THE RED SECTION
nature publishing group
Sessile Serrated Polyps at Screening Colonoscopy: Have They Been Under Diagnosed? Jill Tinmouth, MD, PhD1–3, Pauline Henry, MD, PhD3,4, Eugene Hsieh, MDCM3,4, Nancy N. Baxter, MD, PhD1,5–7, Robert J. Hilsden, MD, PhD8, S. Elizabeth McGregor, PhD8,9, Lawrence F. Paszat, MD, MS1,3,7,10,11, Arlinda Ruco, MPH3, Refik Saskin, MSc1, Andrew J. Schell, MD12,13, Emina E. Torlakovic, MD, PhD4,14 and Linda Rabeneck, MD, MPH1,2,7,10,11 OBJECTIVES:
The sessile serrated adenoma/polyp (SSA/P) is increasingly recognized as an important precursor to colorectal cancer (CRC) and may contribute to proximal postcolonoscopy CRCs. Hyperplastic polyps (HPs) generally follow a more benign course than do SSA/Ps, but they have a similar histologic appearance. Our aims were to identify patient and polyp factors associated with reclassification of HPs as SSA/Ps during a central pathology review and to characterize and compare their subsequent clinical management with other polyps.
METHODS:
From 2003 to 2008, we prospectively enrolled asymptomatic persons aged 50–74 years in a study of screening colonoscopy. Because criteria for SSA/P diagnosis evolved over our study period, we initiated a second review of all HPs > 5 mm in size in 2011, with reclassification of polyps if indicated. Rates of subsequent colonoscopies, polypectomies, and CRCs were identified.
RESULTS:
We enrolled 2,527 persons who underwent colonoscopy in whom 111 had HPs > 5 mm. Thirty-two of the 111 participants (28.8%) with HPs > 5 mm had their polyps reclassified as SSA/Ps. There were no significant differences in patient characteristics between those with reclassified SSA/Ps and those who had HPs > 5 mm. SSA/Ps were more likely to be proximal (P < 0.001) and larger (P < 0.007) than the HPs. In all, 48.3% of those with high-risk adenomas received appropriate follow-up compared with 26.1% of those with high-risk SSA/Ps.
CONCLUSIONS:
Almost 1/3 of recently diagnosed HPs > 5 mm were reclassified as SSA/Ps. Patients previously diagnosed with larger HPs in the proximal colon may benefit from a pathologic review to ensure appropriate diagnosis and follow-up.
Am J Gastroenterol advance online publication, 8 July 2014; doi:10.1038/ajg.2014.78
Previously, hyperplastic polyps (HPs) were the only recognized polyps with serrated morphology. HPs were thought to have no malignant potential (1), whereas it was recognized that adenomas may develop into colorectal cancers (CRCs) through the adenoma–carcinoma pathway (2–4). However, since the late 1980s, the potential neoplastic nature of some serrated polyps has been recognized (2–6). More recently, Torlakovic
et al. (4,6) have identified and reported on the distinct architectural and histological features of another serrated polyp, the “sessile serrated adenoma/polyp” (SSA/P). SSA/Ps are flat, sessile polyps with lateral serration, branching, and dilated distorted crypts (1,4,7). Over the past decade, SSA/Ps have been increasingly recognized as an important precursor to CRC through the serrated neoplastic
1
Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; 2Cancer Care Ontario, Toronto, Ontario, Canada; 3Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 4Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada; 5Department of Surgery, University of Toronto, Toronto, Ontario, Canada; 6Department of Surgery and Li Ka Shing Research Institute, St. Michael’s Hospital, Toronto, Ontario, Canada; 7 Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada; 8Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada; 9Alberta Health Services Cancer Care, Calgary, Alberta, Canada; 10Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada; 11Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada; 12Calgary Laboratory Services, Calgary, Alberta, Canada; 13Department of Pathology and Laboratory Medicine, University of Calgary, Calgary, Alberta, Canada; 14Department of Pathology, UHN Toronto General Hospital, Toronto, Ontario, Canada. Correspondence: Jill Tinmouth, MD, PhD, Cancer Care Ontario, Scientist and Staff Gastroenterologist, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue (Room HG40), Toronto, Ontario, Canada M4N 3M5. E-mail: [email protected] © 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
1
2
THE RED SECTION
pathway (3,4,8). This pathway involves CpG island hypermethylation and mutation of the oncogene BRAF resulting in tumors with microsatellite instability (1,3). SSA/Ps have indistinct borders, which makes endoscopic recognition and excision of these polyps challenging; therefore, it is possible that the SSA/P may account for a significant proportion of postcolonoscopy CRCs (1,3,4,8). Because SSA/Ps are histologically similar to HPs (1,2,4–10) and because they were until recently an underappreciated entity, in the past, SSA/Ps may have been misdiagnosed and reported as HPs. Patients with polyps that were misdiagnosed in this manner may have suboptimal surveillance, putting them at an increased risk of CRC. The objectives of our study were twofold. First, we examined patient and polyp factors associated with reclassification of HPs as SSA/Ps during a second pathologic review of HPs > 5 mm. Second, we characterized and compared the subsequent clinical management of SSA/Ps with other polyps in light of current surveillance guidelines (11).
Figure 1. Sessile serrated adenoma/polyp with irregular serration extending the full length of the crypts, and a focal inverted T-shaped crypt (arrow).
Methods
From 2003 to 2008, we prospectively enrolled 2,527 asymptomatic persons aged 50–74 years in a study of screening colonoscopy (termed “original cohort”). The study was approved by the Research Ethics Board at Sunnybrook Health Sciences Centre and Women’s College Hospital. Sixteen endoscopists performed the colonoscopies; 2 were surgeons and 14 were gastroenterologists. At colonoscopy, the number, size, shape, location, bowel preparation (good, fair, poor), and removal method of all polyps was collected in a standardized manner. Each polyp was sent to pathology in a separate jar. The initial pathologic interpretation of all biopsied lesions was performed by the designated study pathologist (E.H.). The pathologist classified each mass/polyp as cancer, advanced adenoma (≥10 mm or with villous histology or high grade dysplasia), tubular adenoma, traditional serrated adenoma, SSA/P, HP, or other (for example, leiomyoma or neuroendocrine tumor). Demographic information, family history, body mass index, smoking history, alcohol consumption, and nonsteroidal anti-inflammatory drug use was obtained before colonoscopy. The adenoma detection rate in this cohort was 23.5%. In 2011, we initiated a second pathologic review of tissue specimens from all persons in the original cohort with HPs > 5 mm in size (n = 111) that was overseen by the same pathologist. Those who were diagnosed with SSA/Ps at the initial pathologic report (n = 4) were not included in the study cohort, as we were interested in factors associated with reclassification. Diagnostic criteria were based on the 2010 WHO Classification of Tumours of the Digestive System (7) and a paper by Torlakovic et al. (10). Polyps were reclassified as SSA/Ps if they were sessile and had a serrated surface with distorted crypt bases, usually in a boot or ‘L’ shape (1,4,10). The reliability of the histologic diagnosis by our pathologist was tested by having a second pathologist (A.J.S.) review a subset of 20 samples. Of these, there was disagreement in just one case (data not shown). The polyps were divided into two categories accordThe American Journal of GASTROENTEROLOGY
Figure 2. Hyperplastic polyp with serration predominantly in the upper half of the crypts.
