Skip Navigation


Rheumatology Advance Access originally published online on August 9, 2006
Rheumatology 2007 46(3):446-453; doi:10.1093/rheumatology/kel262
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
46/3/446    most recent
kel262v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (8)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lequerré, T.
Right arrow Articles by Vittecoq, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lequerré, T.
Right arrow Articles by Vittecoq, O.
Related Collections
Right arrow Rehabilitation
Right arrow Rheumatoid Arthritis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Autoantibodies, metalloproteinases and bone markers in rheumatoid arthritis patients are unable to predict their responses to infliximab

T. Lequerré1,2, F. Jouen3, M. Brazier4, S. Clayssens5, N. Klemmer1, J.-F. Ménard6, O. Mejjad1, A. Daragon1,2, F. Tron2,3, X. Le Loët1,2 and O. Vittecoq1,2

1Rheumatology Department, Rouen University Hospital, 2Inserm U519, IFR 23, Faculté de Médecine, 3Immunology Department, Rouen University Hospital, Rouen, 4Laboratory of Bone Biology, Amiens University Hospital, Amiens, 5Biochemical Department and 6Biostatistics Department, Rouen University Hospital, Rouen, France.

Correspondence to: Dr T. Lequerré, Department of Rheumatology, Rouen University Hospital, 76031 Rouen Cedex, France. E-mail: thierry.lequerre{at}univ-rouen.fr


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Conclusion
 Acknowledgement
 References
 
Objectives. To identify biochemical, immunological and bone markers as predictors of rheumatoid arthritis (RA) patients’ responses to infliximab.

Methods. A total of 76 patients with active RA (American College of Rheumatology criteria), refractory to disease-modifying anti-rheumatic drugs, including methotrexate, received infliximab (3 mg/kg) infusions at weeks 0, 2, 6, and then every 8 weeks in combination with methotrexate or leflunomide. At week 14, infliximab efficacy was evaluated using disease activity score (DAS)28. A serum sample, collected just before starting infliximab, was tested by ELISA (unless stated otherwise) for the following immunological markers: rheumatoid factor by agglutination and ELISA (IgA, IgG and IgM isotypes); anti-cyclic citrullinated protein; autoantibodies recognizing calpastatin domain I and its 27 C-terminal fragment, glucose-6-phosphate isomerase, {alpha}-enolase; anti-keratin and anti-perinuclear factor antibodies (immunofluorescence); biochemical markers: C-reactive protein (nephelometry), metalloproteinase-1 and -3, tissue inhibitors of metalloproteinases-1 and -2, antioxidants (vitamins A and E; selenium); bone resorption markers: pyridinoline, deoxypyridinoline, osteoprotegerin, soluble receptor activator of nuclear factor-{kappa}B ligand, cartilage oligomeric matrix protein. Each parameter's predictive value of the response to infliximab was analysed using Fisher's exact, Mann–Whitney and chi-square tests. Hierarchical clustering was performed with The Institute for Genomic Research (TIGR) multiple experiment viewer software.

Results. Good, moderate and non-responder rates were 6.5, 61.8 and 31.5%, respectively. No significant difference was observed between responders and non-responders, regardless of the serum parameters considered. Analysis of dichotomous or continuous variables failed to identify markers predictive of a good or poor response to infliximab.

Conclusion. The search for soluble markers in RA patients’ sera likely to predict response to infliximab because of their involvement in RA pathogenesis seems disappointing. However, because of the limited power to detect smaller differences in biomarkers, the present study is a preliminary exploratory analysis.

KEY WORDS: Rheumatoid arthritis, Infliximab, Predictive factors of response, TNF-{alpha} blocking agent, Prognosis


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Conclusion
 Acknowledgement
 References
 
The combination of tumour necrosis factor-{alpha} (TNF-{alpha}) blocking agents and methotrexate demonstrated real therapeutic efficacy by decreasing inflammation and slowing and stopping joint destruction in 70–80% of rheumatoid arthritis (RA) patients [1–6]. Regardless of the blocking agent prescribed, about 30% of RA patients failed to respond, but we have not yet been able to identify them before initiating treatment. Even though the pathophysiological heterogeneity of RA and the existence of subgroups of patients susceptible to respond better to one molecule rather than another have been well-demonstrated, we are still unable to predict the efficacy of one treatment or another in a given patient. In addition, use of TNF-{alpha} blocking agents exposes patients to certain risks (infectious, neoplastic, allergic, etc.) that seem disproportionate, especially if the patient might not respond to one of these therapies [3, 4]. Moreover, the range of biotherapies will soon be extended, as highly promising, innovative, new agents become available (interleukin-6 receptor, cytotoxic T-lymphocyte-associated antigen 4, anti-CD20), but their respective places in therapeutic strategies have not yet been defined [7, 8]. Without markers or indexes, it will be difficult to choose—among all the available agents—the molecule that will achieve immediate efficacy and will be the best adapted to a given patient. This decision is rendered even more difficult by the fact that the therapeutic targets of the various molecules are different. Because of all these unknown risks and limiting factors, it would be highly desirable to be able to identify patients susceptible to respond—or not—to a given agent. One way to achieve this objective is to identify factors predictive of a response to therapy.

