BACKGROUND/AIMS: Serrated adenomas (SAs) of the colorectum can be broadly divided into two subtypes: type I more closely mimicking hyperplastic polyps, and type II unequivocal traditional adenomas. The aim of this study was to clarify their differential clinicopathologic and colonoscopic features. METHODOLOGY: A total of 127 SAs (53 type I, 52 type II and 22 admixed type I+II) were investigated and colonoscopic surface patterns were divided into three categories: speckled, granular and cerebriform. RESULTS: The cerebriform pattern was most frequently observed in all SA types. Types I+II (median size, 7.5 mm) or type II SAs (median size, 10 mm) were generally sessile or pedunculated polyps in the rectosigmoid colon whereas some type I lesions (median size, 5 mm) demonstrated a flat-elevated morphology and were found in the ascending colon and cecum. Co-existing (2/127: 1.6%) invasive carcinomas were only detected with type II SAs. In contrast, synchronous invasive carcinomas distant from SAs were more frequently observed with type I (31%) than types I+II (5%) or II (12%). CONCLUSIONS: Clinicopathologic differences are apparent among the types of SAs. A type II SA-invasive carcinoma sequence might exist. We stress recognition of type I SA as a neoplastic, rather than a hyperplastic lesion, often accompanying invasive carcinomas at a distance from the SA.
BACKGROUND/AIMS: Serrated adenomas (SAs) of the colorectum can be broadly divided into two subtypes: type I more closely mimicking hyperplastic polyps, and type II unequivocal traditional adenomas. The aim of this study was to clarify their differential clinicopathologic and colonoscopic features. METHODOLOGY: A total of 127 SAs (53 type I, 52 type II and 22 admixed type I+II) were investigated and colonoscopic surface patterns were divided into three categories: speckled, granular and cerebriform. RESULTS: The cerebriform pattern was most frequently observed in all SA types. Types I+II (median size, 7.5 mm) or type II SAs (median size, 10 mm) were generally sessile or pedunculated polyps in the rectosigmoid colon whereas some type I lesions (median size, 5 mm) demonstrated a flat-elevated morphology and were found in the ascending colon and cecum. Co-existing (2/127: 1.6%) invasive carcinomas were only detected with type II SAs. In contrast, synchronous invasive carcinomas distant from SAs were more frequently observed with type I (31%) than types I+II (5%) or II (12%). CONCLUSIONS: Clinicopathologic differences are apparent among the types of SAs. A type II SA-invasive carcinoma sequence might exist. We stress recognition of type I SA as a neoplastic, rather than a hyperplastic lesion, often accompanying invasive carcinomas at a distance from the SA.