Current Colorectal Cancer Reports

, Volume 7, Issue 1, pp 42–49

Hyperplastic Polyps: Are They Completely Innocent?

Authors

    • Gastroenterology DepartmentInstitut de Malalties Digestives i Metabòliques, CIBERehd, IDIBAPS, Hospital Clínic
  • Antoni Castells
    • Gastroenterology DepartmentInstitut de Malalties Digestives i Metabòliques, CIBERehd, IDIBAPS, Hospital Clínic
Article

DOI: 10.1007/s11888-010-0080-z

Cite this article as:
Balaguer, F. & Castells, A. Curr Colorectal Cancer Rep (2011) 7: 42. doi:10.1007/s11888-010-0080-z
  • 48 Views

Abstract

Until recently, all serrated polyps were classified simply as hyperplastic polyps and were thought to represent benign lesions with no progression to malignancy. However, the recognition during the last decade that some serrated polyps may progress to colorectal cancer through the “serrated neoplastic pathway” has changed the understanding of the significance of these lesions. Classification of serrated polyps is evolving and underscores the fact that morphological changes in serrated lesions are linked to a preneoplastic behavior. This fact has been observed not only in the hyperplastic polyposis syndrome, characterized by multiple or large serrated polyps, but also in the sporadic setting. These findings are particularly important for colorectal cancer prevention, and adequate surveillance programs tailored for patients with serrated polyps are urgently needed.

Keywords

Hyperplastic polypsSerrated polypsSerrated pathwayColorectal cancerSurveillanceHyperplastic polyposisBRAFCIMP

Introduction

Hyperplastic polyps (HP) in the colon have traditionally been thought to represent benign lesions with no progression to malignancy. However, it is now recognized that HPs are only a part of the serrated polyp spectrum, which includes several subtypes of lesions with a common histological feature, a saw-tooth appearance, with potential for transformation to colorectal carcinomas through the so-called “serrated pathway” [1••]. Thus, during the last decade, there has been a shift in the paradigm that only adenomas are precursors of colorectal cancer (CRC), and pathologists and clinicians are paying more attention to the diagnosis and characterization of serrated polyps.

Three main reasons have linked serrated polyps to colorectal tumorigenesis. First, hyperplastic polyposis syndrome (HPS), a rare condition characterized by numerous or large hyperplastic polyps, has been definitively associated with CRC development in several studies [2]. Indeed, a recent study including a large HPS cohort has shown that up to 35% of these patients develop a CRC and, more importantly, that the number of HP and sessile serrated adenomas (SSA) correlates with the presence of CRC [3•]. Secondly, a clinical and molecular link between serrated polyps and sporadic CRC has been established, especially with those with microsatellite instability (MSI) [4, 5]. Serrated polyps, in particular SSAs, are usually located in the proximal colon and share similar molecular features with sporadic tumors with MSI, which are caused by somatic methylation of the MLH1 gene. These features mainly include an increased frequency of non-random somatic methylation of several tumor suppressor genes (the so-called CpG island methylator phenotype, or CIMP) [68] and the presence of mutations in the BRAF gene [9, 10]. These two somatic features, which can be traced from cancer to their serrated precursors, constitute the molecular basis of the “serrated pathway” [11•]. Finally, patients with serrated polyps have been recognized to be at risk for colorectal neoplasia [1214]. The presence of large and proximal serrated polyps has been recently shown to constitute risk factors for CRC development, and this might have implication for surveillance [15•, 16•].

This review article focuses on the current pathologic description and classification of serrated polyps and the clinical implications of the diagnosis of serrated polyps.

