Total polyp number may be more important than size and histology of polyps for prediction of metachronous high-risk colorectal neoplasms - BMC Gastroenterology - BMC Gastroenterology

There are two main types of neoplastic polyps that have the malignant potential to progress to colorectal cancer—conventional adenomatous polyps and serrated polyps. Adenomatous polyps can be further categorized according to the following features: size, degree of dysplasia, and proportion of villous components. Serrated polyps can be further categorized into hyperplastic polyps, sessile serrated lesions, and traditional serrated polyps [13]. Many guidelines use these intrinsic factors and the number of polyps to suggest the adequate follow-up interval for surveillance colonoscopy after colorectal polypectomy. However, in clinical practice, determining the proper follow-up interval for each patient is not so straightforward. For example, many patients have both adenomatous polyps and serrated polyps [14]. In case of a patient with both advanced adenomatous polyps and advanced serrated polyps, it is difficult to determine which polyp poses a higher risk of metachronous neoplasm and should therefore be prioritized for determining the follow-up interval of surveillance colonoscopy. In addition, it is also unclear whether the risk of metachronous colorectal neoplasm is as high in patients with multiple (≥ 3) but small (< 10 mm in size) adenomatous polyps without any high-risk histologic features as in patients with advanced adenomatous polyps. Therefore, we compared the risk of metachronous HR-CRN among the ACRN, ASP, and HRP groups and found that it was not significantly different among the three groups. Instead, the number of total polyps, regardless of size and other histologic features of each polyp, was independently associated with the risk of metachronous HR-CRN. The percentage of patients who had metachronous HR-CRN increased as the total number of polyps at index colonoscopy increased—metachronous HR-CRN was found in more than half (57%) of patients with more than 10 polyps. After adjusting for other factors, patients with multiple polyps (5 or more) had a higher risk of metachronous HR-CRN (OR 2.575). This result corresponds very well with that of a recent Spanish study, in which the presence of multiple polyps (3 or more adenomas and/or serrated polyps) was found to be a strong predictor of HR-CRN after index polypectomy [15]. The study also did not find any histological characteristics that increased the risk of metachronous HR-CRN. However, contrary to our study, bowel preparation was not evaluated.

A recent meta-analysis reported that there was no significant difference in the risk of metachronous ACRN between patients with serrated polyps and those with conventional adenomas [16]. However, because most studies included in this meta-analysis had not reported the results according to size or number of serrated polyps, this meta-analysis could not assess the comparative risk of metachronous ACRN between ACRN and ASP. In keeping with this, the U.S. Multi-Society Task Force on Colorectal Cancer recommends a 3-year surveillance interval not only for both ACRN and ASP, but also in cases with 5–10 adenomas and 5–10 serrated lesions. In the present study, ACRN and ASP were not found to be independent risk factors for metachronous HR-CRN when surveillance colonoscopy was performed mostly within 3 years of the index colonoscopy. Therefore, applying the current U.S. recommendations for ACRN and ASP to Korean subjects seems reasonable. It is not clear whether one of these two main categories of advanced neoplastic polyps should be prioritized over the other. Nevertheless, as the risk for metachronous HR-CRN was higher in patients with multiple polyps (5 or more) regardless of the size and histology of each polyp, more intensive surveillance (at intervals shorter than 3 years) deserves consideration for patients who have 5 or more polyps when counted by combining adenomas and serrated lesions. However, because no trend in the proportion of patients who had metachronous HR-CRN was observed depending on the time interval of surveillance colonoscopy in the 1–3-year range, it is still unclear whether surveillance colonoscopy at 1 year is beneficial in these high-risk patients. Evidence for 1 year follow-up for multiple adenomas (more than 10) in the current U.S. guideline stems from a Korean study [17]. However, in the present study, when we performed additional analysis, the presence of more than 10 polyps was not found to be a statistically significant independent risk factor for metachronous HR-CRN. Although, given that metachronous HR-CRN was found in 57% of patients with more than 10 neoplastic polyps, our study also suggests that shorter surveillance intervals are considerable for this patient group.

Interestingly, contrary to the other two groups, the proportion of patients below 50 years of age in the ASP group was as high as 45% in the present study. The proportions of patients with hypertension and diabetes and of patients taking aspirin were relatively low in the ASP group; this could be attributed to the relatively younger age distribution in the ASP group. Similar trends for age distribution of patients with serrated polyps were also found in a previous study performed at our institution [18], in which the prevalence of conventional adenomas at screening colonoscopy was found to increase sharply with age. In contrast, the prevalence of serrated polyps appeared to be relatively high among patients aged < 50 years. The reasons for these phenomena are unclear; however, considering the recent increase in colorectal cancer in young adults [19], these results suggest that we should pay more attention not to miss serrated polyps in younger individuals.

The present study has several strengths. This was a prospective study, and we analyzed the results after adjusting for important major risk factors for colorectal cancer, including obesity, comorbidities, family history, and drug use. Nevertheless, this study also had certain limitations. First, the quality of the endoscopic procedure may affect the incidence of metachronous HR-CRN. Even though highly experienced endoscopists performed all procedures in this study, we had not monitored the adenoma detection rate of these endoscopists. In addition, we did not use magnifying colonoscopy. Therefore, the appropriateness of polypectomy may not have been thoroughly evaluated. To avoid missed lesions being considered newly developed lesions, we excluded patients who underwent follow-up colonoscopy less than 12 months after index colonoscopy. However, because it is challenging to distinguish between missed polyps and newly developed polyps, it is still possible that some missed polyps were included in metachronous polyps. Second, information was not collected about whether endoscopic mucosal resection for large polyps (≥ 20 mm in size) was performed en-bloc or piecemeal and the morphological characteristics of these large lesions. However, we believe that most patients with incomplete polyp resection were excluded from this study because we excluded patients who underwent follow-up colonoscopy less than 12 months after index colonoscopy, and the main reason due to which they underwent follow-up colonoscopy so early was incomplete resection. Third, although the patients were prospectively enrolled in this study, some parts of the study design are retrospective. The cut-off value for 5 or more polyps as a risk factor for metachronous HR-CRN was not set before study initiation. Therefore, although some studies suggest 5 as a cut-off for the total number of polyps [2, 5, 6], this value cannot be an absolute cut-off to determine the risk of metachronous HR-CRN. Last, because information about previous colonoscopy procedures was not collected, one must be cautious when applying these results to the general population. We started this study to evaluate the risk factors for metachronous HR-CRN. Therefore, we collected various risk factors using a survey. However, we found later that information about the history of colonoscopy was missing from the questionnaire. Therefore, the history of colonoscopy in each patient might affect the results of this study.

Total polyp number may be more important than size and histology of polyps for prediction of metachronous high-risk colorectal neoplasms - BMC Gastroenterology - BMC Gastroenterology