Recurrence and surveillance of colorectal adenoma after polypectomy in a southern Chinese population
- First Online:
- Cite this article as:
- Huang, Y., Gong, W., Su, B. et al. J Gastroenterol (2010) 45: 838. doi:10.1007/s00535-010-0227-3
- 198 Views
Background and aim
Repeat colonoscopy is often performed within a short time after polypectomy due to the fear that colorectal adenomas were missed during the initial colonoscopy or that new adenomas have developed. The aim of this study was to estimate the actual recurrence rate of adenoma and its association with the length of the surveillance interval after polypectomy in a southern Chinese population.
A total of 1356 patients undergoing endoscopic polypectomy and completing three or more surveillence colonoscopies between 1976 and 2007 were retrospectively analyzed. The recurrence rates of adenoma and advanced adenoma and surveillance intervals after polypectomy were identified based on the features of adenomas detected on initial colonoscopy.
The recurrence rates of advanced adenoma in patients with non-advanced adenoma on the initial colonoscopy were 0.9, 3.9, 5.8, and 29.2% during surveillance intervals of 1–3, 3–5, 5–10, and 10–20 years post-initial colonoscopy; for patients with advanced adenoma on the initial colonoscopy, the recurrence rates were 3.8, 13.1, 34.7, and 52.0% during the same surveillance intervals, respectively. Older age (p < 0.05 for trend) and male sex [hazard ratio (HR) 2.11, 95% confidence interval (CI) 1.27–3.53] were significantly associated with recurrence for advanced adenoma, as were the size and number of baseline adenoma (p < 0.05 for trend), tubulovillous, villous adenoma (HR 2.57, 95% CI 1.24–5.32), and high-grade dysplasia (HR 1.61, 95% CI 1.07–2.42). When 5% of patients had recurring advanced adenoma, the surveillance interval was estimated to be 6.9 (95% CI 6.3–12.2) years in the low-risk group and 3.0 (95% CI 2.7–3.2) years in the high-risk group.
Among our patient group, the recurrence of advanced adenoma after polypectomy increased with the length of the surveillance interval. Based on our results, a 3-year follow-up of patients after polypectomy could be effective in preventing the recurrence of advanced adenoma in high-risk patients.
The incidence of colorectal cancer (CRC) in China has increased significantly in recent years and is currently the fourth most common cancer and the fifth leading cause of death by cancer in the country . Colorectal neoplasia screening and removal by colonoscopy is an effective strategy for reducing CRC incidence and mortality [2, 3]. Surveillance colonoscopy is currently recommended after the initial screening and removal of colonic neoplasia due to the possibility of newly developing new neoplasia [4, 5]. Among patients who have adenoma removed during the initial colonoscopy, 20–50% will be found to have recurrent adenoma on a surveillance colonoscopy within 3–5 years; of these, 20% have advanced adenoma [2, 3, 6–9], and a small proportion will be invasive CRC [10–12]. With the aim of detecting early cancers or advanced adenoma, current recommendations provide guidelines for colonoscopy surveillance intervals after polypectomy based on the characteristics of the neoplasia on initial colonoscopy . However, these surveillance intervals are difficult to put into practice because some neoplasms are missed in colonoscopy procedures [13, 14]. Whether Western surveillance guidelines and practices can be directly applied to the Chinese population has always been a question plaguing Chinese endoscopists. At the present time, there are no recommended colonoscopy surveillance guidelines for the Chinese population. A majority of Chinese colonoscopists consider that surveillance colonoscopy for patients with adenoma should be performed within 1 year, which may be a possible waste of medical resources. Given this background, the aim of our study was to analyze retrospectively the actual recurrence rate of adenoma and advanced adenoma in a Chinese patient population with adenoma on the initial colonoscopy and assess the relationship between the characteristics of baseline adenoma and the recurrence of adenoma. We also estimated possible colonoscopy surveillance intervals after polypectomy.
Materials and methods
The registry for colorectal adenoma was established in 1976. All cases of colorectal adenoma removed by colonoscopy have therefore been documented prospectively in the endoscopic center of the Gastroenterology Department at NanFang Hospital, Guangzhou since this time. Data on patients in the study interval (January 1976 to December 2007) include the characteristics of the adenoma (size, location, number, shape, and histopathology), age and sex of the patient, and outcomes of surveillance colonoscopy. All colonoscopies were performed by experienced endoscopists. To be included in the study, the patient needed to meet the following criteria: (1) older than 20 years; (2) complete colonoscopy was performed at every examination; (3) at least one surveillance colonoscopy was performed within 6 months after the initial colonoscopy, with the aim of removing missed adenomas; (4) end point data on size, number, and histopathology of adenomas were available on surveillance colonoscopy. Patients with familial adenomatous polyposis, inflammatory bowel disease, or personal history of CRC were excluded. A complete colonoscopy was considered to include: colonoscopy reaching the cecum, a good bowel preparation, and removal of all visualized lesions by colonoscopic polypectomy or endoscopic mucosal resection (EMR). All adenomas detected at subsequent colonoscopies performed within 6 months of the baseline colonoscopy were considered to be adenomas missed on the initial colonoscopy and to belong to the “baseline adenomas” in terms of baseline status. Adenoma detected at the prior polypectomy site, identified based on the description of the previous anatomic location and by the post-polypectomy scar, was regarded to be local residual adenoma. Complete resection was confirmed when no adenoma was found at the site of the post-polypectomy scar, and recurrent adenoma was metachronous adenoma detected at the surveillance colonoscopy. We considered as end point colorectal adenoma or advanced adenoma that were diagnosed during an interval beginning 6 months after the baseline colonoscopies and ending on the date of colonoscopy in which adenoma or advanced adenoma was found. The surveillance interval was at least >1 year and any deviation from the scheduled surveillance interval was less than 0.5 year.
