factors of significant serrated polyps (eg, large and/or 70% of SPS patients have CRC at the time of initial diag- right-sided polyps) and SSAs specifically. nosis or during follow-up evaluation.42,45 However, such studies are prone to referral bias, and some studies report no associated cancer risk,46 and therefore the actual risk Prevalence of malignancy associated with SPS is unknown. Approximately 20% to 40% of adults have at least 1 serrated polyp, including distal HPs.50,51 SSAs represent Histopathology 3% to 22% of serrated lesions, and 75% to 90% of SSAs are right sided.31,32,35,36,52 In average-risk screening pa- SSAs are characterized by distorted crypt bases and tients, the reported SSA prevalence ranges from 2% to crypt dilation caused by migration of the proliferative 7%.37,53 A study that used high-resolution magnifying 26 zonetothesideofthecrypt. These features distin- chromoendoscopy to optimize detection and included guish SSAs from HPs, which have a normal basal pro- some high-risk patients (ie, personal or family history of liferative zone and straight (although serrated) crypts CRC or polyps) reported the prevalence of SSAs was as 17 (Figure 1). Unlike conventional adenomas, SSAs do high as 14%.32 Other investigators using surrogate def- not typically have overt cytologic dysplasia, although initions for significant serrated lesions (eg, large or dysplasia does develop in some lesions as they prog- proximally located serrated polyps) have reported 31 ress. It is important to note that the dysplasia found in prevalence estimates in the 2% to 20% range in average- SSAs has different molecular features than the dysplasia risk screening populations.54–57 A number of pathology seen in conventional adenomas, and hence SSAs with series focused on serrated lesions also have been re- cytologic dysplasia are not simply a combined SSA/ ported in which, typically, colorectal polypectomy spec- 17 tubular adenoma. SSAs also can have foci of intra- imens from a particular pathology laboratory are mucosal carcinoma and/or adjacent invasive reviewed using current pathologic criteria.31,35,36 From 31,47,48 CRC. these series, it is estimated that SSAs represent 1% to 9% of all resected colorectal polyps. Multiplicity of Genetics and Epigenetics serrated polyps is common. A recent study of more than 1000 patients with serrated polyps (including HPs, SSAs, The pathway by which conventional adenomas give and TSAs) reported that 44% of patients with serrated rise to CRC has been well studied and involves mutations polyps had more than 1 serrated polyp, including 21% in APC, KRAS, and p53 among others,3 whereas the mo- with 2 serrated polyps, 10% with 3 serrated polyps, 5% lecular alterations associated with SSAs include BRAF with 4 serrated polyps, and 8% with 5 or more serrated 54 mutation, and CIMP, which can lead to epigenetic silencing polyps. of MLH1, a mismatch repair gene, resulting in MSI (Figure 1).5,17,35 In contrast, BRAF mutations are Risk Factors extremely rare in conventional adenomas.49 There is heterogeneity among serrated lesions, however; some A number of studies have examined the epidemiology CIMP-positive serrated lesions lead to MSS cancers, and risk factors of HPs as a group, but because small possibly owing to methylation of MGMT, tumor suppres- distal HPs represent the majority of serrated polyps, the sors, or apoptosis genes.4,8 In line with their distinct ge- significance of these findings with respect to SSAs is netic and epigenetic profile is a corresponding difference questionable. Although the risk factors for SSAs and HPs in behavior of SSAs compared with conventional ade- likely overlap to some degree, it is important to under- nomas. Because serrated pathway cancers can involve stand factors that predispose persons to develop SSAs MSI, it is thought that they have the potential for rapid specifically. As a surrogate for SSAs, some investigators growth once cytologic dysplasia has developed,47 have examined the epidemiology of proximally located or 16 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1 large (typically 1 cm or larger) serrated polyps. Proximal SSAs arise from serrated aberrant crypt foci and/or serrated polyps (PSP) have been associated with tobacco precursor HPs (namely MVHP), but it also is possible that use in some studies,58,59 but not others.60 Nonsteroidal they arise de novo (Figure 1).5,17,66,67 As mentioned anti-inflammatory drug use also has been associated earlier, cytologic dysplasia is not common in SSAs, with a decreased risk of PSPs.59,60 although they do have malignant potential. In one large With regard to SSAs specifically, the published cross-sectional study of 2416 SSAs, 85% were nondys- research on risk factors is limited. Several studies have plastic, 14% had low- or high-grade dysplasia, and 1% identified tobacco as a risk factor for SSAs.61–63 There had adenocarcinoma.31 In addition, there have been case have been mixed results with respect to obesity because reports of rapid progression of SSAs to cancer in less one study did find an association between increased than 1 year.47 body mass index and SSAs,61 however, 2 other studies 60,62 did not. Although conventional adenomas are Synchronous Neoplasia consistently more common in men, this does not appear to be true of SSAs. Colonoscopic series have reported Although the risk of synchronous neoplasia for SSAs that SSAs are at least as common in women as they are in specifically has not been well studied, there is some ev- men, with women representing 50% to 63% of those idence that proximal and large serrated polyps are with detected SSAs.37,62 Similarly, pathology series associated with an increased risk of synchronous report that 49% to 65% of those with removed SSAs are advanced neoplasia and cancer.54,58,64,68,69 Furthermore, women.31,32,36,37,61,64,65 Interestingly, one study reported persons with concomitant SSAs and conventional ade- that women comprised progressively higher proportions nomas appear to be at high risk of synchronous CRC.64 of those with more advanced serrated lesions (women Persons with SSAs also frequently have synchronous represented 53%, 57%, 69%, and 76% of those with serrated polyps.70 nondysplastic SSAs, SSAs with low- and high-grade dysplasia, and SSAs with invasive adenocarcinoma, respectively).31 One US study reported an association Metachronous Neoplasia between higher levels of education and SSAs.62 Further research is needed to clarify whether other factors such A few studies have suggested that SSAs and large HPs as race, alcohol intake, other socioeconomic factors, and may be associated with an increased risk of metachro- diet are potentially significant risk factors for the nous polyps,58,65,71 but these studies generally used development of SSAs. Identifying predisposing factors for outdated pathologic criteria to define the serrated le- these important CRC precursors is important for pre- sions, were not confined specifically to SSAs, or were vention and diagnostic efforts. limited by small sample size and variable follow-up pe- Recently, a risk score for identifying large, proximal, riods. Further research is needed to better quantify the or dysplastic serrated polyps has been reported.59 This risk of synchronous and metachronous neoplasia asso- risk score was based on associated factors identified in a ciated with SSAs specifically. cross-sectional analysis of individuals undergoing elec- tive colonoscopy at a single Dutch hospital. The identified predictive factors based on multivariate analysis Detection and Removal included age older than 50 (odds ratio [OR], 2.2; 95% confidence interval [CI], 1.3–3.8), history of serrated Endoscopic Appearance polyps (OR, 2.6; 95% CI, 1.3–4.9), current smoking (OR, 2.2; 95% CI, 1.4–3.6), and lack of regular nonsteroidal Characteristic features of SSAs include proximal anti-inflammatory drug use (OR, 1.8; 95% CI, 1.1–3.0). It location (>75%),32,36 sessile or flat morphology remains to be seen whether this score will be useful in (>90%),39,72 a resemblance to prominent folds (37%),72 clinical practice, but it does serve to emphasize that the pale color (75%),39 indistinct borders (73%),73 and risk profile for SSAs differs somewhat from that of con- mucus capping (64%–100%)72,74 (Figure 3). All of these ventional adenomas (eg, lack of association with male features contribute to a subtle endoscopic appearance, sex), and endoscopists should be especially vigilant for and thus without careful attention from the endoscopist, SSAs in certain groups of patients (eg, smokers and those even large SSAs can escape detection. In 20% to 50% of with a history of SSAs). SSAs, a rim of bubbles or debris is present that can help delineate the lesions and serve as an identification aid.72 On chromoendoscopy or narrow-band imaging, SSAs Natural History typically have a type II-O or open Kudo pit pattern on a magnified view.40 Several Japanese studies have re- Risk of Dysplasia and Progression ported this pit pattern has good sensitivity and speci- ficity for identifying SSAs (84%–97% and 66%–86%, The risk and rate of progression of SSAs is not well respectively), but this technique is highly operator- documented. Some investigators have postulated that dependent.73,75,76 January 2015 Sessile Serrated Adenomas 17

Figure 3. Endoscopic images of SSA larger than 1 cm from 3 different patients featuring sessile or flat morphology and pale color (all), mucus/debris covering (all), resemblance to prominent folds (A and C), and indistinct borders (A and B). Some SSAs are located behind folds and are best visualized (and removed) from a retroflexed position (B).