ing to size (6–9 mm, ≥10 mm). The pathologist was aware of the size category for each polyp (based on endoscopic evaluation) but not the location in the colon. Participants were then categorized using the following hierarchy based on their most advanced finding: SSA/P > large HP (≥10 mm) > medium HP (6–9 mm). Figures 1–3 illustrate endoscopic and histologic characteristics of SSA/Ps and HPs. To characterize the surveillance of those with reclassified SSA/Ps and to compare it with the management of those with other lesions, participants were allocated into one of six lesion categories. Lesion categories were based on the current United States Multi-Society Task Force on Colorectal Cancer guidelines for colonoscopy VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
a
b
c
d
Figure 3. Endoscopic images of sessile serrated adenoma/polyp with mucous cap (a), under standard white light (b), under narrow band imaging (c), and after the injection of methylene blue (d).
surveillance after screening and polypectomy in average-risk individuals (11) (Table 1). Participants with poor bowel preparation were excluded from this analysis. Information obtained from the Registered Persons Database (RPDB), the Ontario Health Insurance Plan (OHIP) Claims History Database (CHDB), and the Ontario Cancer Registry (OCR) was used to identify subsequent colonoscopies, polypectomies, and CRCs, and to characterize the surveillance received by the study cohort. To address missed colonoscopies, we also searched the databases for deaths and those who might have moved. The RPDB, OHIP CHDB, and the OCR have been previously described (12,13). Data analysis. Descriptive and univariate comparisons were used to explore participant and polyp characteristics of those with lesions reclassified as SSA/P vs. those with lesions that were not reclassified in the study cohort. All tests were two-sided, and a P value of ≤0.05 was considered statistically significant. All analyses were performed in SAS version 9.3 (SAS Institute, Cary, NC).
Results
Patient and polyp characteristics. Of the total 2,527 participants, 111 (4.4%) had an HP > 5 mm in size (with or without other pathology). On the basis of the second pathologic review, 32 (28.8%) of those with HPs > 5 mm had their polyps reclassified as SSA/Ps. Table 2 shows the patient and polyp characteristics of the study cohort. Three polyps © 2014 by the American College of Gastroenterology
reclassified as SSA/Ps were excluded owing to missing information on polyp location and/or removal procedure. Only 17.9% (n = 12) of polyps originally classified as medium HPs (6–9 mm) were reclassified compared with 41.5% (n = 17) of large HPs (≥10 mm). Notably, no SSA/Ps had dysplasia. There were no significant differences in patient characteristics between those whose polyps were reclassified as SSA/Ps and those whose polyps were not reclassified. SSA/Ps were more likely to be proximal (P < 0.001) and larger in size (≥10 mm) (P < 0.007) than HPs. Participants with polyps reclassified as SSA/Ps were also more likely to have another SSA/P/HP lesion (P < 0.001). SSA/Ps ≥10 mm had a predisposition for the proximal colon, whereas the majority of HPs ≥10 mm were located in the distal colon (data not shown). Subsequent clinical management. Table 3 outlines the proportion of participants who received a subsequent colonoscopy/ polypectomy within the current recommended surveillance interval for each lesion category, the median time to subsequent colonoscopy in years, subsequent CRC diagnosis, and the followup time in years for each lesion category. The proportion of participants receiving a subsequent colonoscopy within the recommended interval was higher for those with high-risk adenomas (n = 72; 48.3%) compared with those with high-risk SSA/Ps (n = 6; 26.1%). A similar trend was observed when comparing medium-risk adenomas and SSA/Ps, although the number of persons with medium-risk SSA/Ps was small. Of those who had polyps reclassified, 12 (37.5%) would have had their surveillance intervals changed based on the reclassification. None The American Journal of GASTROENTEROLOGY
3
4
THE RED SECTION
Table 1. Polyp histology and recommended surveillance interval for each lesion categorya Recommended surveillance interval
Lesion category
Definition
High risk—adenoma
Advanced adenoma (≥10 mm, with any villous or high grade dysplasia) OR ≥3 tubular adenomas±any SSA/P/HP
3 Years
High risk—SSA/P
SSA/P (any location) ≥10 mm OR HP (proximal only) ≥10 mm OR SSA/P any size with dysplasia OR TSA±other SSA/P/HP but no other findings
3 Years
Medium risk—adenoma
1–2 tubular adenomas±non-high-risk SSA/P/HP
5 Years
Medium risk—SSA/P
SSA/P (any location) up to 9 mm±non-high-risk SSA/P/HP but no other findings
5 Years
Low/minimal risk—HP
HP (distal) any size OR HP (proximal) < 10 mm
10 Years
Low/minimal risk—no polyps
No polyps
10 Years
CRC, colorectal cancer; HP, hyperplastic polyp; SSA/P, sessile serrated adenoma/polyp; TA/TSA, tubular or traditional serrated adenoma. a Adapted from the United States Multi-Society Task Force on Colorectal Cancer guidelines for colonoscopy surveillance after screening and polypectomy in average risk individuals (13).
of the participants with an SSA/P developed CRC over a median (interquartile range) follow-up time of 6.1 (5.0–7.4) years. Fortynine persons passed away before 31 December 2012 (end of follow-up data), and only three (6.1%) of them had a high-risk or medium-risk SSA/P.
