• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

Evidence table for review question 3When should colonoscopic surveillance be started and what should be the frequency of surveillance?

Study IDStudy designFollow upPopulationPrognostic factors or surveillance programmesOutcomesComments
Kronborg et al. (2006)Randomised surveillance study.

The groups were compared for patient characteristics.

Size was measured immediately after polypectomy

Years of observation were calculated from the first polypectomy to the most recently performed surveillance, or to censoring because of death, refusal to undergo surveillance, or emigration. Proportions were compared as relative risks (RR) with 95% confidence intervals. RR was calculated as the risk in the group with the longest interval of surveillance.
10 yearsBetween 1981 and 1991 a total of 673 patients (382 men, 291 women; age, 28-77 years) with newly diagnosed adenomas were allocated at random to either 24 months (group A) or 48 months years (group B) between surveillance examinations.

From 1981 to 1987, 73 patients with flat and sessile adenomas (more than 5 mm in diameter) and villous adenomas were randomly allocated to either intervals of 6 months (group C) or 12 months (group D) between examinations during the first 5 years and then every year in all.

Finally, 200 patients with similar adenomas to those in groups C and D were randomised to intervals of 12 months (group E) or 24 months (group F), the intake being from 1988 to 2000.

Patients were excluded if colorectal cancer (CRC) was detected at the initial examination, or if they had a history of previous colorectal neoplasia (carcinoma or adenoma), familial adenomatous polyposis (FAP) or hereditary non-polyposis colorectal cancer (HNPCC).
Colonoscopic surveillance:
group A = 2 years, group B = 4 years, group C = 6 months, group D = 12 months, E= 12 months and F= 24 months, between surveillance examinations.

Different surveillance intervals, 6, 12, 24 months.

Double-contrast barium enema (DCBE) was added if colonoscopy was incomplete. In patients with multiple polyps or unsatisfactory bowel preparation, colonoscopy was repeated within 3 months. Surveillance examinations were done mainly by colonoscopy, but DCBE was used if the patient refused colonoscopy. If a surveillance examination was done more than 3 months after the date planned, the examination was considered ‘in between’.

Patients without complete colonoscopy and less than optimal compliance were kept in the study
Colorectal neoplasia and adenoma detection
B versus A
After the first follow-up period (24 months in A and 48 months in B) fewer patients had adenomas detected in group A than in group B but it was not statistically significant (58 of 292 versus 64 of 232; RR = 0.7, 95% CI 0.5 to1.0), and the number of patients with significant neoplasia did not differ (10 of 292 versus 13 of 232; RR = −0.6, 95% CI 0.3 to 1.4). Overall, adenomas were detected in a smaller proportion of surveillance examinations in group A than in group B (123 of 684 versus 83 of 300; RR = 0.7, 95% CI 0.5 to 0.8). The same was true of significant new neoplasia (18 of 684 versus 17 of 300; RR = 0.5, 95% CI 0.2 to -0.9).

In group A the total number of patients having new adenomas and new significant neoplasia was 95 and 16, respectively. In group B the figures were 77 and 17, respectively.

New adenomas tended to be detected more often in group A, but advanced new adenomas appeared equally as frequently in groups A and B. Overall, larger size contributed mainly to the advanced state (19 and 21 patients), whereas severe dysplasia and villousness was seen in 3 patients in both arms. However, CRC was diagnosed significantly more often in group B.
D versus C
The number of patients was limited, but the cumulative number of surveillance years was 10 years on average in both groups.
Advanced new adenomas tended to be more frequent in the D group (p = 0.08), but the one case of cancer was detected in group C at a planned examination 6 months after a ‘clean colon’. The cancer was at an early stage and the patient developed another early CRC more than 5 years later. Nearly all new adenomas were at an advanced stage because of large size alone.
F versus E
The two groups were similar initially and the average time of surveillance was 5 years. The number of colonoscopies was nearly twice as high in group E, but the number of new adenomas regardless of state was similar. There was no significant difference in risk of CRC but the two cancers in group E were both early stage, one being detected 12 months after a ‘clean colon’ (a mucinous tumour), the other, 57 months after a ‘clean colon’ and the patient's refusal to undergo further examinations. In group F the cancers were more advanced. Three of the four patients had a ‘clean colon’ 24 months before the CRC was detected during a planned examination, but one had many recurrences at the site of the original large sessile adenoma in the rectum, before the cancer was detected (Dukes' B).
The age, sex, and polyp characteristics of the patients were distributed evenly in the two groups.

