- Email: [email protected]
Presence of small sessile serrated polyps increases rate of advanced neoplasia upon surveillance compared with isolated low-risk tubular adenomas
ORIGINAL ARTICLE
Presence of small sessile serrated polyps increases rate of advanced neoplasia upon surveillance compared with isolated low-risk tubular adenomas Joshua Melson, MD,1 Karen Ma, MD,1 Saba Arshad, MBBS,1 Michael Greenspan, MD,1 Thomas Kaminsky, MD,1 Vinesh Melvani, MD,1 Faraz Bishehsari, MD,1 Brett Mahon, MD,2 Shriram Jakate, MD2 Chicago, Illinois, USA
Background and Aims: The U.S. Multi-Society Task Force (USMSTF) stratifies patients with sessile serrated polyps (SSPs) without cytologic dysplasia of <10 mm in size as at low risk for metachronous advanced neoplasia and recommends management similar to low-risk conventional tubular adenomas. Evidence supporting the recommended surveillance interval for these low-risk SSPs is limited. We aimed to assess rates of metachronous advanced neoplasia based on the presence of an initial low-risk SSP compared with isolated low-risk tubular adenomas. Methods: Colonoscopy data were retrieved for 2260 patients found to have an adenoma or SSP on pathology records between 2005 and 2011 at an academic medical center. The 788 patients who met study design criteria were stratified into 4 groups based on the presence of a high- or low-risk adenoma (HRA or LRA) and of a synchronous SSP on initial colonoscopy. The rates of advanced neoplasia at surveillance colonoscopy were then compared between groups. Results: The rate of advanced neoplasia at surveillance in the LRA inclusive of SSP group (12/66, 18.2%) was greater than in the LRA without any SSP group (29/370, 7.8%; P Z .019). The rate of advanced neoplasia at surveillance in patients with isolated low-risk SSP (10/56, 17.9%) remained significantly greater than those with isolated low-risk tubular adenomas (29/370, 7.8%; P Z .024). The rate of advanced neoplasia upon surveillance in the LRA inclusive of SSP group (18.2%) was comparable with the rate observed in the index HRA without any SSP group (15.9%) (40/252, P Z .709). Conclusions: The rate of advanced neoplasia upon surveillance in patients with initial low-risk SSPs is higher than in patients with initial isolated low-risk tubular adenomas and more similar to patients with initial highrisk tubular adenomas. These findings suggest that the rate of metachronous advanced neoplasia in patients with what are considered by USMSTF as “low-risk” SSPs is higher than in those without SSPs. Therefore, a surveillance interval that accounts for the presence of SSPs even in small lesions without cytologic dysplasia should be considered. (Gastrointest Endosc 2016;-:1-8.)
Serrated polyps recently have been recognized as an important contributor to interval colorectal cancers.1 The malignant pathogenesis of serrated polyps arises
from a molecular pathway alternate to conventional tubular adenomas known as the serrated neoplasia pathway.2-4 The sessile serrated pathway is associated
Abbreviations: ADR, adenoma detection rate; HRA, high-risk adenoma; LRA, low-risk adenoma; SDR, serrated detection rate; SSP, sessile serrated polyp; USMSTF, U.S. Multi-Society Task Force.
Current affiliations: Department of Medicine, Division of Digestive Diseases (1), Department of Pathology (2), Rush University Medical Center, Chicago, Illinois, USA.
DISCLOSURE: J. Melson has received research grant support from Boston Scientific. All other authors disclosed no financial relationships relevant to this publication.
Reprint requests: Joshua Melson, MD, Department of Medicine, Division of Digestive Diseases, Rush University Medical Center, 1725 W. Harrison Street Suite 206, Chicago, IL 60612.
Copyright ª 2016 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2016.01.064
If you would like to chat with an author of this article, you may contact Dr Melson at [email protected].
Received November 2, 2015. Accepted January 29, 2016.
www.giejournal.org
Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 1
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
with CpG island methylation phenotype (CIMP)-positive tumors and accounts for up to 20% of all sporadic colorectal cancers yet, notably, greater than 30% of interval cancers.1,5 Tumors in this pathway have a high frequency of BRAF mutations, associated with epigenetic silencing of mismatch repair genes (hMLH1).3 The consequent DNA instability in the serrated pathway is proposed to cause neoplastic changes at a rapid rate similar to what is seen in Lynch syndrome.6 The World Health Organization formally defined 3 major subtypes of serrated lesions including hyperplastic polyps (HP), sessile serrated polyps (SSP, with or without cytologic dysplasia), and traditional serrated adenoma (TSA, with or without conventional dysplasia).7 The true prevalence rate of SSPs is not well documented because the reported ranges widely vary from .6% to 13.8%, largely because of variations in serrated detection rates (SDRs).8-12 A recently published study found an overall SSP prevalence of 8.1% by a single experienced colonoscopist.9 Subtle endoscopic features including indistinct borders, asymmetric shape, and obscuration of vascular patterns by adherent mucus often challenge the detection of serrated lesions.13 In addition, endoscopic resection with conventional polypectomy techniques can be difficult and raises the concern for incomplete resections.14 Postpolypectomy guidelines recently incorporated specific recommendations on surveillance intervals for serrated polyps. Prior versions were silent in regard to the impact of serrated polyps on surveillance intervals.15,16 The 2012 U.S. Multi-Society Task Force (USMSTF) on Colorectal Cancer guideline recommends patients with low-risk SSPs (1-2 polyps < 10 mm and without cytologic dysplasia) follow a 5-year surveillance interval similar to low-risk tubular adenomas, although the evidence for this recommendation is low.17,18 Patients with high-risk SSPs are recommended to undergo a 3-year surveillance interval similar to high-risk tubular adenomas. Several studies have confirmed an increased risk for synchronous and metachronous advanced neoplasia in proximal large serrated polyps.19-21 Unlike the well-studied tubular adenomas, the quality of postpolypectomy surveillance evidence for serrated polyps remains limited. To our knowledge, there is no prior surveillance study that specifically assesses surveillance rates in low-risk adenomatous lesions based on the presence or absence of a serrated polyp. The primary aim of this study was to compare rates of advanced neoplasia upon surveillance colonoscopy in patients with initial low-risk polyps in the presence versus the absence of SSPs. The secondary aim was to compare rates of advanced neoplasia upon surveillance colonoscopy in patients with initial low-risk SSPs versus those with initial high-risk conventional adenomas (tubular or tubulovillous). 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
METHODS Study design A natural language search of pathology records for “adenoma” and “sessile serrated” identified 2260 consecutive patients who had a colorectal adenoma or SSP detected during colonoscopy between January 2005 and December 2011 at Rush University Medical Center. The Institutional Review Board of Rush University Medical Center granted study approval. Initial colonoscopy date, name of endoscopist, indication for procedure, bowel preparation quality, endoscopic polyp size, number and colonic location of polyps, and histologic features were recorded. Patients were then stratified into 4 groups as detailed below. Endoscopy and pathology reports for the follow-up surveillance colonoscopy, defined as the next consecutive study, were then reviewed for findings of nonadvanced and advanced neoplasia. Exclusion criteria included surveillance performed at less than 1 year, current or prior colorectal cancer, polyposis syndromes such as Lynch syndrome, inflammatory bowel disease, bowel preparation of poor quality, incomplete colonoscopy without cecal intubation, and lack of surveillance colonoscopy. Any patients with 10 adenomatous polyps found at the time of index colonoscopy were also excluded. The presence of hyperplastic polyps did not alter patient stratification. Traditional serrated adenomas and those with associated dysplasia were considered SSPs with features of a high-risk lesion.4 However, because they warrant more intensive surveillance,17 traditional serrated adenomas (n Z 6) were excluded during the analysis of metachronous advanced neoplasia. Only surveillance cases completed at Rush University Medical Center with reviewed pathology were included. The adenoma detection rate (ADR) of the 14 faculty gastroenterologists ranged from 22.2% to 44.7%.22 The mean of the top and lowest quintiles was 38.3% and 24.5%, respectively. The gastroenterologists achieved a mean sessile SDR of 2.4%. The mean SDR of the top quintile was 3.9%, whereas that of the lowest quintile was .5%.
Procedures Colonoscopies were performed using high-definition Olympus CF-H180AL and PCF-H180AL series colonoscopes (Olympus America, Center Valley, Pa). Narrow-band imaging usage was individualized. Experienced pathologists specialized in GI diseases (B.M. and S.J.) reviewed all polypectomy specimens.
