Prevalence, histology, and natural history of polyps according to lesion size

Based on a large number of clinical trials and experience, anywhere from 35% to 50% of adults more than 50 years of age may harbor at least 1 colorectal polyp. This figure may increase even further with the implementation of more advanced endoscopic techniques. In most cases, the largest lesion will be diminutive. Because of the broad differences in the detection rates of diminutive lesions and their relative lack of clinical importance, polyp prevalence at the 6-mm and 10-mm size thresholds are much more reproducible and relevant values to consider. Recent colonoscopy screening studies have shown a remarkably narrow prevalence range for polyps greater than or equal to 6 mm of 13% to 16% ( Table 1 ). Similarly, the prevalence for large polyps is 5% to 6%, which results in about 8% of individuals in whom the largest polyp will lie within the 6- to 9-mm range. As a general rule, approximately one-third of diminutive lesions will be adenomatous (almost exclusively tubular adenomas) and two-thirds will be nonadenomatous, predominately consisting of nonneoplastic mucosal tags and hyperplastic polyps. In polyps larger than 6 mm, the ratio of adenomatous to nonadenomatous polyps reverses, with neoplastic lesions representing approximately two-thirds of nondiminutive lesions.

The ideal screening target for prevention of CRC is the advanced adenoma, which is defined as an adenoma that is large (≥10 mm) or contains histologic findings of high-grade dysplasia or a prominent villous component. Although largely unproven, most experts believe that high-grade dysplasia is a more concerning feature than villous histology. The serrated polyp pathway, which is distinct from the classic adenoma-carcinoma sequence, may account for about 15% of CRC cases. For this particular pathway, sessile serrated adenomas less than 10 mm without dysplasia should not be considered as histologically advanced lesions, but serrated adenomas that are large (≥10 mm) or exhibit dysplasia should also be categorized as advanced (Michael J. O’Brien, MD, personal communication, 2009). The term “advanced neoplasia” encompasses advanced (but still benign) adenomas and invasive adenocarcinoma. This term is useful for CRC screening because it combines the key features of prevention and detection.

Although large adenomas (≥10 mm) comprise about 90% of all advanced neoplasia in the screening setting, approximately 4% of 6- to 9-mm adenomas will show advanced histology, with a reported range of 2.7% to 5.3% (see Table 1 ). Assuming an 8% screening prevalence of 6- to 9-mm polyps and a 4% frequency of advanced histology, the overall screening prevalence of small advanced adenomas is approximately 0.3%, with a reported range of 0.17% to 0.56% (see Table 1 ). The presence of high-grade dysplasia in 6- to 9-mm adenomas is even more uncommon, with an overall prevalence of about 0.05% (see Table 1 ). Although the overall prevalence of diminutive polyps is many times higher than small 6- to 9-mm polyps, the prevalence of diminutive advanced neoplasia is considerably lower than that for small polyps.

One striking feature of the recent screening data is the lower rate of cancer according to lesion size compared with the high-risk, symptomatic, and/or surgical cohorts in the older literature. For example, a commonly quoted historical figure for the cancer rate among small 6- to 9-mm adenomas is 0.9%. However, when the recent large screening studies are tallied, the frequency of cancer decreases to 0.1% or lower, ranging from 0% to 0.5% (see Table 1 ), with most of the reported small cancers concentrated within one Korean series. The percentage falls even lower if all 6- to 9-mm polyps, and not just small adenomas, are considered in the denominator. We have yet to encounter a subcentimeter invasive cancer in our combined CTC and OC experience, including more than 1000 6- to 9-mm polyps. Even for large 1- to 2-cm lesions, the cancer rate seems to be only about 1% (see Table 1 ), which is considerably lower than the commonly quoted historical range of 5% to 10%, which is again based on high-risk cohorts and not screening populations. Given that about 30% to 40% of large polyps are nonadenomatous and that some large lesions detected at CTC may be false-positives, the actual cancer risk for a 1- to 2-cm lesion detected at CTC is considerably less than 1%, lower than the frequency of significant complications at OC referral for therapeutic polypectomy.

The natural history of small colorectal polyps has become an issue of critical importance in CRC screening. One reason for this is that CTC is an efficacious and cost-effective approach to population screening if only large polyps (≥10 mm) were considered appropriate to trigger polypectomy. If all small 6- to 9-mm CTC-detected polyps were to be referred to therapeutic colonoscopy for polypectomy, the usefulness of CTC as an intermediate filter would be diminished, but likely still useful. Although CTC provides an ideal tool for in vivo surveillance of small unresected polyps, there are several older studies that have followed these lesions using other colorectal examinations, including endoscopy and barium enema. Contrary to the general perception, many of the data on polyp natural history already exist from these older longitudinal trials. As a group, these longitudinal studies have repeatedly shown the benign, indolent nature of unresected subcentimeter colorectal polyps, with no study showing that leaving 6- to 9-mm polyps in place is a harmful practice.

