Synovitis

In a synovial biopsy study performed by Kruithof et al, comparing patients with RA, ankylosing spondylitis (AS), PsA and undifferentiated spondyloarthropathy (uSpA), the synovial phenotype was similar in AS-uSpA-PsA and distinct from RA.3 The patients with AS-uSpA-PsA had less sublining layer infiltrate, a greater number of polymorphonuclear leucocytes, CD163 macrophages and vascularity, and no intracellular citrullinated proteins compared with RA. The Dublin group has also shown unique differences in vascularity in PsA synovium.4 However, in a more recent study by van Kuijk et al, it was observed that there was no difference between RA and PsA synovium relative to CD4 and CD8 lymphocytes, CD68 and CD163 macrophages, adhesion molecules, matrix metalloproteinase 1 (MMP-1), MMP-3, MMP-13 and tissue inhibitor of MMP, von Willebrand factor, αvβ3, vascular endothelial growth factor, tumour necrosis factor α (TNFα), interleukin (IL)-1β, IL-6 or IL-18, suggesting more similarity than previously thought.5 Factors that may influence synovial phenotypic variation include disease duration, background treatment and sampling location. Many of these cellular constituents and cytokines decrease with effective treatment with methotrexate (MTX) and anti-TNF agents, including T cells, macrophages, adhesion molecules, metalloproteinases, angiogenic factors and intracellular kinases.6,–,11

Several biomarkers have been noted to denote psoriasis disease compared with controls, including receptor activator of nuclear factor-κB ligand, TNF superfamily 14 (TNFSF14), MMP-3 and cartilage oligomeric matrix protein. Biomarkers which distinguish PsA from psoriasis include high sensitivity C-reactive protein, osteoprotegerin, MMP-3 and the ratio of CPII/C2C.2

Enthesitis

Enthesitis, inflammation at the site of attachments of tendons, ligaments and joint capsule fibres, is a unique and important clinical feature of SpAs, in general, and PsA, in particular, and is rarely seen in RA. Most clinical studies have pegged the frequency of enthesitis in PsA cohorts in the 30–50% range,12 13 but this may be an underestimate, as we begin to use assessment instruments in which more numerous enthesial sites are palpated and ultrasonography or MRI demonstrate enthesopathy not appreciated clinically.14 15 As there are few histological studies on enthesitis, McGonagle has helped elucidate pathoanatomical findings via T2 STIR MRI16 17 demonstrating the frequency and widespread nature of this clinical feature in PsA and hypothesising its predilection for lower extremity entheses, such as the Achilles tendon and plantar fascia calcaneal insertions and patellar tendon insertion due to tissue microtrauma.

Osteoimmunology

Osteitis has been frequently demonstrated in patients with PsA and other forms of SpA via MRI.18 19 There is a paucity of histological data providing information about the cellular constituency of these osteitic lesions. Appel et al microscopically analysed the femoral heads of patients with ankylosing spondylitis (AS), RA and osteoarthritis undergoing hip replacement, and noted an increased cellular infiltrate composed of lymphocytes, osteoclasts and increased vascularity at the marrow–endosteal–cartilage interface within bone, not connected to the synovium.20 The presence of these lesions suggests the possibility of a second front of inflammatory attack, in addition to synovitis, on bone and cartilage of the joint. Ritchlin has demonstrated that patients with PsA and with a more erosive phenotype have uniquely activated osteoclast precursor cells in significant numbers which actively home to sites of inflammation, contributing to bone erosions and the striking lytic phenomenon in digits known as arthritis mutilans.21

Spondylitis

Inflammation in the spine is a key distinguishing feature of PsA compared with RA. Unlike RA, in which spine involvement is primarily confined to C1–C2 facets, spine changes in PsA can affect all parts of the spine and the sacroiliac joints, occurring clinically in up to half of patients with PsA and probably a greater percentage can be demonstrated with sensitive imaging techniques.22 Spine lesions include osteitis, enthesitis, facet and sacroiliac arthritis, and syndesmophytes. These clinical features are less severe and uniform than those seen in classic AS. In addition to inflammatory cytokines such as TNFα, osteoproliferative cytokines such as bone morphogenic protein are involved in driving the pathology of PsA spondylitis.23 Research findings derived from subjects with AS are generally used as a surrogate for our understanding of PsA spondylitis. Recently published classification criteria for ‘axial spondylo- arthritis’ developed by the ASAS working group will expand the set of patients assessed in disease state and treatment studies to include PsA spondylitis.24

Skin and nails

Discussion of the pathophysiology, assessment and management of skin and nail disease is beyond the scope of this article and has recently been thoroughly reviewed93.