Patient stratification by polyp histology Patients were stratified into 4 groups based on the presence of a high-risk adenoma (HRA) or low-risk www.giejournal.org
Melson et al
adenoma (LRA) and of a synchronous SSP on initial colonoscopy. Based on the 2012 USMSTF guidelines,17 LRA refers to patients with 1 to 2 adenomas (including SSPs if present) with each polyp <10 mm in size. HRA refers to patients with 3 adenomas (including SSPs if present) or a single adenoma 10 mm in size or containing villous histology or high-grade dysplasia. The 4 groups, stratified according to index colorectal neoplasia, consisted of LRA inclusive of SSP, LRA without any SSP, HRA inclusive of SSP, and HRA without any SSP. The following examples illustrate the grouping of LRA and HRA based on the presence or absence of an SSP. If the baseline colonoscopy showed 1 small tubular adenoma and a synchronous small SSP or, alternatively, 1 or 2 isolated SSPs < 10 mm absent of any conventional adenomas, then these cases were grouped in LRA inclusive of SSP. In contrast, patients with 1 or 2 small tubular adenomas lacking advanced features (villous or tubulovillous morphology and high-grade dysplasia) were grouped as LRA without any SSP. Patients with a high-risk SSP or a traditional serrated adenoma or, alternatively, a low-risk SSP with a synchronous highrisk tubular adenoma were grouped in HRA inclusive of SSP. An isolated high-risk tubular adenoma without coexistent SSPs were considered HRA without any SSP. The rates of metachronous advanced neoplasia at first surveillance colonoscopy were analyzed and contrasted between groups. Advanced neoplasia was defined as an adenoma or serrated polyp 10 mm or villous histology, or high-grade dysplasia, or colorectal cancer anywhere in the colorectum.23
Pathology review Pathology slides were interpreted by 2 board-certified GI pathologists. A concordance analysis to define interobserver agreement was performed with a kappa score calculation. Seventy-five cases of respective consecutive tubular adenomas (n Z 25), hyperplastic polyps (n Z 25), and SSPs (n Z 25) from the study period were reviewed by both pathologists blinded to the other’s interpretation. Among the 25 cases that were initially interpreted as SSPs, there was concordance in 23 cases, with 1 SSP reinterpreted as a tubulovillous adenoma and 1 as a hyperplastic polyp. Thus, the kappa score for SSPs was .92 (standard error, .055). Otherwise, there was concordance among all other reviewed cases with a kappa of .97 (standard error, .019).
Statistical analyses Descriptive statistics were used to describe the study population. The 2-sided Fisher exact test set at P .05 was considered statistically significant. All statistical analyses were performed with SAS software (SAS Institute, Cary, NC). www.giejournal.org
Presence of small SSPs increases rate of advanced neoplasia
Total colonoscopies reviewed → 2,260 Lack of surveillance colonoscopy → 1,286 Current or prior colorectal cancer → 78 Surveillance performed at < 1 year → 77 Poor bowel preparation or lack of cecal intubation → 45 Inflammatory bowel disease → 30 Polyposis syndrome → 7 Total patients included → 788
Figure 1. Flow diagram of patient exclusions (some patients met more than 1 exclusion criterion), resulting in 788 patients included in the study.
RESULTS Cohort demographic characteristics Seven hundred eighty-eight patients met the study design criteria, as depicted in Figure 1. The most common exclusion criteria were lack of surveillance colonoscopy (n Z 1286), current or prior colorectal cancer (n Z 78), surveillance performed at less than a 1-year interval (n Z 77), and poor quality bowel preparation or lack of cecal intubation on any studied colonoscopy (n Z 45). Among the 87 patients with initial LRA inclusive of SSP lesions that were excluded, baseline characteristics were comparable with the 66 included patients with surveillance data on the basis of age (mean, 58.2 years; P Z 1.00), sex (64.4% female; P Z .866), and race (71.3% white; P Z .535). Based on initial colonoscopy findings, the 4 groups consisted of 66 patients (8.4%) in LRA inclusive of SSP, 370 patients (46.9%) in LRA without any SSP, 100 patients (12.7%) in HRA inclusive of SSP, and 252 patients (32.0%) in HRA without any SSP. Among the 166 patients with serrated polyps (groups 1 and 3 combined), 129 patients had at least 1 serrated polyp in the proximal colon (defined as proximal to the splenic flexure). There were 122 patients (73.5%) with serrated polyp(s) isolated to the proximal colon. On a polyp basis, 74.2% of all serrated polyps (173/233) were found in the proximal colon. Each group’s demographic and baseline colonoscopy composition is detailed in Tables 1 and 2. The mean age of each group was similar and ranged between 59 and Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 3
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
TABLE 1. Demographic and baseline colonoscopy characteristics of the USMSTF-defined low-risk polyps
TABLE 2. Demographic and baseline colonoscopy characteristics of the USMSTF-defined high-risk polyps
LRA inclusive of SSP (group 1)
LRA without any SSP (group 2)
HRA inclusive of SSP (group 3)
HRA without any SSP (group 4)
No. of patients
66 (8.4)
370 (46.9)
No. of patients
100 (12.7)
252 (32.0)
Mean age, y ( SD)
59 9.5
60 9.9
.447
Mean age, y ( SD)
61 8.8
61 10.5
.083
Sex
Sex
P value
25 (37.9)
184 (49.7)
Male
47 (47.0)
144 (57.1)
Female
41 (62.1)
186 (50.3)
Female
53 (53.0)
108 (42.9)
.001
Race
.472
White
50 (75.8)
185 (50.0)
White
63 (63.0)
148 (58.7)
African American
11 (16.7)
133 (35.9)
African American
33 (33.0)
87 (34.5)
Hispanic
4 (6.1)
37 (10.0)
Hispanic
4 (4.0)
14 (5.6)
Other
1 (1.5)
15 (4.1)
Other
0
3 (1.2)
Tobacco use of those assessed (current or prior)
.399
Tobacco use of those assessed (current or prior)
.220
Yes
31 (52.5)
167 (45.8)
Yes
47 (47.0)
110 (44.5)
No
28 (47.5)
198 (54.2)
No
43 (43.0)
137 (55.5)
Screening
44 (66.7)
263 (71.1)
Screening
69 (69.0)
175 (69.4)
Surveillance
22 (33.3)
107 (28.9)
Surveillance
31 (31.0)
77 (30.6)
92 (92.0)
223 (88.5)
8 (8.0)
29 (11.5)
Indication for colonoscopy
.467
Bowel preparation quality Good/excellent Fair
.279 62 (93.9)
327 (88.4)
4 (6.1)
43 (11.6)
Indication for colonoscopy
1.000
Bowel preparation quality Good/excellent Fair
Polyp burden
.441
Polyp burden
Isolated SSP
56 (84.8)
N/A
Isolated SSP
44 (44)
N/A
Isolated TA
N/A
370 (100)
Isolated TA
N/A
252 (100)
Mixed SSP þ TA Mean surveillance interval, mo ( SD)
1.000 .097
Male
Race
P value
10 (15.2)
N/A
47.3 16.7
53.9 22.1
Mixed SSP þ TA .021
Values are number of cases with percents in parentheses unless otherwise noted. LRA, Low-risk adenoma; SSP, sessile serrated polyp; SD, standard deviation; N/A, not applicable; TA, tubular adenoma.
61 years old. A significantly greater number of women were observed in the serrated groups collectively at 56.6% (94/166) than in the nonserrated groups at 47.3% (294/622; P Z .036). The racial distribution was whitepredominant, followed by African American and Hispanic in each group. The serrated groups collectively demonstrated a higher rate of current or prior tobacco use when compared with the nonserrated groups, although this did not reach statistical significance (52.3% [78/149] vs 45.3% [277/612]; P Z .143). Forty-four patients (66.7%) in the LRA inclusive of SSP group underwent their initial colonoscopy for screening purposes. A shorter mean surveillance interval was observed in the LRA inclusive of SSP group (47.3 16.7 months) compared with LRA without any SSP (53.9 22.1 months; P Z .021). Within the LRA inclusive of SSP group, 75.8% (50/66) of index and surveillance colonoscopies were performed by the same endoscopist. 4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
TSA Mean surveillance interval, mo ( SD)
50 (50)
N/A
6 (6)
N/A
42.5 17.2
40.1 20.9
.309
Values are number of cases with percents in parentheses unless otherwise noted. HRA, High-risk adenoma; SSP, sessile serrated polyp; SD, standard deviation; TA, tubular adenoma; N/A, not applicable; TSA, traditional serrated adenoma.
Metachronous advanced neoplasia in initial low-risk lesions The rate of advanced neoplasia upon surveillance in the LRA inclusive of SSP group was significantly greater than in the LRA without any SSP group (18.2% vs 7.8%, P Z .019). The composition of advanced neoplasia features upon surveillance is highlighted in Table 3. Patients with isolated low-risk serrated polyps (without a coexistent tubular adenoma) demonstrated a 17.9% rate (10/56) of metachronous advanced neoplasia compared with 7.8% in those with isolated low-risk tubular adenomas (29/370; P Z .024). When assessing the subset of patients undergoing initial colonoscopy only for screening purposes, the rate of metachronous advanced neoplasia remained significantly greater in the LRA inclusive of SSP group (18.2%, 8/44) compared with the LRA without any SSP group (6.5%, 17/263; P Z .015). www.giejournal.org
Melson et al
Presence of small SSPs increases rate of advanced neoplasia
TABLE 3. Findings on follow-up surveillance colonoscopy among all groups LRA inclusive of SSP (group 1) (n [ 66)
LRA without any SSP (group 2) (n [ 370)
P value
HRA inclusive of SSP* (group 3) (n [ 94)
HRA without any SSP (group 4) (n [ 252)
P value
Advanced neoplasia
12 (18.2)
29 (7.8)
.019
30 (31.9)
40 (15.9)
.0007
Size 1 cm
10 (83.3)
20 (69.0)
30 (31.9)
31 (77.5)
Villous
2 (16.7)
8 (27.5)
5 (16.7)
15 (37.5)
HGD
0
2 (6.9)
1 (3.3)
4 (10.0)
CRC
0
2 (6.9)
1 (3.3)
0
Any adenoma
39 (59.1)
179 (48.4)
.141
64 (68.1)
158 (62.7)
.624
SSP
22 (33.3)
16 (4.3)
.001
31 (33.0)
15 (6.0)
.001
Values are number of cases with percents in parentheses. LRA, Low-risk adenoma; SSP, sessile serrated polyp; HGD, high-grade dysplasia; CRC, colorectal carcinoma. *Excluding traditional serrated adenoma cases (n Z 6).