Histological Classification of Serrated Polyps

The concept that serrated polyps are constituted by several lesions with different potential for malignization has encouraged pathologists and clinicians to recognize and report these lesions for patient management and research purposes [1••, 17, 18]. The presence of “saw-tooth” infolding of the crypt epithelium in longitudinal sections is the primary histological feature of all serrated polyps (usually seen as a stellate appearance on cross sections). The nomenclature of serrated polyps has evolved since the first description of a serrated adenoma made by Longacre and Fenoglio-Preiser in 1990 [19] to a more recent nomenclature recommended for reporting serrated polyps in 2005 [17], mainly based on the work by Torlakovic and Snover [2022]. The current classification of serrated polyps includes HPs, SSAs, traditional serrated adenomas (TSA), and the mixed serrated polyps (MP) (Table 1). Although these categories of serrated polyps have reached some consensus [1••, 17, 18], there is still a scientific debate about the morphological criteria for the diagnosis and report of serrated polyps. Therefore, it is not surprising that the interpretation of the literature is complicated by differing interpretations of the morphological changes of these lesions. Moreover, even among gastrointestinal pathologists, there is a significant interobserver variability in classification [2325].
Table 1

Classification of serrated colorectal polyps

Histological classification

Abbreviation

Key histological features

Predominant distribution

BRAF/KRAS mutations

Hyperplastic polyps

 

Normal architecture and proliferation activity confined to the lower third of the crypt

  
  

Serration in the upper-mid crypt

  

 - Microvesicular type

MVHP

Microvesicular mucin in the upper crypt

Colon and rectum

BRAF (30%–80%), KRAS (10%)

 - Goblet cell–rich type

GCHP

Predominance of mature goblet cells in upper crypt

Sigmoid and rectum

KRAS (50%), BRAF (20%)

Sessile serrated adenoma

SSA

Abnormal architecture that extends to the base of the crypts: crypt branching, dilation and serration, horizontal crypt growth (crypts parallel to the muscularis mucosa), goblet cells in the base of the crypt

Proximal colon

BRAF (80%–90%), KRAS (3%–8%)

Abnormal proliferation: focal nuclear stratification can be seen, but not cytologic dysplasia

Traditional serrated adenomas

TSA or SA

Uniform population of epithelial dysplastic cells associated with serration

Distal colon

BRAF (20%–62%), KRAS (20%–25%)

Ectopic crypt formation (crypts not seated in the muscularis mucosa)

Mixed serrated polyp

MP

Features of HP or SSA associated with a dysplastic component resembling a conventional adenoma

Proximal colon

BRAF (40%–100%), KRAS (43%–50%)

Individual components should be specified

Hyperplastic Polyps

HPs account for the majority of serrated polyps (80%–90%). Grossly, they are usually slightly elevated and <5 mm. The main characteristic of HPs is the presence of a normal architecture and proliferation with absence of cytological dysplasia [1••, 17]. Microscopic appearance reveals cellular crowding with serration in the upper to mid crypt with a normal proliferation compartment in the lower crypt. HPs have been divided into subtypes, although the relevance of reporting these subtypes remains questionable. The goblet cell type (GCHP), characterized by a predominance of mature goblet cells, is often located in the distal colon and rectum. The microvesicular type (MVHP) is the most widely distributed HP variant in the colon and rectum, in which the upper crypts show abundant enlarged microvacuolated columnar cells mixed with fewer goblet cells. Unlike GCHP, MVHP have been associated with somatic BRAF mutations (30%–80%) and CIMP, and they are thought to constitute the earliest lesion in the serrated pathway [2628].

Sessile Serrated Adenomas

SSAs are thought to account for approximately 15% to 20% of serrated polyps [27, 29]. Grossly, SSAs are usually flat or slightly elevated lesions typically >5 mm in diameter. Although SSA can be found throughout the colon, there is a predilection for the proximal segments [1••, 29, 30]. The main feature of SSA is the presence of abnormal crypt architecture, which reflects disordered growth. The symmetry present in HP is lost and instead, crypts are dilated, branched, and the alterations extend to the bottom of the crypt. It is very characteristic the presence of crypts that grow parallel to the muscularis mucosa making inverted T or L-shaped crypts, although this feature is not always present [31]. Therefore, it is essential for the diagnosis of SSA that basal mucosa can be evaluated. Unlike the HPs, serration is often seen at the base of the crypts [31]. In addition, SSAs show abnormal proliferation, with proliferation zones migrating upward in the crypt. However, SSAs do not usually demonstrate cytological dysplasia. From a molecular point of view, SSAs are consistently associated with BRAF mutations (>80%–90%) and the CIMP phenotype, and are considered to be the precursor of dysplastic serrated polyps and serrated adenocarcinoma [5, 26, 27, 32].