At present, the only markers that have been shown to be able to predict the response to infliximab are genetic markers identified by analysis of genetic polymorphisms. Mugnier et al. [9] examined the polymorphism of the promoter gene encoding TNF-{alpha} in position –308 in 59 patients treated with infliximab. Patients with the –308 G/G genotype were better responders [disease activity score (DAS) diminution >2.29] than those expressing the A/A or A/G genotype (DAS diminution >1.24). The results of those investigations are highly promising because genetic factors predictive of response can be intimately associated with mechanisms of TNF-{alpha} regulation in RA, as is the case for the polymorphism of the promoter gene coding for TNF-{alpha} that induces quantitative TNF-{alpha} variations. Notably, the –308 A allele is associated with higher levels of TNF-{alpha} transcription/transcripts than the G allele [9].

Nevertheless, other mechanisms are also probably involved because, so far, genetic polymorphism could not explain the common clinical observation of a patient responding to one anti-TNF-{alpha} agent but not necessarily others [10, 11]. In addition, those studies on polymorphism provided information concerning the efficacy of the therapeutic class of agents (TNF-{alpha} blockers) in a patient, but alone, they probably do not contribute to orienting the choice of the most effective molecule among those available. Finally, as Bridges [12] recently emphasized in an editorial in Arthritis and Rheumatism, response to treatment cannot be simplified to a single polymorphism, but rather to several, because the genes interact. Thus, the identification of responders and non-responders to TNF-{alpha} blocking agents cannot be restricted to the study of cytokine polymorphisms.

What other parameters might be pertinent for the prediction of response to a biotherapy? Knowing that TNF-{alpha} blocking agents are prescribed to treat severe and progressing RA, an initial approach is to measure the markers associated with RA severity: erythrocyte sedimentation rate (ESR)/1st h, C-reactive protein (CRP), rheumatoid factors (RF), antibodies to cyclic citrullinated peptide (anti-CCP) and HLA-DRB1*04 and 01 [13–16]. In addition, based on our knowledge of the pathogenic mechanisms implicated in RA, a certain number of potentially informative markers can be deduced. First and foremost, hyperplasia of synovial tissue, a real pseudotumoural process, will engender osteocartilaginous destruction, predominantly caused by metalloproteinase (MMP) 1 and MMP3, whose actions are contained by their respective tissue inhibitors (TIMP1 and TIMP2) and cartilage oligomeric matrix protein (COMP) [17]. Some markers reflect the hyper-resorption of bone, like the collagen type 1-crosslinking molecules [pyridinoline (PYD) and deoxypyridinoline (DPYD)], but also the receptor activator of nuclear factor-{kappa}B–receptor activator of nuclear factor-{kappa}B ligand (RANK–RANKL) system and osteoprotegerin (OPG) [18, 19]. Moreover, the antioxidants—vitamins A and E, and selenium—counterbalance the free radicals that reflect inflammation. Several studies showed negative correlations between RA severity and serum antioxidant concentrations [20–23]. Finally, recent data plead in favour of a major role for B-lymphocytes in the pathogenesis of RA, as attested by the increasing number of autoantibodies (autoAb) found in this pathology. These biological parameters, combined in various associations, might be able to predict the response to TNF-{alpha} blocking agents, especially because some of them are regulated by TNF-{alpha}.

The primary objective of this prospective, longitudinal study on RA patients naïve to all biotherapies was the identification of markers able to predict the response to infliximab given in combination with methotrexate or leflunomide. To identify such indicators in patients whose RA was resistant to disease-modifying anti-rheumatic drugs (DMARDs), soluble markers (proteases, autoAb, antioxidants) known to have a prognostic or pathophysiological value were measured in serum before starting infliximab. In this exploratory analysis, two biostatistical methods—univariate analysis and microarray hierarchical clustering—were applied to attempt to identify potential biochemical, immunological and bone markers as predictors of RA patients’ responses to infliximab combined with methotrexate or leflunomide.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Conclusion
 Acknowledgement
 References
 
Patients
A total of 76 patients with RA satisfying the American College of Rheumatology (ACR) criteria and followed in our Rheumatology Department were included in this prospective, longitudinal study. These patients had active RA refractory to methotrexate and/or leflunomide that necessitated combining infliximab (RemicadeTM, Schering Plough, Levallois-Perret, France) to their treatment regimen. Their disease failed to respond to DMARD (methotrexate had to have been tried unless contraindicated); RA was active with at least three of the following five criteria: ≥3 synovia affected, ≥6 painful joints, duration of morning stiffness ≥45 min, ESR >22 mm/1st h, CRP >20 mg/l; DMARD therapy with methotrexate or leflunomide was used, regardless of the dose prescribed. The sole criterion of non-inclusion was contraindication of infliximab administration. Patients included gave their written consent to participate in this protocol that had been approved by the Upper Normandy Ethics Committee (protocol 2002/061 HP).

Infliximab infusions
Infliximab was administered intravenously by slow perfusion at the manufacturer's recommended dose of 3 mg/kg throughout the study at weeks 0, 2, 6 and 14, and then every 8 weeks. Treatment efficacy was evaluated with response criteria established by the European League Against Rheumatism (EULAR) based on DAS28 [24].

Clinical parameters
At the inclusion visit, the following information was collected: sex, age, RA duration, RA history, DMARDs prescribed and their doses, concomitant regimens (corticosteroids and dose), night time awakenings or not, duration of morning stiffness, pain evaluation using a visual analogue scale (VAS) ranging from 0 to 100 mm, the patient's evaluation of disease activity with a VAS, number of painful joints, number of joints with synovitis and the French version of the Health assessment questionnaire (HAQ) [25].

Then, disease activity was re-evaluated just before each infliximab administration by collection of the following information: duration of morning stiffness, pain evaluation using the VAS, the patient's evaluation of disease activity with the VAS, number of painful joints, number of joints with synovitis, ESR, CRP and HAQ.