TABLE 4. Findings on follow-up surveillance colonoscopy based on the presence or absence of an SSP Presence of SSP* Absence of SSP (groups 1 and 3 (groups 2 and 4 combined) combined) (n [ 160) (n [ 622) P value Advanced neoplasia
42 (26.3)
69 (11.1)
.0001
Any adenoma
103 (64.4)
337 (54.2)
.021
SSP
53 (33.1)
31 (5.0)
.0001
SSP, Sessile serrated polyp. *Excluding traditional serrated adenoma cases (n Z 6).
There was a significant difference in racial distribution between the low-risk groups (groups 1 and 2) with a greater proportion of whites (75.8% vs 50%) compared with African Americans (16.7% vs 35.9%) in the serrated group (P Z .001). The difference in rates of advanced neoplasia on surveillance remained significantly greater when comparing whites in the LRA inclusive of SSP (20%, 10/50) and LRA without SSP (5.9%, 11/185; P Z .004) groups.
Metachronous advanced neoplasia in initial high-risk lesions Next, we compared LRA inclusive of SSP with HRA without any SSP and found no difference in rates of advanced neoplasia upon surveillance between the groups (18.2% vs 15.9%, P Z .709). The rate of advanced neoplasia upon surveillance in the HRA inclusive of SSP group (excluding traditional serrated adenoma) was 31.9% (30/94), which was significantly greater than the 15.9% (40/252) seen in the HRA without any SSP group. In patients with high-risk isolated serrated polyps, there was a 32% rate (16/50) of metachronous advanced neoplasia compared with 15.9% in those with high-risk isolated tubular adenomas (P Z .015).
Metachronous advanced neoplasia in initial serrated polyp presence In patients with an initial serrated polyp (groups 1 and 3 combined), the rates of metachronous advanced www.giejournal.org
neoplasia, adenoma, and serrated recurrence were significantly greater compared with patients without an initial serrated polyp (groups 2 and 4 combined). This is highlighted in Table 4. The presence of an initial serrated polyp increased the overall rate of metachronous advanced neoplasia from 11.1% (69/622) to 26.3% (42/160; P Z .0001) and the rate of a metachronous SSP from 5% (31/622) to 33.1% (53/160; P Z .0001). The rate of advanced neoplasia on surveillance between patients with proximal versus distal SSPs on initial colonoscopy did not differ significantly. Among patients in the LRA inclusive of SSP group, 9 of 54 patients with isolated proximal SSPs (16.7%) and 3 of 12 patients with distal SSPs (25%) had advanced neoplasia on surveillance (P Z .679). Among patients in the HRA inclusive of SSP group, 24 of 68 patients with isolated proximal SSPs (35.3%) and 9 of 32 patients with distal SSPs (28.1%) had advanced neoplasia on surveillance (P Z .505). Isolated serrated polyps without synchronous adenomas were found in 56 patients (84.8%) in the LRA inclusive of SSP group and in 50 patients (50%) in the HRA inclusive of SSP group. The rates of SSPs on surveillance in both the initial LRA inclusive of SSP and the initial HRA inclusive of SSP groups were significantly greater than either respective nonserrated group. Specifically, the rate of SSPs on surveillance in the LRA inclusive of SSP group was 33.3% (22/66), which was significantly greater than that of the LRA without SSP group (4.3%, 16/370, P Z .001) and the HRA without any SSP group (6.0%, 15/252, P Z .0001). The rates of SSPs on surveillance were similar between both the LRA inclusive of SSP and HRA inclusive of SSP groups (33.3% vs 33.0% in HRA inclusive of SSP, P Z 1.00).
DISCUSSION This is the first study to show that the risk of advanced neoplasia during surveillance in patients with initial lowrisk serrated polyps is higher than in patients with lowrisk tubular adenomas without coexistent SSPs. The rate Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 5
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
of advanced neoplasia during surveillance of low-risk serrated polyps was more similar to patients with initial high-risk tubular adenomas. Specifically, the rate of advanced neoplasia on surveillance in the LRA inclusive of SSP group was 18.2%, higher than 7.8% in the LRA without any SSP group (P Z .019) and comparable with the 15.9% in the HRA without any SSP group (P Z .709). Additionally, subanalysis of patients with initial serrated polyps revealed a markedly greater rate of metachronous serrated polyps compared with patients without initial serrated polyps (33.1% vs 5.0%; P Z .0001). To our knowledge, this study is the largest surveillance cohort of low-risk serrated polyps and assessment on rates of metachronous advanced neoplasia. The association of large and proximal serrated polyps with an increased risk for synchronous and metachronous advanced neoplasia has been previously reported.19-21,24,25 However, few studies accurately examine the impact of small SSPs on rates of advanced neoplasia upon surveillance. Schreiner et al19 found an increased rate of neoplasia on surveillance in patients with a proximal serrated polyp; however, there was no significant difference in rate of advanced neoplasia. Their study cohort was gathered from a predominantly male veteran sample with index and surveillance colonoscopies performed in late 1990s and early 2000s, an era preceding the formal classification of serrated polyps.4,7 In addition, the subgroup of patients with serrated lesions, especially low-risk serrated lesions, was small in comparison with our study. These study design factors likely account for the differences between our findings. Another recent study26 found that the presence of a serrated polyp in patients with either an advanced or a nonadvanced adenoma did not increase the risk for metachronous advanced adenomas. However, the serrated cohort was not further divided into high-risk and low-risk lesions, a key stratification point for appropriate screening intervals. Although the authors did attempt to account for interendoscopist variability, SDRs were not addressed. A benchmark for SDR has yet to be established, and it has been shown that detection rates of serrated lesions vary widely between endoscopists.8-10,12,18,27 In a study by Hetzel et al,10 the detection of serrated polyps ranged from 0% to 2.2% among endoscopists at an academic medical center. The mean SDR (excluding hyperplastic polyps) of our study’s colonoscopists was 2.4%, similar to other published values. Kahi et al28 suggested an ADRequivalent proximal detection rate of serrated polyps to be around 4.5%. It is unknown how a higher SDR would have affected study outcomes. However, the rates of advanced neoplasia at surveillance colonoscopy in our conventional LRA patients is similar to rates published in a meta-analysis.29 The above studies demonstrated a strong correlation between ADRs and proximal SDRs. The ADRs of all 14 gastroenterologists studied22 are above established thresholds of satisfactory adenoma detection.30 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
There are several limitations of the study. We do not address colorectal cancer rates upon surveillance but rather the surrogate marker of advanced neoplasia. The study is not powered to define colorectal cancer risk in low-risk lesions based on the presence of SSPs. The sample size was limited by the low rate of surveillance follow-up (43.1%) available. The interval to surveillance colonoscopy was actually significantly shorter in the LRA inclusive of SSP group compared with the LRA without any SSP group. Thus, the comparative increase in advanced neoplasia in the LRA inclusive of SSP group would likely be even larger if similar surveillance intervals occurred. The findings of serrated polyp studies depend on accurate and precise pathologic interpretation of serrated polyps. Several studies have found significant discordance between pathologists in identification of serrated polyps, both among internal pathologists and between various centers.10,31,32 We found the interobserver concordance between our pathologists to be high, with a kappa score for SSPs of .92 (standard error, .055). However, the issue of interobserver variation of serrated polyps remains a limitation of the study, as these were not externally validated. The difficult distinction between SSPs and hyperplastic polyp lesions have prompted some experts to regard all serrated polyps (whether SSPs or hyperplastic polyps) 1 cm or larger as a high-risk SSP.25,33 We did not include index large (1 cm) hyperplastic polyps in the serrated groups, thus leaving the possibility of missed high-risk serrated lesions. However, there were only 2 metachronous hyperplastic polyps 1 cm in the serrated groups. These were not considered metachronous advanced neoplasia findings, because the current USMSTF guidelines17 do not include hyperplastic polyps as an endpoint of advanced neoplasia. Given the high internal concordance between our pathologists and the infrequency of large hyperplastic polyps in the proximal colon, the study’s primary evaluations of metachronous neoplasia would not have been impacted by inclusion of hyperplastic polyps. Incomplete resection of serrated polyps has been reported to be near 30% in the CARE study by Pohl et al14 and is a recognized factor in the development of interval cancers. A surrogate marker would be to focus on the cases with metachronous advanced neoplasia and compare similarities in colonic segment location and pathology between the initial and follow-up lesions. Overall, we found 34 of 114 patients (29.8%) with metachronous advanced neoplasia developed a recurrent lesion of identical histology in the same colonic segment where the initial polypectomy was performed. Specifically, there was a 19.5% recurrence rate (8/41) in serrated polyps and a 35.6% recurrence rate (26/73) in tubular adenomas. The high rate of same-segment recurrence in high-risk tubular adenomas (50%, 20/40) is interesting and may warrant further studies. There are certainly other contributory etiologies to same-segment recurrence, such as missed lesions and local colonic field defects leading to a www.giejournal.org
Melson et al
predisposition for the presence of similar location recurrence of advanced neoplasia.34 The sex,3,10,35 racial,36,37 and smoking19,26 distribution of our patients with serrated polyps is consistent with other studies. A greater proportion of female patients was found in the serrated groups. Also, serrated polyps were more common in whites than African Americans and Hispanics. The higher rate of advanced neoplasia on surveillance in the low-risk serrated group remained significant among whites alone. We did not consider family history of colorectal cancer, use of aspirin or nonsteroidal anti-inflammatory agents, or body mass index, all of which have been shown to affect colorectal cancer risk.38 Overall, our cohort was gathered from a diverse urban community, and the generalizability to other populations is unknown. The USMSTF on Colorectal Cancer published postpolypectomy surveillance guidelines for serrated polyps and reported that these recommendations were based on limited evidence.17,18 Our study found that the risk of interval advanced neoplasia in patients with prior initial low-risk serrated polyps is higher than in patients with initial low-risk tubular adenomas and actually more similar to patients with initial high-risk tubular adenomas. Although the USMSTF recommendations for LRAs with serrated polyps are currently similar to LRAs without serrated polyps, our results reflect a difference in rates of metachronous advanced neoplasia. This study should lead to reconsideration of the recommended surveillance intervals for low-risk serrated polyps.
