S1453 Disease Characteristics Associated With Bowel Urgency in Crohn’s Disease: Results From the Phase 3 VIVID-1 Trial

Introduction: Data on bowel urgency (BU) in Crohn's disease (CD) are limited, and associated factors are not well identified. Here we evaluate the association of BU with baseline (BL) disease characteristics in patients with moderately to severely active CD in the randomized, double‐blind, double‐dummy, treat‐through design Phase 3 VIVID‐1 trial (NCT03926130). Aims & Methods: BU was assessed using the average Urgency Numeric Rating Scale (UNRS) (range 0‐10) from 7 days prior to the visit. Adult patients as previously described1, that had an UNRS ≥3 and no missing values at BL were included in this analysis. The association of UNRS with Crohn's Disease Activity Index (CDAI), abdominal pain (AP), stool frequency (SF), Simple Endoscopic Score for CD (SES‐CD) total score, disease location, count of inflamed intestinal segments, rectal involvement, presence of narrowing, presence of perianal fistulae, number of bowel resections and inflammatory biomarkers were analyzed. Other factors explored included sex, age, age at diagnosis, duration of CD, BL corticosteroid and immunomodulator use, and prior biologic failure. Random forest method (RFM) was applied to rank variables by importance associated with BU score at BL. Pearson correlation coefficients (r‐value) were calculated for continuous baseline variables. Results: At BL, 73.4% (782/1065) patients randomized to placebo (n=141), MIRI (n=433), or ustekinumab (n=208) reported a UNRS ≥3 with a median UNRS of 7.0 Among BL disease characteristics, CDAI total score and AP were the top two important factors based on RFM (CDAI=0.701; AP=0.398) and had moderate positive correlations (CDAI=0.479; AP=0.384) with UNRS (Table). Notably, disease characteristics such as SF (RFM=0.236; r=0.286), SES‐CD total score (RFM=0.056; r=0.118 ), age at diagnosis (RFM=0.037; r=‐0.018), age (RFM=0.033; r=‐0.022), faecal calprotectin (RFM=0.029; r=0.031), C‐reactive protein (RFM=0.028; r=0.050), count of inflamed segments (RFM=0.021; r=0.069), duration of CD (RFM=0.015; r=‐0.007), presence of perianal fistula (RFM=0.009; r=0.090), and number of bowel resections (RFM=0.000; r=‐0.036) all had weak correlations with UNRS. BL corticosteroid (RFM=0.015) and immunomodulator (RFM=0.007) use, disease location (RFM=‐0.005), rectal involvement (RFM=0.011), presence of narrowing (RFM=0.007), prior biologic failure (RFM=0.003), and sex (RFM=0.007) had weak correlations with UNRS implied by RFM. (Table present) Conclusion: Most patients reported BU ≥3 at BL. The only moderate positive correlations were observed between BU, clinical disease severity (CDAI), and AP. Even though the association of BU with other disease characteristics such as extent and severity of endoscopic inflammation, rectal involvement, presence of active perianal disease, disease location, previous resection, and duration of CD were weak, they potentially interact and contribute together resulting in BU.