Accepted Manuscript

Title: Surveillance after positive colonoscopy based on

adenoma characteristics

Authors: Ido Laish, Ilia Seregeev, Timna Naftali, Fred M.

Konikoff

PII: S1590-8658(17)30882-4
DOI: http://dx.doi.org/doi:10.1016/j.dld.2017.05.005
Reference: YDLD 3452

To appear in: Digestive and Liver Disease

Received date: 2-9-2016

Accepted date: 9-5-2017

Please cite this article as: Laish Ido, Seregeev Ilia, Naftali Timna, Konikoff Fred
M.Surveillance after positive colonoscopy based on adenoma characteristics.Digestive
and Liver Disease http://dx.doi.org/10.1016/j.dld.2017.05.005

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 Surveillance after positive colonoscopy based on adenoma
characteristics
Short title: Surveillance after positive colonoscopy
Ido Laish, MD1,2* ido.laish@gmail.com, Ilia Seregeev, MD1, Timna Naftali, MD1,2, Fred M
Konikoff, MD, PhD1,2
1
Gastroenterology and Hepatology Institute, Meir Medical Center, Kfar Saba, Israel
2
Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
*
Corresponding Author at. Gastroenterology and Hepatology Institute, Meir Medical Center,
59 Tschernihovsky St. Kfar Saba 44281 Israel. Tel: +972 (9) 7472523; Fax: +972 (9) 7471320
Abstract

Background: Patients with adenomatous polyps are at increased risk for developing

colorectal cancer based on the characteristics and number of polyps, but less is

known about the individual and combined contribution of these factors. This study

aimed to better characterize the risk of advanced adenoma and cancer in patients

with positive baseline colonoscopy.

Methods: Patients who had polyps at baseline colonoscopy were included in this

retrospective cohort study (N= 1165) and were categorized into 6 groups: 1) 1-2 non-

advanced adenomas (NAA’s), 2) ≥ 3 NAA’s, 3) advanced tubular adenoma, 4) small

tubulovillous adenoma (TVA), 5) large TVA and 6) multiple advanced adenomas

(MAA’s). Findings at surveillance colonoscopy were documented in each group.

Results: The combined incidence of advanced adenoma, ≥ 3 NAA’s, and colorectal

cancer at surveillance colonoscopy was significantly higher in the baseline large TVA

(29.2%) than small TVA groups (13.5%, P< 0.001), as well as in the MAA’s group

(44.1%) compared with large TVA group (P= 0.02). The incidence of colorectal

cancer, however, was not significantly different between the groups.

Conclusions: The size of the polyp and the number of advanced lesions are more

important than its histology for predicting the risk of high-risk metachronous lesions

at follow-up.

Abbreviations
CRC: Colorectal cancer
FAP: Familial adenomatous polyposis
HGD: High grade dysplasia
IMC - Intra-mucosal carcinoma
LGD: Low grade dysplasia
MAA: Multiple advanced adenomas
NAA: Non-advanced adenomas
OR: Odds ratio
TA: Tubular adenoma
TVA: Tubulovillous adenoma

Key words: surveillance colonoscopy, metachronous advanced lesion, colon cancer

Introduction

Patients with adenomatous polyps are at increased risk for developing

colorectal cancer (CRC)(1). Based on the number, size and histology of

adenomatous polyps at baseline colonoscopy, the current US Multisociety

Task Force (USMSTF) guidelines for surveillance ( 2) define two major groups

for surveillance colonoscopy in an average-risk population: (1) low-risk

adenoma , defined as 1-2 tubular adenomas < 10 mm with low-grade

dysplasia (LGD) , and (2) high-risk adenoma, defined as advanced adenoma

(villous histology, high-grade dysplasia (HGD) or > 10 mm), or 3 or more non-

advanced adenomas (NAA). While low-risk adenoma necessitates a follow-up

colonoscopy of 5-10 years, a 3-year interval is recommended for surveillance

for patients with high-risk adenoma. As opposed to the American guidelines,

the European (3) and British guidelines (4) do not include histologic features

in risk stratification of patients with previous adenoma, and propose three

patient classifications (low-, intermediate- and high-risk) based solely on the

number (1-2, 3-4 or ≥ 5) and size (< 10 mm, ≥ 10 mm and < 20mm or ≥

20mm) of polyps. Recommended colonoscopic surveillance intervals are no

surveillance or 5 years for the low-risk group, 3 years for the intermediate-risk

group and 1 year for the high-risk group.