Most longitudinal polyp surveillance studies have focused more on small 6- to 9-mm polyps, although some have focused on diminutive or large lesions. In Norway, Hofstad and colleagues performed serial colonoscopy on unresected subcentimeter polyps and found that only 1 (0.5%) of 189 lesions eclipsed the 10-mm threshold after a 1-year time interval. At the 3-year follow-up, most polyps in this study remained stable or regressed in size, and there was an overall tendency to net regression among the 5- to 9-mm polyps. The investigators of this endoscopic trial concluded that following unresected 5- to 9-mm polyps for 3 years was a safe practice. Longitudinal studies using flexible sigmoidoscopy have also shown the stability of smaller polyps over time. In one study that used serial sigmoidoscopy to follow polyps measuring up to 15 mm over a 3- to 5-year period, Knoernschild reported a significant increase in polyp size in only 4% of patients. In a longitudinal study using barium enemas to follow colorectal polyps, Welin and colleagues showed slow growth rates by studying 375 unresected polyps over a mean interval of 30 months. The high observed adenoma detection rates at surveillance in the National Polyp Study, in conjunction with the low observed CRC incidence, was thought to be explainable only by regression of adenomas. In a high-risk cohort of patients undergoing colonoscopy surveillance following CRC surgery, Togashi and colleagues followed 500 polyps 6 mm or less over an average interval of 3.6 years. They concluded that this practice was safe even in the high-risk setting. In a classic barium enema study by Stryker and colleagues, the cumulative 5-year and 10-year risk of cancer related to large colorectal polyps (≥1 cm) left in place was less than 3% and 10%, respectively.

These reassuring longitudinal endoscopic and barium enema studies have done little to quell the current debate about the clinical management of small polyps detected at CTC screening. Part of the problem may be a simple lack of awareness of these study results. CTC can now be used as the preferred instrument to follow unresected colorectal lesions. CTC provides superior polyp measurement capabilities compared with the other colorectal imaging examinations, including improved accuracy and reproducibility for linear size assessment. In addition, CTC can assess polyp volume, which greatly amplifies interval changes in lesion size compared with linear measurement.

The University of Wisconsin School of Medicine and Public Health and the National Naval Medical Center (NNMC) in Bethesda, Maryland, are currently collaborating on a small-polyp natural history trial that commenced in 2004. The early interim results of CTC surveillance in 128 small colorectal polyps from the initial 100 patients has largely recapitulated the findings from the older endoscopic and barium enema trials. With an average CTC follow-up interval of about 1.5 years, 12 (9.4%) of the small polyps showed interval growth, including 11 proven adenomas (1 polyp was removed but not retrieved at OC). There were no cancers that developed during this short interval, and none of the lesions grew past the 10-mm threshold. Five of the adenomas represented advanced lesions, corresponding to 4% of the total polyp cohort (ie, the expected number of advanced lesions from the entire group of 6- to 9-mm polyps). The remaining 116 polyps (90.6%) did not increase in size at CTC follow-up, and some of them had regressed. These findings suggest that interval growth can predict important histology, allowing for noninvasive identification of the small fraction of polyps for which polypectomy is clearly of benefit.

CTC detection of small 6- to 9-mm colorectal polyps

The accuracy of CTC for detecting large polyps (≥10 mm) and masses (≥3 cm) has been well established, with most studies reporting sensitivity and specificity values of 90% or higher. CTC performance tends to be more robust when three-dimensional (3D) polyp detection is used alongside two-dimensional (2D) evaluation, when oral contrast tagging has been applied, and when automated carbon dioxide delivery is used for colonic distention. When state-of-the-art CTC is undertaken, there is evidence to suggest that CTC sensitivity for large polyps and cancers may exceed that of OC. However, there are a few notable exceptions in which the CTC sensitivity for large-polyp detection was in the 50% to 60% range, but none of these studies used primary 3D detection, oral contrast tagging, or carbon dioxide.

The CTC performance for small 6- to 9-mm polyps is more variable ( Fig. 1 ). One problem is the lack of a reliable reference standard, because the miss rate for small lesions at OC can be 10% or higher when tandem (back-to-back) colonoscopy studies are performed. In addition, several published studies have reported by-patient results at the 6-mm and 10-mm thresholds, but not specifically for the 6- to 9-mm range. Although such results can generally be inferred, the conversion is imperfect related to the use of different polyp-matching algorithms. The patient populations are also somewhat heterogeneous, representing screening and nonscreening cohorts. For most CTC studies that have evaluated at least 100 patients, the per-patient sensitivity for small 6- to 9-mm polyps lies somewhere within the range of 50% to 95% ( Fig. 2 , Table 2 ). The only outlier was the study by Cotton and colleagues, in which the per-patient sensitivity was only 30%.

