Rheumatology Advance Access originally published online on October 13, 2006
Rheumatology 2007 46(2):185-187; doi:10.1093/rheumatology/kel342
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
EDITORIALS |
Very early ‘Rheumatoid’ arthritis cohorts: limited by selection
Istanbul University, Istanbul Medical Faculty, Department of Internal Medicine, Division of Rheumatology, Istanbul, Turkey
Correspondence to: Murat Inanc, Istanbul Medical Faculty, Department of Internal Medicine, Division of Rheumatology, Capa 34390 Istanbul, Turkey. E-mail: drinanc{at}istanbul.edu.tr
The heterogeneous course and prognosis of early arthritis (EA) patients continues to be a major challenge for clinicians involved in the management of these patients. A variable proportion of EA patients can be classified under rheumatoid arthritis (RA) or another disease (e.g. spondylarthritis and lupus) during the follow-up, while others remain undifferentiated or resolve spontaneously. In cohort studies of EA, the cumulative prevalence of RA varied significantly and was reported between 7 and 42%, which indicates that selection factors may influence the outcome of such cohorts [1].
Identifying EA patients at risk for developing persistent and/or erosive arthritis is mandatory for selecting a treatment strategy according to the current early aggressive treatment approach [2]. It has been widely discussed that the current ACR 1987 classification criteria for RA is not an appropriate tool for treatment purposes in the very early phase of the disease mainly because of low sensitivity and lack of exclusions [3]. Instead of classifying a subset of disease with an already proven persistent and destructive nature, a need for a criteria to guide clinicians about selecting treatment strategies in very early patients is apparent [3]. Propensity of arthritis to become persistent and/or destructive would be the focus of these efforts [2, 4]. Recently, an The European League Against Rheumatism (EULAR) expert panel formed a set of recommendations based on evidence-based literature review, including the early referral of patients with arthritis within 6 weeks and the early start of disease-modifying anti-rheumatic drugs (DMARDs), mainly methotrexate, in patients at risk of developing persistent and/or erosive arthritis [4].
In this issue of Rheumatology, Machold et al. [5] report the outcome of patients classified as RA during the follow-up from Austrian Early Arthritis Action, which included patients with arthritis of no longer than 12 weeks. They included 314 patients into the cohort in 5 yrs starting from 1996, and among the regular 229 patients they identified 138 cases of RA. More than half of this cohort was lost to follow-up. They present here the radiological outcome of 55 patients who completed 3 yrs of follow-up. Clinical and serological variables that were collected from this very early stage as predictive factors of erosive disease were analysed.
This cohort has many patients lost to follow-up and the analysis is restricted only to the remaining patients with follow-up data, which limits the validation of the results. Nevertheless, from the data of the remaining patients, their findings and conclusions might have important implications for practising rheumatologists and researchers. For this report, their specific aims were to determine:
- the frequency of erosive arthritis among patients followed up from a very early stage to the end of early stage,
- the features of patients who did and did not develop radiological erosions,
- the pace of erosive changes during the 3 yr period and
- prognostic factors for the development of erosive disease.
Their results showed that almost two-thirds of the patients developed erosive disease in 3 yrs despite being followed up by a specialized arthritis clinic. In almost half of the patients erosions became manifest in the first year of the disease. In a real-life-like situation (treatment was not pre-defined), most patients started DMARD monotherapy (87%) at median 19 weeks from the onset of the symptom. During the observation period, many patients had switched to other DMARDs. Most of the patients used steroids, and continuous use was more frequent in erosive patients. Owing to the avaliability, tumour necrosis factor (TNF)-blockers started in only 11% of the erosive patients.
In this patient group, none of the baseline clinical parameters and acute-phase responses could identify patients who developed erosions, but 91% of patients with rheumatoid factor (RF) positivity and 96% of patients with anti-cyclic citrullinated peptide (anti-CCP) positivity ended up with erosive disease. The authors pointed out that based on a regression model >60% of radiological progression can be explained by RF, anti-CCP, c-reactive protein (CRP), cumulative swolen joint count and total time in low disease activity/remission. On the other hand, one-fourth of the patients who developed erosive disease did not have any positive serology. Strikingly for the core set parameters, the difference between the erosive and non-erosive groups became apparent only during the second or third year. Long-standing clinical (DAS28) remission as an ultimate target was reached in only 29% of the patients who were either in the non-erosive group (69%) or who had initial erosions without progression (31%).
The study by Machold et al. [5], tried to answer very important questions but also raised important issues to be discussed.