Discussion
In this study, almost 1/3 of polyps originally (2003–2008) classified as HPs > 5 mm were reclassified as SSA/Ps (2011). Proximal location and polyp size ≥10 mm were associated with reclassification. Fewer patients with polyps in the serrated pathway had a subsequent colonoscopy within the recommended interval compared with those with polyps in the adenomatous pathway, although compliance with surveillance guidelines was suboptimal in both groups. The proportion reclassified among other studies (9,14–16) examining reclassification of previously diagnosed HPs is highly variable and ranges from 5.8 (15) to 85% (14). These differences are likely attributable to study design and inter-observer variability. For example, 1,402 HPs biopsied between 1980 and 2001 were reviewed by three pathologists and were only reclassified as SSA/Ps (5.8%) when all three agreed (15). The criteria of agreement among three pathologists likely contributed to the lower proportion of polyps being reclassified. Another study (14), which evaluated 40 right-sided HPs larger than 5 mm in 2007 initially diagnosed in 2001, found significant inter-observer variability in reclassification among 3 pathologists. The proportion of polyps classified ranged from 30 to 85%, and only slight agreement was observed among the pathologists (κ-value = 0.16) (14). The use of size and location as criteria for review of previously diagnosed HPs may be considered to reduce the burden of this task. Singh et al. (9) evaluated the proportion of previously reported HPs without concomitant findings that were reclassified as SSA/Ps and found that polyp size and proximal location were associated with reclassification (9). On the basis of these The American Journal of GASTROENTEROLOGY
findings, the authors (9) suggest that review of HPs be limited to those > 5 mm in size, as only two of the polyps that were reclassified in their study were larger than 10 mm. In a similar analysis, Kim et al. (16) restricted the evaluation to HPs > 10 mm and found that only eight (16.4%) were reclassified as SSA/Ps. Although we also found size to be an important factor associated with reclassification, almost half (41.4%) of the reclassified SSA/Ps in our cohort would have gone undetected had we reviewed only those polyps > 10 mm. Our results and those of other studies indicate that SSA/Ps are more likely to be located in the proximal colon (9,14,15). There were very few differences in patient factors associated with the reclassification of HPs as SSA/Ps; there are several possible explanations for this finding, including a persistent misclassification despite the high concordance between two pathologists that we demonstrated in a subsample of specimens and our relatively small sample size. Another possible explanation is that the HP may be a precursor of the SSA/P lesion. Notably, none of the reclassified SSA/Ps in our study had dysplasia, a finding that is not surprising given the very low rates (0.17%) reported by Lash et al., (17) in the largest series to date. Surveillance guidelines are an important tool to guide clinical management, and can help identify patients who are overmanaged or undermanaged. As would be anticipated, those in whom lesions were reclassified as SSA/Ps in our study had suboptimal surveillance according to the current guidelines (11). The need for adequate surveillance in this group may be supported by the high rate of polypectomy among those who did have a subsequent colonoscopy (Table 3). Interestingly, persons with high-risk adenomas had suboptimal surveillance, despite the fact that practitioners should have a much greater understanding of the appropriate management of these polyps. This may reflect a lack of coordinated care among various health professionals or the underutilization of guidelines in routine clinical practice. Our results suggest the need for better adherence to guidelines to ensure that patients are receiving the highest quality care. VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
Table 2. Characteristics of the 108a participants with HPs > 5 mm included in the second pathologic review, classified by final diagnosis SSA/P (n =29)
HP, > 5 mm (n =79)
Total (n =108)
50–54
13 (44.8)
33 (41.8)
46 (42.6)
55–59
10 (34.5)
18 (22.8)
28 (25.9)
60–64
4 (13.8)
12 (15.2)
16 (14.8)
65–69
2 (6.9)
14 (17.7)
16 (14.8)
70–74
0 (0.0)
2 (2.5)
2 (1.9)
Female
17 (58.6)
45 (57.0)
62 (57.4)
Male
12 (41.4)
34 (43.0)
46 (42.6)
No
22 (75.9)
70 (88.6)
92 (85.2)
Yes
7 (24.1)
9 (11.4)
16 (14.8)
No
14 (48.3)
31 (39.2)
45 (41.7)
Yes
15 (51.7)
48 (60.8)
63 (58.3)
23 (79.3)
63 (79.7)
86 (79.6)
6 (20.7)
16 (20.3)
22 (20.4)
No
12 (41.4)
45 (57.0)
57 (52.8)
Yes
17 (58.6)
34 (43.0)
51 (47.2)
Left
13 (44.8)
62 (78.5)
75 (69.4)
Right
16 (55.2)
17 (21.5)
33 (30.6)
1 (3.4)
7 (8.9)
8 (7.4)
Sessile
26 (89.7)
72 (91.1)
98 (90.7)
Missing
2 (6.9)
0 (0.0)
2 (1.9)
Age group in years, n (%)
P value NS
Sex, n (%)
NS
First-degree relative with CRC, n (%)
NS
Smoking (ever), n (%)
NS
Smoking (current), n (%) No Yes
NS
2
BMI > 27 kg/m , n (%)
NS
Polyp location, n (%)
< 0.001
Polyp shape, n (%)
0.043
Pedunculated
Removal procedure, n (%)
NS
Snare polypectomy
25 (86.2)
54 (68.4)
79 (73.1)
Other
4 (13.8)
25 (31.6)
29 (26.9)
Bowel preparation, n (%)
NS
Good
22 (75.9)
50 (63.3)
72 (66.7)
Fair or poor
7 (24.1)
29 (36.7)
36 (33.3)
Polyp size, n (%)
0.007
6–9 mm
12 (41.4)
55 (69.6)
67 (62.0)
≥10 mm
17 (58.6)
24 (30.4)
41 (38.0)
Most advanced other non-SSA/P/HP finding, n (%) b
NS
Advanced adenoma
5 (17.2)
12 (15.2)
17 (15.7)
TA/TSA
8 (27.6)
21 (26.6)
29 (26.9)
Other
3 (10.3)
10 (12.7)
13 (12.0)
No other non-SSA/P/HP
13 (44.8)
36 (45.6)
49 (45.4) Table 2 continued on following page
© 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
5
6
THE RED SECTION
Table 2. Continued SSA/P (n =29)
HP, > 5 mm (n =79)
Total (n =108)
P value
Most advanced other SSA/P/HP finding, n (%)
< .001
SSA/P
5 (17.2)
0 (0.0)
5 (4.6)
HP, large
2 (6.9)
0 (0.0)
2 (1.9)
HP, medium
0 (0.0)
7 (8.9)
7 (6.5)
HP, small
10 (34.5)
33 (41.8)
43 (39.8)
No other SSA/P/HP
12 (41.4)
39 (49.4)
51 (47.2)
BMI, body mass index; CRC, colorectal cancer; HP, hyperplastic polyp; NS, not significant; SSA/P, sessile serrated adenoma/polyp; TA/TSA, tubular or traditional serrated adenoma. a Excludes three participants with missing information on polyp location and/or removal procedure. b Advanced adenoma=adenoma ≥10 mm or with villous histology or high grade dysplasia.
Table 3. Proportion who went onto colonoscopy/polypectomy within recommended surveillance interval, time to this subsequent colonoscopy, and follow-up time among participantsa (n=2,316) classified by most advanced finding Lesion category
Interval
High-risk adenoma (n=149)
High-risk SSA/P (n=23)
Medium-risk adenoma (n=415)
Medium-risk SSA/P (n=7)
Low-risk HP (n=311)
Low risk, no polyps (n=1,411)
3 Years
3 Years
5 Years
5 Years
10 Years
10 Years
Total (n=2,316) P value
Subsequent colonoscopy within recommended surveillance interval, n (%)
< 0.001
Yes
72 (48.3)
6 (26.1)
194 (46.7)
2 (28.6)
100 (32.2)
332 (23.5)
706 (30.5)
No
797 (51.7)
17 (73.9)
221 (53.3)
5 (71.4)
211 (67.8)
1,079 (76.5)
1,610 (69.5)
Subsequent polypectomy within recommended surveillance interval, n (%)
< 0.001
Yes
34 (22.8)
4 (17.4)
80 (19.3)
1 (14.3)
43 (13.8)
72 (5.1)
234 (10.1)
No
115 (77.2)
19 (82.6)
335 (80.7)
6 (85.7)
268 (86.2)
1,339 (94.9)
2,082 (89.9)
Time to subsequent colonoscopy within recommended surveillance interval (years) Median (IQR)
1.0 (0.5–1.7)
0.8 (0.2–2.0)
3.4 (2.9–4.2)
2.4 (1.0–3.8)
< 0.001 4.9 (3.7–5.5)
5.0 (4.1–5.6)
4.2 (2.8–5.2)
Subsequent CRC within recommended surveillance interval, n (%)
0.012
Yes
1 (0.7)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
1 (0.0)
No
148 (99.3)
23 (100.0)
415 (100.0)
7 (100.0)
311 (100.0)
1,411 (100.0)
2,315 (100.0)
Follow-up timeb (years) Median (IQR)
3.0 (3.0–3.0)
< 0.001 3.0 (3.0–3.0)
5.0 (5.0–5.0)
5.0 (4.7–5.0)
6.7 (5.6–7.7)
6.8 (5.7–7.7)
6.1 (5.0–7.4)
CRC, colorectal cancer; HP, hyperplastic polyp; IQR, interquartile range; SSA/P, sessile serrated adenoma/polyp. a Excludes participants diagnosed with CRC (n=5), SSA/P (n=4), or other (n=107) at index colonoscopy. Also excludes participants with missing covariate information (n=11) and poor bowel preparation (n=84). b Follow-up time was determined as the earliest of: date of death, date of last contact, end of claims history data (31 December 2012), or lesion-specific follow-up time.
