Is Remission a More Realistic Goal in Psoriatic Arthritis?

Introduction

Psoriatic arthritis (PsA) is a chronic inflammatory arthritis, usually seronegative for rheumatoid factor associated with psoriasis [Taylor et al. 2006; Reece et al. 1999].

Although PsA was first described and classified in the 1970s it is only in more recent times that its serious consequences have been studied and understood more widely, not only among rheumatologists but also in the broader medical community. The clinical phenotype varies widely, which has led to difficulties with classification, diagnosis and therefore predicting prognosis. Initially, PsA was considered a benign disease, one study suggesting only 11% of patients developed erosions over 7 years [Shbeeb et al. 2000]. However, in the same journal it was highlighted that a number of reports suggested a high occurrence of erosions in as many as 46—62% of patients [Veale, 2000]. The incidence of PsA varies from 5.4 to 42% depending on the report. In a Finnish population-based study 46% of patients developed erosions [Kaipiainen-Seppanen, 1996] and in another study 62% of patients worsened and the pattern of disease changed over time [Jones et al. 1994]. Several recent studies, however, suggest PsA is progressive, often disabling and associated with increased mortality [Gladman, 2008]. In a study of early arthritis clinic patients, PsA accounted for 13% of new patients and progressive, erosive damage occurred in almost 50% of patients in the first 2 years [Kane and Pathare, 2005].

In the absence of evidence from randomized clinical trials, methotrexate is generally accepted to be useful for controlling peripheral arthritis, but it has little efficacy in spinal disease [Van Der Horst-Bruinsma et al. 2002]. In a study of early PsA, however, erosive damage appeared to develop even when methotrexate therapy was commenced early [Kane and Pathare, 2005]. This raises the question, should anti-tumour necrosis factor (TNF) agents be introduced early? Remission implies the reversibility of functional impairment, minimal or no progression to joint destruction, and at least a theoretic potential to heal a damaged joint [Emery et al. 2007]. Recent studies suggest remission may now be attainable in rheumatoid arthritis (RA) with the advent of anti-TNF therapy [Breedveld et al. 2004], however RA remission has been defined by different criteria: a Disease Activity Score using 28 joint counts (DAS28) value of 2.6 or lower [Saleem et al. 2006]; imaging — no progression on X-ray/ultrasound/MRI; or American College of Rheumatology (ACR) criteria [Arnett et al. 1988]. Drug-induced remission may be defined as minimal or no clinically detectable disease activity in the presence of continuing drug treatment, which is not stopped or interrupted, but is required to retain the remission state [Goekoop-Ruiterman et al. 2005]. Drug-free remission persists in the absence of medication. In a recent editorial, de Vlam and Lories highlighted that remission may now be a possible goal in PsA [de Vlam and Lories, 2008] and our group have published our latest PsA cohort analysis of remission following TNF therapy [Saber et al. 2010].

Clinical features

PsA was first described and classified in the 1970s by Moll and Wright [Moll and Wright, 1973]. It is only in the last decade, however, that PsA has been widely accepted as a disease entity, in its own right, with potentially serious consequences similar to RA, such as articular erosions, joint deformity, disability and increased mortality [Kane et al. 2003; Khan et al. 2003]. It is now recognized that by the time the patient first presents at the first consultation they may already be finding it very difficult to carry out even the necessary daily activities let alone leisure and recreational activities [Husted et al. 1996].

In one of the early and most comprehensive clinical series of PsA patient studies in a Canadian clinic, almost two thirds of patients with established disease were noted to already have developed bone erosions visible on plain radiography [Gladman, 1987]. Since then, several studies have confirmed that erosive changes may develop in up to 60% of patients and increase in frequency at a significant annual rate. This suggests that PsA is neither a mild condition nor is it uncommonly associated with joint damage and subsequent disability [Gladman, 2008; Jones et al. 1994].

In a more recent study of patients attending an early arthritis clinic, Kane and colleagues established that PsA may progress, radiologically, despite adequate control of signs and symptoms with disease-modifying antirheumatic drugs (DMARDs) such as methotrexate [Kane and Pathare, 2005]. In addition, this study highlighted that erosions may occur in as many as 50% of patients by the second year. As it became clear that PsA was a significantly frequent problem with significant morbidity and mortality, and that damage may progress despite DMARDs, we were faced with the challenge of finding more effective treatment.

Advances in therapy

The advent of biological therapies, in particular TNF blocking agents, used to effectively treat patients with RA showed a significant benefit over DMARD monotherapy in controlling signs and symptoms and also significantly reducing the progression of bone damage [Antoni et al. 2005; Van Vollenhoven et al. 2003]. Studies followed which identified TNF as a key inflammatory cytokine in psoriasis [Markham et al. 2006] and PsA [Danning et al. 2000], establishing the rationale for trials of these agents in PsA. TNF agents were initially given to patients with severe, longstanding disease, however in view of the dramatic effect in preventing bone erosion it was only natural that studies were designed to examine the effectiveness of these agents in early disease and measuring their impact on disease activity.