The study was randomised by random numbers but no details of concealment or blinding of pathologists is mentioned.

Advanced adenomas were defined as those with severe dysplasia or being at least 10 mm in diameter or villous.
Relative risks of new adenomas and carcinomas during surveillance with 95% CI
B versus AD versus CF versus E
New adenomas0.88 (0.69 to 1.12)0.82 (0.43 to 1.52)0.88 (0.57 to 1.34)
Advanced new adenomas1.15 (0.61 to 2.15)3.12 (0.87 to 14.50)*0.97 (0.40 to 2.35)
Colorectal carcinomas6.22 (1.06 to 117, 48)**-1.93 (0.38-13.94)
 *p = 0.08; **p = 0.04
Adapted from table V in Kronborg (2006)

Adverse events
B versus A
Seven complications to colonoscopy were minor and treated without surgery, six during surveillance. The perforations occurred during surveillance in each of the two groups and were treated successfully with suture alone. A perforation during initial colonoscopy in group A proved fatal, the patient dying of septicemia after inadequate closure of a temporary colostomy. A: two diagnostic perforations and two therapeutic perforations and B: one diagnostic perforation and one polypectomy syndrome.
D versus C
Two severe complications (1 diagnostic perforation and 1 polypectomy syndrome) were seen in the C group, but both patients fully recovered. No severe complications were found in group D.
F versus E
Two colonoscopic perforations were seen, both patients fully recovered after surgery (one diagnostic perforation in each group).
Lieberman et al. (2007)Patients with cancer or adenomas with high-grade dysplasia had follow-up based on clinician decisions.

501 participants with no neoplasia at baseline were matched by age to patients with adenomas ≥10 mm and assigned to surveillance at 5 years.
5.5 yearsParticipants were enrolled in 13 Veterans Affairs Medical Centres between February 1994 and January 1997. 24 centres were selected to achieve geographic and racial diversity.

Among patients who met the eligibility criteria, 1463 (31.4%) declined to participate, 3196 eligible patients were enrolled, and 3121 had complete colonoscopy examinations to the caecum.
Surveillance intervals of 2 or 5 years and adenoma detection in groups based on index colonoscopy results: according to the following hierarchy: no neoplasia, hyperplastic polyp, 1 or 2 tubular adenomas <10 mm, 3 or more tubular adenomas <10 mm, tubular adenoma ≥10 mm, adenoma with villous histology (25% or more), adenoma with high-grade dysplasia, invasive cancer.1171 patients with neoplasia and 501 with no neoplasia at baseline were scheduled to have at least 1 follow-up colonoscopy within 5.5 years.

Neoplasia detection
The relative risk in patients with baseline neoplasia was 1.92 (95% CI 0.83 to 4.42) with 1 or 2 tubular adenomas <10 mm, 5.01 (95% CI 2.10 to11.96) with 3 or more tubular adenomas <10 mm, 6.40 (95% CI 2.74 to14.94) with tubular adenomas >10 mm, 6.05 (95% CI 2.48 to14.71) for villous adenomas, and 6.87 (95% CI 2.61 to18.07) for adenomas with high-grade dysplasia.

The most serious outcome was the finding of invasive cancer or high-grade dysplasia. The rates of interval high-grade dysplasia or cancer per 1000 person-years of follow-up. The risk of high-grade dysplasia or cancer per 1000 person-years of follow-up was 0.7 with no neoplasia at baseline, 1.5 with tubular adenomas <10 mm, 6.4 with large tubular adenomas (>10 mm), 6.2 with villous adenomas, 26.0 with high-grade dysplasia.
All pathology was reviewed locally and sent for blinded central pathology review. When there was a discrepancy, a third referee pathologist reviewed the material.