Several follow-up studies of patients with adenomas at baseline

colonoscopy were published recently (5-10). Martinez et al (5) analyzed data

from 8 prospective studies evaluating the risk for advanced neoplasia at 3-5

years in patients with baseline adenomas. Compared with patients with TAs,

those with villous histology had an increased risk of advanced neoplasia but

the level of risk was lower than that for size > 10 mm or for multiple polyps.
Similarly, in a meta-analysis by Saini et al (6), the relative risk for advanced

adenoma at surveillance colonoscopy among patients with baseline

adenomas was slightly increased for villous histology (1.26) but lower

compared to the other two parameters. In contrast, other studies

demonstrated comparable rates (15.5% -16.1%) of advanced adenoma at 5

years for baseline villous histology, large or ≥ 3 small adenomas (7).

However, the available studies do not separately identify patients whose most

advanced polyp is a TVA < 10 mm in size (2). Since villous histology tends to

arise in larger polyps, and size is presumably a greater risk factor for

advanced neoplasia than is villous histology, these two parameters should be

assessed individually in order to more clearly define their relative contributions

to future neoplasia. We hypothesized that the risk of small TVA is more like

small TA than large TVA. The aim of this study was therefore to stratify the

risk for advanced adenoma and carcinoma based on both pathology and

polyp size at baseline colonoscopy, by comparing these 3 groups. In addition,

we compared the risk of one vs. several advanced adenomas, thus also

including the number of polyps in risk stratification.

Patients and methods

This retrospective cohort study was conducted at Meir Medical Center, Kfar

Saba, Israel and another two affiliated gastroenterology clinics in the

surrounding district, which serve a population of about 1 million. The workload

of our hospital and each of the clinics is around 5,000 colonoscopies per year.

The procedures were performed by 15 senior gastroenterologists, all of them

worked both in the hospital and affiliated clinics. The study was approved by
the Institutional Ethics Committee. Data were collected from medical records

from 2005 through 2014. All adult patients above the age of 18 who had a

documented neoplasia at baseline colonoscopy and a surveillance

colonoscopy, with colonoscopy and pathology reports available for evaluation

were included in the study. According to departmental policy, all polyps,

regardless of size or location, were resected during the procedure and sent to

pathology. The interval between the procedures was at least 1 year and less

than 5 years, which is the longest recommended interval for adenomas.

Cases in which the indication for index or surveillance colonoscopy was other

than screening (e.g. iron deficiency anemia, rectal bleeding or abdominal

pain) were also included.

Patients were excluded if they had previous colonoscopies before the study

period; baseline colorectal cancer more invasive than endoscopically-resected

intra-mucosal carcinoma (IMC); serrated adenoma at baseline; patients with >

10 polyps at baseline, suspicious for hereditary polyposis syndromes;

personal history of inflammatory bowel disease, hereditary non-polyposis

colon cancer , familial adenomatous polyposis (FAP), attenuated FAP and

muyth- associated polyposis; personal history of CRC or bowel resection for

other indications. Patients were also excluded if they had poor bowel

preparation as determined by the endoscopist, or incomplete examinations at

either of the two colonoscopies. Patients with piecemeal polypectomy at

baseline colonoscopy were not routinely excluded and a follow-up

colonoscopy was performed within 6 months to assure complete resection.

However, if complete resection could not be assured (no free pathological

margins), patients were excluded.

Data were obtained from the medical records at baseline colonoscopy and

included demographic information (age, gender,); chronic aspirin use;

smoking habits; obesity (defined as body mass index > 30) and diabetes

(based on A1C hemoglobin > 7 percent or the use of anti-diabetic therapy);

family history of colon cancer in first-degree relatives; indication for

colonoscopy; quality of bowel preparation; and details (number, size and

histology) regarding polyps. When more than one polyp was found, the most

advanced (either in size or histology) was used for categorization.

