Psoriatic arthritis: Current concepts on pathogenesis-oriented therapeutic options
Introduction
Psoriatic arthritis (PsA) is a chronic inflammatory arthropathy of the peripheral joints, spine, and entheses, associated with psoriasis and characterized by diverse phenotypic subtypes and a variable clinical course. Much progress has been made in identifying the distinctive characteristics of this disease since Alibert first described the association between psoriasis and arthritis in 1818 in “Lepre squammeuse,” his discourse on skin diseases (1). Recent insights into the immunopathogenic mechanisms of PsA have revealed disease characteristics in the synovium, vascular structures, entheses, and bone of PsA patients that are similar to, and distinct from, those of rheumatoid arthritis (RA) as well as other forms of spondylarthritis (SpA), including ankylosing spondylitis (AS), reactive arthritis, and enteropathic arthritis.
Such investigations, along with advances in biotechnology and the development of a number of targeted biologic response modifiers (BRMs) with demonstrated effectiveness for both skin and joint manifestations, have led to substantial progress in the treatment of PsA and a renewed interest in the mechanistic processes behind this complex disease. As with RA and SpA, a subset of patients with PsA fail to respond to conventional therapies and to anti–tumor necrosis factor (anti-TNF) agents that are currently approved for the treatment of PsA. A number of novel agents beyond TNF blockade are under investigation for PsA, underscoring the diverse mechanisms likely to be at play in these heterogeneous phenotypic subtypes of PsA.
Classification and diagnostic strategies
Defining PsA—a work in progress.
In any discussion on PsA, acknowledging the difficulties that underlie the classification and diagnosis of this heterogeneous disease is important. In clinical practice and in intervention studies, as well as in designing and executing proper prospective epidemiologic investigations which are sorely needed for this disease, there is a clear unmet need for a valid, accepted, and easy-to-use set of classification criteria for PsA.
It is accepted that PsA, as a member of the SpA family, is characterized by infrequent seropositivity for rheumatoid factor (RF) and anti–cyclic citrullinated peptide (anti-CCP), as well as an association with HLA–B27 alleles, particularly in those patients with axial involvement. Additional clinical features can include enthesitis, dactylitis, iritis, peripheral arthritis (both oligoarticular asymmetric and polyarticular symmetric), spondylitis, and a variable clinical course. It is this heterogeneity in clinical presentation and course that makes PsA a particularly challenging disease to diagnose and classify relative to other forms of SpA and inflammatory arthropathies. Indeed, the variety of classification criteria proposed over the past 30 years by numerous investigators (2-10) highlights this point. To date, the seminal work of Moll and Wright more than 30 years ago, in which they described their experiences with a large case series of patients with PsA (4), still remains the key reference by which PsA is divided into its 5 clinical subtypes (Table 1).
| Polyarticular, symmetric arthritis (RA-like) |
| Oligoarticular (<5 joints), asymmetric arthritis |
| Distal interphalangeal joint predominant |
| Spondylitis predominant |
| Arthritis mutilans |
- * To meet the Moll and Wright 1973 classification criteria for psoriatic arthritis (4), a patient with psoriasis and inflammatory arthritis who is seronegative for rheumatoid arthritis (RA) must present with 1 of the above 5 clinical subtypes. Criteria specificity is 98% and sensitivity is 91%.
CASPAR contributions.
As a result of the international collaborative efforts of the CASPAR (ClASsification criteria for Psoriatic ARthritis) Study Group, a new set of PsA classification criteria were recently published, based on a large prospective study from 30 rheumatology clinics in 13 countries (11). In the CASPAR study, patient-derived data collected on 588 consecutive clinic patients with PsA and 536 control subjects with other inflammatory arthritides (including RA, AS, and connective tissue disorders), who were matched for approximate disease duration, were used to develop the new classification criteria (Table 2).
| Inflammatory articular disease (joint, spine, or entheseal) plus the following: |
| Psoriasis: current (2), history of (1), family history of (1)† |
| Nail dystrophy (1) |
| Negative rheumatoid factor (1) |
| Dactylitis: Current (1), history of (1)‡ |
| Radiographs (hand or foot) with juxtaarticular new bone formation (1)§ |
- * To meet the ClASsification criteria for Psoriatic ARthritis (CASPAR) 2006 classification criteria for psoriatic arthritis, a patient must have inflammatory articular disease and ≥3 points from the remaining categories; the assigned scores are in parentheses. Criteria specificity is 98.7% and sensitivity is 91.4%.