The authors compared demographic factors (age, race) and possible risk factors for advanced neoplasia (family history, smoking, use of non-steroidal anti-inflammatory drugs) to determine whether the surveillance cohort was similar to patients who did not receive surveillance. In the neoplasia group, the rate of active smoking was higher in patients who had no surveillance compared with those with surveillance (33.8% vs 21.7%, respectively, (p < 0.001). There were no significant differences in the control group.
Lieberman et al. (2008)During the study period, the Clinical Outcomes Research Initiative repository (CORI) consortium included 65 practice sites in 25 states.

Ten sites contributed more than 500 reports, 6 sites contributed 100–500 reports, and 1 site contributed less than 100 reports.
Retrospective, registryPatients were asymptomatic adults receiving colonoscopy for screening during 2005 from 17 practice sites, which provide both colonoscopy and pathology reports to the Clinical Outcomes Research Initiative repository. Patients were included in this analysis if they were over age 20 years undergoing screening with no symptoms of lower gastrointestinal pathology.Colonoscopic surveillance for polyps less than 10 mm.

Size of polyp and location of polyp's association with advanced histology.
Three asymptomatic groups were included: average risk, family history of CRC or adenoma, and patients receiving colonoscopy for a positive faecal occult blood test or polyp found at screening sigmoidoscopy. Patients were stratified by indication group.

Among 13,992 asymptomatic patients who had screening colonoscopy, 6360 patients (45%) had polyps, with complete histology available in 5977 (94%) patients.

Advanced histology
The proportion with advanced histology (defined as an adenoma with villous or serrated histology, high-grade dysplasia, or an invasive cancer) was 1.7% in the 1 to 5 mm group, 6.6% in the 6 to 9 mm group, 30.6% in the greater than 10 mm group.

Distal location
Distal location was associated with advanced histology in the 6 to 9 mm group (p = 0.04) and in the greater than 10-mm group (p = 0.002).
Sensitivity analysis was done to determine how misclassification of polyp size would impact the outcome. The analysis assumed that polyps were either overestimated in size by 1 mm (for example, a 10 mm polyp is reclassified as 9 mm) or underestimated (a 9 mm polyp is reclassified as 10 mm).
Advanced histology was defined as an adenoma with villous or serrated histology, high-grade dysplasia, or an invasive cancer.
The risk factors compared were age, sex, race, indication for colonoscopy (that were similar) and location of largest polyp
Lund et al. (2001)RCT to investigate whether regular endoscopic surveillance and polypectomy would decrease the incidence of colorectal cancer and to determine if identification of low - and high-risk groups would allow less frequent surveillance in the low-risk group.Total person years follow up was 5148 yearsIncluded if undergoing colonoscopy for: (i) colorectal symptoms, including rectal bleeding; (ii) possible polyp or other incidental findings on barium enema; or (iii) investigation of positive faecal occult bloods.Those found to have colonic adenomas between June 1984 and January 1995 were considered for recruitment to one of six surveillance strategies involving either colonoscopy every 2 or five years or flexible sigmoidoscopy every year, every 2 years, or every 5 years.NOTE: reported only those outcomes related to interval of surveillance for colonoscopy (other outcomes either included in the Saini 2006 review or not relevant for this question)

Early termination because of low rates of adenoma recurrence meant that the trial was underpowered to detect differences in the effect of the various surveillance intervals. However, the authors reported that ‘follow up endoscopy for colonic adenomas can be reduced safely to five yearly intervals for the vast majority of patients (excluding patients with hereditary non-polyposis colorectal cancer and familial adenomatous polyposis)’.
Significant limitations because of early termination and lack of power.
Martinez et al. (2009)Pooled analysis of eight North American studies (six were randomised controlled trials).