Based on the findings at baseline colonoscopy, the study population was

categorized into 6 groups: (1) 1-2 non-advanced adenomas (NAA’s), (2) ≥ 3

NAA’s, (3) advanced tubular adenoma (TA) (either ≥ 10mm or with HGD

histology), (4) small tubulovillous adenoma (TVA) (< 10 mm) with either LGD

or HGD/IMC, (5) large TVA (≥ 10 mm) with either LGD or HGD/IMC, and (6)

multiple advanced adenomas (MAA's) (more than one advanced lesion either

in size or histology). Polyp characteristics from surveillance colonoscopies

were documented in each group. The outcome was the rate of metachronous

advanced lesions, defined as the combined rate of advanced colorectal

neoplasia (IMC + invasive colon cancer), advanced adenoma and ≥ 3 NAA’s

at surveillance colonoscopy in each of the study groups, with emphasis on the

small TVA group vs. large TVA and 1-2 NAA’s groups.

Assessment of polyp size

The determination of polyp size that was used for polyp stratification, was

based on subjective estimation by the endoscopist, either by the use of open

forceps or unaided visual estimation, and whenever available, also by

objective pathology-based measurements using a ruler. Whenever there was

a discrepancy between the two, we considered the pathologist's measurement

the more reliable. Among 1,165 patients in our study, pathology-based

measurements were available for 734 patients (63%) in whom en-block snare

polypectomy was performed, either for pedunculated or sessile polyps. In the

other 37% of the patients we relied on subjective estimation only: In 233

patients (20%) the polypectomy was achieved with forceps due to their small

size, while 198 patients (17%) underwent piecemeal polypectomy, of whom

116 (10%) had polyp size ≥ 15mm and 82 (7%) a polyp size < 15mm.

Colonoscopy procedure

Bowel preparation included one of three liquid preparation solutions

(polyethylene glycol [PEG], sodium phosphate or sodium pico-sulfate), oral

laxatives and enemas. The cleansing level was graded based on a validated

4-level scale. An excellent, good or fair cleansing level was defined as

adequate preparation, while patients with poor cleansing level were excluded.

Method of polyp resection was selected at the discretion of the endoscopist,

including cold and hot forceps or snares. As a rule, snare polypectomy was

performed for polyp ≥ 5 mm.

Recommended surveillance interval

Recommendations for post-polypectomy surveillance interval agreed with the

accepted AGA guidelines (2), i.e. 3 years for advanced adenoma or ≥3 NAA,

5 years for 1-2 NAA and 10 years when no adenoma was detected. However,

when more than 5 adenomas were detected at first colonoscopy, patients

were followed-up in 1 to 3 years, at the preference of the endoscopist. If

pathological free margins could not be assured after polyp resection, including
piecemeal polypectomy, a "clearing" colonoscopy was performed within 6

months period.

The adenoma-detection rate at screening colonoscopy, which was performed

as a first level assessment for average-risk population, was confirmed to be

25% in our institution between 2005 and 2013.

Statistical analysis

Results were expressed as frequencies and percentage for categorical data

and mean± standard deviation for continuous parameters. Differences

between two groups were analyzed by Chi-Square test or Fisher's exact test

for non-metric variables, each when appropriate. Odds ratio (OR) and 95%

confidence Interval (CI) were calculated. For more than two groups

comparisons, differences were calculated by One-Way ANOVA or Kruskal

Wallis non-parametric test with Bonferroni post –hoc comparisons, each as

fitted. Multivariate analysis was performed by entering all potential factors that

are known to affect the development of adenoma into a backwards logistic

regression model.

A P value <0.05 was considered statistically significant. All statistical analyses

were done with SPSS-23 software (IBM, NY)

Results

After ineligible patients were excluded, 1,185 patients who appeared to satisfy

the inclusion criteria were evaluated (Figure 1). Of these, 20 were excluded

because of incomplete data for the baseline or surveillance colonoscopy,

resulting in 1,165 patients in the final study population. The mean age (± SD)
of the patients was 62.4 ± 9.6 years, and the cohort consisted of 632 men

(54.2%) and 533 woman (45.8%).

All patients were categorized into the 6 above-mentioned groups, based on

findings at baseline colonoscopy. Size of groups, demographics, indications

for colonoscopy and background diseases at baseline colonoscopy are

displayed in Table 1.

In univariate analysis, background factors that were significantly associated

with advanced adenoma or metachronous advanced lesions were advanced

age (p<0.001) and the use of aspirin (p<0.001), while BMI ≥ 30 (p= 0.04),

family history of CRC (p=0.06) and diabetes (p=0.06) were marginally

associated. In multivariate analysis, only advanced age (OR 1.02, 95% CI 1.0-

1.03; p value – 0.01) and the use of aspirin (OR 1.44, 95% CI 1.07-1.94; p

value – 0.01) were associated with advanced adenoma / lesion.