Assessment

The key clinical features of PsA which should be assessed in order to determine disease severity and effect of treatment, include peripheral arthritis, skin and nail psoriasis, axial disease, enthesitis and dactylitis.25,–,27 Effective treatment leads to improvement of pain, fatigue and depression, inhibition of structural damage to joints and the significant impairment of function and quality of life (QoL) which results from combined skin and musculoskeletal disease. A potential additional benefit will be reduction of inflammation-induced atherogenesis and early mortality from cardiovascular disease (CVD). Assessment of PsA has generally been accomplished by adapting measures used in clinical trials for RA, psoriasis and, to a lesser extent, AS, as well as general measures such as the Short Form 36, which measures function and QoL (table 1).28,–,36 Although generally not validated in PsA, these measures have proved to be reliable and to show adequate discrimination and responsiveness characteristics in therapeutic trials. Several studies have documented the effectiveness of ultrasound and MRI in detecting inflammation in the joints and enthesial sites of patients with PsA, as well as the extent of structural damage.14 32 As these tools become more refined and widely accessible, they will enhance our ability to diagnose PsA earlier and assess the effectiveness of treatment in treating inflammation and inhibiting the progression of joint damage.

Since the majority of patients develop the skin lesions of psoriasis long before musculoskeletal clinical features, raising awareness about the identification of PsA among patients with psoriasis and dermatologists who treat psoriasis is desirable for the sake of early detection of the disease and institution of treatment. Current estimates are that approximately 20–30% of patients with psoriasis will develop PsA (range of prevalence estimates is from 6% to 39%).37 Early and accurate diagnosis of PsA has been a focus of educational initiatives for dermatologists and rheumatologists, as well as patient education initiatives developed and presented by patient service leagues internationally. Members of the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) have developed simple screening tools which can be used by dermatologists and other clinicians to screen for the presence of PsA in their patient populations.38 These instruments are currently being assessed for their sensitivity and specificity before being widely promoted for clinical use. Several groups are trying to identify genetic or other biomarkers which can predict the development of PsA.39 40 It is not known if current use of more effective biological agents for psoriasis before arthritis develops will delay or prevent the appearance of PsA. A key element in recognition and proper management of PsA is the quality of the dermatologist–rheumatologist interaction in the co-management of the patient with PsA.37

Treatment

Optimal treatment of PsA involves use of drugs which singly have the ability to improve multiple clinical domains or use of combinations of treatments which can beneficially affect multiple domains and can be used safely together. A comprehensive review of the evidence for effectiveness of various PsA pharmacotherapy approaches has been conducted by GRAPPA in the process of developing international treatment recommendations (table 2).26 27 41,–,47 Since the severity of each domain will vary from patient to patient, treatment should be appropriately customised. Influencing the choice of treatment are factors such as patient preference for mode and frequency of treatment administration, potential adverse events and cost.

It has been demonstrated that early and aggressive control of disease activity in the management of RA results in significantly better clinical and radiographic outcomes.48 49 Although a similar paradigm of ‘treating to target’ of remission or low disease activity state has not been carefully studied in PsA, it is expected that similar benefits would result from early and ‘tight control’ intervention in the patient with risk factors for moderate to severe disease course. In some studies the Disease Activity Score (DAS) has been applied to questions like ‘can remission be achieved?’. This RA scoring system may not be appropriate to determine a quantitative threshold and does not capture the additional clinical domains of PsA, enthesitis, dactylitis, spine and skin disease.50 GRAPPA and other groups are actively trying to ascertain if a composite DAS dealing with all clinical domains of PsA can be developed that can be easily performed and meaningfully capture the impact of each domain on outcomes.51 52 GRAPPA has developed a measure for defining ‘minimal disease activity’ (table 3). Coates et al led an exercise among GRAPPA members, based on reviewing hypothetical cases, which led to the definition of minimal disease activity criteria for PsA (table 3).53 These criteria were validated by assessing patients in Gladman's patient cohort in Toronto54 and in interventional trial datasets.55 The development of this instrument is a step toward ‘treatment to target’ in PsA.

