Background:Production of anti-drug antibodies (ADA) could cause low serum anti-TNF levels and low drug retention rates.Objectives:To assess the relationship between serum adalimumab (ADL) levels and drug retention rates in patients with axial SpA.Methods:Single-center prospective study in patients with axial SpA, according to ASAS criteria, being treated with ADL. In the first part of the study, from December 2010 to June 2016, data was collected and serum samples were taken. In a second part of the study, clinical records were reviewed to find out the dates and reasons of treatment discontinuation. Information was collected on age gender, body mass index (BMI), date of diagnosis of SpA, laboratory data, including HLAB27, erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP), prior biological treatment for SpA, concomitant conventional synthetic disease modifying anti-rheumatic drugs (csDMARDs). SpA disease activity was assessed by BASDAI, ASDAS and PGA scores. A serum sample was taken from all patients just before the next adalimumab dose. ADL and ADA levels were analysed with ELISA Promonitor®, (Proteomika, Progénica, Griffols) and correlated with SpA activity using BASDAI and ASDAS scores. Cox proportional hazards model was performed in order to assess the relation of variables with drug interruption.Results:Up to January 2016, 51 consecutive patients were included. The mean (range) age was 46,9 (18–68) and 47.1% were women. HLAB27 was positive in 82,4% of patients. Mean disease duration was 122,9 months (2-408) and mean duration of treatment with adalimumab was 17,8 months (1-69). ADL was the first biological agent received in 36 patients (70.6%). Eleven patients (21.6%) were on concomitant treatment with receiving concomitant csDMARDs, mainly methotrexate (MTX) (15.7%) and sulfasalazine (5.9%). The mean (SD) activity scores were BASDAI 4,0 ± 2,3; ASDAS-PCR 2,1 ± 1,1 and ASDAS-VSG 2,1 ± 1,0.ADA prevalence was 27.5%, with none detected in the 21.6% taking csDMARDs (p = 0.021). ADL level was normal (> 3 mg/l) in 36 patients (70.6%), all without ADA. Fifteen patients (29.4%) had subtherapeutic ADL levels (< 3 mg/l), with ADA in 14 (93%).Total ADL treatment time was 241,16 patient/years, and mean ADL treatment time was 4,73 years. Cox model results were resumed in table 1. Multivariate study show that ADL level > 3 mg/L was a protective factor for ADL interruption (HR 0.01 (0.00-0.59, p=0.026), while previous etanercept treatment was a risk factor for ADL interruption 9.54 (1.23-74.08, p=0.031).Table 1.Cox model.ContinueDiscon=tinueUnivariateHRMultivariateHRADL level<3 mg/L13 (86,7%)2 (13,3%)0.05 (0.00-0.65, p=0.022)0.01 (0.00-0.59, p=0.026)≥3 mg/L28 (77,8%)8 (22,2%)Concomitant DMARDYes8 (72,7%)3 (27,3%)0.56 (0.14-2.26, p=0.419)0.28 (0.05-1.65, p=0.159)No33 (82,5%)7 (17,5%)BMINormal13 (86,7%)2 (13,3%)2.56 (0.54-12.23, p=0.239)0.48 (0.02-9.75, p=0.634)Overweight/obese27 (77,1%)8 (22,9%)AgeMean (SD)46,4 (12,2)48,7 (11,6)0.93 (0.24-3.60, p=0.912)0.60 (0.10-3.84, p=0.594)GenderMen19 (70,4%)8 (29,6%)0.43 (0.09-2.03, p=0.283)0.58 (0.03-10.06, p=0.705)Women22 (91,7%)2 (8,3%)InfliximabpreviousYes4 (40,0%)6 (60,0%)3.31 (0.92-11.87, p=0.066)1.37 (0.23-8.24, p=0.731)No37(90,2%)4 (9,8%)EtanerceptpreviousYes2 (22,2%)7 (77,7%)8.17 (2.09-31.90, p=0.003)9.54 (1.23-74.08, p=0.031)No39 (92,9%)3 (7,1%)ADL: adalimumab; BMI: body mass index; DMARD: disease modifying antirheumatic drug; SD: standard deviation.Conclusion:Adalimumab drug levels > 3 mg/L is a protective factor against treatment interruption.Etanercept previous treatment was a risk factor for treatment interruption.Acknowledgments:The study was supported by a grant from “Asociación para la Investigación en Reumatología de la Marina Baixa” (AIRE-MB).Disclosure of Interests:None declared