The effect of patient recruitment and lost-to follow-ups in EA cohort studies
The rationale for cohort studies of EA are well-established, and the information gained from these studies on the outcome of these patients is an essential part of our current understanding of RA [6]. However, despite all efforts, fulfilling the reporting requirements for longitudinal studies is difficult [7]. In some of these observational studies, such as the present one [5], treatment before and after the entry to the cohort is not standardized and the assessment of the effect of treatment on the outcome is limited and uncontrolled. The difficulty of keeping the patients in the cohorts is increasing with the longer follow-ups. The statistical analysis of data from observational cohorts needs a multivariate approach considering the use of all relevant data and decreasing number of patients with time. Ideally, methods appropriate for longitudinal analysis should be used, and the validation of results from one data set to another is recommended to test the generalizability of the data [7]. It is crucial to develop methods to keep the patients in the cohorts, to understand the reasons of subject loss and finally to describe the characteristics of those lost to follow-ups to avoid left/right censorship bias.
Use of RA classification criteria in EA cohorts
Machold et al. [5] have chosen the term ‘very recent onset RA’ for their title (instead of VERA) to clearly differentiate their patients from undifferentiated EA. Their key words and running title include VERA, which reflect the difficulty in the terminology. Sixty percent of their EA cohort (with follow-up data, 138 out of 229 patients) consisted of patients fulfilling ACR criteria for RA during the follow-up. The prevalence of RA is one of the highest among similar cohorts, and possibly reflects the effect of their entry criteria plus the longer and regular follow-ups with retaining severe patients [1]. Although the general consensus is <3 months of the symptoms for a the ‘window of opportunity’ in early rheumatoid arthritis (ERA), most of the patients do not meet the ACR criteria in this period. If the criteria is not applied cumulatively, patients might fulfill the criteria temporarily during the course of the disease and some patients satisfy the requirements as late as 5 yrs [8]. The literature on ERA is more difficult, and trials include patients with 6–24 (up to 5 yrs) months of symptom duration [9]. Fulfilling the ACR criteria seems to be a function of time and does not help in predicting patients with persistent and/or erosive arthritis [10, 11]. It should be considered that patients fulfilling ACR criteria with different combinations of individiual cirterion at some point may represent a heterogeneous patient population with different course and outcome.
Anti-CCP test in guiding the management of ERA
A recent systematic review revealed that anti-CCP antibody positivity with a second-generation test is a more specific marker for the diagnosis of RA than RF with comparable sensitivity [12]. A combined analysis of publications concerning more than 2000 patients with early undifferentiated arthritis showed a prevalance of 23% for anti-CCP2 antibodies. The prevalence increased to 51% in more than 1000 patients who fulfilled RA classification criteria after a median follow-up of 18 months. The mean odds ratio was reported to be 25 (confidence interval, 18–35) for the development of RA [12]. It was previously reported that the clinical presentation of anti-CCP-positive and anti-CCP-negative patients were similar, but the anti-CCP-positive patients had more severe radiological destruction during the disease course [13]. The study by Machold et al. [5] also confirms the importance of baseline anti-CCP2 serology in ERA and they reported stable titres during the follow-up. There is a possibility of bias that patients retained in this cohort were selected by their anti-CCP status because of the close association with severe disease. In a study of consecutive patients, in which they showed the importance of the anti-CCP test in predicting radiological and functional outcome, Quinn et al. [14] found a sensitivity of 60% for anti-CCP antibodies in RF-negative RA. In our multicentre study designed to compare anti-CCP status in patients with RF-positive and RF-negative RA, we found that only 20% of our patients were anti-CCP2-positive among established RF-negative RA patients [15]. If positive, the anti-CCP test is an important surrogate marker for especially RF-negative RA and predictive of radiological progression [14, 15]. Additionally, accumulating data indicate that there is a clear difference in genetic background between anti-CCP-positive and anti-CCP-negative patients with RA, and the presence of these autoantibodies is associated with shared epitope exclusively [16]. Cytokine profiles of anti-CCP-positive EA patients anti-CCP-negative patients also showed significant differences [17]. On the other hand, several studies including ours [15] confirmed the existence of seronegative (both for RF and anti-CCP) patients who developed erosive disease.
Radiological outcome of patients with EA
There is a significant variation in the recruitment criteria and methods for the assessment of radiological damage in EA inception cohorts, which may have an impact on the erosion risk. As with similar to the present study [5], erosions may be present in the first year, peak during the second year and increase by almost 70% at the end of the third year. On the contrary, it was previously reported that first erosions may develop after 2 yrs or even longer from the disease onset [18]. It should be noted that several different types of radiological progression have been reported including linear or non-linear developments (reviewed in [6]). The relationship between treatment choices and radiological damage in EA is complex, and TNF antagonists might have important influence on the development of these lesions. We do not know how the advent of imaging techniques and the use of MRI and ultrasonography will effect future data. Early aggressive treatment is expected to improve radiological outcome in EA cohorts at least in patients who could be identified as pro-erosive.