Procedural Factors the existing data linking withdrawal times to serrated polyp detection are relatively sparse. One retrospective Bowel preparation. It is well established that inade- US study of more than 18,000 colonoscopies reported a quate bowel preparation negatively impacts adenoma strong relationship between withdrawal time and overall detection, particularly for smaller lesions.77 SSAs are serrated polyp detection.86 More recently, a prospective likely even more vulnerable to underdetection in persons Dutch study found that longer withdrawal time was with suboptimal preparations. There are multiple rea- associated with significantly better PSP detection (OR, sons for this problem: (1) flat morphology makes SSAs 1.12; 95% CI, 1.10–1.16).56 However, these studies were more vulnerable to being obscured by adherent stool and limited by the small number of endoscopists, and further intestinal fluids; (2) SSAs more commonly are located in research is needed on this topic. the right colon, which is more susceptible to imperfect cleansing; and (3) the mucus cap and ring of debris Endoscopic Detection Rates surrounding SSAs often facilitates identification in well- prepared colons, so poor preparations (and water jet Several studies have documented a wide variability in 37,55 irrigation needed to combat them) may eliminate this endoscopists’ detection of PSPs and SSAs specifically. 37 identification tool. Better bowel preparation is associated Hetzel et al reported results from a single-center with improved detection of flat lesions in general,78 but retrospective cohort study of 7192 patients undergoing one recent study did not find a statistically significant average-risk CRC screening colonoscopies by 13 different association between bowel preparation and PSP detec- endoscopists. In this study, SSA detection ranged from tion.56 However, the effects of bowel preparation on 0% in the lowest detector to 2.2% in the highest detec- detection of SSAs specifically has not been well studied. tor. Interestingly, this same study reported that overall Endoscopic technique. Different investigators have SSA detection improved substantially over time, from reported various techniques to optimize the detection of 0.2% in 2006 to 1.1% in 2008, suggesting either proximally located polyps including intensive inspection increased recognition by endoscopists, pathologists, or or second looks in the right colon, examination with both, over this time period. Kahi et al55 also investigated narrow-band imaging, routine retroflexion in the endoscopic detection of serrated lesions in a single- ascending colon, cap-fitted colonoscopy, and spray center retrospective study of 6681 patients undergoing chromoendoscopy.79–83 However, data that these tech- average-risk CRC screening by 15 different endoscopists. niques improve detection of SSAs specifically are lacking. In this study, detection of PSPs ranged from 1% to 18% Nevertheless, it makes intuitive sense that more careful for individual endoscopists. In a prospective study of and thorough examination of the right colon will help to 1426 people undergoing colonoscopic screening, de optimize SSA detection. In addition, high rates of SSA Wijkerslooth et al56 reported that the PSP detection rate detection have been reported with high-resolution ranged from 6% to 22% among 5 different experienced magnified chromoendoscopy and use of pit-pattern endoscopists. Another single-center (and single endo- recognition as mentioned earlier.32 However, there is scopist) study reported detection of PSPs and/or SSAs in significant variability in endoscopists’ abilities to accu- 23% of screening colonoscopies after using specialized rately identify pit patterns, neither routine chromoendo- lesion-recognition training, cap-fitted colonoscopy, scopy nor endoscopic magnification is in widespread use rigorous cleansing, and intensive inspection.83 Several of currently in the United States, and the cost effectiveness of these studies reported that SSA (or PSP) detection these techniques in average-risk screening populations is generally correlated with adenoma detection rates.37,55 It questionable.84 is clear from these data that when performed skillfully, Withdrawal time. Longer withdrawal time during endoscopy can identify important serrated lesions quite colonoscopic screening has been linked with higher commonly. Distressingly, these data also highlight the detection rates of conventional adenomas.85 However, fact that some endoscopists do not identify SSAs 18 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1 commonly enough, a situation that likely is fairly Pathologic Interpretation pervasive throughout the gastroenterology profession. Individual and practice improvement strategies are Because serrated neoplasia is a recent concept, needed to reduce this variation and improve the detec- pathologic interpretation also is inconsistent. One report tion and removal of important serrated precursor lesions from a single academic center with 12 different pathol- on an individual and systematic level. ogists reported that the SSA classification prevalence varied from 0% to 4% of patients depending on the 55 Detection of Sessile Serrated Adenomas by pathologist. Other studies of tandem pathologist readings have shown variable interobserver agreement Other Methods between pathologists, ranging from poor to good (k ¼ 0.16–0.66), with a substantial number of polyps Although colonoscopy certainly is imperfect, it read as an SSA by one pathologist, and as an HP by likely is superior to alternative CRC screening methods another.41,70,95 Further evidence of pathologic misclas- for the detection of SSAs. Little has been published in sification was provided by a recent Canadian study in this area, but it is known that computed tomography which previously reported HPs were re-reviewed from a (CT) colonography has difficulty detecting flat and city-wide sample by 2 gastrointestinal pathologists ac- sessile lesions because of its reliance on morphology, cording to current criteria.52 In this study, 17% of and absence of information on surface characteristics proximal HPs and 20% of HPs greater than 5 mm were (color, vascular pattern, and so forth).87,88 There are reclassified as SSAs. For these reasons, some in- scant data on the detection of lesions other than con- vestigators advocate treating all right-sided and larger ventional adenomas by CT colonography, but what has (1 cm) HPs the same as SSAs with respect to surveil- been published suggests that this technique lacks lance recommendations.96 However, gastrointestinal pa- adequate sensitivity for these lesions.89 With respect to thologists generally have higher reported SSA fecal occult blood testing and fecal immunochemical classification rates than nonspecialists,37,52 and a recent testing (FIT), serrated polyps are thought to be less European study showed excellent agreement among likely to hemorrhage than conventional adenomas, so expert pathologists (k >0.8) regarding serrated le- screening tests that rely on the detection of blood are sions.97 In addition, there is evidence that pathologic unlikely to be very effective.90,91 The recent develop- diagnoses of SSAs have increased steadily