Clinical predictive factors for progressive disease in PsA include polyarticular involvement, elevated acute phase reactants, evidence of physical disability and erosive joint disease, as well as demonstration of lack of response to initial therapeutic agents.56 Work is also underway to characterise soluble biomarkers which can identify patients at risk for a more severe disease course and serve as markers of treatment response.39 57 58

Pharmacotherapy

Patients with mild forms of musculoskeletal inflammation may use non-steroidal anti-inflammatory drugs (NSAIDs), analgesics, low-dose corticosteroids or receive intra-articular or enthesial injections of steroids.59 60 There is scant trial evidence for the efficacy of these agents in PsA, since this has been primarily developed in RA, AS and osteoarthritis.37 60

Oral disease-modifying antirheumatic drugs

MTX is one of the most commonly used systemic drugs in PsA, despite minimal evidence of efficacy in a small controlled trial.37 60 61 Whereas MTX may improve symptoms of arthritis and psoriasis, its ability to treat enthesitis, dactylitis and to inhibit structural damage has not been prospectively assessed. A 2-year retrospective analysis of matched patients with PsA who were either receiving or not receiving MTX treatment did not show any difference in radiological progression scores in the two groups.62 Using evidence from AS, since this has not been assessed in PsA, MTX is not considered to be effective in treating spinal disease.42 The potential for MTX-induced hepatotoxicity has been a special concern for dermatologists, based on the finding of greater hepatotoxicity in psoriasis than in RA in older serial liver biopsy studies.63 A current hypothesis is that this is at least partly related to the tendency for patients with psoriasis, and to a lesser extent patients with PsA, to be obese and have non-alcoholic hepatic steatosis (‘fatty liver’) as a concomitant liver problem, as well as other factors, such as excess alcohol use. Dermatologists, but not rheumatologists, have historically recommended periodic liver biopsy for monitoring safety. New guidelines for MTX monitoring in dermatology do not require biopsy.64 Although the combination of MTX and TNF inhibitors has in RA been shown to be superior in all clinical parameters of efficacy, including inhibition of structural damage, this has not been systematically assessed in PsA. A recent trial conducted in patients with PsA naïve to MTX, comparing infliximab plus MTX versus MTX monotherapy, demonstrated greater efficacy in clinical measures in the combination group.65 In clinical practice, MTX may sometimes be discontinued after initiation of biological treatment if there is concern about hepatotoxicity, only to be restarted if the patient experiences inadequate control of disease with biological monotherapy.

Although, the largest number of controlled trials of traditional disease-modifying antirheumatic drug treatment in PsA has been conducted with sulfasalazine,59 60 its utility remains limited because of absence of effect on the skin, radiographic evidence and occasional gastrointestinal intolerability.

Leflunomide, a pyrimidine antagonist, has shown effectiveness in PsA and is formally approved for PsA treatment in Europe at a dose of 20 mg/day.59 60

Ciclosporin can achieve rapid improvement of the skin lesions of psoriasis, but evidence for its effectiveness in musculoskeletal disease is scant, and its usefulness is limited by concerns about the adverse effects of hypertension and renal insufficiency.59 60 It has been used in combination with MTX and adalimumab.66

TNFα inhibitors

The efficacy and safety of etanercept in PsA was pivotally established in a phase 3 trial in 205 patients.59 Approximately half of the patients were receiving background MTX and stratified, based on MTX use, to etanercept (50 mg/week) or placebo. Significant improvement was demonstrated in joints, inhibition of structural damage (demonstrated radiographically), skin, function and QoL. Two-year extension data demonstrated sustained efficacy in all domains.60 Background MTX made no difference to outcomes. The drug was well tolerated and no safety concerns emerged apart from those seen in clinical trial and general clinical experience with etanercept in RA.

More recently, the Psoriasis Randomized Etanercept STudy in Subjects with Psoriatic Arthritis (PRESTA) trial assessed 752 patients with highly active PsA and psoriasis (average body surface area involvement with psoriasis 31%) randomised to a standard dose of etanercept approved for PsA, 50 mg/ week (group 1), versus a dose approved for psoriasis, 50 mg twice a week for 12 weeks followed by 50 mg/week thereafter (group 2).68 American College of Rheumatology (ACR) and enthesitis scores improved similarly in both dose arms at 12 and 24 weeks. ACR 20 was 61%/72% at each time point, respectively, in group 1 and 66%/69% in group 2. Of those patients with enthesitis, assessed by Achilles tendon and plantar fascia insertion tenderness, 65% and 66% in groups 1 and 2, respectively, had no enthesitis at week 12 and 76% and 75% had none at week 24. Similarly improved dactylitis scores were noted, demonstrating the ability of etanercept to effectively treat the clinical problems of enthesitis and dactylitis, but no additional advantage was achieved in musculoskeletal domains by using the higher dose initially. Substantial improvement of skin lesions occurred, with Psoriasis Area Severity Index (PASI) 75 response at week 12 seen in 55% versus 36% (p<0.001) in groups 2 and 1, respectively, and 70% and 62% at week 24 (p<0.05). Thus, consideration might be given to initially higher dosing in patients with severe skin disease, followed by standard dosing, an approach commonly used by dermatologists with this and other biological agents in psoriasis.