- † Patient-reported history in a first- or second-degree relative.
- ‡ As recorded by a rheumatologist.
- § Excludes osteophyte formation.
Compared with the other previously published PsA criteria, the CASPAR schematic was found to have better specificity but less sensitivity than the criteria that performed best of the existing classification methods to date (6, 11). Of note, the CASPAR criteria were developed using data from patients with established PsA (mean disease duration ∼12.5 years), and therefore the utility of these criteria for patients presenting at an early stage of disease is unclear; however, this limitation of identifying and classifying patients with early rheumatic disease is clearly not limited to PsA. The work by the CASPAR Study Group provides a timely and critically important discussion on the need to identify accurate, agreed-upon, and clinically applicable classification criteria, since numerous therapeutic agents for PsA currently being studied (and many on the horizon) require rigorous clinical trials which rely upon standardized and validated enrollment criteria.
Pathophysiologic mechanisms
Interplay of genes, environment, and immunologic factors.
Although the etiology of PsA remains unknown, recent research has elucidated the changes occurring in the skin, synovium, enthesium, and bone of patients with PsA, illuminating features of this disease both similar to, and distinct from, other forms of SpA and inflammatory arthritides such as RA. Such basic, translational, and clinical investigations, along with continued advances in biotechnology, have provided a number of attractive and logical targets for the treatment of both individual and collective features of the disease. The interplay between genetic, environmental, and immunologic factors has become more apparent and provides important clues that will enable the further investigation into arenas of potential therapies.
Genetic factors.
Studies have documented that more than 40% of patients with PsA have a first-degree family member with either psoriasis or PsA (5, 12). Moreover, an increased frequency of psoriasis and PsA has been observed in monozygotic and dizygotic twins (13). Additional investigations have proposed several genetic susceptibility loci, with the strongest effect residing within the major histocompatibility complex (MHC) (14). PsA population studies have shown an increased frequency of HLA–B13, B17, B27, B38, B39, DR4, DR7, and Cw6 (12, 15). Genes at the HLA region may be relevant to disease expression in PsA, as suggested by studies showing that HLA–B27 has been associated with radiographic sacroiliitis and axial involvement (16), while HLA–B22 was shown to be protective of damage in a cohort of ∼300 patients with PsA who were followed up prospectively (17). Karason et al (18) performed a genotype analysis of 100 patients with PsA from 39 families, which revealed a paternal transmission associated with chromosome 16q. Linkage to this region has also been observed in other SpA subsets, including AS and Crohn's disease (19-21). Furthermore, a recent study further characterized the association of susceptibility genes in PsA and Crohn's disease, an association independent of both psoriasis and undifferentiated inflammatory arthritis (22) and suggesting that common inflammatory pathways underlie both of these diseases.
Other genes within the MHC region have been explored. MHC class I chain–related A (MICA) A9 and Cw6 have been touted as the strongest genetic susceptibility factors in PsA (16). Previous results have also suggested that TNFα promoter polymorphisms or a gene in linkage disequilibrium with TNFα predispose the patient to psoriasis and PsA (23). In patients treated with interferon-γ (IFNγ), which is a potent inducer of class II MHC expression, PsA can develop (24), a finding that is suggestive of the molecular involvement of class II HLA. In contrast to patients with RA, the prevalence of the HLA–DRB1 shared epitope (SE) is not increased in patients with PsA; however, when present in PsA, the SE is associated with more erosive disease compared with that in patients lacking the SE (19).
Environmental factors.
Trauma and infection have been implicated as causative agents in PsA. Psoriatic lesions arising at areas of trauma (Koebner phenomenon) have been described in a significant percentage of patients with psoriasis (25), and Langevitz et al described PsA following trauma to a joint in 24% of patients in an observational cohort (26). Both bacterial and viral pathogens have been suggested as etiologic agents in PsA. Evidence of a possible link between gram-positive infection and PsA was shown in a study by Vasey and colleagues in which sera from PsA patients showed higher levels of antibody to streptococcal infection (27). Indirect observation of enhanced humoral and cellular immunity to gram-positive bacteria typically found in psoriatic plaques supports the role of bacterial antigens in the pathogenesis of psoriasis and PsA (14).
