Sustained rheumatoid arthritis remission is uncommon in clinical practice

Rheumatoid Arthritis (RA) is characterized by joint inflammation leading to joint destruction. This causes decreased functional capacity, work disability, and reduced quality of life [1]. Advances in the understanding of RA pathogenesis have led to the development of novel therapeutic targets and new treatment guidelines aiming for remission [2]. However, it is not generally accepted how best to define disease remission [3]. At least three remission definitions are in use: the Disease Activity Score (DAS)-28 < 2.6 and < 2.3 score, Simplified Disease Activity Index (SDAI) ≤3.3 score and Clinical Disease Activity Index (CDAI) ≤2.8 score [4‐7]. Recently, the American College of Rheumatology (ACR), the European League against Rheumatism (EULAR), and the Outcome Measures in Rheumatology Initiative (OMERACT) developed new remission criteria [8, 9], the ACR/EULAR provisional definition of remission. All of these remission definitions examine disease activity at a single point in time, making them less useful for long-term follow-up studies [10, 11]. Studies evaluating tight control and treat-to-target strategies advocate that remission should be reached as soon as possible and should be maintained during the course of the disease [12, 13]. The investigators suggest that remission must be sustained to halt joint damage [12, 13]. Relatively little is known about the duration of remission in clinical practice. With remission being a stated goal of RA treatment, it would thus be important to compare remission criteria and to examine the duration of RA remission [2].

To understand remission duration in RA, we examined a large clinical cohort of patients followed up for multiple years. Our aims were (a) to describe the duration of remission, regardless of how these patients achieved remission; and (b) to compare remission duration according to different remission criteria.

Because of advances made in therapy and treatment strategies, remission in RA has become the treatment goal in clinical trials and in clinical practice. However, previous studies have already suggested that remission in clinical practice might not be as common as in clinical trials [22, 23]. The aim of our study was to evaluate whether subjects in remission maintain this state over a long period. Our analyses show that not even half of the RA patients maintained remission beyond 1 year, regardless of the remission definition (see Figure 1). Even after multiple years in remission, a patient's disease can become more active. These findings have clinical implications, in that physicians need to continue to monitor RA patients in remission closely and may need to consider treatment changes in these patients if they flare or present with sustained disease activity. According to the Kaplan-Meier survival-curve analyses, the likelihood of patients experiencing active disease after remission decreases as the years in remission increase. Whether patients were considered to be in remission depended on the criteria used, although differences in remission duration were not statistically significant. We found the SDAI and the new ACR/EULAR criteria to be the most stringent. This is in accordance with the aims of the ACR, EULAR, and OMERACT, who requested a stringent remission definition and proposed the SDAI and newly developed ACR/EULAR remission [9]. Although the ACR/EULAR remission definition was designed and validated for use in clinical trials and not for observational studies evaluating remission in real-life clinical practice, we thought it would be interesting to evaluate the definition in the current study [8‐11].

It should be noted that aging of patients could have played a role in not fulfilling remission criteria, because subjects had a median age of 56 years at entrance to BRASS. A study from Finland evaluated individuals from the general population older than 50 years by the ACR remission criteria and concluded that the majority did not meet the criteria [24]. However, the aim of the current study was to evaluate durability of remission, not the incidence of remission. It is not likely that age could have played a major role in the fact that more than half of subjects did not meet remission criteria any more after only 1 year.

Other studies have described sustained remission in daily practice as uncommon, being reached by only 17% to 36% of RA patients for up to 6 months [19, 22, 25, 26]. These studies did not evaluate time in remission beyond 6 months. A recent study investigated the probability of remaining in remission up to 24 months, according to the ACR/EULAR, SDAI, and CDAI remission criteria in two different cohorts [27]. They also concluded that long-term remission is rare, considering that the probability of a remission lasting 2 years was 6% to 14%.

We evaluated differences in duration of sustained remission according to patient and disease characteristics. We did not find statistically significant differences based on gender, seropositivity, or disease duration, but these stratified analyses were underpowered and should be considered only exploratory. A trend seemed to exist toward longer remission duration in men and subjects with 5 years or less disease duration at start of remission. Schipper et al. [19] recently showed that time to achieve remission correlates with remission duration [19]. Although our data suggest the same, we cannot be sure because most BRASS patients enter the study after the initiation of disease; thus subjects could have been in remission before entrance into BRASS.

In addition to data on remission duration, we investigated patterns of regaining remission for DAS-CRP < 2.6. Fewer than a quarter of subjects who were previously in remission at one visit, and then dropped out of remission, regained remission at the next annual visit. Slightly more subjects regained remission if demonstrating remission at two consecutive visits, but the percentage was still low (31%). Subsequently, we evaluated time spent in remission during the specific follow-up period for each subject in BRASS with at least 2 years of follow-up. These data not only show that sustained remission is uncommon in clinical practice, but also that most subjects spend little follow-up time in remission. In the BRASS cohort, most subjects were in remission during less than half of their follow-up visits.

The ultimate goal of treating RA is to achieve remission and halt the progression of joint damage. Several studies show that one time point of meeting remission criteria does not always mean that the disease does not progress [28‐30]. A longer period of meeting the remission criteria might be an indication of cessation of disease activity, rather than drug-induced disease suppression [31]. In juvenile idiopathic arthritis, patients with inactive disease at two or more yearly visits developed less cumulative joint damage than did those in such a state at only one visit [32]. These data suggest that it is important not only to reach remission, but also to maintain this state. It will be important to determine whether RA patients in sustained remission have less disease destruction compared with those not in sustained remission.

In a recently published study, we showed that an increased number of visits in remission was associated with reduced radiographic damage [30]. The goal of the current study was to describe the duration of remission with different criteria; we did not examine how patients achieved remission in this study. This will be the focus of future work.

The BRASS cohort represents patients from one academic practice setting, and data are gathered over a long period. In this single-center cohort, most RA patients entered the cohort with longstanding disease. Because disease activity was measured yearly in BRASS, it is possible that a patient's disease flared between visits. Thus, this limitation of our data would strengthen the conclusion that one time point of remission in no guarantee for sustained remission. Also this will not affect the comparisons of different remission criteria in the survival analysis, because all remission criteria were measured at similar time points.

Recently published recommendations for the management of RA by EULAR suggest that as long as a target of remission or low disease activity has not been reached, treatment should be adjusted by frequent and strict monitoring [2]. Although remission without respect to duration is a good goal, it is not the most important target for therapy. Data from our study confirm that even after reaching remission, strict monitoring is required, as fewer than half of the patients in clinical practice will not sustain remission for more than 1 year.