Failure to identify SSA/Ps endoscopically or, as in our study, histologically, has been cited as a possible contributor to postcolonoscopy CRCs (1,3,9). Interestingly, we did not observe any subsequent CRCs among those who had their polyps reclassified despite suboptimal surveillance. There are several possible reasons for this finding. First, our sample size of those with SSA/Ps was quite small (n = 32). A larger number of persons with these polyps is likely needed to measure the risk of subsequent CRC. Second, The American Journal of GASTROENTEROLOGY
persons with these lesions were followed up for a relatively short period of time. A cross-sectional study of 290,810 colonoscopic specimens found a 15-year interval between the median age of those with SSA/P and SSA/P with carcinoma, suggesting that progression from SSA/P to carcinoma exceeds the follow-up duration in our study (17). The results of our study must be considered in light of the study strengths and limitations. Our initial study comprised a VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
well-characterized and large cohort of more than 2,500 participants undergoing screening colonoscopy at a large center and is likely reflective of routine clinical practice. Centralized pathologic review of all polyps was conducted initially and during the second review in 2011. The study pathologist was aware of the size range but not of the location of polyps being reassessed, which could have led to bias for reassessment of larger polyps. Our study involved a relatively small number of participants with SSA/Ps; however, most other reported retrospective studies report similar numbers. Almost 1/3 of HPs > 5 mm diagnosed more than 5 years ago were reclassified as SSA/Ps in 2011, and these patients received suboptimal surveillance. Patients previously diagnosed with larger HPs may benefit from pathologic review to ensure appropriate diagnosis and management. The use of standardized histologic criteria, standardized training of pathologists in order to reduce inter-observer variability, careful attention to colonoscopy technique, and knowledge of and adherence to current surveillance guidelines are needed to ensure accurate diagnosis and optimal management of persons with SSA/Ps. ACKNOWLEDGMENTS
We acknowledge the support of the Institute for Clinical Evaluative Sciences, the Ontario Ministry of Health and Long Term Care and Cancer Care Ontario. CONFLICT OF INTEREST
Guarantor of the article: Jill Tinmouth, MD, PhD. Specific author contributions: J.T., P.H., N.N.B., R.J.H., E.H., S.E.M., E.E.T., L.F.P., A.J.S., and L.R. were involved in planning and/or conducting the study. Collection and/or interpreting data was done by all authors. J.T. and A.R. drafted the manuscript. R.S. completed the statistical analysis. J.T. was responsible for study supervision. All authors approved the final draft submitted. J.T. accepts full responsibility for the conduct of the study. Financial support: This research was funded by a grant from the Canadian Institutes of Health Research (FRN 74560). Dr Baxter is supported by a Cancer Care Ontario Health Services Research Chair. Dr Tinmouth is supported by a CIHR New Investigator Award. The funding source had no role in the study design, collection, analysis or interpretation of the data, writing of the report, or in the decision to publish. Potential competing interests: None.
© 2014 by the American College of Gastroenterology
Disclosure: The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by Institute for Clinical Evaluative Sciences, the Ontario Ministry of Health and Long Term Care and Cancer Care Ontario is intended or should be inferred.
REFERENCES 1. Limketkai BN, Lam-Himlin D, Arnold MA et al. The cutting edge of serrated polyps: a practical guide to approaching and managing serrated colon polyps. Gastrointest Endosc 2013;77:360–75. 2. Longacre TA, Fenoglio-Preiser CM. Mixed hyperplastic adenomatous polyps/serrated adenomas. Am J Surg Pathol 1990;14:524–37. 3. Leggett B, Whitehall V. Role of serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138:2088–100. 4. Torlakovic E, Skovlund E, Snover DC. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003;27:65–81. 5. Urbanski SJ, Kossakowska AE, Marcon N et al. Mixed hyperplastic adenomatous polyps: an underdiagnosed entity. Report of a case of adenocarcinoma arising within a mixed hyperplastic adenomatous polyp. Am J Surg Pathol 1984;8:551–6. 6. Torlakovic E, Snover DC. Serrated adenomatous polyposis in humans. Gastroenterology 1996;110:748–55. 7. Bosman FT, Carneiro F, Hruban RH et al. (Eds.), World Health Organization Classification of Tumours of the Digestive System. IARC Press: Lyon, France, 2010. 8. Goldstein NS, Bhanot P, Odish E et al. Hyperplastic-like colon polyps that preceded microsatellite-unstable adenocarcinomas. Am J Clin Pathol 2003;119:778–96. 9. Singh H, Bay D, Ip S et al. Pathological reassessment of hyperplastic colon polyps in a city-wide pathology practice: implications for polyp surveillance. Gastrointest Endosc 2012;76:1003–8. 10. Torlakovic EE, Gomez JDG, Driman DK et al. Sessile serrated adenoma (SSA) vs. traditional serrated adenoma (TSA). Am J Surg Pathol 2008;32:21–9. 11. Lieberman DA, Rex DK, Winawer SJ et al. 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. 12. Robles SC, Marrett LD, Clarke EA et al. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495–501. 13. Bressler B, Paszat L, Chen Z et al. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology 2007;132:96–102. 14. Khalid O, Radaideh S, Cummings OW et al. Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001. World J Gastroenterol 2009;15:3767–70. 15. Lu FI, Niekerk dW, Owen D et al. Longitudinal outcome study of sessile serrated adenomas of the colorectum: an increased risk for subsequent right-sided colorectal carcinoma. Am J Surg Pathol 2010;34:927–34. 16. Kim SW, Cha JM, Lee JI et al. A significant number of sessile serrated adenomas might not be accurately diagnosed in daily practice. Gut Liver 2010;4:498–502. 17. Lash RH, Genta RM, Schuler CM. Sessile serrated adenomas: prevalence of dysplasia and carcinoma in 2139 patients. J Clin Pathol 2010;63:681–6.
