Non-conventional dysplasia in inflammatory bowel disease is more frequently associated with advanced neoplasia and aneuploidy than conventional dysplasia

Aims Several different non-conventional morphological patterns of epithelial dysplasia have been recently described in inflammatory bowel disease (IBD), but there is limited information regarding their clinicopathological and molecular features, as well as potential risk for high-grade dysplasia (HGD) or colorectal cancer (CRC) compared with conventional dysplasia developing in IBD. Methods and results A total of 317 dysplastic lesions from 168 IBD patients were analysed. All lesions were re-reviewed and subtyped as either conventional [including tubular adenoma-like (n = 183) and tubulovillous/villous adenoma-like (n = 56)] or non-conventional dysplasia [including dysplasia with increased Paneth cell differentiation (DPD, n = 40), crypt cell dysplasia (CCD, n = 14), goblet cell deficient (GCD, n = 10), hypermucinous (n = 7), sessile serrated lesion (SSL)-like (n = 4) and traditional serrated adenoma (TSA)-like (n = 3)]. DNA flow cytometry was performed on 70 low-grade conventional (n = 24) and non-conventional (n = 46) dysplastic biopsies to determine their malignant potential and molecular pathways to HGD or CRC. Eleven sporadic tubular adenomas with low-grade dysplasia (LGD) were utilised as controls. Seventy-eight non-conventional dysplastic lesions were identified in 56 (33%) of the 168 patients, whereas 239 conventional dysplastic lesions were identified in 149 (89%) patients. Although both non-conventional and conventional dysplasias were most often graded as LGD at diagnosis (83% and 84%, respectively), non-conventional dysplasia (38%) was more likely to develop HGD or CRC in the same colonic segment than conventional dysplasia (19%) on follow-up (P < 0.001). Almost half (46%) of non-conventional dysplastic samples showed aneuploidy, whereas only 8% of conventional dysplasia (P = 0.002) and 9% of sporadic tubular adenomas (P = 0.037) did. Also, non-conventional dysplasia more frequently presented as a flat/invisible lesion (41%) compared with conventional dysplasia (18%) (P < 0.001). Among the non-conventional subtypes (n = 78), DPD was the most common (n = 40; 51%), followed by CCD (n = 14; 18%), GCD (n = 10; 13%), hypermucinous (n = 7; 9%), SSL-like (n = 4; 5%) and TSA-like (n = 3; 4%) variants. Hypermucinous dysplasia (mean = 2.1 cm) was significantly larger than DPD, SSL-like, TSA-like and GCD variants (mean = 1.0, 1.2, 1.2 and 1.9 cm, respectively) (P = 0.037). HGD or CRC was more likely to be associated with CCD (n = 13; 93%), hypermucinous (n = 4; 57%), GCD (n = 4; 40%) and SSL-like (n = 3; 75%) variants than DPD (n = 6; 15%) and TSA-like dysplasia (n = 0; 0%) on follow-up (P < 0.001). Furthermore, the rate of aneuploidy was significantly higher in CCD (100%), hypermucinous (80%) and GCD (25%) variants than in DPD (12%), SSL-like dysplasia (0%) and TSA-like dysplasia (0%) (P < 0.001). Conclusions Non-conventional morphological patterns of dysplasia are not uncommon in IBD, detected in 33% of the patients. The higher frequencies of advanced neoplasia (HGD or CRC) and aneuploidy in non-conventional dysplasia, in particular CCD, hypermucinous and GCD variants, suggest that they may have a higher malignant potential than conventional dysplasia or sporadic tubular adenomas, and thus need complete removal and/or careful follow-up. Greater than 40% of non-conventional dysplasia presented as a flat/invisible lesion, suggesting that IBD patients may benefit from random biopsy sampling in addition to targeted biopsies. The majority of non-conventional subtypes appear to develop via the chromosomal instability pathway, whereas an alternative serrated pathway may be responsible for the development of at least a subset of SSL-like and TSA-like dysplasias.