Small 6-mm tubular adenoma detected at CTC screening. 3D endoluminal ( A ) and 2D transverse ( B ) CTC images show a well-circumscribed 6-mm sessile polyp ( arrow ) in the ascending colon, which proved to be a tubular adenoma after resection at same-day OC ( C ). With state-of-the-art CTC technique, the diagnostic performance for detecting small 6- to 9-mm polyps likely approaches that for larger lesions. ( From Pickhardt PJ. The colon and rectum. In: Pickhardt PJ, Arluk GM, editors. Atlas of gastrointestinal imaging: radiologic-endoscopic correlation. Philadelphia: Saunders; 2007. p. 212; with permission.)

Bubble graph showing CTC by-patient sensitivity for detecting 6- to 9-mm polyps for published trials that involved 100 or more patients and used OC as the reference standard. The bubble sizes correspond to the size of the patient cohorts. The studies are displayed in chronologic order, with the oldest on the left (see Table 2 ). Note the 1 outlier where the CTC sensitivity decreases to less than 50%.

More recent data from the clinical CTC screening programs at the University of Wisconsin (UW) and the NNMC suggest that the performance for state-of-the-art CTC for 6- to 9-mm polyps is now approaching that for larger lesions (see Fig. 1 ). An ongoing CTC trial at NNMC continues to show sensitivity for small polyps of about 90% (Brooks Cash, personal communication, 2009). At UW, the positive predictive value of 6- to 9-mm CTC-detected polyps is more than 90%, significantly higher than results from the published clinical trials. In routine clinical practice, the positive predictive value is an important quality measure, along with the overall yield of advanced neoplasia, because performance assessments by sensitivity, specificity, and accuracy cannot be measured when negative CTC cases do not go on to OC. The common CTC methodology used at UW and NNMC provide further support for primary 3D interpretation, which has also been shown to improve small-polyp detection compared with 2D detection alone in a phantom study.

Even if CTC has high accuracy for detecting small polyps, it remains unclear whether all such lesions warrant immediate polypectomy. Evaluating the potential benefit (ie, preventing CRC) against the potential risks (eg, perforation, bleeding, sedation-related events) and costs (eg, OC procedure, pathology charges), it becomes clear that the conclusion will be largely driven by the input assumptions. Given the low risk (approximately 4%) that a 6- to 9-mm polyp will be an advanced adenoma and the extremely low risk (<0.1%) of CRC, deferring polypectomy may be an attractive option for individuals who have already decided to undergo a less invasive screening route. To coincide with current standard of care, the current protocol at UW is to offer all patients with any CTC-detected polyp that is larger than or equal to 6 mm same-day OC for polypectomy (see Fig. 1 ). However, individuals with one or two 6- to 9-mm lesions, corresponding to a C-RADS C2 classification, are also offered the option of short-term CTC follow-up in 2 to 3 years. Preliminary results with CTC surveillance (described earlier) suggest that this approach may effectively identify the small subset of lesions for which polypectomy is indicated and avoid the need for colonoscopy in most other cases. However, more data are needed before drawing firm conclusions. Given the published data establishing the risk of future advanced neoplasia related to finding multiple adenomas at the index colonoscopy, the policy at CTC is that patients with 3 or more small polyps are referred for polypectomy. This approach corresponds with a C-RADS C3 categorization, placing 3 or more small polyps detected at CTC at the same level as 1 or more large (≥10 mm) polyps.

Given the limited health care dollars available for expensive resources, it is critical to also consider costs alongside the anticipated health consequence for the various screening strategies. We have studied the theoretical cost-effectiveness of immediate polypectomy versus 3-year CTC surveillance for small 6- to 9-mm polyps detected at CTC screening. Without any intervention, the estimated 5-year CRC death rate for patients with unresected 6- to 9-mm polyps was 0.08%, which already represents a sevenfold decrease from the 0.56% 5-year CRC death rate in the general (unscreened) population, most of whom do not harbor polyps. Therefore, for patients with 6- to 9-mm polyps detected at CTC screening, the exclusion of large polyps (≥10 mm) and masses already confers a low CRC risk. Focusing on a concentrated cohort with only small 6- to 9-mm polyps, the death rate was further reduced to 0.03% with the CTC surveillance strategy and to 0.02% with immediate colonoscopy referral. However, for each additional cancer-related death prevented with immediate polypectomy versus CTC follow-up, 10,000 additional colonoscopy referrals would be needed, resulting in an expected 10 additional perforations and an exorbitant incremental cost-effectiveness ratio (ICER) of $372,853. We therefore concluded that the high costs, additional complications, and low incremental yield associated with immediate polypectomy of 6- to 9-mm polyps support the practice of 3-year CTC surveillance, which allows for selective noninvasive identification of small polyps at risk (as described earlier). CTC surveillance of small unresected polyps should only be undertaken in the context of a dedicated CTC program, in which a reliable mechanism for follow-up is in place and in which the patient understands the relative risks and benefits involved.