Findings at surveillance colonoscopy

The indications for surveillance colonoscopy were surveillance only (71%),

rectal / occult bleeding (10%), abdominal pain / change in bowel habits (14%)

and iron deficiency (5%).

Findings on surveillance colonoscopy are depicted in Table 2 and multivariate

analysis of baseline groups for their association with the development of

metachronous advanced adenoma or combined advanced lesions, is shown

in Table 3. The rate of advanced polyp at surveillance colonoscopy was

higher in the large TVA group (23.6%) compared to the small TVA group

(8.5%) and to the 1-2 NAA’s group (12.3%) (Adjusted OR’s 3.3 and 2.1,

respectively, p value < 0.001). In contrast, the rate of multiple NAA’s at

surveillance colonoscopy was not significantly different in the large TVA group
compared to the other two groups (P = 0.6 for both) (Table 2). The combined

rate of metachronous advanced lesions, the primary outcome of this study,

was 29.2% in the large TVA group. This was significantly higher compared to

the small TVA group (13.5%) and the 1-2 NAA’s group (16.8%) (Adjusted

OR’s 2.6 and 1.9, respectively, p value < 0.001). The difference between the

small TVA and 1-2 NAA groups was not significant (P = 0.16) (Table 3).

In order to evaluate the significance of high-grade dysplasia histology, we

compared the rate of metachronous advanced lesions between those with

LGD and HGD within the small TVA and large TVA groups (Figure 2).

Although this rate was as expected higher in the small TVA-HGD group

compared to small TVA-LGD group, and paradoxically higher in the large

TVA-LGD group compared to the large TVA-HGD group, the differences were

not significant (P= 0.6 and 0.2, respectively). While the polyp size was the

same between small TVA-LGD and HGD and between large TVA-LGD and

HGD (data not shown), there was a difference in the colonoscopy interval

within each group (3.2 ± 0.9 vs. 2.8 ± 1.3 years, respectively, P = 0.05; 3.2 ±

0.9 vs. 2.6 ± 1.2 years, respectively, P = 0.0001).

Within the advanced TA group, 1/9 patients (11.1%) with baseline small

TA-HGD developed advanced adenoma, compared to 17/71 patients (23.9%)

with large TA. If those 9 patients would have been added to the 1-2 NAA

group, yielding a group consisting of all small TA (both low- and high- grade

dysplasia), the incidence of metachronous advanced adenoma would still

have been 12.3% (59/479).

Compared with the single large TVA group, when MAA's were detected at

baseline colonoscopy, the incidences of multiple polyps or advanced

adenoma at surveillance colonoscopy were non-significantly higher (P = 0.13

and 0.09, respectively), while that of combined metachronous advanced

lesion was significant in multivariate analysis (44.1% vs. 29.2%; OR 1.7, 95%

CI: 1.0-2.8; P = 0.02). While the colonoscopy interval was the same in both

groups, approximately 3 years, the mean size of the largest polyp was

significantly smaller in patients with MAA's (P < 0.0001).

Incidence of advanced colorectal neoplasia

A total of 19 patients (1.6%) were found to have advanced colorectal

neoplasia (IMC + invasive colon cancer) at second surveillance colonoscopy

(interval cancer) (Table 2). Its incidence was 1.3%, 1.5%, 1.1% and 4.3% in

the 1-2 NAA, small TVA and large TVA groups and the MAA's group,

respectively. This latter risk was not significantly higher when compared to

each of other the groups (P > 0.2) or to the entire rest of the cohort (P = 0.1).

However, among patients who developed interval cancer, invasive tumors

(beyond T1) were detected in 40% and 33% in the 1-2 NAA and small TVA

groups, respectively, while it was found in all cancer patients (100%) in the

groups with larger, advanced polyps.