in recent ment of stool DNA tests for CRC largely have focused years.98 Taken together, these studies indicate that on markers based on the traditional adenoma- variation in pathologist interpretation may be reduced carcinoma sequence (ie, APC, KRAS), but these tests as the concept of serrated neoplasia becomes better offer some promise of serrated polyp detection, understood and disseminated. particularly if methylated or other markers of the serrated pathway are included.91,92 However, there is a paucity of published data on SSA detection with fecal Proposed Management occult blood testing, FIT, and stool DNA panels because, for the most part, serrated lesions have been Over the past several decades, colonoscopic ignored in CRC screening trials. Therefore, with all of screening in the United States has been tailored to the its limitations, colonoscopy likely represents our best detection, removal, and surveillance of conventional chance of addressing and removing serrated pathway adenomas. SSAs have received considerably less atten- precursor lesions such as SSAs. tion until recently. The problem of serrated neoplasia and SSAs specifically highlights many of the weaknesses Endoscopic Removal of our current colonoscopy screening approach to CRC prevention (Figure 4). Indeed, there are data that prior Few published studies have addressed the issue of colonoscopy offers little or no protection from SSAs, in optimal SSA polypectomy technique. In one recent study, contrast to advanced adenomas.99 From a public health resections of 346 polyps (including 42 SSAs) were standpoint, gastroenterologists must direct their atten- examined for completeness using standardized post- tion to the optimal detection, removal, and appropriate polypectomy biopsy specimens from the resected mar- surveillance of SSAs to effectively prevent sporadic CRC. gins.93 This study found that 31% of SSAs were resected From a patient-care standpoint, endoscopists must be incompletely, compared with 7% of other polyps. More aware of the existence, importance, and identifying strikingly, nearly half (48%) of the large SSAs (between 1 features of SSAs to optimize the management of pa- and 2 cm) were resected incompletely. This alarmingly tients harboring these premalignant lesions. Ongoing high rate of incomplete resection of SSAs highlights the professional education is important in this regard need for extra care when resecting large flat polyps in because this is a burgeoning area of research, and the right colon to avoid residual polyp tissue. Incomplete suggested management likely will be in evolution for resection of polyps is a likely explanation for some in- the foreseeable future. Specific recommendations are terval cancers (Figure 2).94 provided later. January 2015 Sessile Serrated Adenomas 19

Figure 4. The issue of serrated neoplasia magnifies many of the flaws of CRC screening programs, and colonoscopy screening specifically. (1) Risk factors for SSAs are poorly understood. Current risk-based screening recommendations (eg, screen between ages 50–75) may not optimize prevention of serrated pathway cancers. (2) SSAs are less likely to be detected by CT colonography (because of flat appearance) or flexible sigmoidoscopy (because of proximal location). Prevention of serrated pathway cancers may be more difficult in individuals who choose noncolonoscopy screening modalities. (3) SSAs have a subtle endoscopic appearance and can be missed during colonoscopy. There is significant variability in endoscopic detection of SSAs. Poor preparation quality (especially in the right colon) may further impair the detection of SSAs. (4) Because of flat/sessile morphology, size, and proximal location, removal of SSAs requires more polypectomy time and skill, and in- volves more risk. Incomplete resection is common with SSAs. (5) Pathologic understanding of serrated neoplasia is in evo- lution, with variability in interpretation and terminology used. SSAs may be read as HPs by some pathologists. (6) Surveillance recommendations are based largely on expert opinion and conjecture, not data. Some guidelines do not make specific management recommendations for serrated polyps. (7) Because of issues 3, 4, 5, and 6, individuals harboring SSAs may not undergo surveillance colonoscopy at appropriate intervals. f/u, follow up; prep, preparation.

Educate Your Patients About detect polypoid adenomas.102,103 Nevertheless, it goes Bowel Preparation without saying that all endoscopists should strive to meet and exceed published standards for conventional As discussed earlier, adequate purgative preparation adenoma detection rates.104 One group has suggested a before colonoscopy is critical to providing high-quality target PSP detection rate of at least 5% for average-risk endoscopic screening for adenomas, flat lesions, and screening colonoscopies.57 Although optimal or bench- (likely) SSAs. There are now good data to support the use mark SSA detection rates have not yet been firmly of split-dose preparation regimens, which are associated established, existing data suggest that SSAs should be with better patient satisfaction, better colonic cleansing identified in at least 1% of screening colonoscopies.105 (particularly of the right colon), and improved adenoma The true prevalence of SSAs is probably significantly and total polyp detection rates.100,101 Therefore, we higher than this figure, so even endoscopists meeting currently recommend using split-dose preparation regi- these arbitrary metrics likely can improve their detection mens for optimization of both conventional adenoma and rates. Some groups have reported use of standardized serrated polyp detection. education modules or training programs to optimize the detection of flat and subtle lesions.102 Suboptimal SSA detection, especially if out of proportion to PSP detection, Know Your Sessile Serrated Adenoma or also may indicate discordance between local pathologist Proximal Serrated Polyp Detection Rate interpretation and current pathology recommenda- tions,24,27,96 and should prompt a discussion regarding Given the known variation in endoscopic detection of the classification of serrated polyps with the interpreting SSAs, endoscopists should be aware of their SSA (or PSP) pathologist. detection rate. Although there are data that SSA/PSP detection rates and adenoma detection rates are corre- lated,37,55,57 these are not equivalent; the endoscopists Perform Attentive Colonoscopy with the highest adenoma detection rates do not always have the highest SSA or PSP detection rates.37,56 More- Meticulous colonoscopy technique is needed to opti- over, greater endoscopy experience does not necessarily mize SSA detection. In particular, assiduous examination imply a high PSP detection rate.56 Recognizing SSAs and of the right colon is important given that the majority other nonpolypoid and flat neoplastic lesions requires an of SSAs are proximally located. The endoscopist overlapping yet distinct skill set than that required to must perform careful examination of colonic folds and 20 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1 mucus-covered areas for associated mucosal abnormal- rare, and reserved for very large and/or dysplastic