Schatzkin et al. (2000); Baron et al. (1999, 2003); Winawer et al. (1993b); Alberts et al. (2000, 2005); Greenberg et al. (1994); Lieberman et al. (2000)
Median follow-up period of 47.2 monthsIndividual patients:
included people at average with a first-time diagnosis of adenomatous polyps.

Study inclusion studies: (1) 800 or more study participants; (2) complete baseline colonoscopy with removal of one or more adenomas and removal of all visualised lesions; (3) a specified schedule of surveillance follow-up; (4) end point data regarding the number, size, and histopathology of adenomas and colorectal cancers detected.
Determining the actual risk of developing advanced adenomas and cancer after polypectomy or the factors that determine risk.Advanced colorectal neoplasia was diagnosed in 1082 (11.8%) of the patients, 58 of whom (0.6%) had invasive cancer.

Definitions for adenomas were as follows: tubular ≤25% villous component), tubulovillous (26–75% villous component), or villous (>75% villous component). They considered advanced adenomas to be those that had one or more of the following features: 10 mm in diameter or larger, presence of high-grade dysplasia, or greater than 25% villous features (also classified as tubulovillous or villous histology). They then combined advanced adenomas and invasive cancer into an end point of advanced colorectal neoplasia or metachronous advanced neoplasia.

Risk factors for advanced metachronous adenomas
Risk of a metachronous advanced adenoma was higher among patients with 5 or more baseline adenomas (24.1%; standard error, 2.2) and those with an adenoma 20 mm in size or greater (19.3%; standard error, 1.5). Risk factor patterns were similar for advanced adenomas and invasive cancer.

Risk factors for metachronous advanced neoplasia
Multivariate analyses: older age (p < 0.0001 for trend) and male sex (odds ratio [OR], 1.40; 95% confidence interval [CI], 1.19 to 1.65) were significantly associated with an increased risk for metachronous advanced neoplasia, as were the number and size of previous adenomas (p < .0001 for trend), the presence of villous features (OR, 1.28; 95% CI, 1.07 to 1.52), and proximal location (OR, 1.68; 95% CI, 1.43 to 1.98). High-grade dysplasia was not associated independently with metachronous advanced neoplasia after adjustment for other adenoma characteristics.
Patient level data was used from the included studies. Of the 10,021 men and women who were enrolled in the individual studies, we excluded patients who had a colorectal cancer present at baseline (n = 27) and those who did not have a follow-up colonoscopy performed after the first 6 months of the study (n = 827) because these were likely people who were not under typical postpolypectomy surveillance. Thus, data for 9167 (91.5%) patients remained for inclusion in our pooled analyses.
Nusko et al. (2002)Follow-up records of 1159 patients undergoing surveillance examination. The following statistical procedures were performed:
(1) multiple regression analysis; (2) likelihood ratio tests; (3) calculation of the times t0.05, t0.10, and t0.20 for the relevant risk groups based on their hazard functions; (4) 1000 bootstrap samples
Records from 1978 to 1996A total of 3134 patients undergoing endoscopic removal of colorectal adenomas were prospectively recorded on the Erlangen Registry of Colorectal Polyps between 1978 and 1996.

The patients had no previous history of colorectal adenomas or carcinomas.

Patients with a familial history of adenomatous polyposis or hereditary non-polyposis colon cancer syndrome, or inflammatory bowel disease were excluded.
Identifying risk factors determining surveillance intervals for patients with metachronous adenomas of advanced pathologyA total of 3134 patients undergoing endoscopic removal of colorectal adenomas between 1978 and 1996. Single adenomas were found in 1052 patients (53.6%) and 797 (46.4%) had multiple initial lesions. Mean age at the initial clearing examination for patients who were followed up was 57.08 years (SD 11.25) compared with 59.74 (SD 11.61) for those who were not followed up. A total of 1159 patients underwent regular follow-up examinations: 747 (64%) of these patients were males and 412 (36%) were females. 100 patients (8.6%) had a parental history of colorectal carcinoma while in 24 patients (2.1%) the relevant data were not available.