Discussion

Although several studies have been conducted in recent years to address the

yield of surveillance colonoscopy in patients with adenomas at baseline

colonoscopy, there are still some areas of uncertainty in which

recommendations by the various professional societies differ. One of them is

the recommended time interval for resected adenoma < 10 mm with villous
histology or high-grade dysplasia. While the USMSTF (2) and European

Society of Gastrointestinal Endoscopy (ESGE) (11) guidelines endorse a 3-

year interval, the European guidelines (3) consider it only as an "optional"

criterion for surveillance in 3 years. This study is the first to stratify the risk for

metachronous advanced lesions based on both pathology and polyp size at

baseline colonoscopy, by directly comparing small (< 10 mm) and large (≥ 10

mm) TVAs. We found a significant difference in the incidence of

metachronous advanced lesions in these two groups at surveillance

colonoscopy. The findings that this incidence was comparable between small

TA and small TVA (13.5-16.8%), as well as in all large polyps [TA (21/71,

29.5%) and TVA (29.2%)], support the notion that the size of the polyp is

more important than villous histology as a risk factor for future advanced

lesions. Similarly, we found that the presence of HGD histology at baseline

small TA and TVA groups did not significantly alter the risk for further

advanced lesions. This is in accordance with Martinez et al.(6), who did not

find HGD to be independently associated with an increased risk of advanced

neoplasia after adjustments for size and histology (OR 1.05; 95% CI: 0.81-

1.35). Thus, more liberal recommendations for surveillance intervals should

be considered when a small TVA is found (e.g 5 years), in accordance with

the European guidelines.

Another interesting finding of this study is the significant combined risk of

metachronous advanced lesions in patients with MAA's relative to a single

large TVA after adjusting for differences in background diseases. The

combination of multiplicity of polyps and either larger size or advanced

histology probably poses these patients at higher risk level than patients with
a single (size, histology) or the two risk factors. The smaller average polyp

size in this group further underscored the notion that multiplicity of polyps is a

considerable risk factor for further advanced lesions. Thus, patients with

MAA's should probably be followed-up more intensely than patients with

multiple NAA’s or a single advanced lesion. In any case, part of this group is

already categorized as "high-risk" by the European guidelines (11) (patients

with ≥ 5 adenomas or polyps >20 mm) and these patients are recommended

to have a clearing colonoscopy in 1 year.

The aims of screening and surveillance colonoscopy are to reduce

colorectal cancer incidence and mortality. In a pooled analysis of 9,167

patients with previous adenomas, Martinez et al. reported a 0.6% interval

cancer rate after a mean follow-up period of 4 years (5). However, risk of

invasive cancer was 1.2% in patients who had larger (20 mm or greater)

baseline adenomas and 1.3% in those who had prior lesions with high-grade

dysplasia. In another French population-based study (9), the rate of CRC at

7.7 year follow up was 0.8% with baseline low-risk adenoma and 2.8% if

patients harbored previous high-risk adenoma. In our cohort, the rate of

advanced colorectal neoplasia, which includes IMC and invasive cancer, was

of the same magnitude for small and large TVAs (1.1-1.5%) and non-

significantly higher in the group with MAA's when compared separately to

each group or to the rest of the cohort. However, all cancer cases that

occurred in the MAA group, as well as in all other large polyps groups, were

invasive, while most cases in the smaller polyp groups were IMC. This further

supports the need to consider shortening the surveillance interval for this

high-risk group.
Among background diseases, the most dominant factor that was associated

with advanced lesions was aspirin use. Although it seems paradoxical, due to

the well-known preventive effect of aspirin on the development of polyps (12),

this is probably a bias since aspirin is used more frequently in older patients

with metabolic derangements, which are positively associated with polyps and

cancer.

Limitations of the study include a retrospective design with its inherent

limitations and a relatively small cohort. Second, the study was performed in a

single referral center and its affiliated clinics; thus, findings may not be

completely generalizable. Third, although all the procedures were performed

by a small group of senior gastroenterologists, which controls for

heterogeneity at the performance level, we do not have, and therefore did not

adjust for, the endoscopist-related quality measures (e.g. adenoma detection

rate, number of previous colonoscopies).