REFERENCES 1. Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138:2088-100. 2. O’Brien MJ, Yang S, Mack C, et al. Comparison of microsatellite instability, CpG island methylation phenotype, BRAF and KRAS status in serrated polyps and traditional adenomas indicates separate pathways to distinct colorectal carcinoma end points. Am J Surg Pathol 2006;30: 1491-501. 3. Spring KJ, Zhao ZZ, Karamatic R, et al. High prevalence of sessile serrated adenomas with BRAF mutations: a prospective study of patients undergoing colonoscopy. Gastroenterology 2006;131: 1400-7. 4. Torlakovic E, Skovlund E, Snover DC, et al. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003;27:65-81. 5. Arain MA, Sawhney M, Sheikh S, et al. CIMP status of interval colon cancers: another piece to the puzzle. Am J Gastroenterol 2010;105: 1189-95. 6. Kahi C. How does the serrated polyp pathway alter CRC screening and surveillance? Dig Dis Sci 2015;60:773-80. 7. Snover D, Ahnen DJ, Burt RW, et al. Serrated polyps of the colon and rectum and serrated (“hyperplastic”) polyposis. In: Bozman FT, Carneiro F, Hruban RH, et al, eds. WHO Classification of tumours pathology and genetics tumours of the digestive system, 4th ed. Berlin: Springer-Verlag; 2010, 160-5. 8. East JE, Vieth M, Rex DK. Serrated lesions in colorectal cancer screening: detection, resection, pathology and surveillance. Gut 2015;64:991-1000.
www.giejournal.org
Presence of small SSPs increases rate of advanced neoplasia 9. Abdeljawad K, Vermulapalli KC, Kahi CJ, et al. Sessile serrated polyp prevalence determined by a colonoscopist with a high lesion detection rate and an experienced pathologist. Gastrointest Endosc 2015;81:517-24. 10. Hetzel JT, Huang CS, Coukos JA, et al. Variation in the detection of serrated polyps in an average risk colorectal cancer screening cohort. Am J Gastroenterol 2010;105:2656-64. 11. Lash RH, Genta RM, Schuler CM. Sessile serrated adenomas: prevalence of dysplasia and carcinoma in 2139 patients. J Clin Pathol 2010;63:681-6. 12. Kahi CJ, Hewett DG, Norton DL, et al. Prevalence and variable detection of proximal colon serrated polyps during screening colonoscopy. Clin Gastroenterol Hepatol 2011;9:42-6. 13. Hazewinkel Y, Lopez-Ceron M, East JE, et al. Endoscopic features of sessile serrated adenomas: validation by international experts using high-definition white-light endoscopy and narrow-band imaging. Gastrointest Endosc 2013;77:916-24. 14. Pohl H, Srivastava A, Bensen SP, et al. Incomplete polyp resection during colonoscopy- results of the Complete Adenoma Resection (CARE) study. Gastroenterology 2013;144:74-80. 15. Winawer SJ, Zauber AG, Fletcher RH, et al. Guidelines for colonoscopy surveillance after polypectomy: a consensus update by the US MultiSociety Task Force on colorectal cancer and the American Cancer Society. Gastroenterology 2006;130:1872-85. 16. Levin B, Lieberman DA, McFarland B, et al. Screening and surveillance for early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US MultiSociety Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570-95. 17. Lieberman DA, Rex DK, Winawer SJ, et al. United States Multi-Society Task Force on Colorectal Cancer. Guidelines for colonoscopy surveillance after screening and polypectomy: a consensus update by the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2012;143:844-57. 18. Rex DK, Ahnen DJ, Baron JA, et al. Serrated lesions of the colorectum: review and recommendations from an expert panel. Am J Gastroenterol 2012;107:1315-29. 19. Schreiner MA, Weiss DG, Lieberman DA. Proximal and large hyperplastic and nondysplastic serrated polyps detected by colonoscopy are associated with neoplasia. Gastroenterology 2010;139: 1497-502. 20. Hiraoka S, Kato J, Fujiki S, et al. The presence of large serrated polyps increases risk for colorectal cancer. Gastroenterology 2010;139: 1503-10. 21. Li D, Jim C, McCulloch C, et al. Association of large serrated polyps with synchronous advanced colorectal neoplasia. Am J Gastroenterol 2009;104:695-702. 22. Greenspan M, Rajan KB, Baig A, et al. Advanced adenoma detection rate is independent of nonadvanced adenoma detection rate. Am J Gastroenterol 2013;108:1286-92. 23. Imperiale TF, Monahan PO, Stump TE, et al. Derivation and validation of a scoring system to stratify risk for advanced colorectal neoplasia in asymptomatic adults. Ann Intern Med. 2015;163:339-46. 24. Hazewinkel Y, de Wijkerslooth TR, Stoop EM, et al. Prevalence of serrated polyps and association with synchronous advanced neoplasia in screening colonoscopy. Endoscopy 2014;46:219-24. 25. Holme O, Bretthauer M, Eide TJ, et al. Long-term risk of colorectal cancer in individuals with serrated polyps. Gut 2015;64:929-36. 26. Macaron C, Vu HT, Lopez R, et al. Risk of metachronous polyps in individuals with serrated polyps. Dis Colon Rectum 2015;58:762-8. 27. Rex DK, Schoenfeld PS, Cohen J, et al. Quality indicators for colonoscopy. Gastrointest Endosc 2015;81:31-53. 28. Kahi CJ, Li X, Eckert GJ, et al. High colonoscopic prevalence of proximal colon serrated polyps in average-risk men and women. Gastrointest Endosc 2012;75:515-20. 29. Saini SD, Kim HM, Schoenfeld P. Incidence of advanced adenomas at surveillance colonoscopy in patients with a personal history of colon
Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 7
Presence of small SSPs increases rate of advanced neoplasia
30.
31.
32.
33.
Melson et al
adenomas: a meta-analysis and systemic review. Gastrointest Endosc 2006;64:614-26. Kaminski M, Regula J, Kraszewska E, et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010;362: 1795-803. Khalid O, Radaideh S, Cummings O, et al. Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001. World J Gastroenterol 2009;15:3767-70. Payne SR, Church TR, Wandell M, et al. Endoscopic detection of proximal serrated lesions and pathologic identification of sessile serrated adenomas/polyps vary on the basis of center. Clin Gastroenterol Hepatol 2014;12:1119-26. Hassan C, Quintero E, Dumonceau JM, et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy Guidelines (ESGE) Guideline. Endoscopy 2013;45:842-51.
