Background: Due to its wide accessibility, fast availability and proven correlation with disease activity, C-reactive protein (CRP) remains an essential tool in the management of Crohn's disease (CD), namely when it comes to decide the appropriate treatment in a given situation. However, the correlation of CRP with CD activity is not perfect. It is therefore of great importance to identify the group of patients with active disease and negative CRP.
Methods: We performed a retrospective case-control study, with inclusion of CD patients with proven active disease as demonstrated by endoscopic and/or radiologic examinations. The CRP's cut-off value used to separate patients in two groups (cases and controls) was 1 mg/dL. Demographic, phenotypic and clinical characteristics were collected. Statistical analysis was performed with SPSS 20.0.
Results: We included 88 patients (42 men, 46 women) with a mean age of 34±10.8 years and median of disease duration of 1.23 years (interquartile range 0.07–6.16). Twelve (13.6%) of these patients had negative CRP. There weren't statistically significant differences in CD activity between cases and controls, as evaluated by Harvey-Bradshaw index. Upon exploratory analysis, there were statistically significant differences regarding gender as 21.4% of men vs 6.5% of women had a negative CRP (p=0.04). Even though location was not a significant predictor, all patients with a negative CRP had ileal involvement. On multivariate analysis, gender remained a significant predictor, with an Odds ratios (OR) of 5.64 [Confidence Interval (CI) 95% 1.30–24.39; p=0.02). There was also a tendency to a higher probability of negative CRP in isolated ileal disease (OR 3.87; CI95% 0.97–15.34; p=0.055). There were no differences in age, behaviour, disease duration, previous abdominal surgery or smoking status.
Conclusions: Despite being a useful tool CRP has some limitations and it can be negative in cases of active disease. In patients with the identified characteristics (ie men with ileal disease) other methods should be used to exclude with confidence the presence of inflammatory activity.