SSAs, ities. In our own practice, we recommend that endo- multiple SSAs, or those located in difficult locations (eg, scopists learn to retroflex the endoscope and use this appendiceal orifice).109 technique in the ascending colon to better visualize the backs of folds in the right colon. This practice has been Talk With Your Pathologist Colleagues shown to be safe and effective in experienced hands.80 Examination of the right colon more than once also is As discussed earlier, there is known variation in recommended by some experts, and this practice has pathologic interpretation of serrated polyps. Before the been shown to improve proximal polyp detection.79,106 unique histopathologic features of SSAs were identified, When suspicious lesions or mucosal abnormalities are these lesions generally were read by pathologists as found it is important to accurately record their size and HPs. It is likely that some systematic under-reading of location so that the site can be re-assessed at a later date SSAs continues to occur today. Gastroenterologists if needed. The use of narrow-band imaging or chro- should have a discussion with their pathology col- moendoscopy techniques may facilitate the identification leagues regarding the concept of serrated neoplasia and and delineation of SSAs, especially if the endoscopist is should partner to discuss cases in which the endoscopic trained in pit-pattern recognition.75,84 When a suspected and pathologic determinations differ (eg, a large right- SSA is found, heightened attention throughout the sided sessile polyp with typical endoscopic features of remainder of the examination is warranted because of an SSA read as a HP). Providing relevant clinical infor- the significant risk of synchronous serrated and con- mation, en bloc resections, and properly orientated ventional adenomas, and an increased risk of specimens also helps pathologists to interpret polyps CRC.58,68,70,107 accurately.110 If this is performed, the pathologist should be able to determine if the margins of the polyp are free of residual neoplasia and that information can Resect Carefully aid in follow-up recommendations. As with any neoplastic colorectal polyp, complete resection is important to reduce the risk of dysplastic or Be Aware of Surveillance Guidelines cancerous tissue left in situ, and to avoid regrowth of neoplastic tissue at the polypectomy site. As discussed A key question with respect to the management of earlier, incomplete resection of SSAs, even in experi- patients with SSAs pertains to the recommended interval enced hands, is relatively common.93 Most SSAs are for surveillance colonoscopy. Several investigators and smaller than 2 cm (average size, 5–7 mm)24,35 and can be groups have published recommendations on surveillance resected safely and effectively at the time of a screening of individuals with serrated polyps (Table 2). Most or surveillance colonoscopy. Optimal technique varies recently, the US Multisociety Task Force guidelines for depending on the individual characteristics of the SSA. surveillance after colonoscopic polypectomy recom- For flat lesions, submucosal injection often is useful to mended 1 to 5 years of follow-up evaluation of SSAs, provide a cushion for resection, and to allow for better depending on the number and size of lesions found, and visualization and seating of the polypectomy snare. the presence or absence of dysplasia.111 These guidelines Chromoendoscopy and/or including a dye (eg, indigo largely mirror those for conventional adenomas. A carmine) in the submucosal injection fluid also can help consensus panel of international experts on serrated better define the margins of the polyp and assist com- neoplasia also recently published similar, but more plete resection. In light of the data showing an inordi- detailed, recommendations.96 Individual investigators nately high incomplete resection rate for SSAs larger also have made various recommendations regarding than 1 cm, it seems reasonable to tattoo such lesions and surveillance of these lesions that vary from 6 months to repeat an endoscopy in 3 to 6 months to assess the site 10 years.5,17,64,91,112,113 Some current colonoscopy and and remove any residual polyp. CRC screening guidelines do not make any specific rec- An expanded armamentarium is required for ommendations regarding surveillance intervals for those removing SSAs; certain snares may be more effective for with serrated lesions, or recommend that management removal of these types of polyps including stiffer snares, be analogous to that of conventional adenomas.114–116 or spiral or crescent/duckbill-shaped snares.106 If the We favor the consensus panel guidelines, given their endoscopist is inexperienced in performing endoscopic more detailed management recommendations,96 but it is mucosal resection, or is uncomfortable removing a sus- important to recognize that surveillance recommenda- pected SSA because of its morphology (eg, flat or laterally tions for SSAs likely will be in evolution as our under- spreading), size (eg, >2 cm), or location (behind a fold, standing of the behavior and significance of these lesions abutting the ileocecal valve, and so forth), referral to a improves. One area of ambiguity relates to SSAs with colleague or center with appropriate advanced endos- dysplasia (SSAD), for which the consensus panel guide- copy expertise in complex polypectomies is recom- lines recommend a 1- to 3-year surveillance interval. mended.108,109 The need for surgical resection of SSAs is Although a 3-year surveillance examination seems January 2015 Sessile Serrated Adenomas 21

Table 2. Published Recommendations for Endoscopic Surveillance in Persons Harboring SSAs

Author, year Recommended surveillance for SSAs

Groups/guidelines US consensus panel, 201296 <3 SSA, all <10 mm: 5 y 3 SSAs: 3 y SSA 10 mm: 3 y 2 SSAs 10 mm: 1–3y SSAD: 1–3y US Multi-society Task Force, 2012111 SSA <10 mm: 5 y SSA 10 mm: 3 y SSAD: 3 y Australian Guidelines, 2011115 No specific recommendation British Society of Gastroenterology Guidelines, 2010114 No specific recommendation European Union Guidelines, 2010116 No specific recommendation Individual authors Huang, 20115 <3 SSAs, all <10 mm: 5 y 3 SSAs: 3 y SSA 10 mm: 3 y SSAD: 3 y Vu, 201164 <3 SSA: 5–10 y SSA <10 mm: 5–10 y 3 SSAs: 3 y SSA 10 mm: 3 y SSAD: 3 y Snover, 201117 2 SSAs 10 mm: 1 y Terdiman, 2010112 Proximal SSA/HP: 3 y Distal SSA/HP <10 mm: 10 y SSA/HP 10 mm: 3 y SSAD: 3 y Groff, 2008113 <3 SSA: 5 y 3 SSAs: 3 y SSAD: 3 y East, 200891 SSA 10 mm: 3 y 3 SSAs: 3 y 3 Proximal HP: 5 y Proximal HP 10 mm: 5 y