The American Journal of GASTROENTEROLOGY
7
nature publishing group
Sessile Serrated Polyps at Screening Colonoscopy: Have They Been Under Diagnosed? Jill Tinmouth, MD, PhD1–3, Pauline Henry, MD, PhD3,4, Eugene Hsieh, MDCM3,4, Nancy N. Baxter, MD, PhD1,5–7, Robert J. Hilsden, MD, PhD8, S. Elizabeth McGregor, PhD8,9, Lawrence F. Paszat, MD, MS1,3,7,10,11, Arlinda Ruco, MPH3, Refik Saskin, MSc1, Andrew J. Schell, MD12,13, Emina E. Torlakovic, MD, PhD4,14 and Linda Rabeneck, MD, MPH1,2,7,10,11 OBJECTIVES:
The sessile serrated adenoma/polyp (SSA/P) is increasingly recognized as an important precursor to colorectal cancer (CRC) and may contribute to proximal postcolonoscopy CRCs. Hyperplastic polyps (HPs) generally follow a more benign course than do SSA/Ps, but they have a similar histologic appearance. Our aims were to identify patient and polyp factors associated with reclassification of HPs as SSA/Ps during a central pathology review and to characterize and compare their subsequent clinical management with other polyps.
METHODS:
From 2003 to 2008, we prospectively enrolled asymptomatic persons aged 50–74 years in a study of screening colonoscopy. Because criteria for SSA/P diagnosis evolved over our study period, we initiated a second review of all HPs > 5 mm in size in 2011, with reclassification of polyps if indicated. Rates of subsequent colonoscopies, polypectomies, and CRCs were identified.
RESULTS:
We enrolled 2,527 persons who underwent colonoscopy in whom 111 had HPs > 5 mm. Thirty-two of the 111 participants (28.8%) with HPs > 5 mm had their polyps reclassified as SSA/Ps. There were no significant differences in patient characteristics between those with reclassified SSA/Ps and those who had HPs > 5 mm. SSA/Ps were more likely to be proximal (P < 0.001) and larger (P < 0.007) than the HPs. In all, 48.3% of those with high-risk adenomas received appropriate follow-up compared with 26.1% of those with high-risk SSA/Ps.
CONCLUSIONS:
Almost 1/3 of recently diagnosed HPs > 5 mm were reclassified as SSA/Ps. Patients previously diagnosed with larger HPs in the proximal colon may benefit from a pathologic review to ensure appropriate diagnosis and follow-up.
Am J Gastroenterol advance online publication, 8 July 2014; doi:10.1038/ajg.2014.78
Previously, hyperplastic polyps (HPs) were the only recognized polyps with serrated morphology. HPs were thought to have no malignant potential (1), whereas it was recognized that adenomas may develop into colorectal cancers (CRCs) through the adenoma–carcinoma pathway (2–4). However, since the late 1980s, the potential neoplastic nature of some serrated polyps has been recognized (2–6). More recently, Torlakovic
et al. (4,6) have identified and reported on the distinct architectural and histological features of another serrated polyp, the “sessile serrated adenoma/polyp” (SSA/P). SSA/Ps are flat, sessile polyps with lateral serration, branching, and dilated distorted crypts (1,4,7). Over the past decade, SSA/Ps have been increasingly recognized as an important precursor to CRC through the serrated neoplastic
1
Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; 2Cancer Care Ontario, Toronto, Ontario, Canada; 3Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 4Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada; 5Department of Surgery, University of Toronto, Toronto, Ontario, Canada; 6Department of Surgery and Li Ka Shing Research Institute, St. Michael’s Hospital, Toronto, Ontario, Canada; 7 Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada; 8Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada; 9Alberta Health Services Cancer Care, Calgary, Alberta, Canada; 10Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada; 11Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada; 12Calgary Laboratory Services, Calgary, Alberta, Canada; 13Department of Pathology and Laboratory Medicine, University of Calgary, Calgary, Alberta, Canada; 14Department of Pathology, UHN Toronto General Hospital, Toronto, Ontario, Canada. Correspondence: Jill Tinmouth, MD, PhD, Cancer Care Ontario, Scientist and Staff Gastroenterologist, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue (Room HG40), Toronto, Ontario, Canada M4N 3M5. E-mail: [email protected] © 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
1
2
THE RED SECTION
pathway (3,4,8). This pathway involves CpG island hypermethylation and mutation of the oncogene BRAF resulting in tumors with microsatellite instability (1,3). SSA/Ps have indistinct borders, which makes endoscopic recognition and excision of these polyps challenging; therefore, it is possible that the SSA/P may account for a significant proportion of postcolonoscopy CRCs (1,3,4,8). Because SSA/Ps are histologically similar to HPs (1,2,4–10) and because they were until recently an underappreciated entity, in the past, SSA/Ps may have been misdiagnosed and reported as HPs. Patients with polyps that were misdiagnosed in this manner may have suboptimal surveillance, putting them at an increased risk of CRC. The objectives of our study were twofold. First, we examined patient and polyp factors associated with reclassification of HPs as SSA/Ps during a second pathologic review of HPs > 5 mm. Second, we characterized and compared the subsequent clinical management of SSA/Ps with other polyps in light of current surveillance guidelines (11).
Figure 1. Sessile serrated adenoma/polyp with irregular serration extending the full length of the crypts, and a focal inverted T-shaped crypt (arrow).
Methods
From 2003 to 2008, we prospectively enrolled 2,527 asymptomatic persons aged 50–74 years in a study of screening colonoscopy (termed “original cohort”). The study was approved by the Research Ethics Board at Sunnybrook Health Sciences Centre and Women’s College Hospital. Sixteen endoscopists performed the colonoscopies; 2 were surgeons and 14 were gastroenterologists. At colonoscopy, the number, size, shape, location, bowel preparation (good, fair, poor), and removal method of all polyps was collected in a standardized manner. Each polyp was sent to pathology in a separate jar. The initial pathologic interpretation of all biopsied lesions was performed by the designated study pathologist (E.H.). The pathologist classified each mass/polyp as cancer, advanced adenoma (≥10 mm or with villous histology or high grade dysplasia), tubular adenoma, traditional serrated adenoma, SSA/P, HP, or other (for example, leiomyoma or neuroendocrine tumor). Demographic information, family history, body mass index, smoking history, alcohol consumption, and nonsteroidal anti-inflammatory drug use was obtained before colonoscopy. The adenoma detection rate in this cohort was 23.5%. In 2011, we initiated a second pathologic review of tissue specimens from all persons in the original cohort with HPs > 5 mm in size (n = 111) that was overseen by the same pathologist. Those who were diagnosed with SSA/Ps at the initial pathologic report (n = 4) were not included in the study cohort, as we were interested in factors associated with reclassification. Diagnostic criteria were based on the 2010 WHO Classification of Tumours of the Digestive System (7) and a paper by Torlakovic et al. (10). Polyps were reclassified as SSA/Ps if they were sessile and had a serrated surface with distorted crypt bases, usually in a boot or ‘L’ shape (1,4,10). The reliability of the histologic diagnosis by our pathologist was tested by having a second pathologist (A.J.S.) review a subset of 20 samples. Of these, there was disagreement in just one case (data not shown). The polyps were divided into two categories accordThe American Journal of GASTROENTEROLOGY
Figure 2. Hyperplastic polyp with serration predominantly in the upper half of the crypts.