Traditional Serrated Adenomas

TSAs, also called serrated adenomas, are rare lesions, constituting 1% to 6% of serrated lesions [1••, 18, 26, 27]. TSAs are usually pedunculated and more commonly found in the left side of the colon (60%). It is well accepted that the main difference between SSA and TSA remains on the uniform population of cytological dysplastic cells seen in TSA. So, the dysplastic epithelial cells constitute the major population in TSA, along with prominent serration (Fig. 1) [33]. Ectopic crypt formation (aberrant crypts with lost orientation toward the lamina propria) occurs in TSA but not in SSA [22]. TSA harbor frequently BRAF mutations (30%–60%) and display a CIMP phenotype and interestingly, contiguous SSA is encountered in approximately 50% of BRAF-mutated serrated adenomas [26, 28, 32].
https://static-content.springer.com/image/art%3A10.1007%2Fs11888-010-0080-z/MediaObjects/11888_2010_80_Fig1_HTML.jpg
Fig. 1

Traditional serrated adenoma. This lesion combines crypt serration with adenomatous dysplasia with nuclei stratification

Mixed Serrated Polyp

According to the World Health Organization, MPs are combinations of conventional (tubular, tubulovillous, and villous) adenomas, with different grades of dysplasia and serrated features. Although the concept can be misleading because it does not reveal the potential malignancy of these lesions, “mixed polyp” is an internationally used term to define these kinds of polyps and it is recommended to describe in detail the individual components [1••]. An example of MP would be an SSA with the presence of cytological dysplasia [17, 18].

Clinical Implications of Serrated Polyps

Hyperplastic Polyposis Syndrome

HPS, a condition characterized by the presence of multiple and/or large HPs, was originally perceived as lacking any cancer risk [34]. However, multiple studies have shown that HPS predisposes to CRC development [2, 4, 3538]. In 2000, the World Health Organization proposed a working definition of HPS consisting of the following criteria: 1) at least five histologically diagnosed hyperplastic polyps proximal to sigmoid colon, of which two are >10 mm in diameter; or 2) any number of hyperplastic polyps that occur proximal to the sigmoid colon in an individual who has a first-degree relative with HPS; or 3) >30 hyperplastic polyps of any size that are distributed throughout the colon. Although arbitrary, this definition has been useful in standardizing clinical diagnosis and performing research studies. However, HPS has been shown to be a clinical and genetic heterogeneous disease [37, 38].

Colorectal Cancer Risk in HPS

It is now accepted that HPS confers a substantial risk for CRC, although the magnitude of such a risk remains to be established. Small series initially reported the presence of CRC in up to 50% of patients with HPS. However, in a recent multicenter series of 77 HPS patients followed-up endoscopically, the overall prevalence of CRC was 35%, and more importantly, even in patients under surveillance colonoscopies, the cumulative risk was 7% at 5 years [3•]. The median interval of CRC development was 11 months and most of the tumors were detected in diminutive serrated polyps (4–16 mm). Interestingly, in this study the number of HPS and serrated adenomas were significantly associated with CRC, although the increased risk was marginal.

Clinical Management

Management of HPS remains empirical and involves polyp removal, surveillance, and counseling of the patient and relatives. If the presentation is a CRC, it is recommended to perform a colectomy with ileorectal anastomosis to remove the risk of metachronous lesions [33]. However, flexible sigmoidoscopy every 6 to 12 months to survey the rectal remnant is needed. If the presentation is the presence of multiple or large polyps, colonoscopy every 1 to 2 years is recommended, with the aim of removing all polyps >5 mm in diameter. If this is not possible because of the size or polyp burden, colectomy and ileorectal anastomosis should be considered. If a patient declines surgery or is a poor surgery candidate, colonoscopic surveillance with biopsy/removal of the largest and more adenomatous lesions may be performed, although patients need to understand the risk of CRC development. Patients with HPS should be referred to tertiary centers for surveillance and treatment for several reasons: 1) HPs and SSA often seen in this syndrome are usually flat and difficult to see, which increases the duration of the procedure and probably increases the polyp miss rate [27]; and 2) the use of chromoendoscopy and narrow band imaging increases the detection of HPs, and these techniques, although not proved, may be of value in the polyp detection in HPS [39, 40].