appropriate for patients with SSADs if complete resec- Uncertainties and Future Directions tion was achieved, endoscopists should have a low threshold to perform repeat colonoscopy at a short in- Pathogenesis terval (3–6 mo) for assessment of an SSAD polypectomy site when piecemeal resection was performed, or when The understanding of serrated neoplasia continues to residual polyp tissue is suspected on the basis of endo- advance, but many uncertainties remain. Further scopic appearance or pathology evaluation. research is needed in the area of pathogenesis of An important consideration is patients with sus- serrated neoplasia, so the genetic and epigenetic steps pected SPS, for whom the recommended surveillance along the serrated pathway to CRC (particularly for interval is 1 year. It is worth noting that WHO criteria CIMP-MSS carcinomas) can be better delineated. Dis- 1 and 3 for SPS (see earlier) are the most common phe- 42 covery of the fundamental cause of CIMP and uniformity notypes, and given that multiplicity of serrated polyps is fi 54 in its de nition would help clarify this important feature common, both scenarios can be underdiagnosed, leading of the serrated pathway. In addition, further study of to inappropriately long (ie, >1 y) surveillance recom- 117 patients with SPS and its phenotypes and inheritance mendations in such patients. Interestingly, in patients patterns offers the opportunity to identify the genetic with a history of SSAs, some investigators recommend basis of this hereditary syndrome. It is likely that these considering optimizing surveillance detection by use of 91 genes also would be relevant to sporadic serrated chromoendoscopy. Furthermore, extending colono- neoplasia and SSAs specifically. It also is possible that the scopic surveillance beyond the traditional cut-off point of current diagnostic criteria for SPS may become mean- 75 years of age also has been recommended because of the 91 ingless if the rate of recognition of SSAs continues to presumed potential for rapid neoplastic growth. How- increase, and/or if more meaningful phenotypic patterns fi ever, evidence to support the bene t of these recommen- emerge. dations currently is lacking. 22 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1

Epidemiology and Natural History fully understand the implications of SSAs for screening guidelines and test selection. Despite growing recognition of the problem of Because resection of SSAs often is inadequate,93 serrated neoplasia, as described earlier, the epidemi- further research is needed to optimize complete resec- ology and natural history of SSAs remain poorly under- tion of these lesions to mitigate the risk of polyp stood. A better understanding of the risk factors for SSAs regrowth and interval cancers (Figure 2). It is worth may help to determine not only which populations have noting that the currently available tools for polypectomy the highest risk of harboring these lesions, but also could (eg, standard oval snares) largely were developed with help inform studies of pathogenesis. There have been polypoid adenomas in mind, and other snare shapes and limited longitudinal studies of these lesions, and little is polypectomy tools may need to be appropriated or known about the dwell time, features associated with developed to optimize resection of SSAs and other large rapid growth or recurrence, and the true prognostic and/or flat colonic lesions. significance of finding SSAs on colonoscopy. As such, the current guidelines for the management of serrated lesions96,111 are based more on expert opinion rather Conclusions than a substantial evidence base. Additional prospective studies are necessary to better inform clinical decision SSAs represent a new type of combatant in the war making and surveillance recommendations. Along those against CRC, and it is important for clinicians and re- lines, some investigators have recommended random- searchers to understand their importance. These lesions ized trials of SSA polypectomy and surveillance vs undoubtedly play a role in interval cancer development watchful waiting or serial biopsies to generate evidence (Figure 2). It is clear that unless gastroenterologists regarding the benefits and harms of different manage- direct their attention to the detection, removal, and ment strategies.118 Furthermore, the clinical implications appropriate surveillance of SSAs, optimal prevention of of genetic and epigenetic alterations of SSAs largely are sporadic CRC will not be possible. The current CRC unknown. It is possible that a more comprehensive screening approach also must adapt to address serrated semimolecular classification of colorectal polyps will neoplasia (Figure 4). Additional research is needed in help better risk-stratify patients in the future.17 this area, particularly with regard to pathogenesis, epidemiology, behavior, and endoscopic management of these important CRC precursors. Detection and Resection Supplementary Material Because variable detection (and underdetection spe- cifically) is so pervasive, it seems that initial efforts should be directed at improving endoscopic detection Note: To access the supplementary material accom- and removal with existing technologies, personnel, and panying this article, visit the online version of Clinical infrastructure. However, because the detection of SSAs is Gastroenterology and Hepatology at www.cghjournal.org, affected disproportionately by the inherent technical and at http://dx.doi.org/10.1016/j.cgh.2013.10.035. limitations of visual inspection with standard white-light References colonoscopy, novel or adjunctive approaches may be 1. GLOBOCAN 2008. Section of cancer information. Lyon, France: needed. Several existing endoscopic tools offer the po- International Agency for Research on Cancer (IARC). Available tential to improve SSA detection including magnification at: http://globocan.iarc.fr. Accessed August 26, 2011. chromoendoscopy,32 cap-fitted colonoscopy,82 and 2. American Cancer Society. Cancer facts & figures 2013. Atlanta: retroflexion in ascending colon80 as detailed earlier. In American Cancer Society, 2013. addition, emerging technologies such as wide-angle co- 3. Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations lonoscopy,119 use of the third eye retroscope,120 and during colorectal-tumor development. N Engl J Med 1988; – others also are promising in this regard. However, more 319:525 532. fi research is needed to determine whether these devices 4. Jass JR. Classi cation of colorectal cancer based on correlation and techniques (alone or in combination) are effective, of clinical, morphological and molecular features. Histopathol- ogy 2007;50:113–130. easily learned, practical, and cost effective with respect 5. Huang CS, Farraye FA, Yang S, et al. The clinical significance of to improving SSA detection. As mentioned earlier, stool serrated polyps. Am J Gastroenterol 2011;106:229–240, quiz 241. DNA tests do offer some promise of serrated lesion 6. Leggett B, Whitehall V. Role of the serrated pathway in colorectal detection, particularly if serrated pathway markers are cancer pathogenesis. Gastroenterology 2010;138:2088–2100. included. Endomicroscopy, spectroscopy, and molecular 7. Noffsinger AE, Hart J. Serrated adenoma: a distinct form of non- imaging also are intriguing developments that potentially polypoid colorectal neoplasia? Gastrointest Endosc Clin N Am 121–123 could impact SSA detection in the future. 2010;20:543–563. Furthermore, better quantitative estimates of SSA 8. Snover DC, Jass JR, Fenoglio-Preiser C, et al. Serrated polyps detection by alternative screening methods (eg, FIT, CT of the large intestine: a morphologic and molecular review of an colonography, flexible sigmoidoscopy) are needed to evolving concept. Am J Clin Pathol 2005;124:380–391. January 2015 Sessile Serrated Adenomas 23