ing to size (6–9 mm, ≥10 mm). The pathologist was aware of the size category for each polyp (based on endoscopic evaluation) but not the location in the colon. Participants were then categorized using the following hierarchy based on their most advanced finding: SSA/P > large HP (≥10 mm) > medium HP (6–9 mm). Figures 1–3 illustrate endoscopic and histologic characteristics of SSA/Ps and HPs. To characterize the surveillance of those with reclassified SSA/Ps and to compare it with the management of those with other lesions, participants were allocated into one of six lesion categories. Lesion categories were based on the current United States Multi-Society Task Force on Colorectal Cancer guidelines for colonoscopy VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
a
b
c
d
Figure 3. Endoscopic images of sessile serrated adenoma/polyp with mucous cap (a), under standard white light (b), under narrow band imaging (c), and after the injection of methylene blue (d).
surveillance after screening and polypectomy in average-risk individuals (11) (Table 1). Participants with poor bowel preparation were excluded from this analysis. Information obtained from the Registered Persons Database (RPDB), the Ontario Health Insurance Plan (OHIP) Claims History Database (CHDB), and the Ontario Cancer Registry (OCR) was used to identify subsequent colonoscopies, polypectomies, and CRCs, and to characterize the surveillance received by the study cohort. To address missed colonoscopies, we also searched the databases for deaths and those who might have moved. The RPDB, OHIP CHDB, and the OCR have been previously described (12,13). Data analysis. Descriptive and univariate comparisons were used to explore participant and polyp characteristics of those with lesions reclassified as SSA/P vs. those with lesions that were not reclassified in the study cohort. All tests were two-sided, and a P value of ≤0.05 was considered statistically significant. All analyses were performed in SAS version 9.3 (SAS Institute, Cary, NC).
Results
Patient and polyp characteristics. Of the total 2,527 participants, 111 (4.4%) had an HP > 5 mm in size (with or without other pathology). On the basis of the second pathologic review, 32 (28.8%) of those with HPs > 5 mm had their polyps reclassified as SSA/Ps. Table 2 shows the patient and polyp characteristics of the study cohort. Three polyps © 2014 by the American College of Gastroenterology
reclassified as SSA/Ps were excluded owing to missing information on polyp location and/or removal procedure. Only 17.9% (n = 12) of polyps originally classified as medium HPs (6–9 mm) were reclassified compared with 41.5% (n = 17) of large HPs (≥10 mm). Notably, no SSA/Ps had dysplasia. There were no significant differences in patient characteristics between those whose polyps were reclassified as SSA/Ps and those whose polyps were not reclassified. SSA/Ps were more likely to be proximal (P < 0.001) and larger in size (≥10 mm) (P < 0.007) than HPs. Participants with polyps reclassified as SSA/Ps were also more likely to have another SSA/P/HP lesion (P < 0.001). SSA/Ps ≥10 mm had a predisposition for the proximal colon, whereas the majority of HPs ≥10 mm were located in the distal colon (data not shown). Subsequent clinical management. Table 3 outlines the proportion of participants who received a subsequent colonoscopy/ polypectomy within the current recommended surveillance interval for each lesion category, the median time to subsequent colonoscopy in years, subsequent CRC diagnosis, and the followup time in years for each lesion category. The proportion of participants receiving a subsequent colonoscopy within the recommended interval was higher for those with high-risk adenomas (n = 72; 48.3%) compared with those with high-risk SSA/Ps (n = 6; 26.1%). A similar trend was observed when comparing medium-risk adenomas and SSA/Ps, although the number of persons with medium-risk SSA/Ps was small. Of those who had polyps reclassified, 12 (37.5%) would have had their surveillance intervals changed based on the reclassification. None The American Journal of GASTROENTEROLOGY
3
4
THE RED SECTION
Table 1. Polyp histology and recommended surveillance interval for each lesion categorya Recommended surveillance interval
Lesion category
Definition
High risk—adenoma
Advanced adenoma (≥10 mm, with any villous or high grade dysplasia) OR ≥3 tubular adenomas±any SSA/P/HP
3 Years
High risk—SSA/P
SSA/P (any location) ≥10 mm OR HP (proximal only) ≥10 mm OR SSA/P any size with dysplasia OR TSA±other SSA/P/HP but no other findings
3 Years
Medium risk—adenoma
1–2 tubular adenomas±non-high-risk SSA/P/HP
5 Years
Medium risk—SSA/P
SSA/P (any location) up to 9 mm±non-high-risk SSA/P/HP but no other findings
5 Years
Low/minimal risk—HP
HP (distal) any size OR HP (proximal) < 10 mm
10 Years
Low/minimal risk—no polyps
No polyps
10 Years
CRC, colorectal cancer; HP, hyperplastic polyp; SSA/P, sessile serrated adenoma/polyp; TA/TSA, tubular or traditional serrated adenoma. a Adapted from the United States Multi-Society Task Force on Colorectal Cancer guidelines for colonoscopy surveillance after screening and polypectomy in average risk individuals (13).
of the participants with an SSA/P developed CRC over a median (interquartile range) follow-up time of 6.1 (5.0–7.4) years. Fortynine persons passed away before 31 December 2012 (end of follow-up data), and only three (6.1%) of them had a high-risk or medium-risk SSA/P.
