Objective This study was aimed at comparing high sensitivity C reactive protein (hsCRP) measurement with routine C reactive protein (CRP) evaluation as disease activity parameters in patients with ankylosing spondylitis (AS) and non-radiographic axial spondyloarthritis (nrSpA).
Methods A total of 269 patients (153 with AS and 116 with nrSpA) were included. Level of hsCRP was measured using particle-enhanced immunoturbidimetric method with the lowest detected level of 0.1 mg/litre. The hsCRP values were compared to results of routine turbidimetric CRP test with the lowest detected level of 6 mg/litre.
Results In the whole group of patients with AS, hsCRP showed a better than routine CRP correlation with clinical parameters. In the whole group of patients with nrSpA, hsCRP correlated with the level of enthesitis-related tenderness only. In the AS subgroup with a negative routine CRP (<6 mg/litre) there was a clear trend for an increased level of pain, stiffness and functional impairment in patients with higher hsCRP concentration. Such a trend was less pronounced in patients with nrSpA.
Conclusions hsCRP correlates better than routine CRP with clinical disease activity parameters in patients with axial SpA, especially in AS. Therefore, hsCRP could be superior to standard CRP in assessing disease activity in axial SpA.
Statistics from Altmetric.com
In ankylosing spondylitis (AS), only 40% to 60% of patients have elevated serum concentrations of C reactive protein (CRP).1,–,3 Patients with AS having ‘normal’ values of CRP may well have active disease according to clinical parameters, which accounts for the poor correlation between CRP levels and clinical disease activity in AS in general.1 Routine CRP assays often do not quantify the CRP concentrations below the upper normal limit, which may be relevant for disease activity assessment.
This study aimed at comparing high sensitivity C reactive protein (hsCRP) measurement with routine CRP evaluation as disease activity parameters in patients with AS and non-radiographic axial spondyloarthritis (nrSpA).
Patients and methods
A total of 269 patients (131 men and 138 women) from the German Spondyloarthritis Inception Cohort (GESPIC) were included in the current study. Design of the GESPIC, inclusion and exclusion criteria have been reported elsewhere previously.3 In brief, patients were included in the GESPIC if they had either AS fulfilling the modified New York criteria5 with a maximal duration of AS symptoms ≤10 years, or nrSpA fulfilling a modification of the European Spondylarthropathy Study Group (ESSG) criteria (human leucocyte antigen B27 (HLA-B27) positivity, acute anterior uveitis and dactylitis were added to the list of parameters of which at least one must be present in addition to inflammatory back pain3),6 but not fulfilling the modified New York criteria for AS, with a maximal duration of symptoms ≤5 years. Accordingly, 153 patients were classified as having AS and 116 as patients with nrSpA.
Disease activity and functional status assessment was performed using the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) and the Bath Ankylosing Spondylitis Functional Index (BASFI). General pain and nocturnal pain levels were measured on a 0–10 numerical rating scale. The presence of peripheral arthritis, enthesitis (Berlin score including 12 enthesitis points only of the lower limbs3 7), uveitis, as well as spinal mobility (by means of the Bath Ankylosing Spondylitis Metrology Index (BASMI)), were evaluated.
The median age of the patients was 33 (IQR 28–41) years (33 (IQR 29–41) years for patients with AS and 33.5 (IQR 26–41) years for patients with nrSpA); the median duration of disease was 4 (IQR 2–6) years (5 (IQR 3–7.5) years for AS and 3 (IQR 1–4.75) years for nrSpA). In all, 82.2% of the patients were HLA-B27 positive (83% of the patients with AS and 81% of the patients with nrSpA).
At the moment of examination 14.5% of the patients had peripheral arthritis, 22.7% enthesitis and 1.9% uveitis. In all, 48.3% of patients were treated with non-steroidal anti-inflammatory drugs, 17.1% received sulfasalazin, 8.2% methotrexate, 1.9% of patients were treated with anti-tumour necrosis factor α drugs and 7.4% received systemic steroids.
C reactive protein measurements
HsCRP levels were measured using particle-enhanced immunoturbidimetric method (COBAS Integra C reactive protein (Latex) test, Roche, Basel, Switzerland) with the lowest detected level of 0.1 mg/litre. HsCRP values were compared to results of routine turbidimetric CRP test with the lowest detected level of 6 mg/litre.
