THU0154 Biomarker Sets PREDICT Therapeutic Response to Tnf-Inhibitors in Rheumatoid Arthritis and Spondyloarthritis Patients: A Theragnostic Approach in A Multicenter Cohort

Background TNF inhibitors (TNFi) are first-line treatments for patients suffering from rheumatoid arthritis (RA) and spondyloarthritis (SpA). However, it is currently impossible to predict response to Infliximab (IFX), Adalimumab (ADA) or Etanercept (ETN). Objectives To predict response to IFX, ADA...

Full description

Saved in:
Bibliographic Details
Published inAnnals of the rheumatic diseases Vol. 73; no. Suppl 2; p. 233
Main Authors Baillet, A., Trocmé, C., Marotte, H., Soubrier, M., Tébib, J., Thomas, T., Miossec, P., Pellot-Prades, B., Grange, L., Toussaint, B., Juvin, R., Morel, F., Drouet, C., Gaudin, P.
Format Journal Article
LanguageEnglish
Published London BMJ Publishing Group LTD 01.06.2014
Online AccessGet full text

Cover

Loading…
More Information
Summary:Background TNF inhibitors (TNFi) are first-line treatments for patients suffering from rheumatoid arthritis (RA) and spondyloarthritis (SpA). However, it is currently impossible to predict response to Infliximab (IFX), Adalimumab (ADA) or Etanercept (ETN). Objectives To predict response to IFX, ADA and ETN in RA and SpA patients with insufficient response to disease modifying anti-rheumatic drugs by measuring 8 biomarkers. Methods Serum levels of apolipoprotéineA1 (ApoA1), Platelet Factor (PF4), Prealbumin, Haptoglobin (Hapto), C-reactive protein (CRP), transthyretin, α1 anti-trypsin (α1-AT), S100A12 and S100A8/A9 proteins were measured in a prospective French multicenter cohort before treatment with IFX, ETN or ADA. For RA patients, EULAR response was assessed at 6 months. A decrease<20% of BASDAI was considered a poor response and a decrease>50% of BASDAI as a good response at 6 months for SpA patients. Mann Whitney tests evaluated differences in biomarker levels between a good response and a moderate/no reponse at 6 months. A multivariate analysis was conducted to characterize biomarker sets which predict response to TNFi at 6 months with the best sensitivity and specificity. Robustness of biomarker sets was assessed by a “bootstrapping” re-sampling approach. Results 101 patients (53 RA and 48 SpA) treated with TNFi (32 IFX, 30 ADA and 39 ETN) were included. RA patients consisted of 75% of women against 29% in the SpA group. The median age [interquartile range] was 54 [46-61] years in RA and 40 [33-78] years in SpA. At 6 months, the therapeutic response was good for 21 RA and 20 SpA patients, moderate in 18 RA and 11 SpA patients while 13 RA and 17 SpA patients did not respond. Univariate analysis showed that individual biomarkers had a moderate diagnostic value when taken separately: Levels of pre-albumin was lower among patients who responded to the ADA compared to patients with moderate or no response (223 [208-248] mg/l vs.260 [247-321] mg/l, p=0.02). ApoA1 level was lower in responders to ETN than in non responders (1.2 [1.1-1.4] g/l vs. 1.4 [1.3-1.6] g/l, p=0.02). Multivariate analysis identified several sets of 2 to 8 biomarkers predictive of therapeutic response at 6 months. Metrological properties of seven of these combinations are shown in Table 1. For female RA patients treated with IFX or ADA, female SpA patients treated with IFX or ADA and male RA patients treated with IFX, the numbers were insufficient to generate a set of biomarkers. Bootstrapping re-sampling approach suggested a good external validity of biomarker sets. Table 1. Biomarker sets predict response to TNFi Biomarker sets Metrological properties, % S100A8/A9 S100A12 PF4 Apo A1 Pre albumin α1 AT Hapto Se Sp PPV NPV RA male ETN – – – X X – – 80 100 100 89 RA female ETN X – X – X – – 100 100 100 100 RA male ADA – X – – – – – 100 100 100 100 SpA male ETN X – X X – X – 80 96 89 92 SpA female ETN X – – X X – – 100 92 75 100 SpA male ADA – X – X – – – 100 97 67 100 SpA male IFX X X X – – – – 80 100 100 97 Conclusions Sets of 2 to 8 biomarkers predicted of good clinical response at 6 months with a good sensitivity and specificity. Disclosure of Interest : A. Baillet Consultant for: Pfizer, C. Trocmé: None declared, H. Marotte: None declared, M. Soubrier: None declared, J. Tébib: None declared, T. Thomas: None declared, P. Miossec: None declared, B. Pellot-Prades: None declared, L. Grange: None declared, B. Toussaint: None declared, R. Juvin: None declared, F. Morel: None declared, C. Drouet: None declared, P. Gaudin: None declared DOI 10.1136/annrheumdis-2014-eular.5750
ISSN:0003-4967
1468-2060
DOI:10.1136/annrheumdis-2014-eular.5750