Discussion
In this study, almost 1/3 of polyps originally (2003–2008) classified as HPs > 5 mm were reclassified as SSA/Ps (2011). Proximal location and polyp size ≥10 mm were associated with reclassification. Fewer patients with polyps in the serrated pathway had a subsequent colonoscopy within the recommended interval compared with those with polyps in the adenomatous pathway, although compliance with surveillance guidelines was suboptimal in both groups. The proportion reclassified among other studies (9,14–16) examining reclassification of previously diagnosed HPs is highly variable and ranges from 5.8 (15) to 85% (14). These differences are likely attributable to study design and inter-observer variability. For example, 1,402 HPs biopsied between 1980 and 2001 were reviewed by three pathologists and were only reclassified as SSA/Ps (5.8%) when all three agreed (15). The criteria of agreement among three pathologists likely contributed to the lower proportion of polyps being reclassified. Another study (14), which evaluated 40 right-sided HPs larger than 5 mm in 2007 initially diagnosed in 2001, found significant inter-observer variability in reclassification among 3 pathologists. The proportion of polyps classified ranged from 30 to 85%, and only slight agreement was observed among the pathologists (κ-value = 0.16) (14). The use of size and location as criteria for review of previously diagnosed HPs may be considered to reduce the burden of this task. Singh et al. (9) evaluated the proportion of previously reported HPs without concomitant findings that were reclassified as SSA/Ps and found that polyp size and proximal location were associated with reclassification (9). On the basis of these The American Journal of GASTROENTEROLOGY
findings, the authors (9) suggest that review of HPs be limited to those > 5 mm in size, as only two of the polyps that were reclassified in their study were larger than 10 mm. In a similar analysis, Kim et al. (16) restricted the evaluation to HPs > 10 mm and found that only eight (16.4%) were reclassified as SSA/Ps. Although we also found size to be an important factor associated with reclassification, almost half (41.4%) of the reclassified SSA/Ps in our cohort would have gone undetected had we reviewed only those polyps > 10 mm. Our results and those of other studies indicate that SSA/Ps are more likely to be located in the proximal colon (9,14,15). There were very few differences in patient factors associated with the reclassification of HPs as SSA/Ps; there are several possible explanations for this finding, including a persistent misclassification despite the high concordance between two pathologists that we demonstrated in a subsample of specimens and our relatively small sample size. Another possible explanation is that the HP may be a precursor of the SSA/P lesion. Notably, none of the reclassified SSA/Ps in our study had dysplasia, a finding that is not surprising given the very low rates (0.17%) reported by Lash et al., (17) in the largest series to date. Surveillance guidelines are an important tool to guide clinical management, and can help identify patients who are overmanaged or undermanaged. As would be anticipated, those in whom lesions were reclassified as SSA/Ps in our study had suboptimal surveillance according to the current guidelines (11). The need for adequate surveillance in this group may be supported by the high rate of polypectomy among those who did have a subsequent colonoscopy (Table 3). Interestingly, persons with high-risk adenomas had suboptimal surveillance, despite the fact that practitioners should have a much greater understanding of the appropriate management of these polyps. This may reflect a lack of coordinated care among various health professionals or the underutilization of guidelines in routine clinical practice. Our results suggest the need for better adherence to guidelines to ensure that patients are receiving the highest quality care. VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
Table 2. Characteristics of the 108a participants with HPs > 5 mm included in the second pathologic review, classified by final diagnosis SSA/P (n =29)
HP, > 5 mm (n =79)
Total (n =108)
50–54
13 (44.8)
33 (41.8)
46 (42.6)
55–59
10 (34.5)
18 (22.8)
28 (25.9)
60–64
4 (13.8)
12 (15.2)
16 (14.8)
65–69
2 (6.9)
14 (17.7)
16 (14.8)
70–74
0 (0.0)
2 (2.5)
2 (1.9)
Female
17 (58.6)
45 (57.0)
62 (57.4)
Male
12 (41.4)
34 (43.0)
46 (42.6)
No
22 (75.9)
70 (88.6)
92 (85.2)
Yes
7 (24.1)
9 (11.4)
16 (14.8)
No
14 (48.3)
31 (39.2)
45 (41.7)
Yes
15 (51.7)
48 (60.8)
63 (58.3)
23 (79.3)
63 (79.7)
86 (79.6)
6 (20.7)
16 (20.3)
22 (20.4)
No
12 (41.4)
45 (57.0)
57 (52.8)
Yes
17 (58.6)
34 (43.0)
51 (47.2)
Left
13 (44.8)
62 (78.5)
75 (69.4)
Right
16 (55.2)
17 (21.5)
33 (30.6)
1 (3.4)
7 (8.9)
8 (7.4)
Sessile
26 (89.7)
72 (91.1)
98 (90.7)
Missing
2 (6.9)
0 (0.0)
2 (1.9)
Age group in years, n (%)
P value NS
Sex, n (%)
NS
First-degree relative with CRC, n (%)
NS
Smoking (ever), n (%)
NS
Smoking (current), n (%) No Yes
NS
2
BMI > 27 kg/m , n (%)
NS
Polyp location, n (%)
< 0.001
Polyp shape, n (%)
0.043
Pedunculated
Removal procedure, n (%)
NS
Snare polypectomy
25 (86.2)
54 (68.4)
79 (73.1)
Other
4 (13.8)
25 (31.6)
29 (26.9)
Bowel preparation, n (%)
NS
Good
22 (75.9)
50 (63.3)
72 (66.7)
Fair or poor
7 (24.1)
29 (36.7)
36 (33.3)
Polyp size, n (%)
0.007
6–9 mm
12 (41.4)
55 (69.6)
67 (62.0)
≥10 mm
17 (58.6)
24 (30.4)
41 (38.0)
Most advanced other non-SSA/P/HP finding, n (%) b
NS
Advanced adenoma
5 (17.2)
12 (15.2)
17 (15.7)
TA/TSA
8 (27.6)
21 (26.6)
29 (26.9)
Other
3 (10.3)
10 (12.7)
13 (12.0)
No other non-SSA/P/HP
13 (44.8)
36 (45.6)
49 (45.4) Table 2 continued on following page
© 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
5
6
THE RED SECTION
Table 2. Continued SSA/P (n =29)
HP, > 5 mm (n =79)
Total (n =108)
P value
Most advanced other SSA/P/HP finding, n (%)
< .001
SSA/P
5 (17.2)
0 (0.0)
5 (4.6)
HP, large
2 (6.9)
0 (0.0)
2 (1.9)
HP, medium
0 (0.0)
7 (8.9)
7 (6.5)
HP, small
10 (34.5)
33 (41.8)
43 (39.8)
No other SSA/P/HP
12 (41.4)
39 (49.4)
51 (47.2)
BMI, body mass index; CRC, colorectal cancer; HP, hyperplastic polyp; NS, not significant; SSA/P, sessile serrated adenoma/polyp; TA/TSA, tubular or traditional serrated adenoma. a Excludes three participants with missing information on polyp location and/or removal procedure. b Advanced adenoma=adenoma ≥10 mm or with villous histology or high grade dysplasia.
Table 3. Proportion who went onto colonoscopy/polypectomy within recommended surveillance interval, time to this subsequent colonoscopy, and follow-up time among participantsa (n=2,316) classified by most advanced finding Lesion category
Interval
High-risk adenoma (n=149)
High-risk SSA/P (n=23)
Medium-risk adenoma (n=415)
Medium-risk SSA/P (n=7)
Low-risk HP (n=311)
Low risk, no polyps (n=1,411)
3 Years
3 Years
5 Years
5 Years
10 Years
10 Years
Total (n=2,316) P value
Subsequent colonoscopy within recommended surveillance interval, n (%)
< 0.001
Yes
72 (48.3)
6 (26.1)
194 (46.7)
2 (28.6)
100 (32.2)
332 (23.5)
706 (30.5)
No
797 (51.7)
17 (73.9)
221 (53.3)
5 (71.4)
211 (67.8)
1,079 (76.5)
1,610 (69.5)
Subsequent polypectomy within recommended surveillance interval, n (%)
< 0.001
Yes
34 (22.8)
4 (17.4)
80 (19.3)
1 (14.3)
43 (13.8)
72 (5.1)
234 (10.1)
No
115 (77.2)
19 (82.6)
335 (80.7)
6 (85.7)
268 (86.2)
1,339 (94.9)
2,082 (89.9)
Time to subsequent colonoscopy within recommended surveillance interval (years) Median (IQR)
1.0 (0.5–1.7)
0.8 (0.2–2.0)
3.4 (2.9–4.2)
2.4 (1.0–3.8)
< 0.001 4.9 (3.7–5.5)
5.0 (4.1–5.6)
4.2 (2.8–5.2)
Subsequent CRC within recommended surveillance interval, n (%)
0.012
Yes
1 (0.7)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
0 (0.0)
1 (0.0)
No
148 (99.3)
23 (100.0)
415 (100.0)
7 (100.0)
311 (100.0)
1,411 (100.0)
2,315 (100.0)
Follow-up timeb (years) Median (IQR)
3.0 (3.0–3.0)
< 0.001 3.0 (3.0–3.0)
5.0 (5.0–5.0)
5.0 (4.7–5.0)
6.7 (5.6–7.7)
6.8 (5.7–7.7)
6.1 (5.0–7.4)
CRC, colorectal cancer; HP, hyperplastic polyp; IQR, interquartile range; SSA/P, sessile serrated adenoma/polyp. a Excludes participants diagnosed with CRC (n=5), SSA/P (n=4), or other (n=107) at index colonoscopy. Also excludes participants with missing covariate information (n=11) and poor bowel preparation (n=84). b Follow-up time was determined as the earliest of: date of death, date of last contact, end of claims history data (31 December 2012), or lesion-specific follow-up time.
