Endoscopic detection of aberrant crypt foci by magnification chromoscopy: A blinded comparison against dissecting microscopy and histology in an Ex Vivo model

Introduction: Aberrant crypt foci (ACF), the putative precursors of colorectal adenomas, may be relevant markers for assessing cancer risk and monitoring preventive interventions. The clinical utility of ACF, however, will depend on valid techniques to enumerate them in vivo. ACF cannot be reliably seen by conventional colonoscopy, and recent efforts using magnification chromoscopy (MC) have yielded mixed results. Aim: The purpose of this study was to evaluate the accuracy of MC for detection of ACF compared to dissecting microscopy and histology in an ex vivo model. Methods: To assess the accuracy of MC to detect ACF, surgical specimens from 14 subjects were studied. The colonic mucosa was washed with 10% acetylcysteine to remove mucus prior to staining with 0.2% methylene blue for 3 minutes. Each specimen was then cut into 4 cm² segments and examined with the Olympus CF-200Z magnifying colonoscope at a distance of 2 to 3 mm (magnification 30X-40X). Dissecting light microscopy was performed by a blinded investigator after discarding the muscularis propria of the methylene blue stained tissue. For histological examination, 4 mm portions were embedded in paraffin and 5 p.m sections were cut parallel to the mucosal surface, stained with hematoxylin-eosin, and examined by a blinded pathologist. The ACF density was calculated by dividing the number of ACF visualized by the surface area. The correlation coefficient was determined by linear regression analysis. Results: A total of 572 cm² of the colonic mucosa (range 10 to 56 cm², mean±SEM 40.9±4.8, median 52 cm²) from 14 subjects (9 colorectal cancer, 2 diverticulosis, 1 volvulus, 1 angiodysplasia, 1 adenoma) was studied. The age ranged from 37 to 93 years (median 74, mean ± SEM 70.6 ± 3.9 years) and the ACF density from 0 to 0.25 ACF/cm². With a correlation coefficient of 0.94, the ACF density determined by dissecting microscopy correlated well with histology. The correlation coefficients were 0.8 and 0.6 for chromoscopy compared to dissecting microscopy and histology, respectively. MC had a specificity of 76.2% and a sensitivity of 47,5% to detect ACF. Summary: Magnification chromoscopy appears to be a moderately sensitive and relatively specific technique for detection of ACF ex vivo. ACF density values obtained by dissecting microscopy and histology correlated well with each other. The correlation between chromoscopy and dissecting microscopy, or histology, may be less than ideal. Conclusion: Magnification chromoscopy is a moderately sensitive technique. It can detect ACF with relatively good accuracy ex vivo and may become a useful technique for estimating the ACF density in vivo if a well definable region can be examined thoroughly. Dissecting light microscopy may be used for confirmation of lesions thought to be ACF by chromoscopy.