Conclusion
‘Austrian Early Arthritis Action’ is one of the important efforts that aims to form an inception cohort of EA patients to investigate the long-term outcomes and predictive factors for bad prognosis of these patients. These observational studies have been productive and have resolved many issues in EA. It is important to understand that EA cohort studies have limitations in terms of recruiting target population, retaining patients in the study, data collection and analysing treatment effects. These limitations may be different among countries and health systems, which makes extrapolations of data difficult, and hence should be studied further. The emerging concept of anti-CCP-positive patients as a distinct patient population with some similarities to other EA patients, but poor outcomes and different pathogenesis may improve clinical management of those. On the other hand, there exist pro-erosive patients without any clinical or serological clue during the course of the disease with varying patterns of joint destruction reflecting the heterogeneity of RA patients fulfilling current classification criteria.
Acknowledgement
The author is grateful to Professor Alan Silman for his helpful suggestions and comments on the manuscript.
The author received honoraria/sponsorship for consultancies, scientific meetings, educational and research purposes from Abbot laboratories, Sanofi-Aventis, Schering Plough and Wyeth Laboratories.
References
- Hitchon CA, Peschken CA, Shaikh S, El-Gabalawy HS. (2005) Early undifferentiated arthritis. Rheum Dis Clin N Am 31:605–26.[CrossRef][Web of Science][Medline]
- Smolen JS, Aletaha D, Machold KP. (2005) Therapeutic strategies in early rheumatoid arthritis. Best Practice & Research Clinical Rheumatology 19:163–77.[CrossRef][Medline]
- Aletaha D, Breedveld FC, Smolen JS. (2005) The need for new classification citeria for rheumatoid arthritis. Arthritis Rheum 52:3333–6.[CrossRef][Web of Science][Medline]
- Combe B, Landewe R, Lukas C, et al. (2006) EULAR recommendations for the management of early arthritis. Report of a task force of the European Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis (e-pub).
- Machold KP, Stamm TA, Nell VPK, et al. (2006) Very recent onset rheumatoid arthritis: clinical and serological patient characteristics associated with radiographic progression over the first years of disease. Rheumatology (in press).
- Symmons DP, Hazes JM, Silman AJ. (2003) Cases of early inflammatory polyarthritis should not be classified as having rheumatoid arthritis. J Rheumatol 30:902–4.
[Free Full Text] - Silman A and Symmons D. (1999) Reporting requirements for longitudinal observational studies in rheumatology. J Rheumatol 26:481–3.[Web of Science][Medline]
- Symmons DPM and Silman AJ. (2006) Aspects of early arthritis. What determines the evolution of early undifferentiated arthritis and rheumatoid arthritis? An update from the Norfolk Arthritis Register. Arhritis Research & Therapy 8:214.
- Sokka T, Hannnen P, Möttönen T. (2005) Conventional disease modifying antirheumatic drugs in early rheumatoid arthritis. Rheum Dis Clin N Am 31:729–44.[CrossRef][Web of Science][Medline]
- Young A. (2005) Early rheumatoid arthritis. Rheum Dis Clin N Am 31:659–79.[CrossRef][Web of Science][Medline]
- Visser H, le Cessie S, Vos K, et al. (2002) How to diagnose rheumatoid arthritis early: a prediction model for persistant (erosive) arthritis. Arthritis Rheum 46:357–65.[CrossRef][Web of Science][Medline]
- Avouac J, Gossec L, Dougados M. (2006) Diagnostic and predictive value for anti-cyclic citrullinated protein antibodies in rheumatoid arthritis: a systematic literature review. Ann Rheum Dis 65:845–51.
[Abstract/Free Full Text] - van der Helm-van Mil AHM, Verpoort KN, Breedveld FC, Toes REM, Huizinga TWJ. (2005) Antibodies to citrullinate proteins and differences in clinical progression of rheumatoid arthritis. Arthritis Res Therapy 7:R949–58.[CrossRef]
- Quinn MA, Gough AKS, Green MJ, et al. (2006) Anti-CCP antibodies measured at disease onset help identify seronegative rheumatoid arthritis and predict radiological and functional outcome. Rheumatology 45:478–80.
[Abstract/Free Full Text] - Inanc N, Dalkilic E, Kamali S, et al. (2006) Anti-CCP antibodies in rheumatoid arthritis and psoriatic arthritis. Clin Rheumatol (e-pub).
- de Vries RRP, Huizinga TWJ, Toes REM. (2005) Redefining the HLA and RA association: to be or not to be anti-CCP positive. J Autoimmune 25:21–5.[CrossRef]
- Hitchon CA, Alex P, Erdile LB, et al. (2004) A distinct multicytokine profile is associated with anti-cyclical citrullinated peptide antibodies in patients with early untreated inflammatory arthritis. J Rheumatol 31:2336–46.
[Abstract/Free Full Text] - Bukhari M, Harrison B, Lunt M, Scott DGI, Symmons DPM, Silman AJ. (2001) Time to first occurrence of erosions in inflammatory polyarthritis. Arthritis Rheum 44:1248–53.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K P Liao, K L Batra, L Chibnik, P H Schur, and K H Costenbader Anti-cyclic citrullinated peptide revised criteria for the classification of rheumatoid arthritis Ann Rheum Dis, November 1, 2008; 67(11): 1557 - 1561. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||