Abstract
Background
Optimal management of patients with ulcerative colitis (UC) requires the accurate assessment of disease activity. Endoscopic evaluation is considered the gold standard approach, but it is invasive. We aimed to determine how strong patient reported outcomes, clinical scores and symptoms correlate with endoscopy for assessment of disease activity in UC patients.
Methods
171 patients were included prospectively and consecutively (age: 49 (IQR: 38-61) years, duration 12 (4-19)years, 79 females (46.2%), 57.3% extensive disease, 42.7% on biologicals) at the time of the colonoscopy. The 2 item patient reported outcome (PRO), partial MAYO, Simple Clinical Colitis Activity Index (SCCAI), Mayo endoscopic subscore (MES), Baron and Ulcerative Colitis Endoscopic Index of Severity (UCEIS) scores were calculated. C reactive Protein (CRP) and fecal calprotectin (FCAL) was available in 83 and 45.6% of patients. 17.0% had clinical flare, treatment was escalated in 14.6% of patients. Sensitivity, specificity, PPV and NPV values were calculated, ROC analysis and K-statistics were performed.
Results
Rectal bleeding (RBS), stool frequency (SF) subscore of 0, or total PRO2 remission (RBS 0 and SF ≤1), partial MAYO (≤2) and SCCAI (≤2.5) remission were similarly associated to mucosal healing defined by MES (0 or ≤1) or Baron (0 or ≤1) scores (Table 1).
PRO2 (AUCMES0/Baron0: 0.770/0.740, AUCMES0-1/Baron0-1: 0.868/0.858), SF (AUCMES0/Baron0:0.751/0.724, AUCMES0-1/Baron0-1:0842/0.820), RBS (AUCMES0/Baron0: 0.718/0.698, AUCMES0-1/Baron0-1: 0.814/0.845) partial Mayo (AUCMES0/Baron0: 0.823/0.788, AUCMES0-1/Baron0-1: 0.927/0.902) and SCCAI (AUCMES0/Baron0: 0.767/0.752, AUCMES0-1/Baron0-1:0.888/0.867) were similarly associated with mucosal healing in a ROC analysis.
There was a strict association between MES 0 and Baron 0 (k=0.917) and UCEIS <4 and MES 0-1 (k=0.813), while moderate to fair agreement between UCEIS <4 and MES 0 (K=0.471) or Baron 0 (K=0.414)/Baron 0-1 (K=0.353), and between MES 0-1 and Baron 0-1 (K= 0.350) scores.
Agreement between CRP and clinical remission or endoscopic healing (MES/Baron) was poor (K~0.2), while agreement between FCAL (>100 or >250) and RBS-PRO2 remission (K>100 or >250: 0.44-0.60) or pMAYO (K>100 or >250: 0.41-0.59) or MES/Baron 0 was moderate to good (K>100:0.53-0.52 and K>250:0.57-0.53).
Conclusion
We found no difference across accuracy of RBS, SF, PRO2, partial Mayo and SCCAI in predicting endoscopic healing. A strong association was found with high PPV for MES/Baron ≤1 and high NPV for MES/Baron 0. FCAL, but not CRP was associated to clinical and endoscopic remission.
Letter: histological assessment of disease activity in ulcerative colitis - the problem of score evaluation and validation
