Introduction Terminal ileitis (TI) when identified at endoscopy is generally assumed to be secondary to Crohn’s disease (CD). However, TI may be due to other causes including infection, malignancy, radiation, vasculitis and autoimmune disease. In the presence of histological findings such as granulomas, cryptitis, crypt abscesses, fissuring and ulceration, a diagnosis of CD can be inferred. In the absence of such histological correlation, a diagnosis of CD may be reinforced by finding inflammation on colonoscopy or small bowel studies in other parts of the gut, suggestive of skip lesions. We hypothesise that patients diagnosed with TI on endoscopy do not all have CD, but may be inaccurately labelled as such.
Methods Single-centre retrospective analysis of 56 patients diagnosed with TI on endoscopy at a district general hospital in London between 2007–2013. Data obtained from endoscopic reporting tool and patient records were correlated with histology and imaging. The number of patients with the finding of ileitis on endoscopy subsequently diagnosed with CD was studied and the basis of this diagnosis was evaluated.
Results 68% of patients (38/56) with TI on endoscopy had histological confirmation of ileitis: 11 of whom had characteristic features of Crohn’s on histology, whilst the remaining 27 had mildly active ileitis on histology. 22 of these 27 patients underwent additional investigations to confirm (19/22) or exclude IBD (3/22). 5 of the 27 patients with histological ileitis that was not typical of CD were diagnosed with IBD, without further evidence to demonstrate this. In total 35/38 (92.1%) patients with histological and endoscopic evidence of inflammation were diagnosed with CD. Of the 18/56 patients (32%) with TI on endoscopy but normal histology, 9 had further investigations including MRI, barium studies and US abdomen to exclude CD. 4 patients were diagnosed with CD despite normal histology. 6/18 patients with normal histology were lost to follow up, although all patients diagnosed with IBD were followed up in GI clinic.
Conclusion We conclude that although the majority of patients with TI on endoscopy have CD, 1/3 of our patients had no histological correlation of inflammation. 16% of patients with endoscopic TI with mildly active ilieits on histology had no further imaging to authenticate a diagnosis of CD and 8% of patients with normal histology with no further investigations were labelled as CD. The latter group of patients may be inaccurately labelled with a diagnosis of CD and along with this the potential stigma associated with a chronic ailment, risk of escalation therapy with immununosuppressives/biologics and occasionally unnecessary surgery.
Disclosure of Interest None Declared.