Background Calprotectin and TNF antagonist (TNFa) trough serum levels are associated with disease activity.

Objectives To determine if calprotectin serum levels or drug trough serum levels predict relapse in RA and PsA patients with low levels of disease activity during TNFa therapy.

Methods Prospective 1-year follow-up, single center study (INMUNOREMAR cohort). RA (ACR 1987) and PsA (CASPAR) patients in clinical remission (CR) (DAS28-ESR <2.6) or low disease activity (LDA) (DAS28-ESR <3.2) in ≥2 consecutive visits treated with adalimumab (ADA), etanercept (ETN) or infliximab (IFX) for ≥3 months were included. Demographic data, disease duration, time to initiate csDMARD and bDMARD, time to achieve remission, remission duration, autoantibody status, radiological data, concomitant csDMARD therapy, dose and duration of biological therapy were collected. Clinical and laboratory assessment was made every 4 months and at the time of flare. Calprotectin serum levels (ELISA Kit Calpro AS®) and TNFa trough serum levels (ELISA Kit Promonitor®, Progenika) were determined. Disease flare was defined as DAS28-ESR>3.2 and increase in ΔDAS28-ESR >0.6. Univariate and multivariate regression models were used to identify predictors of disease flare.

Results 103 patients (47RA, 56PsA) were included. Mean Age 57 (30–81) years. 78 (75.8%) had CR and 25 (24.2%) LDA. Mean CR/LDA duration was 58 (4–163) months. 36 patients had received treatment with ADA, 50ETN, and 17IFX. Mean biologic treatment duration 61 (7–166) months. 47.4% patients were on monotherapy, 47% on reduced dose, and 17.9% received steroids. 84 patients (91%) remained in CR/LDA for 12 months, and 12 patients (13%; 8 RA, 4 PsA) experienced flares. Patients with flares had a longer time to achieve CR/LDA [CR/LDA (n=81) 3.1 (1–6) months vs. Flare (n=12) 20 (5–31) months, p<0.001], a shorter CR/LDA duration [CR/LDA (n=81) 59 (5–163) months vs. Flare (n=12) 25 (3–134) months, p=0.031], higher calprotectin levels [CR/LDA (n=81) 1.4 (0.6–3.7) μg/mL vs. Flare (n=12) 6 (4.5–7.9) μg/mL, p<0.001], and lower drug trough serum levels [CR/LDA (n=83) 2.6 (0.6–12) μg/mL vs. Flare (n=12) 0.6 (1–1.2) μg/mL; p<0.001], even when analyzed by biologic [ADA CR/LDA (n=30) 7 (0.2–12) μg/mL vs. Flare (n=4) 0.5 (0.4–1) μg/mL; p=0.003; ETN CR/LDA (n=44) 1.5 (0.7–4.7) μg/mL vs. Flare (n=5) 0.8 (0.9–1.2)μg/mL; p=0.039; IFX CR/LDA (n=14) 3.1 (0.5–7.7) μg/mL vs. Flare (n=3) 0.1 (0–1) μg/mL; p=0.021] at baseline. According to the cut-off developed by our group [1], patients with flares had sub-therapeutic baseline drug serum trough levels [CR/LDA (n=22) 26% vs. Flare (n=12) 100%, p<0.001]. Baseline TNFi trough serum levels [hazard ratio (HR) = 0.47], steroid treatment (HR=3.21), time to CR/LDA (HR=1.17), and baseline calprotectin levels (HR=2.38) significantly predicted flare in the univariate analysis but only baseline calprotectin levels significantly predicted flare in the multivariate analysis [HR=2.74 (1.74–4.31); p<0.0001].

Conclusions Calprotectin and TNF trough serum levels are promising biomarkers to predict flare in RA and PsA patients with low disease activity during anti-TNF treatment.

1. Sanmarti et al. Ann Rheum Dis. 2015 Aug;74(8):e42.

Disclosure of Interest J. Inciarte-Mundo Grant/research support from: Hospital Clinic de Barcelona (Premi E. Letang 2013); Catalan Society of Rheumatology (Research Grant 2013), M. Hernández: None declared, V. Ruiz-Esquide: None declared, J. Ramírez: None declared, A. Cuervo: None declared, S. Cabrera-Villalba: None declared, M. Pascal: None declared, J. Yagüe: None declared, J. Cañete: None declared, R. Sanmarti Grant/research support from: This Study was supported by an Investigator Initiated Research (IIR) from Pfizer (WS2392358)