Introduction Inflammatory bowel disease-associated colorectal cancers (IBD-CRCs) occur infrequently but are associated with a higher mortality than sporadic CRCs. The poorly defined genetic and molecular pathogenesis of IBD-CRCs limits our ability to develop effective prevention, detection and treatment strategies. Since molecular changes identified at the time of surgery are most likely to represent informative biomarkers we elected to study primary IBD-CRC.

Method 6The study was approved by the Lothian NHS Research Scotland BioResource (ref:15/ES/0094). Whole-exome sequencing and a comprehensive mutational analysis was performed on 35 FFPE IBD-CRCs and matched normal lymph node pairs from 32 patients (16 CD). Two of the patients had synchronous cancers.

Results Ten (29%) of 35 IBD-CRCs were hypermutated and had a significantly higher 10 year survival than the non-hypermutator cancers (p=0.04). All hypermutator cancers were in the proximal colon; seven had loss of expression of MLH1, of which 5 had MLH1 promoter methylation. Two had somatic mutations in the proof-reading domain of DNA POLE. The hypermutated IBD-CRCs had a predicted increased neo-epitope load suggesting that these cancers may benefit from immune checkpoint blockade. Six distinct IBD-CRC mutational signatures were identified with four corresponding to known mutational mechanisms. Five genes including TP53, APC, PIK3CA, and KRAS were significantly mutated in the non-hypermutator IBD-CRCs and fourteen genes in the hypermutator IBD-CRCs, including RNF43 and AIM2, which are clinically actionable. Eleven of the 32 patients (34%) had germline mutations that may confer an inherited susceptibility to colorectal cancer; only two of these specific variants have a clinical pathogenic effect.

Conclusion The hypermutated IBD-CRCs are associated with defects in MMR and DNA POLE; with a predicted higher neo-epitope load, which could be exploited using immunotherapies. Loss of MLH1 expression could be evaluated in potential colonic dysplastic or cancerous lesions detected in IBD patients. The identification of novel significantly mutated genes in hypermutated IBD-CRCs could be used to stratify therapy. These approaches would complement and individualise current surveillance and treatment programs for IBD-CRC.

Disclosure of Interest None Declared