Comparisons of the two independent samples were performed using the Mann–Whitney U test. For the comparison of more than two independent samples Kruskal–Wallis H test was used. For correlation analysis Spearman's ρ coefficients were calculated. p Values of <0.05 were considered significant. Statistical analysis was performed using SPSS V.17.0 for Windows software (SPSS, Chicago, Illinois, USA).
The median level of a routine CRP (considering level of CRP <6 mg/litre as negative) in the whole group of the patients was 6.0 (IQR 0–16.0) mg/litre, while the median level of hsCRP was 4.8 (IQR 1.2–13.4) mg/litre. Levels of routine and hsCRP showed good correlation with each other: Spearman's ρ=0.809 (p<0.001) in the whole group, 0.852 (p<0.001) in the AS group and 0.736 (p<0.001) in the group of nrSpA. Levels of routine CRP and hsCRP were significantly higher in patients with AS in comparison to nrSpA: 8.0 (IQR 0–20.0) mg/litre versus 0 (IQR 0–10.0) mg/litre, respectively, for routine CRP (p=0.02) and 6.6 (IQR 1.6–15.2) mg/litre versus 2.7 (IQR 0.7–8.6) mg/litre, respectively, for hsCRP (p<0.001).
In the group of patients with AS (n=153) hsCRP showed somehow a better correlation than routine CRP with clinical parameters: general pain level (ρ=0.340, p<0.001 and 0.253, p=0.002 for hsCRP and routine CRP respectively), spinal pain level (ρ=0.247, p=0.002 and 0.177, p=0.029 respectively), night pain level (ρ=0.325, p<0.001 and 0.253, p=0.002 respectively), intensity of morning stiffness (ρ=0.213, p=0.008 and 0.168, p=0.039 respectively), BASFI value (ρ=0.271, p=0.001 and 0.238, p=0.003 respectively) and BASMI value (ρ=0.229, p=0.005 and 0.232, p=0.004 respectively). Only hsCRP demonstrated significant correlation with the level of enthesitis-related discomfort in tender areas (BASDAI question 4, ρ=0.173, p=0.032) and the overall BASDAI value (ρ=0.173, p=0.034). Levels of hsCRP and routine CRP both correlated well with the erythrocyte sedimentation rate (ρ=0.720, p<0.001 and 0.684, p<0.001 respectively).
In the group of patients with nrSpA (n=116) neither hsCRP nor routine CRP demonstrated a significant correlation with clinical parameters, except for the correlation between the hsCRP level and the level of enthesitis-related discomfort in tender areas (ρ=0.224, p=0.018). A good correlation with erythrocyte sedimentation rate was observed also in this subgroup: ρ=0.667, p<0.001 for hsCRP and 0.586, p<0.001 for routine CRP.
There were no statistically significant differences in the level of hsCRP and routine CRP between patients with (n=39) and without arthritis, as well as between patients with (n=61) and without enthesitis. In five patients with active uveitis (three with AS and two with nrSpA) median levels of hsCRP (37.5 mg/litre (IQR 12.2–71.2)) and routine CRP (22.0 mg/litre (IQR 20.0–88.5)) were significantly higher than in patients without uveitis (4.3 mg/litre (IQR 1.2–13.2), p=0.018 and 6.0 mg/litre (IQR 0–15.0), p=0.002 respectively).
A total of 126 patients (46.8% of the whole group), 65 patients with AS (42.5%) and 61 patients with nrSpA (52.6%), had a level of routine CRP <6 mg/litre. Of these 126, 120 patients (95.2%) also had a level of hsCRP <6 mg/litre; the level of hsCRP=0 mg/litre was detected in 4 patients (3.2%). The distribution of hsCRP values in patients with the level of routine CRP <6 mg/litre is presented in figure 1. According to the level of hsCRP patients with routine CRP level <6 mg/litre were further divided into three subgroups (tertiles): (1) hsCRP 0–1.9 mg/litre, (2) hsCRP 2.0–3.9 mg/litre and (3) hsCRP ≥4.0 mg/litre; see Table 1.