In a recent serial skin biopsy study in patients with psoriasis using etanercept, transcripts of cytokines in the Th17 cell pathway were diminished, presumably based on decreased stimulation of IL-23 by TNF.69 Traditionally anti-TNFs are considered to effect Th1 pathway cells and cytokines, so these results demonstrate the immunological cross-talk that occurs among various inflammatory cell pathways.

The effectiveness of infliximab, a chimeric monoclonal antibody administered at 5 mg/kg intravenously, was demonstrated in PsA in a pivotal trial of 200 patients.61 As with etanercept, efficacy in all clinical domains of PsA, including inhibition of joint damage, was established.

In a study performed in Russia, 115 patients with relatively early PsA (mean disease duration 2.8–3.7 years) were randomised to receive MTX monotherapy (15 mg/week) or a combination of MTX (15 mg/week) and infliximab, 5 mg/kg, in the standard infusion regimen employed for this agent.65 At 16 weeks, patients in the combination arm had better outcomes, with ACR 20/50/70, DAS28 remission and PASI 75/90 responses in 86%/73%/49%, 69% and 97%/71%, respectively, while in the MTX monotherapy arm, these results were 67%/40%/19%, 29% and 54%/29%, respectively. These results suggest that earlier intervention in PsA can result in very substantial improvements of disease activity, especially with the combination of anti-TNF and MTX treatment, and also provides a demonstration of the potential effectiveness of MTX monotherapy in a cohort with such early disease.

Adalimumab, a fully human anti-TNFα monoclonal antibody administered subcutaneously, 40 mg, every other week, was studied in the Adalimumab Effectiveness in Psoriatic Arthritis Trial (n=313).70 As with other anti-TNFs, significant benefit in joints, skin, function, QoL, inhibition of radiographic damage and fatigue was demonstrated. A 2-year extension study demonstrated sustained ACR and PASI responses and persistent inhibition of x-ray progression.71

In a recent study, the combination of adalimumab, 40 mg subcutaneously every other week, and ciclosporin up to 3 mg/ kg/ day versus monotherapy with each agent was studied in PsA. For skin lesions, each agent alone showed similar efficacy as combination treatment. However, combination of the two showed superior efficacy in joint response than monotherapy with either agent.66

Golimumab is a fully human anti-TNFα monoclonal antibody that is approved in a 50 mg monthly subcutaneous application for PsA, based on a study of 405 patients.72 At this dose, at the primary end point of 14 weeks, ACR 20 was achieved by 51% versus 9% in the placebo group (p<0.001), and ACR 50/70 achieved by 30% and 12%, respectively. PASI 75 response was achieved by 40% at week 14 in 109 patients and 56% at week 24 in 102 patients with at least 3% body surface area involvement evaluable for PASI. Of those patients with enthesitis, assessed by the Maastricht enthesitis index, significantly more patients showed resolution of enthesitis than with placebo. Nail changes also significantly improved as did measures of physical function. These improvements were sustained at 104 weeks in an open extension phase of this trial.73 Inhibition of progressive joint damage at 1 year has been reported.74 Safety experience was commensurate with that of other anti-TNF agents in PsA.

A new anti-TNFα agent, certolizumab pegol, now approved for treatment of RA, is a subcutaneously administered pegylated Fab fragment which is being studied in PsA (Mease, unpublished data).

Although there is scant trial evidence,75 experience in management of PsA with currently available anti-TNF agents suggests that when a clinician switches from one of these agents to another, if the first has not had, or has lost, efficacy or caused side effects, a substantial percentage of patients will respond to another drug in this class.

Inflammatory spine disease has not been formally assessed in PsA clinical trials owing to a number of factors, including the variability of expression of this clinical domain in PsA and uncertainty about the validity, reliability and discriminant ability of clinical and radiographic measures developed for AS when used in PsA. Anti-TNF drugs have shown significant efficacy for axial manifestations of AS.76 Although NSAIDs have been shown to be beneficial for axial symptoms of AS, MTX, sulfasalazine and leflunomide have not,76 suggesting that anti-TNFs are the preferred class of drug to be used in those for whom NSAIDs produce an inadequate response. We do not have controlled evidence to know if the same holds true in PsA, although extrapolation of the AS experience to PsA seems reasonable and has been adopted in the GRAPPA treatment recommendations.26

In summary, the anti-TNFα drugs have shown the greatest efficacy of any treatment to date in the various clinical aspects of PsA. Their efficacy in treating joint disease activity, inhibiting structural damage, and improving function and QoL are similar, and effects on skin are similar, depending on the dose used. These agents tend to be well tolerated and patients generally acclimatise to their parenteral administration, especially when they experience significant efficacy. Safety concerns are present, such as a risk of infection, but no new concerns have arisen in the PsA population compared with the more extensively studied RA patient population.

