Efficacy of magnifying endoscopy in the differential diagnosis of neoplastic and non-neoplastic polyps of the large bowel.

Abstract

We have introduced magnifying colonoscopy into clinical practice and analyzed its diagnostic efficacy, especially regarding the ability to distinguish neoplastic from non-neoplastic polyps. The materials consisted of 923 polyps. After identifying the lesions during normal colonoscopy, a dye was sprayed, and then the zoom apparatus of the colonoscope was used to make a magnified observation at a maximum 100 times magnification. We classified the crypt orifices into six categories and labeled them A to F as follows: A, a medium round appearance; B, an asteroid appearance; C, an elliptic appearance; D, a small, round appearance; E, a cerebriform appearance; F, no apparent structural appearance. Forty-two of 923 polyps did not reveal any clear images of crypt patterns. The percentage of histologically neoplastic change in the lesions classified as A, B, C, D, E, and F were 10, 15.9, 93.7, 100, 94.8, and 87.5 percent, respectively. When we considered types A and B to represent a crypt pattern of non-neoplastic lesions, and types C, D, E, and F to represent neoplastic lesions, and when the lesions that did not show any clear images were classified as a misjudgment, the diagnostic accuracy of neoplastic lesions (sensitivity) was 92 percent and that of non-neoplastic lesions (specificity) was 73.3 percent. Overall, the diagnostic accuracy in differentiating neoplastic from non-neoplastic lesions was 88.4 percent. Twenty-three neoplastic lesions that were misjudged to be non-neoplastic were histologically adenoma with mild atypia in 22 and adenoma with moderate atypia in 1. Magnifying colonoscopy was considered to be useful in determining the indications for colonoscopic removal.