Background: Inflammatory bowel diseases (IBD) affects 0.7% of Americans and includes ulcerative colitis (UC) or Crohn’s disease (CD). 5-15% of patients have findings that overlap between UC and CD and are labeled IBD-unclassified (IBDU). Because patients with IBDU are routinely excluded from clinical trials for UC or CD, knowledge of therapeutic safety and effectiveness are limited.
Objectives: We sought to examine reasons that patients are given a diagnosis of IBDU and differences in treatment, endoscopic disease severity, and symptoms between the three diseases.
Methods: We used data from the Study of a Prospective Adult Research Cohort with IBD (SPARC IBD), part of the Crohn’s & Colitis Foundation’s IBD Plexus. SPARC IBD is a prospective, longitudinal, multicenter cohort study of patients which has enrolled over 5000 patients with IBD from 21 sites. The CD cohort was limited to patients with isolated colonic disease. Investigators listed reasons for IBDU diagnosis from a defined set of options. Medications and presence of 8 key symptoms were measured at the most recent encounter. Endoscopic disease activity based on most recent colonoscopy. Endoscopic Mayo score >1 or Simple Endoscopic Score for CD (SES-CD) >2 were considered evidence of active disease. We used logistic regression adjusted for age, years since diagnosis, sex and medications in use at the time of disease activity assessment to separately compare IBDU versus UC and CD. Bonferroni corrected statistical significance was defined as P<.0017.
Results: The study included 109 participants with IBDU, 1558 with UC, and 412 with CD colitis. The median age was 40 years (IQR 30-55) and 52.3% were female. Endoscopic appearance of ulcerations (32.5%) and skip areas (26.0%) were the most common reasons for IBDU diagnosis. Patients with IBDU were more often treated with 5ASA (34.9% IBDU, 37.3% UC, 19.0% CD), targeted small molecule drugs (12.8% IBDU, 8.7% UC, 1.9% CD) and less often with biologics (IBDU 52.3%, UC 49.2%, CD 69.2%) when compared to CD (P <.001 for each), but not UC (P>0.14 for each). Use of immunomodulator drugs, systemic and rectal steroids did not differ between diseases. Active inflammation anywhere in the colon was more common in IBDU (34.7%) than UC (22.6%, nominal P=.03), but not different than CD (35.9%, P=0.38). None of the symptoms met the Bonferroni adjusted threshold for statistical significance, however fecal incontinence was more common in IBDU (16.3%) than UC (9.3%, nominal P=.03), but not CD (13.3%).
Conclusions: Treatment of IBDU more closely reflected UC than CD, but disease activity that was more similar to CD than UC, suggesting that IBDU is more refractory to treatment than UC. Prospective studies are needed to define optimal treatment for IBDU.