Immunologic factors.
Although studies of the genetic and environmental aspects of PsA have provided key insights into potential pathophysiologic mechanisms, they are not, at this time, a primary target for developing therapies. In contrast, a number of investigations have elucidated the immunologically mediated processes at play in both the skin and joints of PsA patients, and these findings serve as a foundation ripe for potential therapeutic strategies. Both the skin and the joint in PsA possess a prominent lymphocytic infiltrate, comprising activated T cells localized to dermal papillae in skin and to the sublining stroma in the joint as well as the inflamed enthesis (28-30). An abundance of B lymphocytes forming primitive germinal centers is also present, the significance of which is not clear, because neither psoriasis nor PsA is associated with high levels of circulating antibodies (e.g., RF and anti-CCP) (28). T cell–derived cytokines, including interleukin-1β (IL-1β), IL-2, IL-10, IFNγ, and TNFα, dominate the psoriatic synovium and skin (31).
There is an association between human immunodeficiency virus (HIV) infection and severe PsA (32-35), providing some suggestion of immunologic mechanisms at play. HIV preferentially depletes T helper (CD4+) cells; diseases dependent on these cells, such as RA and systemic lupus erythematosus, typically improve in the presence of HIV. Thus, an interaction of class I HLA molecules and CD8+ T cells may be fundamental to the pathogenesis of PsA (36, 37). In further support of this notion, a predominance of CD8+ T lymphocytes has been noted in both the synovial fluid and entheses of patients with PsA (38). Interestingly, this CD4:CD8 ratio observed in PsA synovial fluid is reversed in RA synovial fluid.
Proinflammatory cytokines, including TNFα, activate endothelial cells, leading to expression of a variety of adhesion molecules, including vascular cell adhesion molecule 1, intercellular adhesion molecule 1 (ICAM-1), and E-selectin, resulting in lymphocyte migration to sites of inflammation (23, 28, 39). TNFα also plays a role in cartilage degradation via increased production of matrix metalloproteinases (MMPs), which are thought to mediate cartilage erosion (40) and up-regulation of angiogenic factors such as vascular endothelial growth factor (VEGF) and transforming growth factor β (TGFβ), providing the distinctive tortuous vessels noted in psoriatic skin and synovium (23, 28, 41).
In addition, TNFα mediates a number of biologic processes that can result in joint damage, including inhibition of bone formation, inhibition of proteoglycan synthesis, and stimulation of bone resorption (14). Interestingly, in a study of 20 PsA patients treated with infliximab monotherapy, serum levels of TNFα measured during the initial (<12 weeks) phases of infliximab treatment did not correlate with the prompt resolution of the cutaneous and synovial symptoms observed. However, there were notable decreases in serum levels of IL-6, VEGF, E-selectin, and MMP-2 after the initial infusions of infliximab (42). Such observations of temporal cytokine behavior in response to anti-TNFα therapy provide further insight into which specific mediators may contribute more fully to specific disease characteristics and, consequently, provide the foundation for the development of biomarkers to allow a tailored individualized therapeutic regimen.
Distinctions from RA, including clinical correlates.
When examining the immunohistologic characteristics of involved structures (synovium, bone, skin, vascular beds) in PsA compared with those observed in RA, a number of important distinctions can be made. PsA synovial lining cells exhibit less hyperplasia with fewer monocyte/macrophages than those in RA, whereas PsA synovium is more vascular than RA synovium and has a characteristic tortuosity of the vascular beds (28). Similary, and likely related, VEGF, along with TGFβ, is more highly expressed in PsA synovium as compared with RA synovium (43). Although the levels of TNFα are not significantly different between PsA (joint and skin lesions) and RA, IL-1β levels are significantly higher in PsA (31). Compared with healthy controls, the serum levels of osteoclast precursors (OCPs) are higher in PsA patients, particularly those with radiographically evident erosive changes, and the numbers of OCPs are reduced following anti-TNF therapy (44).