Biological parameters tested
Before the first infusion, a serum library was constituted for all the patients to determine various biological parameters (immunological, biochemical, osteocartilaginous) with the aim of identifying markers predictive of the response to treatment. These sera were stored at –30°C until evaluation. The IgM-class RFs were detected using latex agglutination and Waaler–Rose tests. The other immunological markers were measured using enzyme-linked immunosorbent assays (ELISA) devised in the immunology laboratory, RF isotypes IgG, IgA and IgM; autoAb directed against calpastatin domain I (ACAST-DI) and its 27-mer C-terminal fragment (ACAST-C27), glucose-6-phosphate isomerase (G6PI)) or with a commercially available anti-CCP ELISA kit (Bio Avance, Emerainville, France) according to the manufacturer's instructions [26–29]. Anti-perinuclear factor (APF) and anti-keratin antibodies (AKA) were detected by direct immunofluorescence labelling [26]. AutoAb to {alpha}-enolase were identified using an immunoblotting technique, with human placenta as the antigen source. Sera recognizing a 50 kDa polypeptide band were considered to have anti-{alpha}-enolase activity [30]. The CRP concentration was determined nephelometrically. MMP1 and MMP3 and their respective inhibitors, TIMP1 and TIMP2, were measured with ELISA-type immunoenzymatic assays (Human BiotrakTM Assay, Amersham Biosciences, Buckinghamshire, UK). Vitamins A and E and selenium concentrations were measured by high-performance liquid chromatography (Dionex, Voisins le Bretonneux, France) after vitamin/selenium extraction from serum with reagent kits (Chromosystem, Munich, Germany). The osteoarticular markers indicative of bone/cartilage destruction were measured using ELISA kits according to the manufacturer's instructions: for PYD and DPYS (respectively, Metra Serum PYD EIA kit and Metra Total DPD EIA kit Serum; Quidel, San Diego, CA, USA), for RANKL and OPG (sRANKL and Osteoprotegerin, Biomedica, Wien, Germany) and for COMP (COMP® ELISA, AnaMar Medical, Uppsala, Sweden).

Statistical analyses
All the clinical and biological parameters collected before the onset of infliximab were analysed, except those used to evaluate treatment efficacy. Data analysis was conducted according to the criteria established by EULAR that defined three levels of response: good, moderate and non-responders. Each patient's response was determined using the DAS 28 calculated just before the fourth infliximab infusion. Because of the small number of good responders among these patients, moderate and good responders were pooled and called the responders. The predictive ability of the pretherapeutic parameters for the principal judgement criterion (dichotomous variable: responder or non-responder) was subjected to univariate test analysis. The search for a relationship between initial criteria comprising dichotomous (presence or absence of autoAb, ...) and continuous variables (CRP median, autoAb titre, ...) and the judgement criterion was conducted using Fisher's exact test or the Mann–Whitney U-test. An association was considered statistically significant when P < 0.05. For quantitative variables, analysis of variance and Kruskal–Wallis rank-sum tests were applied to compare medians. When the global test was significant, multiple comparisons were conducted between responders and non-responders (Kruskal–Wallis multiple-comparison Z-value test), with P < 0.05 retained as significant.

The power of this study was calculated and ranged from 5 to 20% for the different tests used for each qualitative or quantitative variable. To obtain a difference of 10% between responders and non-responders, which is that observed in the present study, 500 patients are required to have a power of 80% with 5% {alpha}-risk (two-tailed) of a qualitative variable, when the percentage of responders is 60–70%.

Hierarchical clustering analysis of data
Microarray hierarchical clustering is a method taken from the tools used for transcriptome analysis. It allows simultaneous comparison of one patient's biological profile including all the parameters considered to the individual profiles of the other patients. Hierarchical clustering regroups the patients according to the similarities of their biological profiles and identifies the combinations of variables classifying the patients. For each patient and each biological variable, the value determined for one patient is normalized as a function of the median of all the values obtained from this population. Microarray data analysis was then conducted using TIGR Multiple Experiment Viewer (TMEV) http://www.tigr.org/software/tm4/mev.html with the hierarchical clustering function [31–33].


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Conclusion
 Acknowledgement
 References
 
Patient characteristics
Just prior to the first infliximab infusion, the 76 patients (62 women and 14 men) had a mean ± S.D. age of 53.8 ± 12.4 yrs, with RA progressing for 10.5 ± 8.6 yrs; all of them had a DAS28 > 3.2. Their clinical and biological characteristics are summarized in Table 1. Their clinical and biological characteristics are summarized in Table 1. Responders and non-responders did not differ significantly for the number of previous DMARDs taken.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Clinical and biological characteristics of the 76 RA patients included in the study

 
Evaluation of the response to infliximab
The judgement criterion applied to assess the response to infliximab was the DAS28 value obtained just before the fourth infusion. However, to limit the errors of judgement of therapeutic efficacy based on that single DAS28 evaluation, we also verified that the mean of three DAS28 evaluations (made just before the third, fourth and fifth infusions) classified the patients similarly according to their response levels (data not shown). The percentages of responders were comparable (McNemar test = 0.18) and concordant (kappa test: 0.6 < P < 0.8), regardless of the way in which the response was evaluated (one or three evaluations). In addition, Fisher's test was highly significant (n = 10–6), thereby indicating a strong statistical association between the two types of evaluations (i.e. the mean of one or three DAS). Therefore, in accordance with EULAR recommendations, the evaluation at 14 weeks of treatment was finally retained for this prospective study.

