Background/Purpose: The average delay in spondyloarthritis (SpA) diagnosis after symptom onset is one of the longest among the inflammatory rheumatic diseases. Elevated C-reactive protein (CRP) has been incorporated as one of the features for ASAS axial SpA criteria and in the Berlin diagnostic algorithm. However, CRP levels are elevated in only a minority of early SpA patients which limits its potential diagnostic value. More sensitive tests called high-sensitivity CRP (hsCRP) have been developed and can detect lower concentrations of CRP compared with traditional methods. Therefore, hsCRP could be more sensitive than traditional CRP for diagnosis in axial SpA. The aim of this study was to assess if hsCRP measurement contributes to early axial SpA diagnosis compared with traditional CRP measurement.

Methods: Baseline data from 648 patients with inflammatory back pain (IBP) duration >3 months but <3 years from DESIR cohort was used. Design and inclusion criteria have previously been reported (1). Levels of CRP and hsCRP were measured in serum at baseline. The cut-off values selected to define positive hsCRP and CRP were ≥ 2 mg/l and ≥ 5mg/l, respectively.

Results: Based on the ASAS axial SpA criteria, 444 (69%) patients were classified as SpA and 203 (31%) patients as no SpA. Patients’ characteristics and lab results are shown in table 1. Serum levels of CRP and hsCRP were higher in SpA versus no SpA. In the subgroup of patients with a negative CRP, mean serum levels of hsCRP were also higher in SpA patients compared with no SpA patients (1.7 mg/L vs 1.5 mg/L, p 0.03). Moreover, after dichotomizing hsCRP, more patients within the SpA group had a positive hsCRP versus the no SpA group, although this difference was not statistically significant (p=0.06) (table 2). 

Finally, we investigated how many extra patients from the no SpA group (n=203) would be classified as SpA substituting the traditional CRP by hsCRP in the clinical arm (HLA-B27 arm) of the ASAS axial SpA criteria. Only 4 (2%) extra patients had 2 axial SpA features instead of only 1 feature (IBP) substituting hsCRP by CRP (195 vs 191 patients), but none of them was HLA-B27 positive. Consequently, none of the no SpA patients met ASAS criteria applying this modification.

Conclusion: In patients with a normal CRP, hsCRP is increased in axial SpA patients compared with patients without SpA. However, hsCRP measurement in patients with IBP may not add any extra value for early axial SpA diagnose compared with CRP measured by traditional method.

(1) Dougados M et al. Joint Bone Spine. 2011 Dec;78(6):598-603.

Table 1: Characteristics of patients

Table 2: Serum levels of hs-CRP in patients with normal CRP values (<5 mg/L)