P393. Patient-reported symptom measures differ in their association with mucosal healing in adults with moderately to severely active ulcerative colitis: Results from ULTRA 1 and 2
J.-F. Colombel1, W.J. Sandborn2, W. Reinisch3, A.M. Robinson4, W. Wang4, B. Huang4, A. Lazar5, R. Thakkar4, 1Icahn School of Medicine at Mount Sinai, Division of Gastroenterology, New York, United States, 2University of California, San Diego, Division of Gastroenterology, La Jolla, United States, 3McMaster University, Department of Medicine, Hamilton, Canada, 4AbbVie Inc, GPRD, North Chicago, United States, 5AbbVie Deutschland GmbH & Co, KG, GPRD, Ludwigshafen, Germany
The efficacy of adalimumab (ADA) for the induction and maintenance of clinical remission and mucosal healing (MH) in adults with moderate-to-severe ulcerative colitis (UC) has been demonstrated in the randomised double-blind (DB) placebo (PBO)-controlled trials ULTRA 1  and ULTRA 2 . Clinical symptoms may serve as measures of response to treatment, but may not accurately reflect MH status. This post-hoc analysis evaluated the association of the Mayo score components rectal bleeding subscore (RBS) and stool frequency subscore (SFS) with endoscopy subscore (ES) in the 52-week (wk) studies ULTRA 1 and ULTRA 2.
In ULTRA 1 and ULTRA 2, patients (pts) with a Mayo score of 6–12 and ES ≥2 were randomised to PBO or ADA (160/80 mg or 80/40 mg, ULTRA 1 only) at wks 0/2 followed by PBO or 40 mg ADA every other wk (eow). Pts in ULTRA 1 received open-label (OL) ADA from wks 8 or 12 to wk 52. Pts could move to weekly ADA (after OL eow ADA in ULTRA 2) for inadequate response. RBS and SFS were assigned using the worst values of the 3 days prior to each study visit from pt diaries. RBS = 0 indicated no visible blood in stool; SFS = 0 indicated “normal” stool frequency for that pt. All data (DB or OL) from ADA-160/80 mg-treated pts who had subscores at wks 8, 32, and 52 were used to assess the association of RBS and SFS values of 0 and MH (ES = 0–1 or ES = 0) at each visit using the parameters sensitivity (sens; probability that RBS/SFS = 0 when MH = yes), specificity (spec; probability that RBS/SFS >0 when MH = no), positive predictive value (PPV; probability that MH = yes when RBS/SFS = 0), and negative predictive value (NPV; probability that MH = no when RBS/SFS >0).
Of 470 ADA-160/80 mg-treated pts, 145 had subscores at each visit. Sens, spec, PPV, and NPV analyses (Table) showed that RBS = 0 was likely (>75% of the time) when MH was present (ES = 0–1 or ES = 0), while normal SFS was less frequent (∼25% of the time). This relationship did not change with time: among pts with MH (ES = 0–1) at all 3 visits, a substantially higher proportion had RBS = 0 (80.4%, 82.1%, and 89.3% at wks 8, 32, and 52, respectively) compared with SFS = 0 (28.6%, 26.8%, and 33.9%).
|Subscore||Time point||Sensitivity, %||Specificity, %||PPV, %||NPV, %|
|ES = 0–1||ES = 0||ES = 0–1||ES = 0||ES = 0–1||ES = 0||ES = 0–1||ES = 0|
|RBS = 0||wk 8||76.9||80.6||70.2||51.8||75.0||31.2||72.3||90.8|
|SFS = 0||wk 8||24.4||29.0||95.5||88.6||86.4||40.9||52.0||82.1|
Patient-reported typical symptoms of UC differed in their association with MH. Absence of rectal bleeding was associated with MH, whereas stool frequency may remain elevated even in pts with sustained MH. RBS may be a more reliable indicator of MH than SFS.
1. Reinisch W et al. Gut 2011; 60: 780–7.
2. Sandborn WJ et al. Gastroenterology 2012; 142: 257–65.