Good, moderate and non-responders represented 7.8 (n = 6), 53.9 (n = 41) and 38.1% (n = 29), respectively. Because of the small number of good responders, good and moderate responders were pooled (n = 47, 61.8%) for analysis and this group of responders was compared with non-responders. Infliximab and methotrexate led to significant improvement of all the clinical and biological parameters (morning stiffness, VAS for pain, VAS for disease activity, number of painful joints, number of joints with synovitis, ESR, CRP, DAS28 and HAQ) for the responders, while significant improvement for the non-responders was limited to the number of joints with synovitis, DAS28 and HAQ. In the group of responders, each variable improved significantly under treatment. Moreover, the extent of improvement of all clinical and biological parameters obtained during therapy differed significantly between responders and non-responders.

Univariate analyses
As reported in Table 2, univariate analysis of dichotomous variables did not identify any parameter as being predictive of the response to infliximab. Only the DMARD prescribed (methotrexate vs leflunomide) tended to differ significantly between responders and non-responders, with 82.6% of the responders as opposed to 62.1% of non-responders treated with methotrexate. Patients taking methotrexate had a better response on infliximab than those prescribed leflunomide (P = 0.059).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Comparisons of qualitative variables between responders and non-responders in search for markers predictive of the response

 
As reported in Table 3, univariate analysis of quantitative variables failed to identify markers predictive of the response to infliximab. The medians of all these variables did not differ significantly between responders and non-responders (Table 3).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Comparisons of quantitative variables between responders and non-responders in search for markers predictive of response

 
Hierarchical clustering analysis
Analysis of biological data by microarray hierarchical clustering failed to elucidate characteristic profiles of responders and non-responders, and regroup them into distinct subgroups (Fig. 1).


Figure 1
View larger version (56K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. Microarray hierarchical clustering of the biological parameters of the 76 patients with active longstanding RA. Each column represents a biological parameter tested and each line represents a patient [R: (good or moderate) responder; NR: non-responder]. The red–green colour spectrum is the scale of values centred on 0 (black): red signifies higher values while green corresponds to lower values. Each variable was normalized to the median of this value for the study population. Anti-CCP, anti-cyclic citrullinated peptide antibodies; anti-G6PI, anti-glucose-6-phosphate isomerase antibodies; COMP, cartilage oligomeric matrix protein; CRP, C-reactive protein; DPYD, deoxypyridinoline; MMP1/3, metalloproteinase 1 or 3; OPG, osteoprotegerin; PYD, pyridinoline; RF-IgG/IgA/IgM, rheumatoid factor isotype IgG, IgA or IgM; sRANKL, soluble receptor activator of nuclear factor-{kappa}B ligand; TIMP1/2, tissue inhibitor of metalloproteinase 1 or 2; Vit, vitamin. Autoantibodies recognizing the 27 C-terminal fragment or calpastatin domain I were not included in this analysis because too few patients had abnormal values. Anti-keratin antibodies and anti-perinuclear factor are dichotomous variables and thus were not included in the analysis.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Conclusion
 Acknowledgement
 References
 
The main objective of this exploratory study was to identify clinical and/or biological markers able to predict a good or poor response to infliximab in patients with active RA refractory to conventional DMARDs (methotrexate and leflunomide). This investigation concerned 76 patients with longstanding RA (10.5 ± 8.5 yrs), 80.2% of whom were RF-positive. They received infliximab infusions (3 mg/kg) according to the standard therapeutic schedule. Just prior to the first infusion, blood was drawn from all patients to determine serum levels of biological markers implicated in the pathophysiology of RA and/or associated with its severity.

The judgement criterion for therapeutic efficacy in this prospective study was, as recommended by EULAR, based on DAS28 before starting infliximab and 14 weeks later just prior to the fourth infusion at the end of the 8-week interval, when the majority of patients generally reported the reappearance of symptoms [24]. The patients were classified as responders or non-responders, because these classes take into consideration the DAS28-based improvement (compared with the patient's initial status) and RA activity at the time of evaluation, unlike ACR criteria that only consider improvement.

The percentage of non-responders to infliximab (38.2%) observed in our study is close to that generally reported in phase III/IV therapeutic trials (30–40%) [1–4]. Using the DAS44 as the judgement criterion calculated at the sixth infusion, Sidiropoulos et al. [34] obtained respective good, moderate and non-responder rates of 18.8, 37.5 and 43.7% of 68 patients, compared with our respective rates of 6.5, 61.8 and 31.5%. This discordance probably reflects the difference between using DAS44 and 28 as the judgement criterion, and the time of evaluation at the sixth or fourth infusions. In addition, in their study, the methotrexate and/or infliximab dose was adapted to the response at the fourth infusion.

The selection of serum parameters that could potentially be markers predictive of the response to infliximab was based on pathophysiological and clinical arguments. The biological parameters retained can be associated with the progression of bone/cartilage destruction (RF and isotypes) and/or implicated in RA pathophysiology. These latter include: autoAb to CCP, G6PI, calpastatin, etc.; the bone/cartilage-destruction by products COMP, sRANKL, OPG, ...; indicators of inflammation, such as MMPs, their inhibitors (TIMPs), antioxidants and CRP. Moreover, many of these markers are already known to be direct targets of TNF-{alpha} blocking agents because their serum concentrations decline during treatment, as is the case for OPG, sRANKL, COMP, MMP1, MMP3, MMP9 and their inhibitors TIMP1 and TIMP2, nitric oxide (known to destroy joints) and RF; but not always anti-CCP antibodies [35–41]. These biological parameters were measured in serum taken prior to infliximab administration in an attempt to identify markers predictive of the response to this TNF-{alpha} blocking agent.