Failure to identify SSA/Ps endoscopically or, as in our study, histologically, has been cited as a possible contributor to postcolonoscopy CRCs (1,3,9). Interestingly, we did not observe any subsequent CRCs among those who had their polyps reclassified despite suboptimal surveillance. There are several possible reasons for this finding. First, our sample size of those with SSA/Ps was quite small (n = 32). A larger number of persons with these polyps is likely needed to measure the risk of subsequent CRC. Second, The American Journal of GASTROENTEROLOGY
persons with these lesions were followed up for a relatively short period of time. A cross-sectional study of 290,810 colonoscopic specimens found a 15-year interval between the median age of those with SSA/P and SSA/P with carcinoma, suggesting that progression from SSA/P to carcinoma exceeds the follow-up duration in our study (17). The results of our study must be considered in light of the study strengths and limitations. Our initial study comprised a VOLUME 104 | XXX 2014 www.amjgastro.com
THE RED SECTION
well-characterized and large cohort of more than 2,500 participants undergoing screening colonoscopy at a large center and is likely reflective of routine clinical practice. Centralized pathologic review of all polyps was conducted initially and during the second review in 2011. The study pathologist was aware of the size range but not of the location of polyps being reassessed, which could have led to bias for reassessment of larger polyps. Our study involved a relatively small number of participants with SSA/Ps; however, most other reported retrospective studies report similar numbers. Almost 1/3 of HPs > 5 mm diagnosed more than 5 years ago were reclassified as SSA/Ps in 2011, and these patients received suboptimal surveillance. Patients previously diagnosed with larger HPs may benefit from pathologic review to ensure appropriate diagnosis and management. The use of standardized histologic criteria, standardized training of pathologists in order to reduce inter-observer variability, careful attention to colonoscopy technique, and knowledge of and adherence to current surveillance guidelines are needed to ensure accurate diagnosis and optimal management of persons with SSA/Ps. ACKNOWLEDGMENTS
We acknowledge the support of the Institute for Clinical Evaluative Sciences, the Ontario Ministry of Health and Long Term Care and Cancer Care Ontario. CONFLICT OF INTEREST
Guarantor of the article: Jill Tinmouth, MD, PhD. Specific author contributions: J.T., P.H., N.N.B., R.J.H., E.H., S.E.M., E.E.T., L.F.P., A.J.S., and L.R. were involved in planning and/or conducting the study. Collection and/or interpreting data was done by all authors. J.T. and A.R. drafted the manuscript. R.S. completed the statistical analysis. J.T. was responsible for study supervision. All authors approved the final draft submitted. J.T. accepts full responsibility for the conduct of the study. Financial support: This research was funded by a grant from the Canadian Institutes of Health Research (FRN 74560). Dr Baxter is supported by a Cancer Care Ontario Health Services Research Chair. Dr Tinmouth is supported by a CIHR New Investigator Award. The funding source had no role in the study design, collection, analysis or interpretation of the data, writing of the report, or in the decision to publish. Potential competing interests: None.
© 2014 by the American College of Gastroenterology
Disclosure: The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by Institute for Clinical Evaluative Sciences, the Ontario Ministry of Health and Long Term Care and Cancer Care Ontario is intended or should be inferred.
REFERENCES 1. Limketkai BN, Lam-Himlin D, Arnold MA et al. The cutting edge of serrated polyps: a practical guide to approaching and managing serrated colon polyps. Gastrointest Endosc 2013;77:360–75. 2. Longacre TA, Fenoglio-Preiser CM. Mixed hyperplastic adenomatous polyps/serrated adenomas. Am J Surg Pathol 1990;14:524–37. 3. Leggett B, Whitehall V. Role of serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138:2088–100. 4. Torlakovic E, Skovlund E, Snover DC. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003;27:65–81. 5. Urbanski SJ, Kossakowska AE, Marcon N et al. Mixed hyperplastic adenomatous polyps: an underdiagnosed entity. Report of a case of adenocarcinoma arising within a mixed hyperplastic adenomatous polyp. Am J Surg Pathol 1984;8:551–6. 6. Torlakovic E, Snover DC. Serrated adenomatous polyposis in humans. Gastroenterology 1996;110:748–55. 7. Bosman FT, Carneiro F, Hruban RH et al. (Eds.), World Health Organization Classification of Tumours of the Digestive System. IARC Press: Lyon, France, 2010. 8. Goldstein NS, Bhanot P, Odish E et al. Hyperplastic-like colon polyps that preceded microsatellite-unstable adenocarcinomas. Am J Clin Pathol 2003;119:778–96. 9. Singh H, Bay D, Ip S et al. Pathological reassessment of hyperplastic colon polyps in a city-wide pathology practice: implications for polyp surveillance. Gastrointest Endosc 2012;76:1003–8. 10. Torlakovic EE, Gomez JDG, Driman DK et al. Sessile serrated adenoma (SSA) vs. traditional serrated adenoma (TSA). Am J Surg Pathol 2008;32:21–9. 11. Lieberman DA, Rex DK, Winawer SJ et al. 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. 12. Robles SC, Marrett LD, Clarke EA et al. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495–501. 13. Bressler B, Paszat L, Chen Z et al. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology 2007;132:96–102. 14. Khalid O, Radaideh S, Cummings OW et al. Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001. World J Gastroenterol 2009;15:3767–70. 15. Lu FI, Niekerk dW, Owen D et al. Longitudinal outcome study of sessile serrated adenomas of the colorectum: an increased risk for subsequent right-sided colorectal carcinoma. Am J Surg Pathol 2010;34:927–34. 16. Kim SW, Cha JM, Lee JI et al. A significant number of sessile serrated adenomas might not be accurately diagnosed in daily practice. Gut Liver 2010;4:498–502. 17. Lash RH, Genta RM, Schuler CM. Sessile serrated adenomas: prevalence of dysplasia and carcinoma in 2139 patients. J Clin Pathol 2010;63:681–6.
The American Journal of GASTROENTEROLOGY
7
