Nonpolypoid Colorectal Neoplasms -- Prevalence and Association With Colorectal Cancer?

David A. Johnson, MD, FACG, FACP


April 09, 2008

Prevalence of Nonpolypoid (Flat and Depressed) Colorectal Neoplasms in Asymptomatic and Symptomatic Adults

Soetikno RM, Kaltenbach T, Rouse RV, et al
JAMA. 2008;299:1027-1035


Colonic adenomas are recognized as the precursor to colon cancer in the vast majority of cases. The value of screening and/or surveillance colonoscopy has been related to the detection and removal of precancerous polyps. These polyps are typically classified by morphologic description as pedunculated or sessile. Both of these descriptions, however, imply some element of an intraluminal tissue elevation away from the colonic wall. There have been reports, primarily from Japan in the 1980s and early 1990s, suggesting that nonpolypoid (ie, flat or depressed) colorectal neoplasms (NP-CRNs) were common and had implications for cancer that were more ominous than raised CRNs.[1,2] These studies had purported that depressed lesions more commonly harbored in-situ or submucosal carcinoma. However, this was not evident in similar studies by western investigators evaluating North American patient populations.

This current study involved the consecutive evaluation (performed in a US veterans hospital) of 1819 veterans who underwent colonoscopy for screening, surveillance, or evaluation of symptoms suspect for colorectal disease. The majority of the cohort was white (79%) and only 4% were Asian. The mean age of the patients was 64 years and 95% were male.

The majority of the indications for the colonoscopic examination were screening (34%) and surveillance (36%). Only 2% of the exams were done for patients with inflammatory bowel disease. The remaining exams were done for clinically directed testing. All patients received 4 L of polyethylene glycol and 1 bottle of magnesium citrate or 90 mL of sodium phosphate taken orally the night before the procedure. Colonoscopy to the cecum was achieved in 94% of patients. The examinations were performed by 4 board-certified gastroenterologists using standard endoscopic equipment. In addition, chromoendoscopy was performed with indigo-carmine spray to highlight neoplastic lesions. Five years prior to the study, the investigators had participated in an exchange program with the National Cancer Hospital in Japan to develop the expertise and proficiency for diagnosing NP-CRNs. Furthermore, they had incorporated the use of indigo-carmine chromoendoscopy into routine colonoscopic practice for 3 years prior to beginning this study.

A total of 764 patients (42% of the study cohort) was identified as having at least 1 superficial CRN. Nonpolypoid CRNs were diagnosed in 9.35% (95% confidence interval [CI], 8.05%-10.78%). The prevalence of flat and depressed types of NP-CRNs was 8.58% (95% CI, 7.33%-9.96%) and 0.99% (95% CI, 0.59%-1.56%), respectively. The morphology of the neoplasms detected was reported as follows: only nonpolypoid lesions -- 5% (95% CI, 3.94%-5.99%); nonpolypoid and polypoid -- 4.4% (95% CI, 3.55%-5.50%); and only polypoid -- 33%(95% CI, 30.5%-34.9%).

Although nonpolypoid lesions accounted for only 15% of neoplasms overall, they contributed to 54% (95% CI, 35%-71%) of superficial carcinomas. Nonpolypoid morphology was strongly associated with findings of in-situ or submucosal invasive carcinoma (odds ratio [OR], 11.1; 95% CI, 4.98-24.8). After adjusting for the size of the lesion, the association of a nonpolypoid lesion with in-situ or submucosal invasive carcinoma continued to be significant compared with polypoid-shaped lesions when evaluated as: entire study cohort (OR, 9.78; 95% CI, 3.93-24.4), or screening (OR, 2.01; 95% CI, 0.27-15.3) and surveillance (OR, 63.7; 95% CI, 9.41-431) subpopulations.

Evaluation by morphologic category and after size adjustment revealed an association between flat lesions and in-situ or submucosal invasive carcinoma (OR, 5.18; 95% CI, 1.84-14.6). In addition, depressed morphology also showed a positive OR (OR 209; 95% CI, 44-1002), although there were too few depressed CRNs to statistically assess the magnitude of the association. The cancers seen in the small or flat lesions were smaller in diameter than those in the polypoid lesions. There was no apparent confounding effect of the endoscopist performing the exam.


This study provides supporting evidence that NP-CRNs are a relatively common finding among white patients in a single-center veteran's population, with a prevalence of about 9.3% The likelihood that NP-CRNs harbor serious pathology (in-situ or submucosal carcinoma) was more than 5 times greater than the rate seen in polypoid lesions after adjusting for polyp size. These data are consistent with reports in Asian populations and suggest that these lesions are not only prevalent but important, and may differ biologically from polypoid neoplasia. In previous colorectal cancer screening studies, male veterans tended to have more precancerous colon growths than other populations, thus, the rate of flat lesions in females or the general population may not be quite as high as reported in this study.

Japanese researchers became concerned about these flat lesions in the 1980s and 1990s, but North American studies had mixed results. This study challenges an inherent bias by US endoscopists who may have assumed, on the basis of prior available data, that flat growths were less common and less "dangerous." However, this present study provides some very provocative data that may serve as a wake-up call for endoscopists.

Before jumping on this bandwagon, data are needed to evaluate whether the biologic fingerprint of NP-CRNs indeed differs from that of polypoid lesions. If this is true, the prognosis may also differ. The answer can only come from longitudinal follow-up of patients with NP-CRNs. At this time, there is no basis for modifying the current surveillance recommendations for patients with NP-CRNs. In addition, the optimal method for enhancing colonoscopic imaging of NP-CRNs is uncertain. Chromoendoscopy with indigo-carmine was effective in this study, but other methods that may be technically easier to perform may also be equally effective, such as narrow-band imaging or fluorescence endoscopy, which do not require time for spraying the mucosa.

Computed tomography colonography (CTC; virtual colonoscopy) studies have shown that the sensitivity of detection for polyps ≥ 1 cm is approximately 90%.[3] Because this technology targets intraluminal extensions, it is very unlikely that it will detect flat lesions. Thus, any use of CTC for colorectal cancer screening will need to keep this in perspective, in addition to considering other issues such as radiation exposure and variable specificity.

There are 2 key take-home messages from this study. First, the message for physicians is that there is a large amount of data showing that these flat and depressed lesions, which are precursors of cancer and which were always believed to be a Japanese-predominant disease, are actually prevalent in the US population and perhaps on par with the cited prevalence from Japan. Therefore, the emphasis on quality colonoscopy has never been more apparent -- or justified. Second, the message for patients should be to focus on the importance of adequate colonic preparation for the colonoscopy. Although the quality of preparation was not a factor assessed in this study, previous studies have shown the importance of quality preparation and optimization of the colonoscopy examination.[4] Thus, patient adherence to the colonic preparation instructions should have a key role in allowing the endoscopist to maximize detection of all colonic lesions.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.