For this group of patients with active RA, no studied parameter was able to predict a good or poor response to infliximab. In this study, multivariate analysis was not performed because of the lack of statistically different soluble markers between responders and non-responders with the univariate analysis. Moreover, P-values between responders and non-responders in analysis univariate were much dismissed of 0.05. Although the difference was not significant, the only clinical factor susceptible to distinguish responders and non-responders was the associated DMARD, methotrexate or leflunomide. The percentages indicate a better response under methotrexate. Even though associating infliximab with methotrexate seems to have better efficacy than with leflunomide, methotrexate cannot be considered a marker predictive of a good response, especially because 54.2% of the non-responders were taking methotrexate. Furthermore, 75% of our RA patients were taking methotrexate vs 25% on leflunomide. Unlike conventional statistical methods, microarray analysis with hierarchical clustering is able to compare one patient's profile with those of all the others included in the study. The analysis of pre-infliximab serum parameters using the biological profiles of each patient also failed to distinguish between responders and non-responders. Therefore, parameters shown to be predictive of severity or implicated in the pathophysiology of RA do not appear to be relevant for the identification of markers predictive of the therapeutic response.

The results of this study are completely compatible with those of the rare reported investigations in this field. The ATTRACT (Anti-TNF Trial in Rheumatoid Arthritis with Concomitant Therapy) study also failed to identify a clinical parameter able to predict the response to infliximab, among the following: disease duration, age, presence of rheumatoid nodules, DAS28 and the presence of erosions on standard radiographs [42]. In addition, among 20 RA patients treated with infliximab for 1 yr (13 responders and 7 non-responders) Hrycaj et al. [43] tried to identify markers predictive of the response by measuring the following before and 2 weeks after the onset of treatment: ESR, RF-IgM, CRP, C3 and C4 fractions of complement, {alpha}1-acid glycoprotein, {alpha}1 anti-trypsin, serum concentrations of interleukins 4, 6 and 13, TNF-{alpha} types I and II. The pre-treatment determinations were not predictive of response. Their search for a statistical relationship between the response and the difference between pre-treatment and 2 weeks post-infusion 1 was also unsuccessful [43].

Wolbink et al. [44] observed a statistical relationship between the clinical response and two biological parameters, the serum infliximab concentration and the CRP level, measured in 105 patients. The DAS28 was calculated before each infusion as in our study and patients were classified in to three groups according to their initial serum CRP concentration (A: 0–10 mg/l; B: 11–39 mg/l; C: 40–120 mg/l). The serum infliximab concentration at 14 weeks was higher in responders than non-responders, indicating better neutralization of TNF-{alpha} in responders, and it was higher in patients with low CRP concentrations before starting treatment (group A). These patients also had improved DAS28 when the infusions were given at 8-week intervals. The authors concluded that serum infliximab and CRP concentrations determined the capacity of infliximab to neutralize TNF-{alpha}. Higher CRP levels justified shorter intervals between infusions or higher doses of infliximab. In contrast to their findings, the median serum CRP concentration did not differ significantly between our responders and non-responders.

Our conclusions should be viewed with caution. Indeed, as we observed in our study, all response levels can be seen, which means that the response to TNF-{alpha} blocking agents is not binary. To identify markers predictive of the response, it might have been necessary to compare only good vs non-responders, which we could not do because of the small number of good responders. The relatively small sample size is explained by the characteristics of this prospective study. First, the retained population is consistent with the recruitment capacity of the Rheumatology Department of the Rouen University Hospital during this well-defined period (January 2002–October 2003) and the number of patients satisfying the disease criteria given earlier. Second, this inclusion strategy obtained a homogeneous population enabling comparison of potential differences. Indeed, patients receiving infliximab monotherapy, intolerant of methotrexate or leflunomide, or prescribed azathioprine were excluded from this study to achieve such a homogeneous population. Moreover, at the end of 2003, etanercept had just become available in France and competed with infliximab. To offset this difficulty in subsequent studies, more patients must be included to have sufficiently large responder and non-responder groups for comparisons.

In addition, the TNF-{alpha} blocking agent studied probably influenced the judgement criterion. The DAS28 calculated at 14 weeks of treatment is probably more reliable for biotherapies regularly administered subcutaneously (etanercept, adalimumab, interleukin-1 receptor antagonist) than infliximab given intravenously every 6 or 8 weeks, thereby highlighting the potential contribution of evaluation during the between-infusion interval that was not programmed in this study. Managing patients in a hospital–private practice network by means of file sharing, for example, can easily rectify this weakness. Furthermore, infliximab was only evaluated at 3 mg/kg, while it is well-established that a substantial percentage of patients fail to respond to this dosage. Therefore, true infliximab (INF) non-responders would be RA patients who have failed to respond to escalating dosages of infliximab. Several of the non-responders might have been classified as responders if higher doses of INF had been given. This is certainly a limitation of this study but we respected recommendations of the French authorities. Finally, the criterion of DAS28 response after 1 yr of treatment was not retained because some patients were withdrawn from the study due to side-effects or treatment failures and the dose infused or intervals between infusions were adapted to the response after the fifth infusion. As a consequence, the methodology applied here (non-randomization of DMARD, absence of intermediate evaluation, number of patients) probably contributed to skewing the analysis of the data preventing the identifications of markers predictive of the response. In addition, radiographs of the hands and feet, and their assessment with the Sharp score modified by van der Heijde, were not analysed as criteria of a structural response and/or a potential predictive marker. Indeed, patients with high Sharp scores might respond better to infliximab than patients with low scores. Pertinently, the ASPIRE (active controlled study of patients receiving infliximab for the treatment of RA of early onset) study patients who responded best to the methotrexate–inflixmab combination had more severe structural involvement at inclusion [45]. In our study, the Sharp score was not included into the analysis because our objective was to identify factor(s) predicting the response to infliximab and methotrexate or leflunomide that might easily be used in routine practice. Other parameters could have been considered: pro-angiogenic molecules, like vascular endothelial growth factor (VEGF), which are being evaluated in another ongoing trial. Measurement of serum infliximab concentrations has not yet become a routine laboratory test. Finally, because of their high cost, genetic polymorphism analyses could not be carried out for all the patients included in this study. Although genetic markers (genetic polymorphisms) seem to coincide better with an approach detecting markers predictive of a good therapeutic response, the reasons listed above do not enable us to formally exclude the relevance of analysing soluble molecules, especially because the TNF-{alpha} blockers have an impact on some of them.