8 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
34. Adler J, Robertson DJ. Interval colorectal cancer after colonoscopy: exploring explanations and solutions. Am J Gastroenterol 2010;110: 1657-64. 35. Burnett-Hartman AN, Passarelli MN, Adams SV, et al. Differences in epidemiologic risk factors for colorectal adenomas and serrated polyps by lesion severity and anatomical site. Am J Epidemiol 2013;177: 625-37. 36. Lieberman DA, Prindiville S, Weiss DG, et al. Risk factors for advanced colonic neoplasia and hyperplastic polyps in asymptomatic individuals. JAMA 2003;290:2959-67. 37. Wallace K, Grau MV, Ahnen D, et al. The association of lifestyle and dietary factors with the risk for serrated polyps of the colorectum. Cancer Epidemiol Biomarkers Prev 2009;18:2310-7. 38. Din FV, Theodoratou E, Farrington SM, et al. Effect of aspirin and NSAIDs on risk and survival from colorectal cancer. Gut 2010;59:1670-9.
www.giejournal.org
Presence of small sessile serrated polyps increases rate of advanced neoplasia upon surveillance compared with isolated low-risk tubular adenomas Joshua Melson, MD,1 Karen Ma, MD,1 Saba Arshad, MBBS,1 Michael Greenspan, MD,1 Thomas Kaminsky, MD,1 Vinesh Melvani, MD,1 Faraz Bishehsari, MD,1 Brett Mahon, MD,2 Shriram Jakate, MD2 Chicago, Illinois, USA
Background and Aims: The U.S. Multi-Society Task Force (USMSTF) stratifies patients with sessile serrated polyps (SSPs) without cytologic dysplasia of <10 mm in size as at low risk for metachronous advanced neoplasia and recommends management similar to low-risk conventional tubular adenomas. Evidence supporting the recommended surveillance interval for these low-risk SSPs is limited. We aimed to assess rates of metachronous advanced neoplasia based on the presence of an initial low-risk SSP compared with isolated low-risk tubular adenomas. Methods: Colonoscopy data were retrieved for 2260 patients found to have an adenoma or SSP on pathology records between 2005 and 2011 at an academic medical center. The 788 patients who met study design criteria were stratified into 4 groups based on the presence of a high- or low-risk adenoma (HRA or LRA) and of a synchronous SSP on initial colonoscopy. The rates of advanced neoplasia at surveillance colonoscopy were then compared between groups. Results: The rate of advanced neoplasia at surveillance in the LRA inclusive of SSP group (12/66, 18.2%) was greater than in the LRA without any SSP group (29/370, 7.8%; P Z .019). The rate of advanced neoplasia at surveillance in patients with isolated low-risk SSP (10/56, 17.9%) remained significantly greater than those with isolated low-risk tubular adenomas (29/370, 7.8%; P Z .024). The rate of advanced neoplasia upon surveillance in the LRA inclusive of SSP group (18.2%) was comparable with the rate observed in the index HRA without any SSP group (15.9%) (40/252, P Z .709). Conclusions: The rate of advanced neoplasia upon surveillance in patients with initial low-risk SSPs is higher than in patients with initial isolated low-risk tubular adenomas and more similar to patients with initial highrisk tubular adenomas. These findings suggest that the rate of metachronous advanced neoplasia in patients with what are considered by USMSTF as “low-risk” SSPs is higher than in those without SSPs. Therefore, a surveillance interval that accounts for the presence of SSPs even in small lesions without cytologic dysplasia should be considered. (Gastrointest Endosc 2016;-:1-8.)
Serrated polyps recently have been recognized as an important contributor to interval colorectal cancers.1 The malignant pathogenesis of serrated polyps arises
from a molecular pathway alternate to conventional tubular adenomas known as the serrated neoplasia pathway.2-4 The sessile serrated pathway is associated
Abbreviations: ADR, adenoma detection rate; HRA, high-risk adenoma; LRA, low-risk adenoma; SDR, serrated detection rate; SSP, sessile serrated polyp; USMSTF, U.S. Multi-Society Task Force.
Current affiliations: Department of Medicine, Division of Digestive Diseases (1), Department of Pathology (2), Rush University Medical Center, Chicago, Illinois, USA.
DISCLOSURE: J. Melson has received research grant support from Boston Scientific. All other authors disclosed no financial relationships relevant to this publication.
Reprint requests: Joshua Melson, MD, Department of Medicine, Division of Digestive Diseases, Rush University Medical Center, 1725 W. Harrison Street Suite 206, Chicago, IL 60612.
Copyright ª 2016 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2016.01.064
If you would like to chat with an author of this article, you may contact Dr Melson at [email protected].
Received November 2, 2015. Accepted January 29, 2016.
www.giejournal.org
Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 1
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
with CpG island methylation phenotype (CIMP)-positive tumors and accounts for up to 20% of all sporadic colorectal cancers yet, notably, greater than 30% of interval cancers.1,5 Tumors in this pathway have a high frequency of BRAF mutations, associated with epigenetic silencing of mismatch repair genes (hMLH1).3 The consequent DNA instability in the serrated pathway is proposed to cause neoplastic changes at a rapid rate similar to what is seen in Lynch syndrome.6 The World Health Organization formally defined 3 major subtypes of serrated lesions including hyperplastic polyps (HP), sessile serrated polyps (SSP, with or without cytologic dysplasia), and traditional serrated adenoma (TSA, with or without conventional dysplasia).7 The true prevalence rate of SSPs is not well documented because the reported ranges widely vary from .6% to 13.8%, largely because of variations in serrated detection rates (SDRs).8-12 A recently published study found an overall SSP prevalence of 8.1% by a single experienced colonoscopist.9 Subtle endoscopic features including indistinct borders, asymmetric shape, and obscuration of vascular patterns by adherent mucus often challenge the detection of serrated lesions.13 In addition, endoscopic resection with conventional polypectomy techniques can be difficult and raises the concern for incomplete resections.14 Postpolypectomy guidelines recently incorporated specific recommendations on surveillance intervals for serrated polyps. Prior versions were silent in regard to the impact of serrated polyps on surveillance intervals.15,16 The 2012 U.S. Multi-Society Task Force (USMSTF) on Colorectal Cancer guideline recommends patients with low-risk SSPs (1-2 polyps < 10 mm and without cytologic dysplasia) follow a 5-year surveillance interval similar to low-risk tubular adenomas, although the evidence for this recommendation is low.17,18 Patients with high-risk SSPs are recommended to undergo a 3-year surveillance interval similar to high-risk tubular adenomas. Several studies have confirmed an increased risk for synchronous and metachronous advanced neoplasia in proximal large serrated polyps.19-21 Unlike the well-studied tubular adenomas, the quality of postpolypectomy surveillance evidence for serrated polyps remains limited. To our knowledge, there is no prior surveillance study that specifically assesses surveillance rates in low-risk adenomatous lesions based on the presence or absence of a serrated polyp. The primary aim of this study was to compare rates of advanced neoplasia upon surveillance colonoscopy in patients with initial low-risk polyps in the presence versus the absence of SSPs. The secondary aim was to compare rates of advanced neoplasia upon surveillance colonoscopy in patients with initial low-risk SSPs versus those with initial high-risk conventional adenomas (tubular or tubulovillous). 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
METHODS Study design A natural language search of pathology records for “adenoma” and “sessile serrated” identified 2260 consecutive patients who had a colorectal adenoma or SSP detected during colonoscopy between January 2005 and December 2011 at Rush University Medical Center. The Institutional Review Board of Rush University Medical Center granted study approval. Initial colonoscopy date, name of endoscopist, indication for procedure, bowel preparation quality, endoscopic polyp size, number and colonic location of polyps, and histologic features were recorded. Patients were then stratified into 4 groups as detailed below. Endoscopy and pathology reports for the follow-up surveillance colonoscopy, defined as the next consecutive study, were then reviewed for findings of nonadvanced and advanced neoplasia. Exclusion criteria included surveillance performed at less than 1 year, current or prior colorectal cancer, polyposis syndromes such as Lynch syndrome, inflammatory bowel disease, bowel preparation of poor quality, incomplete colonoscopy without cecal intubation, and lack of surveillance colonoscopy. Any patients with 10 adenomatous polyps found at the time of index colonoscopy were also excluded. The presence of hyperplastic polyps did not alter patient stratification. Traditional serrated adenomas and those with associated dysplasia were considered SSPs with features of a high-risk lesion.4 However, because they warrant more intensive surveillance,17 traditional serrated adenomas (n Z 6) were excluded during the analysis of metachronous advanced neoplasia. Only surveillance cases completed at Rush University Medical Center with reviewed pathology were included. The adenoma detection rate (ADR) of the 14 faculty gastroenterologists ranged from 22.2% to 44.7%.22 The mean of the top and lowest quintiles was 38.3% and 24.5%, respectively. The gastroenterologists achieved a mean sessile SDR of 2.4%. The mean SDR of the top quintile was 3.9%, whereas that of the lowest quintile was .5%.
Procedures Colonoscopies were performed using high-definition Olympus CF-H180AL and PCF-H180AL series colonoscopes (Olympus America, Center Valley, Pa). Narrow-band imaging usage was individualized. Experienced pathologists specialized in GI diseases (B.M. and S.J.) reviewed all polypectomy specimens.