Risk factors for advanced metachronous adenomas
Considering only patients with tubular adenomas at the initial clearing procedure, a multivariate model for related observations revealed that adenoma size (p < 0.0001), multiplicity (p = 0.021), parental history of colorectal carcinoma (p = 0.0168), and an interactive effect between size and sex (p = 0.00392) were significant predictive variables. Male patients with large adenomas had a significantly higher risk of developing advanced metachronous adenomas than other patients.

Low-risk group containing patients with no parental history of colorectal carcinoma and with only small (<10 mm) tubular adenomas at the initial clearing examination: 12.2 (95% CI 10.1 to 15.2) years were needed for advanced adenomas to develop in more than 10% of patients. The estimate for 5% was 10.4 years (95% CI 4.1 to 13.2) and for 20% was16.2 years (95% CI 10.5 to 19.2).
High-risk group containing all other patients: those with multiple or large adenomas, tubulovillous or villous adenomas, or a parental history of colorectal carcinoma: 6.1 (95% CI 3.2 to 11.5) years were needed for advanced adenomas to develop in more than 10% of patients. The estimate for 5% was 0.5 years (95% CI 0.1 to 1.6) and for 20% was15.6 years (95% CI 11.5 to 18.2).
Large registry data, studying risk factors. All patients were offered a chance to participate in a scheduled follow-up programme, however 1849 patients either refused follow-up or underwent examinations at other endoscopy departments.

There were no statistically significant differences in baseline patient or adenoma characteristics between patients who underwent surveillance and those who did not. Bivariate analyses done apart from univariate analyses to adjust for confounding covariates. Sensitivity analyses done using bootstrapping.

Kept despite Saini et al. (2006) as the outcomes used in their study did not include the ones extracted from this primary paper.
Saini et al. (2006)Systematic review and meta analysis

Studies included:
Baron et al. (1999), Bonithon-Kopp et al. (2000), Cordero et al. (1999), Fornasarig et al. (1998), Fossi et al. (2001), Hixson et al. (1994), Jørgensen et. al. (1995), Lund et al. (2001), Martinez et al. (2001), Noshirwani et al. (2000), Nusko et al. (2002), Paspatis et al. (1995), Schatzkin et al. (2000), Van Stolk et al. (1998), Winawer et al. (1993b)
Three electronic databases (MEDLIN, PREMEDLINE, and EMBASE) were searched from January 1980 to January 2003Study population was patients with a personal history of adenomas.

Studies enrolling patients with a personal history of hereditary nonpolyposis colorectal cancer (HNPCC), familial adenomatous polyposis (FAP), CRC, or inflammatory bowel disease (IBD) were excluded.
Nine hundred seventy-one references were identified but fifteen primary studies were included.

Identifying risk factors associated with advanced adenomas.
Bonithon-Kopp et al. (2000) showed that the only RR that was statistically significant was for number of adenomas only: RR 3.26 (95% CI 1.81 to 5.89).

Martinez et al. (2001) showed that the only RR that was statistically significant was for size only: RR 1.77 (95% CI 1.30 to 2.41)

Van Stolk et al. (1998) did not find any statistically significant RR for any factors.

Winawer et al. (1993) found the incidence of advanced adenomas at 3-year surveillance colonoscopy was 1.4% in the low-risk patients versus 5–4% in the high-risk patients: RR 3.87 (95% CI 1.09 to13.66). Advanced adenomas defined as adenomas ≥1 cm, villous histological features, or with cancer.

Number and size
Four trials: Bonithon-Kopp et al. (2000), Martinez et al. (2001), Van Stolk et al. (1998), Winawer et al. (1993): provided adequate data to determine the incidence of recurrent advanced adenomas at surveillance colonoscopy on the basis of: (1) the number of adenomas at index colonoscopy (>3 vs 1 or 2) the pooled RR was 2.52 (95% CI 1.07 to 5.97), and the pooled absolute risk difference was 5% (95% CI 1% to 10%); and (2) the size of the largest adenoma at index colonoscopy (≥1 cm [large] vs <1 cm [small]) the pooled RR was 1.39 (95% CI 0.86 to 2.26), and the pooled absolute risk difference was 2% (95% CI −2% to 6%)
The heterogeneity was significant for both cases, p < 0.001 and p < 0.05.