9. Weisenberger DJ, Siegmund KD, Campan M, et al. CpG island 29. Clark JC, Collan Y, Eide TJ, et al. Prevalence of polyps in an methylator phenotype underlies sporadic microsatellite insta- autopsy series from areas with varying incidence of large-bowel bility and is tightly associated with BRAF mutation in colorectal cancer. Int J Cancer 1985;36:179–186. cancer. Nat Genet 2006;38:787–793. 30. Su MY, Hsu CM, Ho YP, et al. Comparative study of conven- 10. Jass JR. Serrated adenoma of the colorectum and the DNA- tional colonoscopy, chromoendoscopy, and narrow-band im- methylator phenotype. Nat Clin Pract Oncol 2005;2:398–405. aging systems in differential diagnosis of neoplastic and 11. Baxter NN, Goldwasser MA, Paszat LF, et al. Association of nonneoplastic colonic polyps. Am J Gastroenterol 2006; colonoscopy and death from colorectal cancer. Ann Intern Med 101:2711–2716. 2009;150:1–8. 31. Lash RH, Genta RM, Schuler CM. Sessile serrated adenomas: 12. Brenner H, Hoffmeister M, Arndt V, et al. Protection from right- prevalence of dysplasia and carcinoma in 2139 patients. J Clin and left-sided colorectal neoplasms after colonoscopy: Pathol 2010;63:681–686. population-based study. J Natl Cancer Inst 2010;102:89–95. 32. Spring KJ, Zhao ZZ, Karamatic R, et al. High prevalence of 13. Nishihara R, Wu K, Lochhead P, et al. Long-term colorectal- sessile serrated adenomas with BRAF mutations: a prospective cancer incidence and mortality after lower endoscopy. N Engl study of patients undergoing colonoscopy. Gastroenterology J Med 2013;369:1095–1105. 2006;131:1400–1407. 14. Arain MA, Sawhney M, Sheikh S, et al. CIMP status of interval 33. O’Brien MJ, Yang S, Mack C, et al. Comparison of microsatellite colon cancers: another piece to the puzzle. Am J Gastroenterol instability, CpG island methylation phenotype, BRAF and KRAS 2010;105:1189–1195. status in serrated polyps and traditional adenomas indicates 15. Sawhney MS, Farrar WD, Gudiseva S, et al. Microsatellite separate pathways to distinct colorectal carcinoma end points. instability in interval colon cancers. Gastroenterology 2006; Am J Surg Pathol 2006;30:1491–1501. 131:1700–1705. 34. Bettington M, Walker N, Clouston A, et al. The serrated pathway 16. Samadder NJ, Curtin K, Tuohy T, et al. Prevalence and pre- to colorectal carcinoma: current concepts and challenges. His- dictors of missed or interval colorectal cancer: a population- topathology 2013;62:367–386. based study in Utah. Gastroenterology 2013;144:S1–S2. 35. Higuchi T, Sugihara K, Jass JR. Demographic and pathological 17. Snover DC. Update on the serrated pathway to colorectal car- characteristics of serrated polyps of colorectum. Histopathology cinoma. Hum Pathol 2011;42:1–10. 2005;47:32–40. 18. Cooper HS, Patchefsky AS, Marks G. Adenomatous and carci- 36. Carr NJ, Mahajan H, Tan KL, et al. Serrated and non-serrated nomatous changes within hyperplastic colonic epithelium. Dis polyps of the colorectum: their prevalence in an unselected Colon Rectum 1979;22:152–156. case series and correlation of BRAF mutation analysis with the 19. Goldman H, Ming S, Hickock DF. Nature and significance of diagnosis of sessile serrated adenoma. J Clin Pathol 2009; – hyperplastic polyps of the human colon. Arch Pathol 1970; 62:516 518. 89:349–354. 37. Hetzel JT, Huang CS, Coukos JA, et al. Variation in the detection 20. Urbanski SJ, Kossakowska AE, Marcon N, et al. Mixed hy- of serrated polyps in an average risk colorectal cancer screening – perplastic adenomatous polyps–an underdiagnosed entity. cohort. Am J Gastroenterol 2010;105:2656 2664. Report of a case of adenocarcinoma arising within a mixed 38. Fu B, Yachida S, Morgan R, et al. Clinicopathologic and genetic hyperplastic adenomatous polyp. Am J Surg Pathol 1984; characterization of traditional serrated adenomas of the colon. 8:551–556. Am J Clin Pathol 2012;138:356–366. 21. Jass JR. Relation between metaplastic polyp and carcinoma of 39. Oka S, Tanaka S, Hiyama T, et al. Clinicopathologic and endo- the colorectum. Lancet 1983;1:28–30. scopic features of colorectal serrated adenoma: differences fi 22. Longacre TA, Fenoglio-Preiser CM. Mixed hyperplastic adeno- between polypoid and super cial types. Gastrointest Endosc – matous polyps/serrated adenomas. A distinct form of colorectal 2004;59:213 219. neoplasia. Am J Surg Pathol 1990;14:524–537. 40. Limketkai BN, Lam-Himlin D, Arnold CA, et al. The cutting edge 23. Torlakovic E, Snover DC. Serrated adenomatous polyposis in of serrated polyps: a practical guide to approaching and man- humans. Gastroenterology 1996;110:748–755. aging serrated colon polyps. Gastrointest Endosc 2013; – 24. Torlakovic E, Skovlund E, Snover DC, et al. Morphologic reap- 77:360 375. praisal of serrated colorectal polyps. Am J Surg Pathol 2003; 41. Farris AB, Misdraji J, Srivastava A, et al. Sessile serrated ade- 27:65–81. noma: challenging discrimination from other serrated colonic – 25. Goldstein NS, Bhanot P, Odish E, et al. Hyperplastic-like colon polyps. Am J Surg Pathol 2008;32:30 35. polyps that preceded microsatellite-unstable adenocarcinomas. 42. Rosty C, Buchanan DD, Walsh MD, et al. Phenotype and polyp Am J Clin Pathol 2003;119:778–796. landscape in serrated polyposis syndrome: a series of 100 pa- – 26. Torlakovic EE, Gomez JD, Driman DK, et al. Sessile serrated tients from genetics clinics. Am J Surg Pathol 2012;36:876 882. adenoma (SSA) vs. traditional serrated adenoma (TSA). Am J 43. Guarinos C, Sanchez-Fortun C, Rodriguez-Soler M, et al. Surg Pathol 2008;32:21–29. Serrated polyposis syndrome: molecular, pathological and – 27. Snover D, Ahnen D, Burt R, et al. Serrated polyps of the colon clinical aspects. World J Gastroenterol 2012;18:2452 2461. and rectum and serrated (“hyperplastic”) polyposis. In: 44. Young J, Jass JR. The case for a genetic predisposition to serrated Bozman F, Carneiro F, Hruban R, et al, eds. WHO classification neoplasia in the colorectum: hypothesis and review of the litera- – of tumours of the digestive system. 4th ed. Berlin: Springer- ture. Cancer Epidemiol Biomarkers Prev 2006;15:1778 1784. Verlag, 2010. 45. Boparai KS, Mathus-Vliegen EM, Koornstra JJ, et al. Increased 28. Johannsen LG, Momsen O, Jacobsen NO. Polyps of the large colorectal cancer risk during follow-up in patients with hyper- intestine in Aarhus, Denmark. An autopsy study. Scand J Gas- plastic polyposis syndrome: a multicentre cohort study. Gut – troenterol 1989;24:799–806. 2010;59:1094 1100. 24 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1

