C-reactive protein thresholds for discriminating active disease in axial spondyloarthritis: should we lower them?

Although C-reactive protein (CRP) is the main inflammatory biomarker used for axial spondyloarthritis (axSpA) assessment, its sensitivity for detecting active disease is low. We aimed to analyse CRP thresholds capable of discriminating across disease activity states in axSpA.

Two hundred consecutive patients with axSpA were recruited (with/without biologics). Discriminative CRP thresholds were determined by the Youden index, while their sensitivity/specificity was evaluated by the area under the ROC curve (AUROC). Sensitivity analyses were performed based on exposure to biologic drugs.

One hundred and twenty-two men and 78 women were included, mean age 43.5±11.6 years, mean age at diagnosis 34±10.7 years, average disease duration 8.6±6.5 years. Median CRP 0.20 mg/dl (IQR 0.10-0.40). A CRP ≥0.25 mg/dl discriminated the population with high/very high disease activity [AUROC 0.71; OR 3.9, p<0.001]. This threshold rose to CRP ≥0.35 mg/dl [AUROC 0.72; OR 10.4, p<0.001] among patients without biological therapy, remaining at ≥0.25 mg/dl [AUROC 0.70; OR 4.02, p<0.001] in those exposed to these therapies. The standard inflammatory CRP value (≥0.5 mg/dl) was highly specific (0.90) but poorly sensitive (0.35) in detecting high/very high disease activity states, making it less discriminatory than previous thresholds.

Our results suggest adopting lower than standard CRP cut-off values for a better assessment of both the global activity of the disease and a better interpretation of the composite activity indices used in axSpA.