Baseline autoantibody profile in rheumatoid arthritis is associated with early treatment response but not long-term outcomes

Abstract

Background: The autoantibody profile of seropositive rheumatoid arthritis (RA) is very diverse and consists of various isotypes and antibodies to multiple post-translational modifications. It is yet unknown whether this varying breadth of the autoantibody profile is associated with treatment outcomes. Therefore, we investigated whether the composition of the autoantibody profile in RA, as a marker of the underlying immunopathology, influences initial and long-term treatment outcomes.

Methods: In serum from 399 seropositive patients with RA in the IMPROVED study, drawn at baseline and at the moment of drug tapering, we measured IgG, IgM, and IgA isotypes for anti-cyclic citrullinated peptide-2 and anti-carbamylated protein antibodies, IgM and IgA rheumatoid factor, and reactivity against four citrullinated and two acetylated peptides (anti-modified protein antibodies (AMPAs)). We investigated the effect of the breadth of the autoantibody profile on (1) change in disease activity score (DAS)44 between 0 and 4 months, (2) initial drug-free remission (DFR, drug-free DAS44 < 1.6) achieved between 1 and 2 years of follow up, and (3) long-term sustained DFR until last follow up.

Results: Patients with a broad autoantibody profile at baseline had a significantly better early treatment response: ΔDAS 0-4 months of 1-2, 3-4, and 5-6 vs 7-8 isotypes, -1.5 (p < 0.001), -1.7 (p = 0.03), and -1.8 (p = 0.04) vs -2.2. Similar results were observed for AMPA number. However, patients with a broad baseline autoantibody profile achieved less initial DFR. For long-term sustained DFR there was no longer an association with the breadth of the autoantibody response. When assessing autoantibodies at the moment of tapering, similar trends were observed.

Conclusions: A broad baseline autoantibody profile is associated with a better early treatment response. The breadth of the baseline autoantibody profile, reflecting a break in tolerance against several different autoantigens and extensive isotype switching, may indicate a more active humoral autoimmunity, which could make the underlying disease processes initially more suppressible by medication. The lack of association with long-term sustained DFR suggests that the relevance of the baseline autoantibody profile diminishes over time.

Trial registration: ISRCTN11916566 . Registered on 7 November 2006. EudraCT, 2006- 06186-16. Registered on 16 July 2007.

Keywords: Ant-CCP; Autoantibodies; Disease activity; Rheumatoid arthritis; Rheumatoid factor.

Fig. 1

Extended patient selection algorithm. ACR, American College of Rheumatology; EULAR, European League Against Rheumatism; RA, rheumatoid arthritis; anti-CCP2, anti-citrullinated protein 2; RF, rheumatoid factor; anti-CarP, anti-carbamylated protein antibodies; AMPA, anti-modified protein antibodies; Cit, citrullinated

Fig. 2

Disease activity score (DAS) (mean +/- 95% confidence intervals) over 4 months of treatment and mean initial change in DAS from baseline to 4 months (ΔDAS 0–4 months), separated by baseline serological status and breadth of autoantibody response. a, b DAS over time and ΔDAS 0–4 months separated by baseline autoantibody seropositivity based on anti-citrullinated protein 2 (anti-CCP2) IgG, rheumatoid factor (RF) IgM, or anti-carbamylated protein (anti-CarP) IgG positivity. Based on availability of antibody data, the total number of patients included in a and b is 472. c, d Within baseline seropositive patients, DAS over time and ΔDAS 0–4 months separated by the total number of isotypes present (anti-CCP2 IgG, IgM, IgA; RF IgM, IgA; and anti-CarP IgG, IgM, IgA). Due to the technical success rate of isotype measurements, and some seropositive patients testing negative upon re-measurement (see Additional file 1: Figure S1), the total number of patients included in c and d is 325. e, f Within patients seropositive at baseline, DAS over time and ΔDAS 0–4 months separated by the total number of anti-modified peptide antibodies (AMPAs) present (anti-CCP2 IgG, anti-CarP IgG, anti-citrullinated-vimentin 59-74 IgG, anti-citrullinated-fibrinogen β 36-52 IgG and α 27-43 IgG, anti-citrullinated-enolase 5-20 IgG, anti-acetylated-lysine IgG, and anti-acetylated-ornithine IgG). Thirty-eight patients were RF IgM positive but had no AMPAs (not shown). Due to the technical success rate of isotype measurements and some anti-CCP2 IgG and anti-CarP IgG positive patients testing negative upon re-measurement (see Additional file 1: Figure S1), the total number of patients included in e and f is 318. g Within baseline seropositive patients, ΔDAS 0–4 months separated by the number of isotypes present per antibody family and for the number of antibodies present to citrullinated or acetylated peptides. Reported p values are adjusted for multiple testing using Holmes-Bonferroni methods. ns, not significant (p ≥ 0.05); *p < 0.05; **p < 0.01, ***p < 0.001. Anti-cit. pept. Abs, anti-citrullinated peptide antibodies; Anti-acetyl. pept. Abs, anti-acetylated peptide antibodies