In general, there was only a weak correlation between clinical parameters and hsCRP level in patients with AS and nrSpA who has a negative CRP in the routine test (<6 mg/litre). This weak correlation is probably related to a highly skewed distribution of hsCRP values, with the majority of patients being in the first tertile. In patients with AS, a significant correlation with hsCRP was found only for the level of night back pain (ρ=0.309, p=0.012) and in the subgroup of patients with nrSpA hsCRP level correlated only with enthesitis-related level of local tenderness (ρ=0.284, p=0.031). Applying the Kruskal–Wallis test revealed significant differences between hsCRP subgroups for the level of night back pain (p=0.011) and a borderline significant difference for the level of spinal pain (p=0.067) in patients with AS (see Table 1). Pairwise comparisons with the Mann–Whitney U test showed that patients in the second tertile (hsCRP level 2.0–3.9 mg/litre) had significantly higher intensity of total pain, spinal pain and night pain than patients with the lowest CRP levels (from 0 to 1.9 mg/litre). Similarly, the intensity of nocturnal spinal pain in patients in the third tertile was significantly higher than in patients in the first tertile. A similar trend was observed in the group of patients with nrSpA who showed a negative routine CRP, although differences were not statistically significant.
CRP as a marker of systemic inflammation is an attractive parameter for disease activity assessment in patients with axial SpA (AS and nrSpA). However, the results of earlier studies demonstrating the lack of a correlation between CRP values and clinical parameters of AS activity1 2 made the role of this parameter for disease activity assessment uncertain. Routine methods of CRP detection (turbidimetric, nephelometric) had poor sensitivity in detecting concentrations of CRP below 6–10 mg/litre. At the same time, measurement of CRP concentrations below lowest routinely detected values (or below recommended by the World Health Organization (WHO) cut-off for the normal CRP level of 5 mg/litre8) could be relevant. An increased cardiovascular risk, an increased risk for non-traumatic osteoporotic fractures and a correlation with clinical parameters in rheumatoid arthritis were found in patients with slightly elevated hsCRP in comparison to low hsCRP.4 9,–,11
Accordingly, we compared highly sensitive and routine CRP assays in the current study for their ability to assess disease activity in patients with AS and nrSpA. There was a good correlation between the results of both assays, but the correlation with clinical parameters was better for hsCRP. Levels of CRP were significantly higher in patients with AS in comparison to nrSpA, as we reported before.3 In the AS subgroup we found a significant correlation of the CRP levels with clinical parameters (pain, morning stiffness, enthesitis-related local discomfort, BASDAI, BASFI, BASMI). In nrSpA patients a significant correlation was found only between hsCRP and enthesitis-related discomfort. Thus, in contrast to earlier reports,1 2 we showed significant correlation between CRP levels and clinical parameters reflecting disease activity in patients with AS, but not for nrSpA. Such a correlation was generally better for hsCRP.
Even more interesting was the relation between clinical parameters and hsCRP levels in patients with negative standard CRP (<6 mg/litre). Subgroup analyses according to the level of hsCRP demonstrated a clear trend for an increase of fatigue, pain and functional disability in patients with higher hsCRP levels in the AS subgroup. Moreover, levels of general pain, spinal pain and night pain were significantly higher in patients with AS, with levels of hsCRP 2–4 mg/litre in comparison to patients with hsCRP levels <2 mg/litre. This trend was also observed in the subgroup of patients with a level of hsCRP >4 mg/litre, but only the difference in the level of night pain reached significant level, possibly due to a small number of patients in the third tertile. A similar trend for an increase in clinical symptoms in patients with higher hsCRP values within the ‘normal’ range of CRP concentration was also observed in patients with nrSpA, although these differences were not significant.
Thus, hsCRP correlated better than standard CRP with clinical disease activity parameters in patients with AS and nrSpA. These data clearly indicate that hsCRP should be included in future axial SpA studies on correlation of CRP with clinical3 and MRI parameters,12 on the role of CRP in prediction of response to tumour necrosis factor α blockers13 and in the investigation of the recently proposed new AS disease activity score,14 which includes CRP. The results of such studies should be awaited for before the real advantage of hsCRP testing compared to conventional CRP in patients with axial SpA can be assessed. Notably, in many laboratories worldwide hsCRP testing has already replaced low sensitivity methods and is used routinely. The data presented here might help doctors in using and interpreting hsCRP values in daily clinical practice.
Competing interests None.
Ethics approval The study protocol was approved by the Berlin Ethics Committee and by the regional ethics committees of all participating centres. Writing informed consent was obtained from all patients.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.