Regarding the genetic counseling in HPS, the model of inheritance remains unknown, and since both autosomal recessive and autosomal dominant inheritance have been described, it is very likely that HPS constitutes a highly heterogeneous genetic condition. In this sense, a small proportion of patients with HPS are caused by biallelic MUTYH germline mutations [41]. Overall, up to 50% of HPS patients have a first-degree relative (FDR) with CRC [2], and HPS has been described in multiple family members in rare occasions [37, 42]. In a recent report [43••], the relative risk (RR) of CRC in FDRs compared to the general population was 5.4 (95% CI, 3.7–7.8). Based on the estimated prevalence of HPS in the general population (1:3000), the RR of HPS in FDRs was 39 (95% CI, 13–121). Thus, as long as no genetic cause is identified, screening colonoscopies for all FDRs is recommended, starting from the age of 35 to 40 years, or 5 to 10 years younger than the index case [11•, 33, 43••].

Sporadic Serrated Polyps

Risk of Colorectal Cancer in Serrated Polyps

Although there are considerable data linking serrated polyps, especially SSA, to CRC with MSI, the natural history of these lesions remains unknown, especially regarding the rate of growth, and the incidence of progression to CRC. It has been suggested that rate of progression could be increased in some serrated adenomas [44]. In this sense, although it constitutes indirect evidence, interval cancers occurring during colonoscopic surveillance are approximately four times more likely to be MSI than no interval cancers [45]. However, Goldstein et al. [31] looked at SSA that preceded the development of MSI CRC in 91 patients, and the mean interval between polypectomy and subsequent adenocarcinoma was 7.3 years (range, 1.2–19.3 years; standard deviation, 4.4 years). The interval from diagnosis of SSA to diagnosis of carcinoma was greater than 3 years in 90% of cases.

The incidence of cancerous transformation of serrated polyps remains unknown and the hypothesis that the presence of serrated polyps constitutes a risk factor for CRC has been based on small studies until very recently (Table 2) [1214, 46]. The first report showing a correlation between serrated polyps and development of adenomatous polyps in a large cohort was published in 2009 by Li et al. [47]. In this study, 4714 asymptomatic subjects undergoing screening colonoscopy were included, and independent predictors of synchronous advanced neoplasia (including tubular adenoma >1 cm, villous histology, high-grade dysplasia, and invasive adenocarcinoma) were determined. In total, 109 (2.3%) of the study population had large serrated polyps (LSP, ≥1 cm), and the presence of these polyps was an independent predictor of advanced neoplasia (OR, 3.24; 95% CI, 2.05–5.13; P < 0.0001). Of note, LSP conferred and increased risk for neoplasia independently of the polyp location. These results have been recently confirmed by a larger study including 10,199 subjects who underwent first-time colonoscopy [15•]. In this study, 140 (1.4%) patients had LSP, and multivariate analysis showed that LSP is a risk factor not only for synchronous advanced neoplasia but also for CRC (OR, 3.34; 95% CI, 2.16–5.03), especially in the proximal colon.
Table 2

Studies evaluating the risk of colorectal cancer associated with serrated polyps

Study, year

Aim

Population

Design

Serrated lesion

Summary of the finding

Schreiner et al. [16•], 2010

Analyze whether ND-SP at screening colonoscopies is associated with advanced neoplasiaa

Screening colonoscopy; Veteran’s Affairs Cooperative Study Group

Longitudinal n = 3121 patients (1371 patients had at least 1 follow-up colonoscopy)

HP, SSA

• Patients with proximal or large ND-SP are more likely to have synchronous advanced neoplasiaa compared with patients without these lesions (OR, 1.9; 95% CI, 1.33–2.70 and OR, 3.37; 95% CI, 1.71–6.65, respectively).