46. Ferrandez A, Samowitz W, DiSario JA, et al. Phenotypic char- 64. Vu HT, Lopez R, Bennett A, et al. Individuals with sessile acteristics and risk of cancer development in hyperplastic pol- serrated polyps express an aggressive colorectal phenotype. yposis: case series and literature review. Am J Gastroenterol Dis Colon Rectum 2011;54:1216–1223. 2004;99:2012–2018. 65. Teriaky A, Driman DK, Chande N. Outcomes of a 5-year follow- 47. Oono Y, Fu K, Nakamura H, et al. Progression of a sessile up of patients with sessile serrated adenomas. Scand J Gas- serrated adenoma to an early invasive cancer within 8 months. troenterol 2012;47:178–183. Dig Dis Sci 2009;54:906–909. 66. Suehiro Y, Hinoda Y. Genetic and epigenetic changes in aber- 48. Goldstein NS. Small colonic microsatellite unstable adenocar- rant crypt foci and serrated polyps. Cancer Sci 2008;99: cinomas and high-grade epithelial dysplasias in sessile serrated 1071–1076. adenoma polypectomy specimens: a study of eight cases. Am J 67. Rosenberg DW, Yang S, Pleau DC, et al. Mutations in BRAF and Clin Pathol 2006;125:132–145. KRAS differentially distinguish serrated versus non-serrated 49. Kambara T, Simms LA, Whitehall VL, et al. BRAF mutation is hyperplastic aberrant crypt foci in humans. Cancer Res 2007; associated with DNA methylation in serrated polyps and can- 67:3551–3554. cers of the colorectum. Gut 2004;53:1137–1144. 68. Hiraoka S, Kato J, Fujiki S, et al. The presence of large serrated 50. Vatn MH, Stalsberg H. The prevalence of polyps of the large polyps increases risk for colorectal cancer. Gastroenterology intestine in Oslo: an autopsy study. Cancer 1982;49:819–825. 2010;139:1503–1510. 51. Williams AR, Balasooriya BA, Day DW. Polyps and cancer of the 69. Li D, Jin C, McCulloch C, et al. Association of large serrated large bowel: a necropsy study in Liverpool. Gut 1982; polyps with synchronous advanced colorectal neoplasia. Am J 23:835–842. Gastroenterol 2009;104:695–702. 52. Singh H, Bay D, Ip S, et al. Pathological reassessment of hy- 70. Pai RK, Hart J, Noffsinger AE. Sessile serrated adenomas perplastic colon polyps in a city-wide pathology practice: im- strongly predispose to synchronous serrated polyps in non- plications for polyp surveillance recommendations. Gastrointest syndromic patients. Histopathology 2010;56:581–588. Endosc 2012;76:1003–1008. 71. Lu FI, van Niekerk DW, Owen D, et al. Longitudinal outcome 53. Kumbhari V, Behary J, Hui JM. Prevalence of adenomas and study of sessile serrated adenomas of the colorectum: an sessile serrated adenomas in Chinese compared with Cauca- increased risk for subsequent right-sided colorectal carcinoma. sians. J Gastroenterol Hepatol 2013;28:608–612. Am J Surg Pathol 2010;34:927–934. 54. Alvarez C, Andreu M, Castells A, et al. Relationship of 72. Tadepalli US, Feihel D, Miller KM, et al. A morphologic analysis colonoscopy-detected serrated polyps with synchronous of sessile serrated polyps observed during routine colonoscopy advanced neoplasia in average-risk individuals. Gastrointest (with video). Gastrointest Endosc 2011;74:1360–1368. Endosc 2013;78:333–341. 73. Hazewinkel Y, Lopez-Ceron M, East JE, et al. Endoscopic fea- 55. Kahi CJ, Hewett DG, Norton DL, et al. Prevalence and variable tures of sessile serrated adenomas: validation by international detection of proximal colon serrated polyps during screening experts using high-resolution white-light endoscopy and colonoscopy. Clin Gastroenterol Hepatol 2011;9:42–46. narrow-band imaging. Gastrointest Endosc 2013;77:916–924. 56. de Wijkerslooth TR, Stoop EM, Bossuyt PM, et al. Differences in 74. Rex DK, Rahmani EY. New endoscopic finding associated with proximal serrated polyp detection among endoscopists are hyperplastic polyps. Gastrointest Endosc 1999;50:704–706. associated with variability in withdrawal time. Gastrointest 75. Kimura T, Yamamoto E, Yamano HO, et al. A novel pit pattern Endosc 2013;77:617–623. identifies the precursor of colorectal cancer derived from 57. Kahi CJ, Li X, Eckert GJ, et al. High colonoscopic prevalence of sessile serrated adenoma. Am J Gastroenterol 2012;107: proximal colon serrated polyps in average-risk men and women. 460–469. Gastrointest Endosc 2012;75:515–520. 76. Ishigooka S, Nomoto M, Obinata N, et al. Evaluation of magni- 58. Schreiner MA, Weiss DG, Lieberman DA. Proximal and large fying colonoscopy in the diagnosis of serrated polyps. World J hyperplastic and nondysplastic serrated polyps detected by Gastroenterol 2012;18:4308–4316. colonoscopy are associated with neoplasia. Gastroenterology 77. Harewood GC, Sharma VK, de Garmo P. Impact of colonoscopy 2010;139:1497–1502. preparation quality on detection of suspected colonic neoplasia. 59. Bouwens MW, Winkens B, Rondagh EJ, et al. Simple clinical risk Gastrointest Endosc 2003;58:76–79. score identifies patients with serrated polyps in routine practice. 78. Parra-Blanco A, Nicolas-Perez D, Gimeno-Garcia A, et al. The Cancer Prev Res (Phila) 2013;6:855–863. timing of bowel preparation before colonoscopy determines the 60. Wallace K, Grau MV, Ahnen D, et al. The association of lifestyle quality of cleansing, and is a significant factor contributing to the and dietary factors with the risk for serrated polyps of the col- detection of flat lesions: a randomized study. World J Gastro- orectum. Cancer Epidemiol Biomarkers Prev 2009;18: enterol 2006;12:6161–6166. 2310–2317. 79. Stoffel EM, Turgeon DK, Stockwell DH, et al. Chromoendoscopy 61. Anderson JC, Rangasamy P, Rustagi T, et al. Risk factors for detects more adenomas than colonoscopy using intensive in- sessile serrated adenomas. J Clin Gastroenterol 2011;45: spection without dye spraying. Cancer Prev Res (Phila) 2008; 694–699. 1:507–513. 62. Burnett-Hartman AN, Passarelli MN, Adams SV, et al. Differ- 80. Hewett DG, Rex DK. Miss rate of right-sided colon examination ences in epidemiologic risk factors for colorectal adenomas and during colonoscopy defined by retroflexion: an observational serrated polyps by lesion severity and anatomical site. Am J study. Gastrointest Endosc 2011;74:246–252. Epidemiol 2013;177:625–637. 81. Pohl J, Schneider A, Vogell H, et al. Pancolonic chromoendo- 63. Rustagi T, Rangasamy P, Myers M, et al. Sessile serrated ade- scopy with indigo carmine versus standard colonoscopy for nomas in the proximal colon are likely to be flat, large and occur detection of neoplastic lesions: a randomised two-centre trial. in smokers. World J Gastroenterol 2013;19:5271–5277. Gut 2011;60:485–490. January 2015 Sessile Serrated Adenomas 25

82. Hewett DG, Rex DK. Cap-fitted colonoscopy: a randomized, meta-analysis of randomized controlled trials. Gastrointest tandem colonoscopy study of adenoma miss rates. Gastrointest Endosc 2011;73:1240–1245. Endosc 2010;72:775–781. 101. Gurudu SR, Ramirez FC, Harrison ME, et al. Increased adenoma 83. Raju GS, Vadyala V, Slack R, et al. Adenoma detection in pa- detection rate with system-wide implementation of a split-dose tients undergoing a comprehensive colonoscopy screening. preparation for colonoscopy. Gastrointest Endosc 2012; Cancer Med 2013;2:391–402. 76:603–608.e1. 84. Rastogi A, Pondugula K, Bansal A, et al. Recognition of surface 102. Rondagh EJ, Bouwens MW, Riedl RG, et al. Endoscopic mucosal and vascular patterns of colon polyps by using narrow- appearance of proximal colorectal neoplasms and potential band imaging: interobserver and intraobserver agreement and implications for colonoscopy in cancer prevention. Gastrointest prediction of polyp histology. Gastrointest Endosc 2009; Endosc 2012;75:1218–1225. 69:716–722. 103. Soetikno RM, Kaltenbach T, Rouse RV, et al. Prevalence of non- 85. Barclay RL, Vicari JJ, Doughty AS, et al. Colonoscopic with- polypoid (flat and depressed) colorectal neoplasms in asymp- drawal times and adenoma detection during screening colo- tomatic and symptomatic adults. JAMA 2008;299:1027–1035. noscopy. N Engl J Med 2006;355:2533–2541. 