Another potential limitation in this study, which is in common with previous

studies (5-9) as well, is the method of measurement of polyp size, which was

either pathology-based measurement or estimation by the endoscopist, each

has its own limitations. Because of the two-dimensional view, estimating polyp

size by endoscopy has quite high inter-observer variability. The accuracy has

been questioned by several reports, suggesting overestimation of polyp size

compared to pathology measurements (13-14). In a recent study (15), 46% of

polyps estimated as ≥ 1 cm on endoscopy, were actually 1 cm on pathology,

while only 3.9% of polyps estimated at < 1 cm on endoscopy were actually

under-called. The most significant factor influencing overcall rates was flat

polyp configuration. Other small studies (16) reported opposite results, i.e.
underestimation by endoscopists. Despite this limitation, in clinical practice

(as well as in our study), the endoscopic assessment is the most frequent and

accepted method for determination of polyp size and for surveillance

recommendations. Although pathology-based measurement is probably more

accurate, piecemeal polypectomy of sessile polyps (which occurred in 17% of

our study population), forceps polypectomy of very small polyps (20% of the

study population) and fragmentation during polyp retrieval are the main

reasons for unavailability of pathology measurements. Polyps around 1cm

size are the most challenging for subjective estimation, especially being the

cut-off for advanced lesions, and therefore are naturally subjects for over- and

under-estimation. In our study, only 7% of the patients had polyps 5-15mm

with no pathology-based measurement, after excluding diminutive polyps and

piecemeal polypectomy of polyps ≥ 15 mm. Thus, although we did not

quantify the amount of discrepancies between the two methods (in terms of

over- and under- estimation), this proportion is too small to significantly impact

the results. Nonetheless, even when measuring an ex situ polyp, inaccuracy

may still occur due to potential shrinkage when electrocauterization is used for

polypectomy or from formalin fixation. However, most studies have

demonstrated no significant difference between fresh (post-excision) and fixed

polyp measurements (13,16-17).

In addition to clinical recommendations for the surveillance intervals based

on the individual risk of developing metachronous lesions, the question is

whether there is a significant impact to the medical system of changing

surveillance practices in these groups. Although it was shown that an

advanced histology is found in only about 10% of small adenomas (i.e about
 2-3% of total colonoscopies) (18), and the potential impact of an alternative

surveillance recommendation is expected to be low in terms of colon cancer

burden or resource demand (19), the small TVA group in our study comprised

17.1% (199/1165) of the study cohort. This poses a potentially significant

burden. The crude estimated annual rate of colonoscopies in Israel is

250,000. Assuming an ADR of 25% of which 17% are small TVA polyps

(according to our study), this yields a total of 10,625 annual colonoscopies

which can be put off to 5-year interval. Taking a lower estimate of small

polyps with advanced pathology of 2-3% of total colonoscopies (19), still

yields 7500 annual colonoscopies to be postponed and decrease the burden

of colonoscopies. Thus, optimization of their surveillance interval from 3 to 5

years is not only justified but also beneficial, and counterbalances the

apparent need for more intense surveillance whenever MAA's are found at

baseline.

In conclusion, we found in this study that while size and multiplicity appear to

be dominant risk factors, histology by itself appears to be less significant. Yet,

additional large-scale prospective surveillance colonoscopy studies should be

conducted in order to clearly define the risk estimates for these lesions.

Supported in part by the Josefina Maus and Gabriela Cesarman Chair for Research
in Liver Diseases, the Tel Aviv University

Author contribution
IL – study concept and design; acquisition of data; analysis and interpretation of
data; drafting of the manuscript
IS – acquisition of data
TN - critical revision of the manuscript for important intellectual content
FMK - study concept and design; critical revision of the manuscript for important
intellectual content

Potential competing interests: none

Acknowledgments

We thank Nava Jellin, MA for performing the statistical analysis.

References

[1] Zauber AG, Winawer SJ, O’Brien MJ, et al. Colonoscopic polypectomy and long-term prevention
of colorectal-cancer deaths. N Engl J Med 2012;366: 687–96.

[2] 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.

[3] Atkin WS, Valori R, Kuipers EJ, et al. European guidelines for quality assurance in colorectal
cancer screening and diagnosis. First Edition--Colonoscopic surveillance following adenoma removal.
Endoscopy. 2012;44:SE151-63.

[4] Cairns SR, Scholefield JH, Steele RJ, et al. British Society of Gastroenterology: Association of
Coloproctology for Great Britain and Ireland. Guidelines for colorectal cancer screening and
surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666-90. .

[5] Martinez ME, Baron JA, Lieberman DA, et al. A pooled analysis of advanced colorectal neoplasia
diagnoses following colonoscopic polypectomy. Gastroenterol 2009;136:832-41. .

[6] Saini SD, Kim HM, Schoenfeld P. Incidence of advanced adenomas at surveillance colonoscopy in
patients with a personal history of colon adenomas: a meta-analysis and systematic review.
Gatrointest Endosc 2006;64:614-26. .