Patient stratification by polyp histology Patients were stratified into 4 groups based on the presence of a high-risk adenoma (HRA) or low-risk www.giejournal.org
Melson et al
adenoma (LRA) and of a synchronous SSP on initial colonoscopy. Based on the 2012 USMSTF guidelines,17 LRA refers to patients with 1 to 2 adenomas (including SSPs if present) with each polyp <10 mm in size. HRA refers to patients with 3 adenomas (including SSPs if present) or a single adenoma 10 mm in size or containing villous histology or high-grade dysplasia. The 4 groups, stratified according to index colorectal neoplasia, consisted of LRA inclusive of SSP, LRA without any SSP, HRA inclusive of SSP, and HRA without any SSP. The following examples illustrate the grouping of LRA and HRA based on the presence or absence of an SSP. If the baseline colonoscopy showed 1 small tubular adenoma and a synchronous small SSP or, alternatively, 1 or 2 isolated SSPs < 10 mm absent of any conventional adenomas, then these cases were grouped in LRA inclusive of SSP. In contrast, patients with 1 or 2 small tubular adenomas lacking advanced features (villous or tubulovillous morphology and high-grade dysplasia) were grouped as LRA without any SSP. Patients with a high-risk SSP or a traditional serrated adenoma or, alternatively, a low-risk SSP with a synchronous highrisk tubular adenoma were grouped in HRA inclusive of SSP. An isolated high-risk tubular adenoma without coexistent SSPs were considered HRA without any SSP. The rates of metachronous advanced neoplasia at first surveillance colonoscopy were analyzed and contrasted between groups. Advanced neoplasia was defined as an adenoma or serrated polyp 10 mm or villous histology, or high-grade dysplasia, or colorectal cancer anywhere in the colorectum.23
Pathology review Pathology slides were interpreted by 2 board-certified GI pathologists. A concordance analysis to define interobserver agreement was performed with a kappa score calculation. Seventy-five cases of respective consecutive tubular adenomas (n Z 25), hyperplastic polyps (n Z 25), and SSPs (n Z 25) from the study period were reviewed by both pathologists blinded to the other’s interpretation. Among the 25 cases that were initially interpreted as SSPs, there was concordance in 23 cases, with 1 SSP reinterpreted as a tubulovillous adenoma and 1 as a hyperplastic polyp. Thus, the kappa score for SSPs was .92 (standard error, .055). Otherwise, there was concordance among all other reviewed cases with a kappa of .97 (standard error, .019).
Statistical analyses Descriptive statistics were used to describe the study population. The 2-sided Fisher exact test set at P .05 was considered statistically significant. All statistical analyses were performed with SAS software (SAS Institute, Cary, NC). www.giejournal.org
Presence of small SSPs increases rate of advanced neoplasia
Total colonoscopies reviewed → 2,260 Lack of surveillance colonoscopy → 1,286 Current or prior colorectal cancer → 78 Surveillance performed at < 1 year → 77 Poor bowel preparation or lack of cecal intubation → 45 Inflammatory bowel disease → 30 Polyposis syndrome → 7 Total patients included → 788
Figure 1. Flow diagram of patient exclusions (some patients met more than 1 exclusion criterion), resulting in 788 patients included in the study.
RESULTS Cohort demographic characteristics Seven hundred eighty-eight patients met the study design criteria, as depicted in Figure 1. The most common exclusion criteria were lack of surveillance colonoscopy (n Z 1286), current or prior colorectal cancer (n Z 78), surveillance performed at less than a 1-year interval (n Z 77), and poor quality bowel preparation or lack of cecal intubation on any studied colonoscopy (n Z 45). Among the 87 patients with initial LRA inclusive of SSP lesions that were excluded, baseline characteristics were comparable with the 66 included patients with surveillance data on the basis of age (mean, 58.2 years; P Z 1.00), sex (64.4% female; P Z .866), and race (71.3% white; P Z .535). Based on initial colonoscopy findings, the 4 groups consisted of 66 patients (8.4%) in LRA inclusive of SSP, 370 patients (46.9%) in LRA without any SSP, 100 patients (12.7%) in HRA inclusive of SSP, and 252 patients (32.0%) in HRA without any SSP. Among the 166 patients with serrated polyps (groups 1 and 3 combined), 129 patients had at least 1 serrated polyp in the proximal colon (defined as proximal to the splenic flexure). There were 122 patients (73.5%) with serrated polyp(s) isolated to the proximal colon. On a polyp basis, 74.2% of all serrated polyps (173/233) were found in the proximal colon. Each group’s demographic and baseline colonoscopy composition is detailed in Tables 1 and 2. The mean age of each group was similar and ranged between 59 and Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 3
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
TABLE 1. Demographic and baseline colonoscopy characteristics of the USMSTF-defined low-risk polyps
TABLE 2. Demographic and baseline colonoscopy characteristics of the USMSTF-defined high-risk polyps
LRA inclusive of SSP (group 1)
LRA without any SSP (group 2)
HRA inclusive of SSP (group 3)
HRA without any SSP (group 4)
No. of patients
66 (8.4)
370 (46.9)
No. of patients
100 (12.7)
252 (32.0)
Mean age, y ( SD)
59 9.5
60 9.9
.447
Mean age, y ( SD)
61 8.8
61 10.5
.083
Sex
Sex
P value
25 (37.9)
184 (49.7)
Male
47 (47.0)
144 (57.1)
Female
41 (62.1)
186 (50.3)
Female
53 (53.0)
108 (42.9)
.001
Race
.472
White
50 (75.8)
185 (50.0)
White
63 (63.0)
148 (58.7)
African American
11 (16.7)
133 (35.9)
African American
33 (33.0)
87 (34.5)
Hispanic
4 (6.1)
37 (10.0)
Hispanic
4 (4.0)
14 (5.6)
Other
1 (1.5)
15 (4.1)
Other
0
3 (1.2)
Tobacco use of those assessed (current or prior)
.399
Tobacco use of those assessed (current or prior)
.220
Yes
31 (52.5)
167 (45.8)
Yes
47 (47.0)
110 (44.5)
No
28 (47.5)
198 (54.2)
No
43 (43.0)
137 (55.5)
Screening
44 (66.7)
263 (71.1)
Screening
69 (69.0)
175 (69.4)
Surveillance
22 (33.3)
107 (28.9)
Surveillance
31 (31.0)
77 (30.6)
92 (92.0)
223 (88.5)
8 (8.0)
29 (11.5)
Indication for colonoscopy
.467
Bowel preparation quality Good/excellent Fair
.279 62 (93.9)
327 (88.4)
4 (6.1)
43 (11.6)
Indication for colonoscopy
1.000
Bowel preparation quality Good/excellent Fair
Polyp burden
.441
Polyp burden
Isolated SSP
56 (84.8)
N/A
Isolated SSP
44 (44)
N/A
Isolated TA
N/A
370 (100)
Isolated TA
N/A
252 (100)
Mixed SSP þ TA Mean surveillance interval, mo ( SD)
1.000 .097
Male
Race
P value
10 (15.2)
N/A
47.3 16.7
53.9 22.1
Mixed SSP þ TA .021
Values are number of cases with percents in parentheses unless otherwise noted. LRA, Low-risk adenoma; SSP, sessile serrated polyp; SD, standard deviation; N/A, not applicable; TA, tubular adenoma.
61 years old. A significantly greater number of women were observed in the serrated groups collectively at 56.6% (94/166) than in the nonserrated groups at 47.3% (294/622; P Z .036). The racial distribution was whitepredominant, followed by African American and Hispanic in each group. The serrated groups collectively demonstrated a higher rate of current or prior tobacco use when compared with the nonserrated groups, although this did not reach statistical significance (52.3% [78/149] vs 45.3% [277/612]; P Z .143). Forty-four patients (66.7%) in the LRA inclusive of SSP group underwent their initial colonoscopy for screening purposes. A shorter mean surveillance interval was observed in the LRA inclusive of SSP group (47.3 16.7 months) compared with LRA without any SSP (53.9 22.1 months; P Z .021). Within the LRA inclusive of SSP group, 75.8% (50/66) of index and surveillance colonoscopies were performed by the same endoscopist. 4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
TSA Mean surveillance interval, mo ( SD)
50 (50)
N/A
6 (6)
N/A
42.5 17.2
40.1 20.9
.309
Values are number of cases with percents in parentheses unless otherwise noted. HRA, High-risk adenoma; SSP, sessile serrated polyp; SD, standard deviation; TA, tubular adenoma; N/A, not applicable; TSA, traditional serrated adenoma.
Metachronous advanced neoplasia in initial low-risk lesions The rate of advanced neoplasia upon surveillance in the LRA inclusive of SSP group was significantly greater than in the LRA without any SSP group (18.2% vs 7.8%, P Z .019). The composition of advanced neoplasia features upon surveillance is highlighted in Table 3. Patients with isolated low-risk serrated polyps (without a coexistent tubular adenoma) demonstrated a 17.9% rate (10/56) of metachronous advanced neoplasia compared with 7.8% in those with isolated low-risk tubular adenomas (29/370; P Z .024). When assessing the subset of patients undergoing initial colonoscopy only for screening purposes, the rate of metachronous advanced neoplasia remained significantly greater in the LRA inclusive of SSP group (18.2%, 8/44) compared with the LRA without any SSP group (6.5%, 17/263; P Z .015). www.giejournal.org
Melson et al
Presence of small SSPs increases rate of advanced neoplasia
TABLE 3. Findings on follow-up surveillance colonoscopy among all groups LRA inclusive of SSP (group 1) (n [ 66)
LRA without any SSP (group 2) (n [ 370)
P value
HRA inclusive of SSP* (group 3) (n [ 94)
HRA without any SSP (group 4) (n [ 252)
P value
Advanced neoplasia
12 (18.2)
29 (7.8)
.019
30 (31.9)
40 (15.9)
.0007
Size 1 cm
10 (83.3)
20 (69.0)
30 (31.9)
31 (77.5)
Villous
2 (16.7)
8 (27.5)
5 (16.7)
15 (37.5)
HGD
0
2 (6.9)
1 (3.3)
4 (10.0)
CRC
0
2 (6.9)
1 (3.3)
0
Any adenoma
39 (59.1)
179 (48.4)
.141
64 (68.1)
158 (62.7)
.624
SSP
22 (33.3)
16 (4.3)
.001
31 (33.0)
15 (6.0)
.001
Values are number of cases with percents in parentheses. LRA, Low-risk adenoma; SSP, sessile serrated polyp; HGD, high-grade dysplasia; CRC, colorectal carcinoma. *Excluding traditional serrated adenoma cases (n Z 6).