Indeed, analysis of the polymorphism of certain genes seems promising but the identification of responders and non-responders to TNF-{alpha} blocking agents cannot be restricted to the study of cytokine polymorphisms, even though these molecules are currently thought to play major roles in RA pathophysiology. Hence, other approaches are needed to identify markers predictive of the response. Examination of gene-expression levels is one possible strategy to identify target genes or candidates for polymorphism studies, but also genetic markers that a priori have no apparent association with the pathophysiology and/or mode of action of the molecule tested. Transcriptomic and proteomic approaches also represent tools that could be used to identify these sought after markers. In the long-term, these bio-informatics tools should enable the construction of algorithms integrating all the genetic, transcriptomic and proteomic data to identify a combination of markers able to predict the response to a molecule and thereby improve treatment of RA.


   Conclusion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Conclusion
Acknowledgement
 References
 
The principal objective of our study on 76 patients with severe and longstanding RA was to identify marker(s) predictive of the response to infliximab combined with methotrexate or leflunomide. The clinical and biological parameters routinely used, the parameters associated with RA severity and some biological parameters implicated in RA pathophysiology are not able to predict the response to infliximab. In this study, the too small sample size and lack of good responders (a consequence of the former) probably contributed to limiting the detection of potential markers of responsiveness to infliximab. This approach was used as a preliminary and original exploratory analysis and, based on its results, we cannot yet preclude the existence of predictive markers. Other studies including more patients are required to conclude as to the contribution of soluble markers in this context.

Formula


   Acknowledgement
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Conclusion
 Acknowledgement
References
 
The authors are grateful to Janet Jacobson for her valuable advice in editing the manuscript.

The authors have declared no conflicts of interest.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Conclusion
 Acknowledgement
 References
 

