High-sensitivity C-reactive protein for identification of disease phenotype, active disease, and clinical relapses in Crohn's disease: a marker for patient classification?
Lajos Sandor Kiss,Maria Papp,Barbara Dorottya Lovasz,Zsuzsanna Vegh,Petra Anna Golovics,Eszter Janka,Eva Varga,Miklos Szathmari,Peter Laszlo Lakatos
DOI: https://doi.org/10.1002/ibd.21933
Abstract:Background: C-reactive protein (CRP) is a traditional nonspecific marker of inflammation, with Crohn's disease (CD) being associated with a strong CRP response. Thus far, no clear cutoff values have been determined. The authors' aim was to investigate whether high-sensitivity (hs)-CRP is useful for the identification disease phenotype, active disease, and relapse during follow-up, using a classification based on the hs-CRP value at diagnosis. Methods: In all, 260 well-characterized, unrelated, consecutive CD patients (male/female: 120/140; duration: 7.0 ± 6.1 years), with a complete clinical follow-up, were included. Hs-CRP, clinical activity according to the Harvey-Bradshaw Index, and clinical data (disease phenotype according to the Montreal Classification, extraintestinal manifestations, smoking habits, medical therapy, and surgical events) were prospectively collected between January 1, 2008 and June 1, 2010. Medical records prior to the prospective follow-up period were analyzed retrospectively. Results: In all, 32.3% of CD patients had normal hs-CRP at diagnosis. Elevated hs-CRP at diagnosis was associated with disease location (P = 0.002), noninflammatory disease behavior (P = 0.058), and a subsequent need for later azathioprine/biological therapy (P < 0.001 and P = 0.024), respectively. The accuracy of hs-CRP for identifying patients with active disease during prospective follow-up was good (area under the curve [AUC]: 0.82, cutoff: 10.7 mg/L). AUC was better in patients with an elevated hs-CRP at diagnosis (AUC: 0.92, cutoff: 10.3 mg/L). In Kaplan-Meier and Cox-regression analyses, hs-CRP was an independent predictor of 3- (P = 0.007) or 12-month (P = 0.001) clinical relapses for patients in remission who had elevated hs-CRP at diagnosis. In addition, perianal involvement (P = 0.01) was associated with the 12-month relapse frequency. Conclusions: Our data suggest that hs-CRP positivity at diagnosis is associated with disease location and behavior, and in patients who are hs-CRP positive at diagnosis, is an accurate marker of disease activity and a predictor of short- and medium-term clinical flare-ups during follow-up.