TABLE 4. Findings on follow-up surveillance colonoscopy based on the presence or absence of an SSP Presence of SSP* Absence of SSP (groups 1 and 3 (groups 2 and 4 combined) combined) (n [ 160) (n [ 622) P value Advanced neoplasia
42 (26.3)
69 (11.1)
.0001
Any adenoma
103 (64.4)
337 (54.2)
.021
SSP
53 (33.1)
31 (5.0)
.0001
SSP, Sessile serrated polyp. *Excluding traditional serrated adenoma cases (n Z 6).
There was a significant difference in racial distribution between the low-risk groups (groups 1 and 2) with a greater proportion of whites (75.8% vs 50%) compared with African Americans (16.7% vs 35.9%) in the serrated group (P Z .001). The difference in rates of advanced neoplasia on surveillance remained significantly greater when comparing whites in the LRA inclusive of SSP (20%, 10/50) and LRA without SSP (5.9%, 11/185; P Z .004) groups.
Metachronous advanced neoplasia in initial high-risk lesions Next, we compared LRA inclusive of SSP with HRA without any SSP and found no difference in rates of advanced neoplasia upon surveillance between the groups (18.2% vs 15.9%, P Z .709). The rate of advanced neoplasia upon surveillance in the HRA inclusive of SSP group (excluding traditional serrated adenoma) was 31.9% (30/94), which was significantly greater than the 15.9% (40/252) seen in the HRA without any SSP group. In patients with high-risk isolated serrated polyps, there was a 32% rate (16/50) of metachronous advanced neoplasia compared with 15.9% in those with high-risk isolated tubular adenomas (P Z .015).
Metachronous advanced neoplasia in initial serrated polyp presence In patients with an initial serrated polyp (groups 1 and 3 combined), the rates of metachronous advanced www.giejournal.org
neoplasia, adenoma, and serrated recurrence were significantly greater compared with patients without an initial serrated polyp (groups 2 and 4 combined). This is highlighted in Table 4. The presence of an initial serrated polyp increased the overall rate of metachronous advanced neoplasia from 11.1% (69/622) to 26.3% (42/160; P Z .0001) and the rate of a metachronous SSP from 5% (31/622) to 33.1% (53/160; P Z .0001). The rate of advanced neoplasia on surveillance between patients with proximal versus distal SSPs on initial colonoscopy did not differ significantly. Among patients in the LRA inclusive of SSP group, 9 of 54 patients with isolated proximal SSPs (16.7%) and 3 of 12 patients with distal SSPs (25%) had advanced neoplasia on surveillance (P Z .679). Among patients in the HRA inclusive of SSP group, 24 of 68 patients with isolated proximal SSPs (35.3%) and 9 of 32 patients with distal SSPs (28.1%) had advanced neoplasia on surveillance (P Z .505). Isolated serrated polyps without synchronous adenomas were found in 56 patients (84.8%) in the LRA inclusive of SSP group and in 50 patients (50%) in the HRA inclusive of SSP group. The rates of SSPs on surveillance in both the initial LRA inclusive of SSP and the initial HRA inclusive of SSP groups were significantly greater than either respective nonserrated group. Specifically, the rate of SSPs on surveillance in the LRA inclusive of SSP group was 33.3% (22/66), which was significantly greater than that of the LRA without SSP group (4.3%, 16/370, P Z .001) and the HRA without any SSP group (6.0%, 15/252, P Z .0001). The rates of SSPs on surveillance were similar between both the LRA inclusive of SSP and HRA inclusive of SSP groups (33.3% vs 33.0% in HRA inclusive of SSP, P Z 1.00).
DISCUSSION This is the first study to show that the risk of advanced neoplasia during surveillance in patients with initial lowrisk serrated polyps is higher than in patients with lowrisk tubular adenomas without coexistent SSPs. The rate Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 5
Presence of small SSPs increases rate of advanced neoplasia
Melson et al
of advanced neoplasia during surveillance of low-risk serrated polyps was more similar to patients with initial high-risk tubular adenomas. Specifically, the rate of advanced neoplasia on surveillance in the LRA inclusive of SSP group was 18.2%, higher than 7.8% in the LRA without any SSP group (P Z .019) and comparable with the 15.9% in the HRA without any SSP group (P Z .709). Additionally, subanalysis of patients with initial serrated polyps revealed a markedly greater rate of metachronous serrated polyps compared with patients without initial serrated polyps (33.1% vs 5.0%; P Z .0001). To our knowledge, this study is the largest surveillance cohort of low-risk serrated polyps and assessment on rates of metachronous advanced neoplasia. The association of large and proximal serrated polyps with an increased risk for synchronous and metachronous advanced neoplasia has been previously reported.19-21,24,25 However, few studies accurately examine the impact of small SSPs on rates of advanced neoplasia upon surveillance. Schreiner et al19 found an increased rate of neoplasia on surveillance in patients with a proximal serrated polyp; however, there was no significant difference in rate of advanced neoplasia. Their study cohort was gathered from a predominantly male veteran sample with index and surveillance colonoscopies performed in late 1990s and early 2000s, an era preceding the formal classification of serrated polyps.4,7 In addition, the subgroup of patients with serrated lesions, especially low-risk serrated lesions, was small in comparison with our study. These study design factors likely account for the differences between our findings. Another recent study26 found that the presence of a serrated polyp in patients with either an advanced or a nonadvanced adenoma did not increase the risk for metachronous advanced adenomas. However, the serrated cohort was not further divided into high-risk and low-risk lesions, a key stratification point for appropriate screening intervals. Although the authors did attempt to account for interendoscopist variability, SDRs were not addressed. A benchmark for SDR has yet to be established, and it has been shown that detection rates of serrated lesions vary widely between endoscopists.8-10,12,18,27 In a study by Hetzel et al,10 the detection of serrated polyps ranged from 0% to 2.2% among endoscopists at an academic medical center. The mean SDR (excluding hyperplastic polyps) of our study’s colonoscopists was 2.4%, similar to other published values. Kahi et al28 suggested an ADRequivalent proximal detection rate of serrated polyps to be around 4.5%. It is unknown how a higher SDR would have affected study outcomes. However, the rates of advanced neoplasia at surveillance colonoscopy in our conventional LRA patients is similar to rates published in a meta-analysis.29 The above studies demonstrated a strong correlation between ADRs and proximal SDRs. The ADRs of all 14 gastroenterologists studied22 are above established thresholds of satisfactory adenoma detection.30 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
There are several limitations of the study. We do not address colorectal cancer rates upon surveillance but rather the surrogate marker of advanced neoplasia. The study is not powered to define colorectal cancer risk in low-risk lesions based on the presence of SSPs. The sample size was limited by the low rate of surveillance follow-up (43.1%) available. The interval to surveillance colonoscopy was actually significantly shorter in the LRA inclusive of SSP group compared with the LRA without any SSP group. Thus, the comparative increase in advanced neoplasia in the LRA inclusive of SSP group would likely be even larger if similar surveillance intervals occurred. The findings of serrated polyp studies depend on accurate and precise pathologic interpretation of serrated polyps. Several studies have found significant discordance between pathologists in identification of serrated polyps, both among internal pathologists and between various centers.10,31,32 We found the interobserver concordance between our pathologists to be high, with a kappa score for SSPs of .92 (standard error, .055). However, the issue of interobserver variation of serrated polyps remains a limitation of the study, as these were not externally validated. The difficult distinction between SSPs and hyperplastic polyp lesions have prompted some experts to regard all serrated polyps (whether SSPs or hyperplastic polyps) 1 cm or larger as a high-risk SSP.25,33 We did not include index large (1 cm) hyperplastic polyps in the serrated groups, thus leaving the possibility of missed high-risk serrated lesions. However, there were only 2 metachronous hyperplastic polyps 1 cm in the serrated groups. These were not considered metachronous advanced neoplasia findings, because the current USMSTF guidelines17 do not include hyperplastic polyps as an endpoint of advanced neoplasia. Given the high internal concordance between our pathologists and the infrequency of large hyperplastic polyps in the proximal colon, the study’s primary evaluations of metachronous neoplasia would not have been impacted by inclusion of hyperplastic polyps. Incomplete resection of serrated polyps has been reported to be near 30% in the CARE study by Pohl et al14 and is a recognized factor in the development of interval cancers. A surrogate marker would be to focus on the cases with metachronous advanced neoplasia and compare similarities in colonic segment location and pathology between the initial and follow-up lesions. Overall, we found 34 of 114 patients (29.8%) with metachronous advanced neoplasia developed a recurrent lesion of identical histology in the same colonic segment where the initial polypectomy was performed. Specifically, there was a 19.5% recurrence rate (8/41) in serrated polyps and a 35.6% recurrence rate (26/73) in tubular adenomas. The high rate of same-segment recurrence in high-risk tubular adenomas (50%, 20/40) is interesting and may warrant further studies. There are certainly other contributory etiologies to same-segment recurrence, such as missed lesions and local colonic field defects leading to a www.giejournal.org
Melson et al
predisposition for the presence of similar location recurrence of advanced neoplasia.34 The sex,3,10,35 racial,36,37 and smoking19,26 distribution of our patients with serrated polyps is consistent with other studies. A greater proportion of female patients was found in the serrated groups. Also, serrated polyps were more common in whites than African Americans and Hispanics. The higher rate of advanced neoplasia on surveillance in the low-risk serrated group remained significant among whites alone. We did not consider family history of colorectal cancer, use of aspirin or nonsteroidal anti-inflammatory agents, or body mass index, all of which have been shown to affect colorectal cancer risk.38 Overall, our cohort was gathered from a diverse urban community, and the generalizability to other populations is unknown. The USMSTF on Colorectal Cancer published postpolypectomy surveillance guidelines for serrated polyps and reported that these recommendations were based on limited evidence.17,18 Our study found that the risk of interval advanced neoplasia in patients with prior initial low-risk serrated polyps is higher than in patients with initial low-risk tubular adenomas and actually more similar to patients with initial high-risk tubular adenomas. Although the USMSTF recommendations for LRAs with serrated polyps are currently similar to LRAs without serrated polyps, our results reflect a difference in rates of metachronous advanced neoplasia. This study should lead to reconsideration of the recommended surveillance intervals for low-risk serrated polyps.