  1. Elliott MJ, Maini RN, Feldmann M, et al. (1994) Randomised double-blind comparison of chimeric monoclonal antibody to tumour necrosis factor alpha (cA2) versus placebo in rheumatoid arthritis. Lancet 344:1105–10.[CrossRef][Web of Science][Medline]
  2. Maini RN, Breedveld FC, Kalden JR, et al. (1998) Therapeutic efficacy of multiple intravenous infusions of anti-tumor necrosis factor alpha monoclonal antibody combined with low-dose weekly methotrexate in rheumatoid arthritis. Arthritis Rheum 41:1552–63.[CrossRef][Web of Science][Medline]
  3. Maini RN, St Clair EW, Breedveld F, et al. (1999) Infliximab (chimeric anti-tumour necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: a randomised phase III trial. ATTRACT Study Group. Lancet 354:1932–9.[CrossRef][Web of Science][Medline]
  4. Maini RN, Breedveld FC, Kalden JR, et al. (2004) Sustained improvement over two years in physical function, structural damage, and signs and symptoms among patients with rheumatoid arthritis treated with infliximab and methotrexate. Arthritis Rheum 50:1051–65.[CrossRef][Web of Science][Medline]
  5. Weinblatt ME, Keystone EC, Furst DE, et al. (2003) Adalimumab, a fully human anti-tumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial. Arthritis Rheum 48:35–45.[CrossRef][Web of Science][Medline]
  6. Weinblatt ME, Kremer JM, Bankhurst AD, et al. (1999) A trial of etanercept, a recombinant tumor necrosis factor receptor:Fc fusion protein, in patients with rheumatoid arthritis receiving methotrexate. N Engl J Med 340:253–9.[Abstract/Free Full Text]
  7. Olsen NJ and Stein CM. (2004) New drugs for rheumatoid arthritis. N Engl J Med 350:2167–79.[Free Full Text]
  8. O'Dell JR. (1999) Anticytokine therapy – a new era in the treatment of rheumatoid arthritis? N Engl J Med 340:310–2.[Free Full Text]
  9. Mugnier B, Balandraud N, Darque A, Roudier C, Roudier J, Reviron D. (2003) Polymorphism at position –308 of the tumor necrosis factor {alpha} gene influences outcome of infliximab therapy in rheumatoid arthritis. Arthritis Rheum 48:1849–52.[CrossRef][Web of Science][Medline]
  10. van Vollenhoven R, Harju A, Brannemark S, Klareskog L. (2003) Treatment with infliximab (Remicade) when etanercept (Enbrel) has failed or vice versa: data from the STURE registry showing that switching tumour necrosis factor alpha blockers can make sense. Ann Rheum Dis 62:1195–8.[Abstract/Free Full Text]
  11. Brocq O, Plubel Y, Breuil V, et al. (2002) Etanercept–infliximab switch in rheumatoid arthritis 14 out of 131 patients treated with anti TNFalpha. Presse Méd 31:1836–9.[Web of Science][Medline]
  12. Bridges S Jr. (2004) Genetic markers of treatment response in rheumatoid arthritis. Arthritis Rheum 50:1019–22.[CrossRef][Web of Science][Medline]
  13. Vittecoq O, Pouplin S, Krzanowska K, et al. (2003) Rheumatoid factor is the strongest predictor of radiological progression of rheumatoid arthritis in a three-year prospective study in community-recruited patients. Rheumatology 42:939–46.[Abstract/Free Full Text]
  14. Bas S, Genevay S, Meyer O, Gabay C. (2003) Anti-cyclic citrullinated peptide antibodies, IgM and IgA rheumatoid factors in the diagnosis and prognosis of rheumatoid arthritis. Rheumatology 42:677–80.[Abstract/Free Full Text]
  15. Kroot EJ, de Jong BA, van Leeuwen MA, et al. (2000) The prognostic value of anti-cyclic citrullinated peptide antibody in patients with recent-onset rheumatoid arthritis. Arthritis Rheum 43:1831–5.[CrossRef][Web of Science][Medline]
  16. Rantapaa-Dahlqvist S, de Jong BA, Berglin E, et al. (2003) Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis Rheum 48:2741–9.[CrossRef][Web of Science][Medline]
  17. Cunnane G, Fitzgerald O, Beeton C, Cawston TE, Bresnihan B. (2001) Early joint erosions and serum levels of matrix metalloproteinase 1, matrix metalloproteinase 3, and tissue inhibitor of metalloproteinases 1 in rheumatoid arthritis. Arthritis Rheum 44:2263–74.[CrossRef][Web of Science][Medline]
  18. Hofbauer LC and Heufelder AE. (2001) The role of osteoprotegerin and receptor activator of nuclear factor {kappa}B ligand in the pathogenesis and treatment of rheumatoid arthritis. Arthritis Rheum 44:253–9.[CrossRef][Web of Science][Medline]
  19. Furumitsu Y, Inaba M, Yukioka K, et al. (2000) Levels of serum and synovial fluid pyridinium crosslinks in patients with rheumatoid arthritis. J Rheumatol 27:64–70.[Web of Science][Medline]
  20. O'Dell JR, Lemley-Gillespie S, Palmer WR, Weaver AL, Moore GF, Klassen LW. (1991) Serum selenium concentrations in rheumatoid arthritis. Ann Rheum Dis 50:376–8.[Abstract/Free Full Text]
  21. Tarp U, Overvad K, Hansen JC, Thorling EB. (1985) Low selenium level in severe rheumatoid arthritis. Scand J Rheumatol 14:97–101.[Web of Science][Medline]
  22. Heliovaara M, Knekt P, Aho K, Aaran RK, Alfthan G, Aromaa A. (1994) Serum antioxidants and risk of rheumatoid arthritis. Ann Rheum Dis 53:51–3.[Abstract/Free Full Text]
  23. Bandt MD, Grossin M, Driss F, Pincemail J, Babin-Chevaye C, Pasquier C. (2002) Vitamin E uncouples joint destruction and clinical inflammation in a transgenic mouse model of rheumatoid arthritis. Arthritis Rheum 46:522–32.[CrossRef][Web of Science][Medline]
  24. van Gestel AM, Prevoo MLL, van't Hof MA, van Rijswijk MH, van de Putte LB, van Riel PLCM. (1996) Development and validation of the European League Against Rheumatism response criteria for rheumatoid arthritis. Arthritis Rheum 39:34–40.[Medline]
  25. Guillemin F, Braincon S, Pourel J. (1991) Measurement of the functional capacity in rheumatoid polyarthritis: a French adaptation of the Health Assessment Questionnaire (HAQ). Rev Rhum Mal Ostéoartic 58:459–65.[Medline]
  26. Vittecoq O, Jouen-Beades F, Krzanowska K, et al. (2001) Rheumatoid factors, anti-filaggrin antibodies and low in vitro interleukin-2 and interferon-gamma production are useful immunological markers for early diagnosis of community cases of rheumatoid arthritis. A preliminary study. Joint Bone Spine 68:144–53.[CrossRef][Web of Science][Medline]