CTC detection of small 6- to 9-mm colorectal polyps

The accuracy of CTC for detecting large polyps (≥10 mm) and masses (≥3 cm) has been well established, with most studies reporting sensitivity and specificity values of 90% or higher. CTC performance tends to be more robust when three-dimensional (3D) polyp detection is used alongside two-dimensional (2D) evaluation, when oral contrast tagging has been applied, and when automated carbon dioxide delivery is used for colonic distention. When state-of-the-art CTC is undertaken, there is evidence to suggest that CTC sensitivity for large polyps and cancers may exceed that of OC. However, there are a few notable exceptions in which the CTC sensitivity for large-polyp detection was in the 50% to 60% range, but none of these studies used primary 3D detection, oral contrast tagging, or carbon dioxide.

The CTC performance for small 6- to 9-mm polyps is more variable ( Fig. 1 ). One problem is the lack of a reliable reference standard, because the miss rate for small lesions at OC can be 10% or higher when tandem (back-to-back) colonoscopy studies are performed. In addition, several published studies have reported by-patient results at the 6-mm and 10-mm thresholds, but not specifically for the 6- to 9-mm range. Although such results can generally be inferred, the conversion is imperfect related to the use of different polyp-matching algorithms. The patient populations are also somewhat heterogeneous, representing screening and nonscreening cohorts. For most CTC studies that have evaluated at least 100 patients, the per-patient sensitivity for small 6- to 9-mm polyps lies somewhere within the range of 50% to 95% ( Fig. 2 , Table 2 ). The only outlier was the study by Cotton and colleagues, in which the per-patient sensitivity was only 30%.

Small 6-mm tubular adenoma detected at CTC screening. 3D endoluminal ( A ) and 2D transverse ( B ) CTC images show a well-circumscribed 6-mm sessile polyp ( arrow ) in the ascending colon, which proved to be a tubular adenoma after resection at same-day OC ( C ). With state-of-the-art CTC technique, the diagnostic performance for detecting small 6- to 9-mm polyps likely approaches that for larger lesions. ( From Pickhardt PJ. The colon and rectum. In: Pickhardt PJ, Arluk GM, editors. Atlas of gastrointestinal imaging: radiologic-endoscopic correlation. Philadelphia: Saunders; 2007. p. 212; with permission.)

Bubble graph showing CTC by-patient sensitivity for detecting 6- to 9-mm polyps for published trials that involved 100 or more patients and used OC as the reference standard. The bubble sizes correspond to the size of the patient cohorts. The studies are displayed in chronologic order, with the oldest on the left (see Table 2 ). Note the 1 outlier where the CTC sensitivity decreases to less than 50%.

Table 2

Reported per-patient sensitivities for small 6- to 9-mm polyps

Trial	Author, Year	Sensitivity	No. of Patients
1	Fenlon et al, 1999	94	100
2	Yee et al, 2001	93	300
3	Lefere et al, 2002	91	100
4	Ginnerup Pedersen et al, 2003	82	144
5	Pineau et al, 2003	84	205
6	Johnson et al, 2003	52	703
7	Pickhardt et al, 2003	87	1233
8	Innaccone et al, 2004	87	203
9	Cotton et al, 2004	30	600
10	Rockey et al, 2005	51	614
11	Arnesen et al, 2005	60	100
12	Arnesen et al, 2007	56	231
13	Jensch et al, 2008	71	168
14	Kim et al, 2008	62	241
15	Johnson et al, 2008	65	2531
16	Graser et al, 2009	90	307

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Debate: Small (6–9 mm) and Diminutive (1–5 mm) Polyps Noted on CTC: How Should They Be Managed? Improving the Accuracy of CTC Interpretation: Computer-Aided Detection Radiologists Should Read CT Colonography MR Colonography and MR Enterography CT Enterography The Role of CT Colonography in a Colorectal Cancer Screening Program

Stay updated, free articles. Join our Telegram channel

Join