104. Rex DK, Bond JH, Winawer S, et al. Quality in the technical 86. Liang J, Kalady MF, Appau K, et al. Serrated polyp detection performance of colonoscopy and the continuous quality rate during screening colonoscopy. Colorectal Dis 2012; improvement process for colonoscopy: recommendations of the 14:1323–1327. U.S. Multi-Society Task Force on Colorectal Cancer. Am J 87. Pickhardt PJ. Missed lesions at CT colonography: lessons Gastroenterol 2002;97:1296–1308. learned. Abdom Imaging 2013;38:82–97. 105. Sanaka MR, Gohel T, Podugu A, et al. Quality indicators to 88. Park SH, Ha HK, Kim MJ, et al. False-negative results at multi- enhance adenoma detection rate: should there be reconsider- detector row CT colonography: multivariate analysis of causes ation of the current standard? Gastrointest Endosc 2011; for missed lesions. Radiology 2005;235:495–502. 73:AB138. 89. Park SH, Kim SY, Lee SS, et al. Sensitivity of CT colonography 106. Bourke MJ, Rex DK. Tips for better colonoscopy from two ex- for nonpolypoid colorectal lesions interpreted by human readers perts. Am J Gastroenterol 2012;107:1467–1472. and with computer-aided detection. AJR Am J Roentgenol 107. Rondagh EJ, Masclee AA, Bouwens MW, et al. Endoscopic red 2009;193:70–78. flags for the detection of high-risk serrated polyps: an obser- 90. Waldock A, Ellis IO, Armitage NC, et al. Histopathological vational study. Endoscopy 2011;43:1052–1058. assessment of bleeding from polyps of the colon and rectum. 108. Buchner AM, Guarner-Argente C, Ginsberg GG. Outcomes of J Clin Pathol 1989;42:378–382. EMR of defiant colorectal lesions directed to an endoscopy 91. East JE, Saunders BP, Jass JR. Sporadic and syndromic hy- referral center. Gastrointest Endosc 2012;76:255–263. perplastic polyps and serrated adenomas of the colon: classi- 109. Swan MP, Bourke MJ, Alexander S, et al. Large refractory fication, molecular genetics, natural history, and clinical colonic polyps: is it time to change our practice? A prospective management. Gastroenterol Clin North Am 2008;37:25–46, v. study of the clinical and economic impact of a tertiary referral 92. Jin YM, Li BJ, Qu B, et al. BRAF, K-ras and BAT26 mutations in colonic mucosal resection and polypectomy service (with colorectal polyps and stool. World J Gastroenterol 2006; videos). Gastrointest Endosc 2009;70:1128–1136. 12:5148–5152. 110. Morales SJ, Bodian CA, Kornacki S, et al. A simple tissue- 93. Pohl H, Srivastava A, Bensen SP, et al. Incomplete polyp handling technique performed in the endoscopy suite improves resection during colonoscopy-results of the complete adenoma histologic section quality and diagnostic accuracy for serrated resection (CARE) study. Gastroenterology 2013;144:74–80.e1. polyps. Endoscopy 2013;45:897–905. 94. Farrar WD, Sawhney MS, Nelson DB, et al. Colorectal cancers 111. Lieberman DA, Rex DK, Winawer SJ, et al. Guidelines for colo- found after a complete colonoscopy. Clin Gastroenterol Hepatol noscopy surveillance after screening and polypectomy: a 2006;4:1259–1264. consensus update by the US Multi-Society Task Force on 95. Khalid O, Radaideh S, Cummings OW, et al. Reinterpretation of Colorectal Cancer. Gastroenterology 2012;143:844–857. histology of proximal colon polyps called hyperplastic in 2001. 112. Terdiman JP, McQuaid KR. Surveillance guidelines should be World J Gastroenterol 2009;15:3767–3770. updated to recognize the importance of serrated polyps. 96. Rex DK, Ahnen DJ, Baron JA, et al. Serrated lesions of the Gastroenterology 2010;139:1444–1447. colorectum: review and recommendations from an expert panel. 113. Groff RJ, Nash R, Ahnen DJ. Significance of serrated polyps of Am J Gastroenterol 2012;107:1315–1329, quiz 1314, 1330. the colon. Curr Gastroenterol Rep 2008;10:490–498. 97. Ensari A, Bilezikci B, Carneiro F, et al. Serrated polyps of the 114. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colon: how reproducible is their classification? Virchows Arch colorectal cancer screening and surveillance in moderate and 2012;461:495–504. high risk groups (update from 2002). Gut 2010;59:666–689. 98. Gill P, Wang LM, Bailey A, et al. Reporting trends of right-sided 115. Cancer Council Australia Colonoscopy Surveillance Working Party. hyperplastic and sessile serrated polyps in a large teaching Clinical Practice Guidelines for Surveillance Colonoscopy–in hospital over a 4-year period (2009-2012). J Clin Pathol 2013; adenoma follow-up; following curative resection of colorectal 66:655–658. cancer; and for cancer surveillance in inflammatory bowel disease. 99. Burnett-Hartman AN, Newcomb PA, Phipps AI, et al. Colorectal Sydney, Australia: Cancer Council Australia, 2011. endoscopy, advanced adenomas, and sessile serrated polyps: 116. Atkin W, Valori R, Kuipers EJ, et al. Colonoscopic surveillance implications for proximal colon cancer. Am J Gastroenterol after adenoma removal. In: Segnan N, Patnick J, von Karsa L, 2012;107:1213–1219. eds. European guidelines for quality assurance in colorectal 100. Kilgore TW, Abdinoor AA, Szary NM, et al. Bowel preparation cancer screening and diagnosis. 1st ed. Luxembourg: IARC, with split-dose polyethylene glycol before colonoscopy: a 2010. 26 Crockett et al Clinical Gastroenterology and Hepatology Vol. 13, No. 1

117. Vemulapalli KC, Rex DK. Failure to recognize serrated polyposis 122. Roy HK, Backman V. Spectroscopic applications in gastroin- syndrome in a cohort with large sessile colorectal polyps. testinal endoscopy. Clin Gastroenterol Hepatol 2012;10: Gastrointest Endosc 2012;75:1206–1210. 1335–1341. 118. Hoff G, Bretthauer M, Garborg K, et al. New polyps, old tricks: 123. Zavaleta CL, Garai E, Liu JT, et al. A Raman-based endoscopic controversy about removing benign bowel lesions. BMJ 2013; strategy for multiplexed molecular imaging. Proc Natl Acad Sci 347:f5843. U S A 2013;110:E2288–E2297. 119. Gralnek IM, Carr-Locke DL, Segol O, et al. Comparison of 124. Jeevaratnam P, Cottier DS, Browett PJ, et al. Familial giant standard forward-viewing mode versus ultrawide-viewing hyperplastic polyposis predisposing to colorectal cancer: a new mode of a novel colonoscopy platform: a prospective, multi- hereditary bowel cancer syndrome. J Pathol 1996;179:20–25. center study in the detection of simulated polyps in an in vitro colon model (with video). Gastrointest Endosc 2013; Reprint requests 77:472–479. Address requests for reprints to: Seth D. Crockett, MD, MPH, Assistant 120. Waye JD, Heigh RI, Fleischer DE, et al. A retrograde-viewing Professor, Division of Gastroenterology and Hepatology, University of North Carolina, CB 7080, Chapel Hill, North Carolina 27599. e-mail: [email protected]; device improves detection of adenomas in the colon: a pro- fax: (919) 966-6842. spective efficacy evaluation (with videos). Gastrointest Endosc 2010;71:551–556. Conflicts of interest fl 121. Buchner AM, Shahid MW, Heckman MG, et al. Comparison of The authors disclose no con icts. probe-based confocal laser endomicroscopy with virtual chro- fi Funding moendoscopy for classi cation of colon polyps. Gastroenter- This project was supported, in part, by an American College of Gastroenter- ology 2010;138:834–842. ology Junior Faculty Development Award (ACG-JR-000-2012). January 2015 Sessile Serrated Adenomas 26.e1

Supplementary Figure 1. WHO taxonomy of serrated polyps.