Histological diagnosis
Three trials: Bonithon-Kopp et al. (2000), Martinez et al. (2001), Van Stolk et al. (1998): provided adequate data to determine the incidence of recurrent advanced adenomas at surveillance colonoscopy on the basis of adenoma histologic features (tubulovillous/villous vs tubular). The pooled RR was 1.26 (95% CI 0.95 to 1.66), and the pooled absolute risk difference was 2% (95% CI −1% to 4%). The test of heterogeneity for the pooled RR was not significant (p > 0 .2), indicating that the individual studies did not demonstrate significant differences in the RR of recurrent advanced adenomas.

Two studies: Bonithon-Kopp et al. (2000) and Van Stolk et al. (1998) provided adequate data to determine the incidence of recurrent advanced adenomas on the basis of the degree of dysplasia at index colonoscopy (high grade vs no high-grade dysplasia). The pooled RR was 1.84 (95% CI 1.06 to 3.19), and the pooled absolute risk difference was 4% (95%CI 0 to 8%). The test of heterogeneity for the pooled RR was not significant (p > 0 .2)

Risk factors for advanced adenomas at surveillance
Nine studies identified a total of 5 risk factors that were associated with advanced adenomas at surveillance colonoscopy: (1) number of adenomas, (2) size of largest adenoma, (3) incomplete index colonoscopy, (4) concurrent proximal and distal adenomas, and (5) parental history of CRC.

Risk factors for recurrence of adenomas
14 studies reported a total of 6 risk factors: (1) number of adenomas, (2) size of largest adenoma, (3) patient age, (4) tubulovillous/villous features or severe dysplasia, (5) advanced adenoma, and (6) adenoma in the proximal colon.
All Mesh and free key words used for the searches were given in the paper. The PRISMA chart was available.
Winawer et al. (1993b)RCT to compare follow-up colonoscopy at 3 years and follow-up colonoscopy at both 1 and 3 years in people with newly diagnosed adenomatous polyps.Median interval between enrollment and initial follow-up examination was 1.15 years in the two-examination group; 3.15 years in the one-examination group. Follow-up clinical status was determined for 97.2% (1379/1418).9112 patients referred for colonoscopy who had no history of polypectomy, IBD, familial polyposis, or colorectal cancer identified at 7 clinical centres. Of 3778 patients in whom polyps were detected, 2632 (69%) had adenomas and were eligible for randomisation; 1418 (53.9%) of eligible patients with adenomas consented to participate.Participants were randomly assigned to a follow-up examination either 1 and 3 years after colonoscopy (the two-examination group) or 3 years after colonoscopy (the one-examination group). Follow-up colonoscopy 6 years after the examination at entry was also offered to both groups.NOTE: reported only those outcomes related to interval of surveillance for colonoscopy (other outcomes either included in the Saini 2006 review or not relevant for this question)

2-exam group (N=338)1-exam group (N=428)RR (95% CI)
Any adenomas141 (41.7%)137 (32.0%)1.3 (1.1 to 1.6)p = 0.006
Adenoma with advanced pathological feature (<1.0 cm, HGD, or invasive cancer)11 (3.3%)14 (3.3%)1.0 (0.5 to 2.2)p = 0.99

From: Appendix 6, Evidence tables

Cover of Colonoscopic Surveillance for Prevention of Colorectal Cancer in People with Ulcerative Colitis, Crohn's Disease or Adenomas
Colonoscopic Surveillance for Prevention of Colorectal Cancer in People with Ulcerative Colitis, Crohn's Disease or Adenomas.
NICE Clinical Guidelines, No. 118.
Centre for Clinical Practice at NICE (UK).
Copyright © 2011, National Institute for Health and Clinical Excellence.

All rights reserved. This material may be freely reproduced for educational and not-for-profit purposes. No reproduction by or for commercial organisations, or for commercial purposes, is allowed without the express written permission of NICE.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.