In the course of this study, we reviewed all colonoscopy and pathology reports and collected data on the number, size, location, shape and pathology of the adenoma, and age and sex of the patient. The size of the adenoma was estimated with an opened biopsy forceps (6 mm) or measured after resection, and the location was categorized as proximal (cecum, ascending colon, hepatic flexure, transverse colon, and splenic flexure) or distal (descending colon, sigmoid colon, and rectum). The shape of the adenoma was classified as two types: protruded and non-protruded lesions. For non-protruded adenoma, we further classified the lesions into types “0-IIa”, “0-IIb”, and “0-IIc” according to the Paris endoscopic classification when we reviewed all endoscopic pictures of the adenoma . Histopathology features of the adenoma were evaluated by a specialist gastrointestinal pathologist according to the colorectal neoplasia classification of the World Health Organization recommendations . The size, location, and histological type in patients with multiple adenomas were classified according to the largest, the most proximal, and the most advanced lesion, respectively. Advanced adenoma was considered to be adenoma that had one or more of the following features: a tubular adenoma 10 mm or larger in diameter, villous or tubulovillous adenoma, or the presence of high-grade dysplasia.
The cumulative hazard of the recurrence of any adenoma and advanced adenoma stratified according to characteristics of baseline adenoma was determined on the basis of the Kaplan–Meier method. Differences were tested using the log-rank test. To assess the relative risk of recurrence of any adenoma and advanced adenoma based on size, presence of villous, and number of baseline adenomas and to control for the age and sex of the patient, we fitted the data into the Cox proportional hazard model by computing the hazard ratio (HR) and the 95% confidence intervals (CI).
In order to estimate the distribution of the time at which an adenoma becomes advanced, we calculated the times required for 5 (t0.05), 10 (t0.10), and 20% (t0.20) of the patients, stratified on the basis of the risk factors for advanced adenoma recurrence at the surveillance colonoscopy, to develop adenomas of advanced pathology. The Kaplan–Meier method was used to calculate the cumulative recurrence rate of advanced adenoma. We used the bootstrap sampling method  to randomly pick out 3000 samples from eligible patients in each sampling test and calculated the times corresponding to the above-mentioned quantiles. Repeat sampling was performed 1000 times. All calculated times formed new data sets. The median method was used to determine the 95% confidence intervals for these times .
Differences were considered to be significant when the two-sided p value was <0.05. All calculations were performed using the statistical software SAS ver. 9.12 (SAS Institute, Cary, NC).
Patients and findings of the baseline colonoscopies
Of the 1356 eligible patients, 863 were males and 493 were females. The mean (±standard deviation, SD) age at the index colonoscopy was 52.4 (±12.5) years (range 20–84 years). In total, 714 (52.7%) patients had non-advanced adenomas, and 642 (47.3%) had advanced adenomas. At baseline, a single adenoma was found in 808 (59.6%) patients and multiple adenomas in 528 (40.4%) patients.
Recurrence rates of colorectal adenoma
Cumulative recurrence rates of advanced adenoma and of any adenoma at surveillance colonoscopy according to baseline history and time since baseline examination
Surveillance time interval (years)
Any adenoma, n (%)
Advanced adenoma, n (%)
Non-advanced adenoma (n = 442)
Advanced adenoma (n = 368)
Any adenoma (n = 810)
Non-advanced adenoma (n = 152)
Advanced adenoma (n = 160)
Any adenoma (n = 312)
Non-advanced adenoma (n = 86)
Advanced adenoma (n = 95)a
Any adenoma (n = 181)
Non-advanced adenoma (n = 24)
Advanced adenoma (n = 29)b
Any adenoma (n = 53)
Three cancers were diagnosed on surveillance colonoscopies. One cancer was detected at 8 years from baseline; the other two cancers were found at 14 and 18 years post-initial colonoscopy. All three patients with cancers detected on the surveillance examination had advanced adenomas at the baseline examination.