Fig. 3

Association between baseline autoantibody profile and initial drug-free remission (DFR) in patients seropositive for anti-citrullinated protein 2 (anti-CCP2) IgG, rheumatoid factor (RF) IgM, or anti-carbamylated protein (anti-CarP) IgG at baseline that had serum available for re-measurement (n = 155). Pairwise comparisons between each group were not significant after multiple testing (see text). a Percentage of patients with the specified number of isotypes present reaching initial DFR. The composite number of isotypes consists of the positivity count for anti-CCP2 IgG, IgM, IgA; RF IgM, IgA; and anti-CarP IgG, IgM, IgA. Due to the technical success rate of isotype measurements, and some seropositive patients testing negative upon re-measurement (see Additional file 1: Figure S1), the total number of patients included in a is 140. b Percentage of patients with the specified number of anti-modified protein antibodies (AMPAs) present reaching initial DFR. The composite number of AMPAs consists of the positivity count for anti-CCP2 IgG, anti-CarP IgG, anti-citrullinated-vimentin 59-74 IgG, anti-citrullinated-fibrinogen β 36-52 IgG, α 27-43 IgG, anti-citrullinated-enolase 5-20 IgG, anti-acetylated-lysine IgG, and anti-acetylated-ornithine IgG. Eleven patients were RF IgM positive but had no AMPA antibodies (not shown). c Percentage of patients with the specified number of antibodies present reaching initial DFR. Anti-cit. pept. Abs, anti-citrullinated peptide antibodies; Anti-acetyl. pept. Abs, anti-acetylated peptide antibodies. Reported p values are adjusted for multiple testing using Holmes-Bonferroni methods. ns, not significant (p ≥ 0.05); *p < 0.05; **p < 0.01, ***p < 0.001

Fig. 4

Association between 8-month autoantibody profile and initial drug-free remission (DFR) in patients seropositive for anti-citrullinated protein 2 (anti-CCP2) IgG, rheumatoid factor (RF) IgM, or anti-carbamylated protein (anti-CarP) IgG at baseline, who had serum available for re-measurement at 8 months (n = 103). Pairwise comparisons between each group were not significant after multiple testing (see text). a Percentage of patients with the specified number of isotypes present reaching initial DFR. The composite number of isotypes consists of the positivity count for anti-CCP2 IgG, IgM, IgA; RF IgM, IgA; and anti-CarP IgG, IgM, IgA. Due to some seropositive patients testing negative upon re-measurement or seroconverting to negative by 8 months, the total number of patients with any isotypes present was 92. b Percentage of patients with the specified number of anti-modified protein antibodies (AMPAs) present reaching initial DFR. The composite number of AMPAs consists of the positivity count for anti-CCP2 IgG, anti-CarP IgG, anti-citrullinated-vimentin 59-74 IgG, anti-citrullinated-fibrinogen β 36-52 IgG and α 27-43 IgG, anti-citrullinated-enolase 5-20 IgG, anti-acetylated-lysine IgG, and anti-acetylated-ornithine IgG. Thirteen patients were RF IgM positive but had no AMPA antibodies or had seroconverted to negative by 8 months (not shown). c Percentage of patients with the specified number of antibodies present reaching initial DFR. Anti-cit. pept. Abs, anti-citrullinated peptide antibodies; Anti-acetyl. pept. Abs, anti-acetylated peptide antibodies. Reported p values are adjusted for multiple testing using Holmes-Bonferroni methods. ns, not significant (p ≥ 0.05); *p < 0.05; **p < 0.01, ***p < 0.001

Fig. 5

Association of baseline autoantibody profile with long-term sustained drug-free remission (DFR) in patients seropositive for anti-citrullinated protein 2 (anti-CCP2) IgG, rheumatoid factor (RF) IgM, or anti-carbamylated protein (anti-CarP) IgG at baseline (n = 336). Pairwise comparisons between each group were not significant after multiple testing (see text). a Percentage of patients with the specified number of isotypes present reaching long-term sustained DFR. The composite number of isotypes consists of the positivity count for anti-CCP2 IgG, IgM, IgA; RF IgM, IgA; and anti-CarP IgG, IgM, IgA. Due to the technical success rate of isotype measurements, and some seropositive patients testing negative upon re-measurement (see Additional file 1: Figure S1), the total number of patients included in a is 309. b Percentage of patients with the specified number of anti-modified protein antibodies (AMPAs) present reaching long-term sustained DFR. The composite number of AMPAs consists of the positivity count for anti-CCP2 IgG, anti-CarP IgG, anti-citrullinated-vimentin 59-74 IgG, anti-citrullinated-fibrinogen β 36-52 IgG and α 27-43 IgG, anti-citrullinated-enolase 5-20 IgG, anti-acetylated-lysine IgG, and anti-acetylated-ornithine IgG. Thirty-six patients were RF IgM positive but had no AMPA antibodies (not shown). c Percentage of patients with the specified number of antibodies present reaching long-term sustained DFR. Anti-cit. pept. Abs, anti-citrullinated peptide antibodies; Anti-acetyl. pept. Abs, anti-acetylated peptide antibodies

Fig. 6

Summary of results. Colours of the antibodies indicate diversity in antigenic targets, and structures indicate diversity in the isotype usage. A broad baseline profile (left) is favourable for early response, but has no association with long-term outcomes like drug-free remission, as compared to a less broad baseline profile (right). The association with the breadth of the baseline profile diminishes with time