Measuring therapeutic response and remission

The DAS was first developed for RA in 1990 [Van Gestel et al. 1998], before being modified for a 28-joint count (DAS28) and validated [Fransen et al. 2004] to provide a scale that could be calculated using a combination of objective joint scores, inflammatory markers and patient reported outcomes. The DAS28 has now also been validated for use in PsA clinical trials [Fransen et al. 2006] and it has been compared with a DAS using a 68-joint count and shown comparable outcomes in patients with PsA receiving biological therapy [Lynch et al. 2008]. Remission remains poorly defined as mentioned previously, so in view of differences in response criteria, combined ACR/European League Against Rheumatism (EULAR) criteria have recently been agreed and published [Van Tuyl et al. 2009]. Remission cutoff values for patients with RA have been identified and examined in relation to both ACR and EULAR response criteria [Fransen et al. 2006]. Remission can thus be defined in several different ways. The performance of criteria to examine response of PsA was tested by Fransen and colleagues in a study that analysed the individual indices as well as pooled criteria applied to datasets from two clinical trials [Fransen and Van Riel, 2005]. This study concluded that the criteria used for response and the pooled indices, in particular the EULAR response criteria, were superior to the ACR or PSARC pooled indices for differentiating significantly between the placebo drug or placebo. The authors also found that the DAS and DAS28 perform better than individual core set measures. Further studies support and validate the use of the DAS28 as an instrument to measure disease activity of patients with PsA in the setting of biological drug therapy [Gladman et al. 2007; Fransen and Van Riel, 2005]. We therefore applied the DAS28 to assess remission in our prospective study of a cohort of patients with PsA who we treated with biological drug therapy. This study concluded that if the DAS28 is used, we could also successfully compare cohorts of patients with PsA and patients with RA.

TNF blocking agents are more appropriately introduced for patients with PsA of high disease activity. We noted in our own biological treated cohort that 58% of patients with PsA compared with 44% of patients with RA achieve remission after 12 months of biological therapy, using the DAS28 as the criterion to define remission. In our cohort of patients there were significant differences in the DAS28 at baseline so we analysed subgroups of patients with PsA (n = 41) and RA (n = 41) matched for disease activity and found that 63% and 41% achieved remission in PsA and RA at 12months respectively. PsA can affect more peripheral joints because of involvement of the distal interphalangeal joints than RA, therefore we also analysed our cohort to confirm that a 28-joint count was a valid indication of disease response. In previous work we have shown that using a 28-joint or a 68-joint count does not affect the response to treatment [Lynch et al. 2008] but may adversely affect recruitment into clinical trials. Function/ quality of life was studied in our cohort of patients with inflammatory arthritis and we found the mean Health Assessment Questionnaire (HAQ) at 12months to be significantly lower in patients with PsA compared with patients with RA. Therefore, in addition to the DAS28'cutoff', we have found that using patient reported outcome measures of function, a number of remission criteria may be met following treatment with TNF blocking agents. This suggests there is reasonable evidence, according to more than one criterion, that remission should now be the primary aim of therapy for patients with PsA. However, we need to clarify the term ‘remission’ and an agreed measure needs to be clearly defined or perhaps a different term used.

Previous studies have compared response rates in PsA and RA, however the DAS28 response or remission was not studied in this context. In a recent study Coates and colleagues examined EULAR response rates in a cohort of 60 patients with PsA [Coates et al. 2010]. The authors found similar proportions of patients with polyarthritis and oligoarthritis. Treatment was guided by NICE restrictions and British Society for Rheumatology (BSR) guidelines, which may have biased choice or duration of use of biological agents. There were no such restrictions in our Dublin cohort studies. The Dublin cohort may therefore be more representative of routine clinical practice and treatments for PsA. In addition, there were fewer patients on infliximab as the data collection started in November 2004 when the subcutaneous agents were widely available. Coates and colleagues were unable to identify predictors of response. The minimal disease activity (MDA) excludes tender joint count yet this could be a significant cause of patient discomfort. Psoriasis is included in two out of the seven measures used to define MDA, although the skin may not be relevant to PsA activity in many cases. Where skin involvement is significant it may have a major impact on the patient's own perception of disease activity and response to treatment. Entheseal tenderness is one of the seven measures but it is not present in most patients so the scoring system will be hard to apply to a population of patients in a clinic. The HAQ may be useful to assess disease activity and response to treatment. C-reactive protein (CRP) has been used in the DAS28, although the erythrocyte sedimentation rate may be an alternative. However, the CRP is not included in the MDA.

The concept of remission in the treatment of RA is now well established, which may be more achievable if biological therapy is commenced early in the disease duration as in the COMET study [Emery et al. 2008]. EULAR and ACR have now agreed that remission ought to be the aim in RA [Combe et al. 2007]. By applying similar approaches to PsA, it may be possible to identify remission in this disease [De Vlam and Lories, 2008]. In another study, outcome was predicted by quality of life using the SF36. The study reported that improvement in the DAS28 score best predicted the quality of life response in patients with PsA. These results are similar to observations reported in the Norwegian registry [Saad et al. 2010; Kvien et al. 2005].

The skin response can be very unpredictable, however in the Dublin cohort, when a patient responded to treatment, from our experience, enthesitis and spondyloarthropathy, if present, also responded well. Therefore the DAS28 is a simple reliable measure that represents the patient's status when all the swollen and tender joint counts are included, as well as a patient global visual analogue scale (VAS).

On analysing the Dublin database, it appeared that men with PsA attained significantly lower DAS28 scores than women: 1.88 versus 2.65 at 1 year (p <0.05), suggesting a possible gender imbalance in terms of response/remission. The QUEST-RA study also found that HAQ response to treatment differed in men and women with RA [Sokka et al. 2009]. This raises an important question for future research:‘Is there a real gender differential for response to treatment?’ with male gender being a positive predictive factor for better outcome. In conclusion, response and remission rates in patients with PsA being treated with biological therapy appear to be higher than those seen in RA studies. The DAS28, a validated composite measure used to define remission in RA and PsA, is further validated as a useful outcome measure in these patient groups. The available evidence suggests that a greater response to biological therapy may be measured by several criteria in different studies, including routine clinical practice and registries. We conclude that remission is a very real goal and is achievable in patients with PsA. However, a definition for remission in PsA needs to be agreed by the global community. The DAS28 is currently a useful measure that can easily be used in routine clinical practice when following disease progression and response to treatment.