Research article
Hypogalactosylation of serum N-glycans fails to predict clinical response to methotrexate and TNF inhibition in rheumatoid arthritis
Author affiliations
1 Division of Rheumatology, Immunology and Allergy, Brigham and Women's Hospital, One Jimmy Fund Way, Smith 516c, Boston MA, 02115 USA
2 Division of Immunology, Children's Hospital Boston, 300 Longwood Avenue, Boston, MA 02115, USA
3 Robert S Boas Center for Genomics and Human Genetics, Feinstein Institute Medical Research, 350 Community Drive, Manhasset, NY 11030, USA
4 Ludger Ltd, E1 Culham Science Centre, Abingdon, Oxfordshire, OX14 3EB, UK
5 Dublin-Oxford Glycobiology Laboratory, National Institute for Bioprocessing Research and Training (NIBRT), Conway Institute, University College Dublin, Belfield, Dublin 4, Ireland
6 Novartis Institute for Biomedical Research (NIBR) Autoimmunity, Transplantation and Inflammation (ATI) CHBS, WSJ-386.11.05, Novartis Pharma AG, Novartis Campus, CH-4056 Basel, Switzerland
Citation and License
Arthritis Research & Therapy 2012, 14:R43 doi:10.1186/ar3756
Published: 5 March 2012Abstract
Introduction
Rheumatoid arthritis (RA) is associated with hypogalactosylation of immunoglobulin G (IgG). We examined whether a proxy measure for galactosylation of IgG N-glycans could predict response to therapy or was differentially affected by methotrexate (MTX) or TNF blockade.
Methods
Using a previously defined normal phase high-performance liquid chromatography approach, we ascertained the galactosylation status of whole serum N-glycans in two well-defined RA clinical cohorts: the Autoimmune Biomarkers Collaborative Network (n = 98) and Nested I (n = 64). The ratio of agalactosylated to monogalactosylated N-glycans in serum (sG0/G1) was determined before and during therapy with MTX or TNF inhibition and correlated with anticitrullinated peptide antibody (ACPA) status and clinical response as assessed by 28-joint Disease Activity Score utilizing C-reactive peptide and European League Against Rheumatism response criteria.
Results
RA patients from both cohorts exhibited elevation of sG0/G1 at baseline. Improvement in clinical scores correlated with a reduction in sG0/G1 (Spearman's ρ = 0.31 to 0.37; P < 0.05 for each cohort). However, pretreatment sG0/G1 was not predictive of clinical response. Changes in sG0/G1 were similar in the MTX and TNF inhibitor groups. Corrected for disease activity, ACPA positivity correlated with higher sG0/G1.
Conclusions
Baseline serum N-glycan hypogalactosylation, an index previously correlated with hypogalactosylation of IgG N-glycans, did not distinguish patients with rheumatoid arthritis who were likely to experience a favorable clinical response to MTX or TNF blockade. Clinical improvement was associated with partial glycan normalization. ACPA-positive patients demonstrated enhanced N-glycan aberrancy compared with ACPA-negative patients.