• The presence of a proximal ND-SP and no neoplasia on index colonoscopy is associated with an increased risk of any adenoma during surveillance (OR, 3.14; 95% CI, 1.59–6.2)

• Among patients with advanced neoplasiaa at baseline, those with proximal SP are more likely to have advanced neoplasia during surveillance (OR, 2.17; 95% CI, 1.03–4.59)

Hiraoka et al. [15•], 2010

Determine the association between serrated polyps and colorectal neoplasia

First-time colonoscopy

Cross-sectional n = 11,570 patients

HP, TSA, SSA, MP

• Large serrated polyps are independently associated with synchronous advanced neoplasiaa (OR, 4.01; 95% CI, 2.83–5.69) and CRC (OR, 3.34; 95% CI, 2.16–5.03)

• The presence of LSPs is main risk factor for CRC, particularly for proximal CRC (OR, 4.79; 95% CI, 2.54–8.42)

Lu et al. [46], 2010

Analyze the outcome of SSA

Retrospective diagnosis of SSA (single center)

Longitudinal Case-control (55 patients with SSA vs 55 patients with HP and 55 patients with adenomas)

SSA

• The incidence of subsequent CRCs was higher in SSA than in control patients (12.5% vs. 1.8%; P = 0.046)

• All the subsequent tumors were in the proximal colon, and 4/5 showed MSI

Li et al. [47], 2009

Identify risk factors for advanced colorectal neoplasia

Screening colonoscopy

Cross-sectional Case-control (467 patients with advanced neoplasiaa vs 4247 patients without advanced neoplasia)

HP, SSA, TSA

• Large serrated polyps are independently associated with synchronous advanced colorectal neoplasia (OR, 3.24; 95% CI, 2.05–5.13)

• Right-sided and left-sided large serrated polyps had similar association with advanced neoplasiaa

Laiyemo et al. [13], 2009

Analyze whether the coexistence of HP with adenomas increases the risk of adenoma recurrence

Polyp Prevention Trial

Multicenter RCT n = 1637 participants (No HP, 1200; distal HP, 305; proximal HP, 67; proximal and distal HP, 65)

HP

• There is no association between baseline coexistence of HPs with adenomas and recurrence of adenoma (OR, 1.19; 95% CI, 0.94–1.51) or advanced adenoma (OR, 1.25; 95% CI, 0.78–2.03) when compared with patients with only adenomas at baseline

• No association between HPs location and adenoma recurrence

Glazer et al. [12], 2008

Analyze the relationship between SA and metachronous adenomas

Retrospective diagnosis of SA (single center)

Longitudinal Case-control (17 patients with SA vs 17 controls)

SA

• SA at a baseline colonoscopy is associated with adenomatous polyps in subsequent examinations (24% vs 0%; P = 0.01)

Lazarus et al. [14], 2005

Evaluate the risk of metachronous lesions in serrated polyps

Retrospective diagnosis of SA (single center)

Longitudinal Case-control (38 patients with SA vs 119 patients with adenomas)

SSA, SA, MP

• Patients with SA have a similar risk of subsequent CRC compared with patients with conventional adenomas (5% vs 2.2%; P = 0.295)

aAdvanced neoplasia: tubular adenoma >10 mm, villous component, high-grade dysplasia, or invasive adenocarcinoma

CI confidence interval; HP hyperplastic polyp; MP mixed polyps; ND-SP nondysplastic serrated polyp; OR odds ratio; RCT randomized controlled trial; SA serrated adenoma; SSA sessile serrated adenoma

