DNA flow cytometric and interobserver study of crypt cell atypia in inflammatory bowel disease

The pathological features and diagnostic reliability of crypt cell atypia (CCA) arising in inflammatory bowel disease (IBD) and its clinical significance are unknown.

DNA flow cytometry (FCM) was performed on 14 colon biopsies of CCA from seven IBD patients (male-to-female ratio, 5:2; mean age, 53 years; mean IBD duration, 15 years) using paraffin-embedded tissue. Seven gastrointestinal pathologists were asked to diagnose each biopsy as negative for dysplasia (NEG), indefinite for dysplasia (IND), low-grade dysplasia (LGD) or high-grade dysplasia (HGD) by morphology alone, then again with knowledge of FCM results. Aneuploidy was detected in all 14 biopsies, and five of eight biopsies (63%) also showed strong and diffuse nuclear staining for p53 in the areas of CCA. Six (86%) patients developed HGD (n = 5) or adenocarcinoma (n = 1) in the same colonic segment where CCA had been diagnosed within a mean follow-up time of 27 months. No follow-up information was available in the remaining one patient. When diagnoses were grouped as NEG or 'atypical' (including IND, LGD or HGD), the overall agreement rate of 76% (kappa = 0.51) based on morphology alone improved to 90% (kappa = 0.81) with knowledge of FCM results. Even when categorised as NEG or dysplasia (LGD or HGD) with each of the IND diagnoses reclassified into three categories (NEG, LGD or HGD) based on the degree of suspicion for dysplasia, the overall agreement rate of 63% (kappa = 0.25) based on morphology alone improved to 73% (kappa = 0.46) with knowledge of FCM results. However, when grouped as NEG, LGD or HGD, the overall agreement rate was less than 40% (kappa < 0.09) regardless of knowledge of FCM results.

The presence of aneuploidy, p53 positivity and development of HGD or adenocarcinoma on follow-up indicate that CCA likely represents a dysplastic lesion (at least LGD) and is a histological marker of neoplastic progression. Although the grading of CCA, largely based on cytological abnormalities, is subject to significant interobserver variability, CCA can be histologically identified and should lead to a recommendation of increased endoscopic surveillance, especially if aneuploidy is detected.