PsA pharmacotherapy: other biological agents

Alefacept is a fully human fusion protein that blocks interaction between LFA-3 on the antigen-presenting cell and CD2 on the T cell, by attracting natural killer lymphocytes to interact with CD2 to yield apoptosis of particular T-cell clones. It is approved for treatment of psoriasis in the USA and is administered weekly as a 15 mg intramuscular injection, in an alternating 12 weeks on, 12 weeks off regimen in order to allow return of depleted CD4 cells in the off period. A phase 2 controlled trial of alefacept in PsA (n=185) showed that 54% of patients given a combination of alefacept and MTX had an ACR 20 response as compared with 23% in the MTX alone group (p<0.001) at week 24 and showed sustained responses in patients undergoing a second course.77 78 PASI 75 results were 28% and 24%, for the combination group and MTX alone group, respectively. The modest efficacy of this agent has limited its use in PsA, but it is a consideration in patients for whom other drugs have failed or who have had side effects with other drugs.

Abatacept (CTLA4-Ig) is a recombinant human fusion protein that binds to the CD80/86 receptor on an antigen-presenting cell, thus blocking the second signal activation of the CD28 receptor on the T cell. It is administered intravenously monthly and has been approved for use in RA, based on its ability to improve composite joint scores, function, and inhibit radiographic progression. A trial in psoriasis has been conducted and shown efficacy.79 This drug has been evaluated in a phase 2 trial in PsA. In the standard dose arm of 10 mg/kg intravenous monthly (n=40), 48% achieved ACR 20 response compared with 19% in the placebo arm (p=0.006), although there was greater efficacy in the subgroup not previously exposed to anti-TNF treatment.80 This agent has been well tolerated, with the main safety concern being the risk of infection, comparable to the rate seen with other biological agents.

Both IL-12 and IL-23 are overexpressed in psoriasis plaques. IL-23 is a key cytokine which stimulates the proliferation and activation of Th17 lymphocytes, recently appreciated as important inflammatory cells in a variety of inflammatory diseases. Ustekinumab, an IL-12/IL-23 inhibitor, has shown significant efficacy, administered subcutaneously, in psoriasis.81 This agent has also shown efficacy in a preliminary PsA study.82 This was a placebo-controlled crossover study in which group 1 patients received 90 mg weekly for 4 weeks and subsequent placebo injections. At week 12, 42% of this group (n=76) achieved an ACR 20 response compared with 14% in the placebo group (p=0.0002). The originally placebo group 2 (n=70) received 90 mg at weeks 12 and 16 and at week 12 after the first dose, 51% achieved an ACR 20 response. Highly significant skin improvement was noted, as well as significant improvement in enthesitis and function. The drug was generally well tolerated.

ABT-874 is another IL-12/23 inhibitor which shows significant and enduring skin response with monthly subcutaneous dosing and is likely to be studied in PsA.83

A pilot trial of anti-IL-15 compound has shown efficacy in PsA.84 Rituximab, an anti-CD20 agent which ablates B lymphocytes and is approved in the treatment of lymphoma and RA (1000 mg×2 separated by 2 weeks, every 6 months), demonstrated modest efficacy for arthritis, primarily in the subgroup of patients not previously exposed to anti-TNF treatment.85 Apremilast, an oral phosphodiesterase-4 inhibitor that suppresses multiple proinflammatory mediators and cytokines, was tested in a 12-week trial and showed modest efficacy (ACR 20 43.5%) in a cohort receiving 20 mg twice a day. This agent is now being assessed in several phase 3 trials. A monoclonal antibody which inhibits the IL-6 receptor, tocilizumab, is approved for the treatment of RA, and will be tested in PsA. Small molecules such as the JAK inhibitors, administered orally, which have shown efficacy in RA and psoriasis, will soon be tested in PsA.

Another therapeutic approach, not aimed at inflammation, but instead aimed primarily at reduction of bone erosions, is to reduce osteoclastogenesis. One such agent, an inhibitor of rank ligand, denosumab has been shown to improve osteoporosis and also to reduce the erosion rate in RA.86 It is expected that this type of approach will be useful in PsA as well, especially given the known proclivity for excess osteoclast production and activation in this disease.