Factors related to the recurrence of adenoma
Multivariate analysis: risk factors of the recurrence of any adenoma
Characteristics at baseline colonoscopy
Baseline findings, n (%)
Any adenoma recurrence n (%)
HR (95% CI)
Multivariate analysis: risk factors of advanced adenomas recurrence
Characteristic at baseline colonoscopy
Baseline findings, n (%)
Advanced adenoma recurrence, n (%)
HR (95% CI)
Surveillance interval according to the quantile of advanced adenoma recurrence
Surveillance interval based on the quantiles t0.05, t0.10, and t0.20
95% CI (years)
95% CI (years)
Previous studies reported that patients with adenoma at baseline had higher adenoma recurrence rates than those without adenoma [9, 19], indicating that the surveillance of patients with baseline adenoma is important in terms of reducing the incidence of CRC. In our study, we examined the recurrence rate of any adenoma and of advanced adenoma at a variety of time intervals post-initial colonoscopy in a southern Chinese population with adenoma at baseline. More recent studies have reported recurrence rates of adenoma of about 15–60% within 3–4 years after the initial polypectomy, with about 15% of these recurrences being advanced adenomas [2, 6, 20, 21]. The results of Bertario et al.’s study suggest that the recurrence rates of adenoma in patients whose adenoma was removed at baseline were 18, 23, 31, and 40% at 2, 3, 5, and 10 years post-baseline, respectively . Our results are similar to those of these authors. In our study, the recurrence rate of adenoma or advanced adenoma in various surveillance intervals were higher among patients with advanced adenomas at baseline than in those with non-advanced adenomas at baseline. In order to confirm this result, we used the Kaplan–Meier method to evaluate the cumulative hazard of adenoma or advanced adenoma recurrence in patients with non-advanced adenoma or advanced adenoma at baseline and found a significant difference between these two groups of patients during any one surveillance interval. This result is in accordance with data based on populations of other countries [9, 23–25], suggesting that the classification criteria used for non-advanced adenoma and advanced adenoma during surveillance colonoscopy in Western countries is also applicable to the Chinese population.
We also determined the risk factors associated with the recurrence of any adenoma and advanced adenoma in all patients undergoing surveillance colonoscopy. The recurrence of any adenoma and advanced adenoma was associated with the size, number, and histological features of the baseline adenoma as well as with the age and sex of the patients; these results were consistent with those of prior studies. Studies from Western countries have reported that the presence of proximal adenomas at baseline was also related to the recurrence of adenoma. However, in the Chinese population, we did not find any association between the recurrence of any adenoma or advanced adenoma and the location of the baseline adenoma, which might be due to ethnic differences between the Chinese population and those of other countries. In addition, two or more clearing colonoscopies were performed before a formal follow-up in our study, which determined the removal of missed adenoma that was reported to be often present on the proximal colon during colonoscopy [11, 26]. In terms of the size of the baseline adenoma, we found that the recurrence of advanced adenoma was associated with a baseline adenoma >20 mm in diameter. This result differed from that of prior studies in which the recurrence of advanced adenoma was associated with a baseline adenoma >10 mm in diameter [9, 23, 27].
We found that the risk of adenoma recurrence is higher among men than among women. One possible explanation for this may be associated with biological priority. Randomized controlled studies have demonstrated the chemopreventive abilities of estrogens/progesterones for colorectal neoplasia, potentially mediated by the tumor suppressor role of the estrogen receptor . Moreover, risk factors, such as diet or smoking, may be quite distinct between men and women, leading to different manifestations of colorectal adenoma recurrence .
In the latest guideline from the U.S. Multisociety Task Force on Colorectal Cancer and the American Cancer Society, a repeat colonoscopy is recommended at 3 years post-initial colonoscopy for patients with advanced adenoma or patients with three or more adenomas and at 5–10 years for patients with non-advanced adenomas . Whether this Western surveillance guideline and standard practice can be directly applied to the Chinese population has always been the question plaguing Chinese endoscopists. In our study, we categorized all surveillance patients into two groups based on risk factors of advanced adenoma recurrence and evaluated three quantiles of patients with recurrent advanced adenoma and the corresponding surveillance interval using the bootstrap sampling method and the Kaplan–Meier method. Our results indicate that a surveillance colonoscopy performed at 3 and 6.9 years could detect 5% of the patients in the high-risk group and low-risk groups with recurring advanced adenomas. We therefore consider that, in the Chinese population, a surveillance colonoscopy at 3 years for high-risk patients and one at 6.9 years for low-risk patients could result in detecting at least a 5% recurrence of advanced adenoma and all non-advanced adenoma at the surveillance colonoscopy; these intervals are similar to those recommended by professional groups in the USA . However, the surveillance interval after polypectomy should be further investigated based on the combination of stratified risk factors and cost-effective analysis.
Our study has a number of limitations. We did not assess the risk of colorectal neoplasia after polypectomy based on other stratified patient-specific features, with the exception for age and sex, such as family history of CRC or adenoma, diet, physical activity, smoking status, and body mass index, among others, which may affect the recurrence of adenoma [7, 30–33]. Therefore, large Chinese population cohort studies are needed to clarify these patient-specific features on the recurrence of adenomas.
In conclusion, the results of our study demonstrate that the recurrence of any adenoma and of advanced adenoma after polypectomy did increase with prolonged surveillance intervals. The 3-year follow-up after polypectomy may be an effective surveillance interval in preventing the recurrence of advanced adenoma for high-risk patients. Baseline adenoma features and patient age and sex were risk factors associated with the recurrence of adenoma in a southern Chinese population.