REFERENCES 1. Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138:2088-100. 2. O’Brien MJ, Yang S, Mack C, et al. Comparison of microsatellite instability, CpG island methylation phenotype, BRAF and KRAS status in serrated polyps and traditional adenomas indicates separate pathways to distinct colorectal carcinoma end points. Am J Surg Pathol 2006;30: 1491-501. 3. Spring KJ, Zhao ZZ, Karamatic R, et al. High prevalence of sessile serrated adenomas with BRAF mutations: a prospective study of patients undergoing colonoscopy. Gastroenterology 2006;131: 1400-7. 4. Torlakovic E, Skovlund E, Snover DC, et al. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003;27:65-81. 5. Arain MA, Sawhney M, Sheikh S, et al. CIMP status of interval colon cancers: another piece to the puzzle. Am J Gastroenterol 2010;105: 1189-95. 6. Kahi C. How does the serrated polyp pathway alter CRC screening and surveillance? Dig Dis Sci 2015;60:773-80. 7. Snover D, Ahnen DJ, Burt RW, et al. Serrated polyps of the colon and rectum and serrated (“hyperplastic”) polyposis. In: Bozman FT, Carneiro F, Hruban RH, et al, eds. WHO Classification of tumours pathology and genetics tumours of the digestive system, 4th ed. Berlin: Springer-Verlag; 2010, 160-5. 8. East JE, Vieth M, Rex DK. Serrated lesions in colorectal cancer screening: detection, resection, pathology and surveillance. Gut 2015;64:991-1000.
www.giejournal.org
Presence of small SSPs increases rate of advanced neoplasia 9. Abdeljawad K, Vermulapalli KC, Kahi CJ, et al. Sessile serrated polyp prevalence determined by a colonoscopist with a high lesion detection rate and an experienced pathologist. Gastrointest Endosc 2015;81:517-24. 10. Hetzel JT, Huang CS, Coukos JA, et al. Variation in the detection of serrated polyps in an average risk colorectal cancer screening cohort. Am J Gastroenterol 2010;105:2656-64. 11. Lash RH, Genta RM, Schuler CM. Sessile serrated adenomas: prevalence of dysplasia and carcinoma in 2139 patients. J Clin Pathol 2010;63:681-6. 12. Kahi CJ, Hewett DG, Norton DL, et al. Prevalence and variable detection of proximal colon serrated polyps during screening colonoscopy. Clin Gastroenterol Hepatol 2011;9:42-6. 13. Hazewinkel Y, Lopez-Ceron M, East JE, et al. Endoscopic features of sessile serrated adenomas: validation by international experts using high-definition white-light endoscopy and narrow-band imaging. Gastrointest Endosc 2013;77:916-24. 14. Pohl H, Srivastava A, Bensen SP, et al. Incomplete polyp resection during colonoscopy- results of the Complete Adenoma Resection (CARE) study. Gastroenterology 2013;144:74-80. 15. Winawer SJ, Zauber AG, Fletcher RH, et al. Guidelines for colonoscopy surveillance after polypectomy: a consensus update by the US MultiSociety Task Force on colorectal cancer and the American Cancer Society. Gastroenterology 2006;130:1872-85. 16. Levin B, Lieberman DA, McFarland B, et al. Screening and surveillance for early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US MultiSociety Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570-95. 17. Lieberman DA, Rex DK, Winawer SJ, et al. United States Multi-Society Task Force on Colorectal Cancer. Guidelines for colonoscopy surveillance after screening and polypectomy: a consensus update by the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2012;143:844-57. 18. Rex DK, Ahnen DJ, Baron JA, et al. Serrated lesions of the colorectum: review and recommendations from an expert panel. Am J Gastroenterol 2012;107:1315-29. 19. Schreiner MA, Weiss DG, Lieberman DA. Proximal and large hyperplastic and nondysplastic serrated polyps detected by colonoscopy are associated with neoplasia. Gastroenterology 2010;139: 1497-502. 20. Hiraoka S, Kato J, Fujiki S, et al. The presence of large serrated polyps increases risk for colorectal cancer. Gastroenterology 2010;139: 1503-10. 21. Li D, Jim C, McCulloch C, et al. Association of large serrated polyps with synchronous advanced colorectal neoplasia. Am J Gastroenterol 2009;104:695-702. 22. Greenspan M, Rajan KB, Baig A, et al. Advanced adenoma detection rate is independent of nonadvanced adenoma detection rate. Am J Gastroenterol 2013;108:1286-92. 23. Imperiale TF, Monahan PO, Stump TE, et al. Derivation and validation of a scoring system to stratify risk for advanced colorectal neoplasia in asymptomatic adults. Ann Intern Med. 2015;163:339-46. 24. Hazewinkel Y, de Wijkerslooth TR, Stoop EM, et al. Prevalence of serrated polyps and association with synchronous advanced neoplasia in screening colonoscopy. Endoscopy 2014;46:219-24. 25. Holme O, Bretthauer M, Eide TJ, et al. Long-term risk of colorectal cancer in individuals with serrated polyps. Gut 2015;64:929-36. 26. Macaron C, Vu HT, Lopez R, et al. Risk of metachronous polyps in individuals with serrated polyps. Dis Colon Rectum 2015;58:762-8. 27. Rex DK, Schoenfeld PS, Cohen J, et al. Quality indicators for colonoscopy. Gastrointest Endosc 2015;81:31-53. 28. Kahi CJ, Li X, Eckert GJ, et al. High colonoscopic prevalence of proximal colon serrated polyps in average-risk men and women. Gastrointest Endosc 2012;75:515-20. 29. Saini SD, Kim HM, Schoenfeld P. Incidence of advanced adenomas at surveillance colonoscopy in patients with a personal history of colon
Volume
-,
No.
-
: 2016 GASTROINTESTINAL ENDOSCOPY 7
Presence of small SSPs increases rate of advanced neoplasia
30.
31.
32.
33.
Melson et al
adenomas: a meta-analysis and systemic review. Gastrointest Endosc 2006;64:614-26. Kaminski M, Regula J, Kraszewska E, et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010;362: 1795-803. Khalid O, Radaideh S, Cummings O, et al. Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001. World J Gastroenterol 2009;15:3767-70. Payne SR, Church TR, Wandell M, et al. Endoscopic detection of proximal serrated lesions and pathologic identification of sessile serrated adenomas/polyps vary on the basis of center. Clin Gastroenterol Hepatol 2014;12:1119-26. Hassan C, Quintero E, Dumonceau JM, et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy Guidelines (ESGE) Guideline. Endoscopy 2013;45:842-51.
8 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2016
34. Adler J, Robertson DJ. Interval colorectal cancer after colonoscopy: exploring explanations and solutions. Am J Gastroenterol 2010;110: 1657-64. 35. Burnett-Hartman AN, Passarelli MN, Adams SV, et al. Differences in epidemiologic risk factors for colorectal adenomas and serrated polyps by lesion severity and anatomical site. Am J Epidemiol 2013;177: 625-37. 36. Lieberman DA, Prindiville S, Weiss DG, et al. Risk factors for advanced colonic neoplasia and hyperplastic polyps in asymptomatic individuals. JAMA 2003;290:2959-67. 37. Wallace K, Grau MV, Ahnen D, et al. The association of lifestyle and dietary factors with the risk for serrated polyps of the colorectum. Cancer Epidemiol Biomarkers Prev 2009;18:2310-7. 38. Din FV, Theodoratou E, Farrington SM, et al. Effect of aspirin and NSAIDs on risk and survival from colorectal cancer. Gut 2010;59:1670-9.
www.giejournal.org