[7] Lieberman DA, Weiss DG, Harford WV, et al. Five-year colon surveillance after screening
colonoscopy. Gastroenterology 2007;133:1077-85. .

[8] Miller HL, Mukherjee R, Tian J, et al. Colonoscopy surveillance after polypectomy may be
extended beyond five years. J Clin Gastroenterol 2010; 44:e162-e6.

[9] Cottet V, Jooste V, Fournel I, et al. Long-term risk of colorectal cancer after adenoma removal: a
population-based cohort study. Gut 2012;61:1180-86.
[10] Chung SJ, Kim YS, Yang SY, et al. Five-year risk for advanced colorectal neoplasia after initial
colonoscopy according to the baseline risk stratification: a prospective study in 2452 asymptomatic
Koreans. Gut 2011;60:1537-43. .

[11] Hassan C, Quintero E, Dumonceau JM, et al. Post-polypectomy colonoscopy surveillance:

European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy 2013;45:842-51. .

[12] Rothwell PM, Wilson M, Elwin CE, et al. Long-term effect of aspirin on colorectal cancer
incidence and mortality: 20-year follow-up of five randomised trials. Lancet 2010;376:1741–50.

[13] Morales TG, Sampliner RE, Garewal HS, et al. The difference in colon polyp size before and after
removal. Gastrointest Endosc 1996;43:25-8.

[14] Eichenseer PJ, Dhanekula R, Jakate S, et al. Endoscopic mis-sizing of polyps changes colorectal
cancer surveillance recommendations. Dis Colon Rectum 2013;56:315-21.

[15] Anderson BW, Smyrk TC, Anderson KS, et al. Endoscopic overestimation of colorectal polyp size.
Gastrointest Endosc 2016;83:201-8.

[16] Moug SJ, Vernall N, Saldanha J, et al. Endoscopists' estimation of size should not determine
surveillance of colonic polyps. Colorectal Dis 2010;12:646-50.

[17] Schoen RE, Gerber LD, Margulies C. The pathologic measurement of polyp size is preferable to
the endoscopic estimate. Gastrointest Endosc 1997;46:492-6.

[18] Gupta S, Jacobs ET, Baron JA, et al. Risk stratification of individuals with low-risk colorectal
adenomas using clinical characteristics: a pooled analysis. Gut 2015;11:2015-310196.

[19] Ladabaum U, Schoen RE. Post-polypectomy surveillance that would please goldilocks-not too
much, not too little, but just right. Gastroenterol 2016;150:791-6. .
Figure Captions
Figure 1: Flow chart of patient selection in the study
Figure 2: Rates of metachronous advanced lesions at surveillance colonoscopy in
the small TVA and large TVA groups.
TVA - tubulovillous adenoma; HGD – high grade dysplasia; LGD - low grade dysplasia
Tables
Table 1: Background demographics and clinical characteristics of the study groups at
baseline colonoscopy (N=1,165)
1-2 NAA's Advanced ≥3 Small TVA Large TVA MAA's
TA NAA's
N= 470 N= 80 N= 199 N= 233 N= 109
(40.4%) (6.8%) N= 74 (17%) (20%) (9.4%)
Characteristics
(6.4%)
Mean age ± SD, 60.7 ± 62.4 ± 64.9 ± 63.2 ± 62.9 ± 64.6 ± 9.1
years 9.5 10.3 8.5 8.4 10.3
Male sex (%) 235 (50) 48 (60) 52 107 121 (51.9) 69 (63.3)
(70.3) (53.8)
Family history of 190 23 (28.8) 23 64 (32.2) 71 (30.5) 36 (33)
colorectal cancer (40.4) (31.1)
(%)
Indications (%)
No symptoms 252 33 (41.2) 44 101 76 (32.6) 45 (41.3)
(53.6) (59.5) (50.8)
Rectal 52 (11.1) 11 (13.8) 8 (10.8) 24 (12.1) 45 (19.3) 15 (13.8)
bleeding
 Abdominal 25 (5.3) 6 (7.5) 5 (6.8) 13 (6.5) 13 (5.6) 8 (7.3)
pain
Change in 79 (16.8) 13 (16.3) 6 (8.1) 22 (11) 24 (10.3) 15 (13.7)
bowel habits
Weight loss 3 (0.6) 0 1 (1.4) 1 (0.5) 2 (0.9) 1 (0.9)
Iron 15 (3.2) 5 (6.3) 3 (4.1) 15 (7.5) 14 (6) 9 (8.3)
deficiency
anemia
Occult blood 44 (9.4) 12 (15) 7 (9.5) 23 (11.6) 59 (25.3) 16 (14.7)
Aspirin use (%) 137 32 (40) 28 76 (38.2) 85 (38.5) 50 (45.9)
(29.1) (37.8)
Diabetes (%) 72 (15.3) 16 (20) 25 40 (20.1) 43 (18.5) 37 (33.9)
(33.8)
Smoking habits 103 20 (25) 18 47 (23.6) 51 (21.9) 22 (20.1)
(%) (21.9) (24.3)
BMI ≥ 30 (%) 123 24 (30) 27 55 (27.6) 60 (25.8) 34 (31.2)
(26.2) (36.5)
Preparation (%)
Excellent/ 354 62 (77.5) 51 155 188 (80.6) 83 (76.1)
Good (75.3) (68.9) (77.8)
Reasonable 116 18 (22.5) 23 44 (22.2) 45 (19.4) 26 (23.9)
(24.7) (31.1)
Mean polyp size 4.1 ± 2.0 15.7 5.5 ± 6.2 ± 2.0 18.8 ± 7.3 14.4 ± 8.1
± SD (median), (3) ±5.1 (14) 3.2 (5) (6) (16) (13)
mm