  27. Vittecoq O, Salle V, Jouen-Beades F, et al. (2001) Autoantibodies to the 27 C-terminal amino acids of calpastatin are detected in a restricted set of connective tissue diseases and may be useful for diagnosis of rheumatoid arthritis in community cases of very early arthritis. Rheumatology 40:1126–34.[Abstract/Free Full Text]
  28. Saulot V, Vittecoq O, Salle V, et al. (2002) Autoantibodies directed against the amino-terminal domain I of human calpastatin (ACAST-DI Ab) in connective tissue diseases. High levels of ACAST-DI Ab are associated with vasculitis in lupus. J Autoimmun 19:55–61.[CrossRef][Web of Science][Medline]
  29. Jouen F, Vittecoq O, Leguillou F, et al. (2004) Diagnostic and prognostic values of anti glucose-6-phosphate isomerase antibodies in community-recruited patients with very early arthritis. Clin Exp Immunol 137:606–11.[CrossRef][Web of Science][Medline]
  30. Saulot V, Vittecoq O, Charlionet R, et al. (2002) Presence of autoantibodies to the glycolytic enzyme alpha-enolase in sera from patients with early rheumatoid arthritis. Arthritis Rheum 46:1196–201.[CrossRef][Web of Science][Medline]
  31. Lapointe J, Li C, Higgins JP, et al. (2004) Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci USA 101:811–6.[Abstract/Free Full Text]
  32. Nielsen TO, West RB, Linn SC, et al. (2002) Molecular characterisation of soft tissue tumours: a gene expression study. Lancet 359:1301–7.[CrossRef][Web of Science][Medline]
  33. Eisen MB, Spellman PT, Brown PO, Botstein D. (1998) Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95:14863–8.[Abstract/Free Full Text]
  34. Sidiropoulos P, Bertsias G, Kritikos HD, Kouroumali H, Voudouris K, Boumpas DT. (2004) Infliximab treatment for rheumatoid arthritis, with dose titration based on the disease activity score: dose adjustments are common but not always sufficient to assure sustained benefit. Ann Rheum Dis 63:144–8.[Abstract/Free Full Text]
  35. Klimiuk PA, Sierakowski S, Domyslawska I, Chwiecko J. (2004) Effect of repeated infliximab therapy on serum matrix metalloproteinases and tissue inhibitors of metalloproteinases in patients with rheumatoid arthritis. J Rheumatol 31:238–42.[Abstract/Free Full Text]
  36. Crnkic M, Mansson B, Larsson L, Geborek P, Heinegard D, Saxne T. (2003) Serum cartilage oligomeric matrix protein (COMP) decreases in rheumatoid arthritis patients treated with infliximab or etanercept. Arthritis Res Ther 5:181–5.
  37. Ziolkowska M, Kurowska M, Radzikowska A, et al. (2002) High levels of osteoprotegerin and soluble receptor activator of nuclear factor kappa B ligand in serum of rheumatoid arthritis patients and their normalization after anti-tumor necrosis factor alpha treatment. Arthritis Rheum 46:1744–53.[CrossRef][Web of Science][Medline]
  38. Ziolkowska M and Maslinski W. (2003) Laboratory changes on anti-tumor necrosis factor treatment in rheumatoid arthritis. Curr Opin Rheumatol 15:267–73.[CrossRef][Web of Science][Medline]
  39. Vuolteenaho K, Moilanen T, Hamalainen M, Moilanen E. (2002) Effects of TNFalpha-antagonists on nitric oxide production in human cartilage. Osteoarthritis Cartilage 10:327–32.[CrossRef][Web of Science][Medline]
  40. Bobbio-Pallavicini F, Alpini C, Caporali R, Avalle S, Bugatti S, Montecucco C. (2004) Autoantibody profile in rheumatoid arthritis during long-term infliximab treatment. Arthritis Res Ther 6:R264–72.[CrossRef][Web of Science][Medline]
  41. de Rycke L, Verhelst X, Kruithof E, et al. (2005) Rheumatoid factor, but not anti-cyclic citrullinated peptide antibodies, is modulated by infliximab treatment in rheumatoid arthritis. Ann Rheum Dis 64:299–302.[Abstract/Free Full Text]
  42. Antoni C, Maini R, Breedveld F. (1999) Subgroup analyses show consistent benefit in a 30 week double blind trial (ATTRACT) with an anti-TNF{alpha} monoclonal antibody, infliximab, in rheumatoid arthritis patients on methotrexate. Ann Rheum Dis 58:Suppl 6, 217.
  43. Hrycaj P, Korczowska I, Lacki JK. (2003) Infliximab in rheumatoid arthritis—an the response be predicted [abstract]? Ann Rheum Dis 62:Suppl I, 403.
  44. Wolbink GJ, Voskuyl AE, Lems WF, et al. (2005) Relationship between serum trough infliximab levels, pretreatment C reactive protein levels, and clinical response to infliximab treatment in patients with rheumatoid arthritis. Ann Rheum Dis 64:704–7.[Abstract/Free Full Text]
  45. St Clair EW, van der Heijde DM, Smolen JS, et al. (2004) Combination of infliximab and methotrexate therapy for early rheumatoid arthritis: a randomized, controlled trial. Arthritis Rheum 50:3432–3.[CrossRef][Web of Science][Medline]
Submitted 29 October 2005; revised version accepted 27 June 2006.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Rheumatology (Oxford)Home page
C. Bansard, T. Lequerre, M. Daveau, O. Boyer, F. Tron, J.-P. Salier, O. Vittecoq, and X. Le-Loet
Can rheumatoid arthritis responsiveness to methotrexate and biologics be predicted?
Rheumatology, September 1, 2009; 48(9): 1021 - 1028.
[Abstract] [Full Text] [PDF]

Home page
 Rheumatology (Oxford)Home page
N. Sekiguchi, S. Kawauchi, T. Furuya, N. Inaba, K. Matsuda, S. Ando, M. Ogasawara, H. Aburatani, H. Kameda, K. Amano, et al.
Messenger ribonucleic acid expression profile in peripheral blood cells from RA patients following treatment with an anti-TNF-{alpha} monoclonal antibody, infliximab
Rheumatology, June 1, 2008; 47(6): 780 - 788.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
46/3/446    most recent
kel262v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (8)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lequerré, T.
Right arrow Articles by Vittecoq, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lequerré, T.
Right arrow Articles by Vittecoq, O.
Related Collections
Right arrow Rehabilitation
Right arrow Rheumatoid Arthritis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?