Non-Polypoid Colorectal Neoplasms are Relatively Common Worldwide

Flat adenomas are found commonly at colonoscopy throughout the world. Similarly, small, flat submucosally invasive cancers have been described worldwide but are relatively rare, accounting for 5% to 10% of all cancers detected at colonoscopy. Although there appears to be no difference in frequency of non-polypoid colorectal neoplasms between East and West, considerable variation has been reported by individual studies, probably because of lack of consistency when defining a flat lesion. Flat elevated lesions are the most common type of flat lesion and do not appear to have a greatly increased risk of harboring invasive malignancy; however, flat lesions with depression have a significant risk of malignancy and are probably the precursor lesions for most small, flat, or ulcerating cancers.

Polypoid adenomas are considered to be the precursors of most sporadic colorectal cancers (CRCs), following the adenoma-carcinoma sequence. However, up to 10% of CRCs detected at colonoscopy are small (<1.5 cm), flat lesions with little or no adenomatous component and an alternative hypothesis of CRC oncogenesis has been proposed characterized by relatively rapid progression from flat adenoma to invasive cancer. Flat adenomas may in fact be the precursors of the vast majority of ulcerating cancers, particularly in the proximal colon; therefore, identification and complete excision of flat adenomas is a crucial part of effective cancer prevention.

The term “flat adenoma” was first used by Muto in 1985 to describe 33 flat elevated lesions smaller than 1 cm, detected at colonoscopy or in surgically resected specimens. This article was considered largely irrelevant in the West, as flat adenomas were thought to be exclusively a Japanese phenomenon. However, with Japanese endoscopists traveling to many western centers and encouraging western colleagues to adopt careful examination technique, flat adenomas have now been described worldwide, albeit with widely varying prevalence. This is probably attributable to a combination of factors, including case-mix; geographic variation; quality of colonoscopy, both in terms of individual colonoscopist and the equipment used (degree of definition and advanced imaging); and the lack of standardized endoscopic and histologic definitions. The Paris endoscopic classification has been a major step forward in defining superficial neoplastic lesions. Flat adenomas are therefore described as superficial, non-polypoid colorectal neoplasia (NP-CRN) and can be slightly elevated (0–IIa), flush with the mucosa, ie, truly flat (0–IIb), slightly depressed (0–IIc), or a combination of elevation with central depression and vice versa (0–IIa+IIc; 0–IIc+IIa). Those with a depression (true IIc component) are thought to be of particular biologic significance and are more likely to harbor higher grade dysplasia or submucosal cancer when compared with flat elevated or polypoid lesions of the same size.

Worldwide prevalence of NP-CRN

Data from Asia

Early Japanese studies ( Table 1 ) suggested that NP-CRNs are relatively common. In the largest single-center series from Japan, Kudo and Kashida reported that of 21,262 neoplastic lesions resected over a period of 19 years, 43% were non-polypoid. Similarly, a high proportion of NP-CRN (48%) was described in a recent study from Malaysia, where Rajendra and colleagues prospectively studied the prevalence of flat colonic adenomas, defined as mucosal elevations with height less than half of the diameter of the lesion. A lower proportion of NP-CRN was demonstrated, however, in 2 retrospective pathologic studies from Japan. Kubota and colleagues examined 297 adenomas obtained from 300 surgically resected colons and found that 57 adenomas (19%) could be classified as flat (defined as having a flat surface) or depressed lesions. Similarly, Ajioka and colleagues reported that 140 (22%) of 643 adenomas could be described as having non-polypoid morphology (≤3 mm in height). Two prospective colonoscopy studies from Korea found a significantly lower frequency of NP-CRN. Kil Lee and colleagues assessed the morphology of 3263 lesions larger than 5 mm in 1883 patients and found a 7% frequency of NP-CRN, whereas in a study of 3360 patients by Park and colleagues, 6% of adenomas were considered flat.

Table 1

Prevalence of NP-CRN in the East (excludes cancer unless indicated)

Study	Year	No. of Patients Studied	No. of Patients with NP-CRN (% of All Patients)	Overall no. of Adenomas	No. of NP-CRN (% of All Adenomas Detected)	No. of NP-CRN with Depressed Morphology (% of flat adenomas)
Matsumoto et al	1992	—	32 —	—	34 ^b —	26 (76%)
Kudo	1993	—	—	563	225 (40%)	99 (44%)
Karita et al ^,^d	1993	—	29 —	—	27 —	—
Matsumoto et al	1994	895	36 (4%) ^a	634	34 (5%)	—
Minamoto et al	1994	—	—	—	17 —	6 (35%)
Mitooka et al	1995	1152	32 (3%)	—	37 ^b —	28 (76%)
Kubota et al	1996	—	—	297	57 (19%)	25 (44%)
Kudo et al	2000	—	—	13,718	6097 (44%)	229 (4%)
Ajioka et al	2000	—	—	643	140 (22%)	40 (29%)
Rajendra et al ^,^d	2003	426	7 (2%) ^c	29	14 (48%)	—
Kudo and Kashida ^,^d	2005	—	—	21,262	9189 (43%)	505 (5%)
Park et al ^,^d	2008	3360	207 (6%) ^c	—	—	—
LeeKil Lee et al ^,^d	2008	1883	189 (10%)	3128	228 (7%)	—