Whether the finding of a large or proximal serrated polyp identifies patients with an increased risk for future advanced neoplasia remains poorly studied. A small case-control study including 17 patients with SA and 17 controls found that the finding of an SA at baseline colonoscopy was associated with adenomatous polyps in subsequent examinations (24% vs 0%; P = 0.01) [12]. Using a retrospective cohort of previously undiagnosed SSA, Lu et al. [46] showed that the incidence of CRC in patients with SSA was significantly higher than in patients with HPs or adenomatous polyps. Recently, Schreiner et al. [16•] investigated whether the detection of proximal or large nondysplastic serrated polyps (ND-SP, including HPs and SSAs) at screening colonoscopy is associated with synchronous and metachronous neoplasia. The study included 3121 asymptomatic patients who underwent screening colonoscopy, and 1371 cases had subsequent surveillance. At baseline colonoscopy, 248 (7.9%) and 44 (1.4%) patients had at least one proximal or large ND-SP, respectively. In agreement with previous studies [15•, 47], the presence of proximal or large ND-SP was associated with synchronous advanced neoplasia compared with patients with no ND-SP. During surveillance, only the presence of proximal ND-SP in absence of adenomas on index colonoscopy was associated with an increased risk of metachronous adenomas and, notably, among patients with advanced adenomas, those with proximal ND-SP were more likely to have metachronous advanced adenomas. Unfortunately, the risk associated with large ND-SP could not be calculated because of the sample size.

These studies underline the fact that distinction of different serrated polyps has important clinical relevance and may be used to tailor the preventive strategy. Unfortunately, none of these studies analyzed the specific risk in different subtypes of serrated polyps, especially for SSA, and whether the associated neoplasia shared the typical molecular features of the serrated pathway.

Clinical Management

The evidence discussed above suggests that at least some serrated polyps require complete eradication and enrollment of patients into a colonoscopic surveillance program [1••, 17, 31, 44, 48].

HPs in the rectum are not associated with an increased risk of proximal adenomas or CRC, and their complete resection is not formally recommended [49, 50]. However, the presence of a serrated polyp (HP, SSA, TSA, and MP) outside this location should be removed completely. After piecemeal resection of an SSA, patients should undergo repeat colonoscopy at 2 to 6 months to confirm the complete polipectomy. If the lesion cannot be entirely removed, the attitude depends on the presence of dysplasia. For SSA, repeated colonoscopy with biopsies (ie, every year) aimed at detecting the appearance of dysplasia seems a reasonable option. If there is evidence of dysplasia (TSA, MP) and the lesion is endoscopically unresectable, surgical excision should be considered, especially in lesions of the right colon.

The interval to repeat colonoscopy in patients who have had a complete removal of a serrated polyp is difficult to establish based on the current data. However, it has been suggested that for SSA, a strategy based on polyp size and multiplicity should be followed, likewise the current adenoma guidelines [17, 33]. The study by Schreiner et al. [16•] supports the need to consider surveillance in patients with proximal nondysplastic serrated adenomas in absence of neoplasia, recommending the same regimen for patients with one to two small tubular adenomas (ie, colonoscopy every 5–10 years). However, further studies are needed to evaluate the appropriate surveillance protocol for the different subtypes of serrated polyps. The use of chromoendoscopy or narrow-band imaging should be considered in the surveillance of patients with serrated polyps, since these techniques have demonstrated to improve the visualization of HP, and may help to improve the gross determination of a negative margin based on the mucosal pattern surrounding the excision site.

Conclusions

Serrated polyps are morphologic and genetically heterogeneous, and our current knowledge indicates that some of them have malignant potential even in the absence of cytological dysplasia (ie, SSA). Classification of serrated polyps is evolving and underscores the fact that morphological changes in serrated lesions are linked to a preneoplastic behavior, and consequently this may have an impact on patient management. The neoplastic potential of serrated polyps through a “serrated pathway” is supported by several lines of evidence: HPS, a syndrome characterized by multiple or large serrated polyps, is associated with a high risk of CRC; there is a strong clinical and molecular association between serrated polyps and CRCs with MSI/CIMP/BRAF mutation; and finally, recent evidence from large studies suggests that the presence of large or proximal serrated polyps is associated with synchronous and metachronous neoplasia. Further studies are needed to confirm these findings and improve our understanding of the natural history of these lesions, with the final goal of developing evidence-based guidelines for clinical management.

Acknowledgments

This work was supported by grants from the Ministerio de Ciencia e Innovación (SAF2007-64873 and SAF2010-19273), Asociación Española contra el Cáncer (Fundación Científica y Junta de Barcelona), and Agència de Gestió d’Ajuts Universitaris i de Recerca (2009 SGR 849).

Disclosure

No potential conflicts of interest relevant to this article were reported.

Copyright information

© Springer Science+Business Media, LLC 2011