SD – standard deviation; BMI – body mass index; NAA – non advanced adenomas; TA-

tubular adenoma; TVA- tubulovillous adenoma; MAA multiple advanced adenomas

Table 2: Findings at surveillance colonoscopy
1-2 NAA's Advanced ≥3 Small TVA Large TVA MAA's
TA NAA's
Findings
1-2 NAA's (%) 85 (18.1) 15 (19) 11 (14.9) 48 (24.1) 34 (14.6) 25 (22.9)
≥ 3 NAA's (%) 16 (3.4) 1 (1.2) 5 (6.7) 7 (3.5) 10 (4.3) 9 (8.3)
Advanced polyp 58 (12.3) 18 (22.5) 16 (21.6) 17 (8.5) 55 (23.6) 35 (32.1)
(%)
Advanced 5 (1.1) 2 (2.4) 2 (2.7) 3 (1.5) 3 (1.3) 4 (3.7)
colorectal
neoplasia (%)
IMC 3 1 1 2 0 0
Invasive 2 1 1 1 3 4
Combined 79 (16.8) 21 (26.1) 23 (31) 27 (13.5) 68 (29.2) 48 (44.1)
advanced lesion
(%)
Good 395 (84) 67 (83.8) 52 (70.3) 163 (81.9) 195 (83.7) 84 (77.1)
preparation (%)
1st-2nd 4.3 ± 0.9 2.9 ± 1.0 3.3 ± 1.0 3.2 ± 0.9 3.0 ± 1.0 2.8 ± 0.9
colonoscopy
interval ± SD,
years

SD – standard deviation; NAA – non advanced adenoma; TA - tubular adenoma; TVA -

tubulovillous adenoma; MAA - multiple advanced adenomas; IMC- intra-mucosal carcinoma

 Table 3: Multivariate analysis of baseline colonoscopy groups for their association with the
development of metachronous advanced adenoma or combined advanced lesion. The reference
group for calculation of odds ratios was 1-2 NAA’s.
Advanced adenoma Combined advanced lesion
Findings
Groups at Odds 95% P value Odds 95% P value
baseline ratio confidence ratio confidence
colonoscopy interval interval
(lower- (lower-
upper) upper)
1-2 NAA's 1 - - 1 - -
Advanced TA 2.03 1.12 - 3.69 0.01 1.66 0.94 - 2.91 0.07

≥ 3 NAA's 2.32 1.63 – 3.54 0.000 2.23 1.51 - 2.95 0.0001

Small TVA 0.63 0.36 – 1.12 0.12 0.71 0.44 - 1.15 0.16

Large TVA 2.11 1.40 – 3.19 0.000 1.91 1.31 - 2.78 0.001

MAA's 3.11 1.90 – 5.09 0.000 3.39 2.14 - 5.35 0.000

NAA – non advanced adenoma; TA - tubular adenoma; TVA - tubulovillous adenoma; MAA -
multiple advanced adenomas