A number of clinical characteristics differentiating PsA from RA have also been observed. Although erosive disease can be observed in both diseases, new bone formation and the enthesopathy commonly noted in PsA are rarely seen in RA. Asymmetry and a tendency toward oligoarticular involvement are seen in PsA compared with RA. Unlike RA, PsA is not associated with vasculitis or its clinical correlates. PsA affects men and women almost equally (13), compared with an ∼3:1 female:male ratio observed in seropositive RA. As noted, seropositivity for RF and anti-CCP is not characteristic of PsA, although recent investigations indicated that anti-CCP antibodies can be found in a small proportion of PsA patients (5–8%) and, when present, are associated with polyarticular involvement and a propensity for erosive arthritis (45-47).
Overview of therapeutic options
General considerations.
With the advent of targeted BRMs and their successful application in a number of chronic inflammatory disorders, a surge of interest in the pathophysiologic mechanisms of autoimmune diseases has developed. Elucidating the role of key cytokines, including TNFα, IL-1, and IL-6, involved in RA, the most common of the inflammatory rheumatic disorders, has led to the development of a number of BRMs approved or in late-phase clinical trials for the treatment of RA as well as PsA. Beyond inhibition of a single proinflammatory cytokine, other therapeutic targets in PsA are logically based on an improved understanding of disease pathogenesis. Such mechanistically oriented therapeutic principles include the following: inhibition of T cell activation via a number of costimulatory blockades; depletion of pathogenic T cells through targeting molecules expressed by activated T cells; inhibition of leukocyte recruitment; and induction of an immune system deviation shift from Th1 to Th2. These strategies have led to the development of a number of targeted therapies that are currently being investigated in both diseases (Figure 1).
Five general pathogenesis-oriented therapeutic principles targeting the key pathogenic mechanisms of psoriasis and psoriatic arthritis. The first involves inhibition of T cell activation through inhibition of molecules involved in the formation of the immunologic synapse (A). The second principle is depletion of pathogenic T cells through targeting molecules expressed by activated T cells, such as interleukin-2 (IL-2) receptor or CD4 (B). The third approach involves inhibition of leukocyte recruitment to the inflamed skin, such as inhibition of adhesion molecules (C). The fourth principle is inhibition of key inflammatory cytokines, tumor necrosis factor α (TNFα) being the most prominent example (D). The final approach is inducing an immune deviation to shift the cytokine milieu dominated by Th1 cells to a milieu weighted with Th2 cells, as has been demonstrated with IL-10 and IL-4 (E). TCR = T cell receptor; rIL-10 = recombinant IL-10. Adapted, with permission, from Schon MP, Boehncke WH. Psoriasis. N Engl J Med 2005;352:1899–912.
Although RA and PsA share a number of pathophysiologic mechanisms (which may explain the shared success of some of the BRMs in both disorders), the unique clinical and pathologic characteristics of PsA, as previously discussed, necessitate individual investigations of each BRM in the treatment of PsA, as opposed to simple empiric application of RA therapies in PsA patients. Key investigations into the pathophysiologic mechanisms of PsA and the comparative analyses between PsA and RA have shed light on the misconception that PsA is the mild-to-moderate cousin of RA with a rash. Recent studies have suggested that the clinical severity of PsA may be greater than initially observed. Peripheral joint involvement in PsA is often progressive, despite treatment with conventional disease-modifying antirheumatic drugs (DMARDs) (48, 49), with ∼20% of PsA patients developing a severe destructive and deforming arthritis (50). The functional disability and reduced quality of life observed in PsA are comparable with those documented in RA (51, 52).
RA therapies such as methotrexate (MTX) may have not only differing efficacy in PsA, but also different toxicity patterns (53-56). The challenge with treatment of PsA comes with the diversity of the disease phenotypes as well as the degrees of severity of individual disease manifestations in both the skin and joints. Certain patients may have extensive skin involvement with little joint involvement, whereas the converse may be the case; other patients with seemingly benign mono- or oligoarticular joint involvement may require aggressive management to avoid long-term disability. Matching the disease presentation, its severity, and its likelihood of progressive damage with the appropriate therapeutic choices is essential. The cost-to-benefit ratio, as well as the associated toxicities with each therapy, must be considered.
In assessing the efficacy and outcomes of therapeutic interventions in PsA, it is important to note that the majority of the measures currently used were developed and validated in patients with either RA or psoriasis; these have, for the most part, not been validated in PsA. Such measures include the following: the American College of Rheumatology (ACR) response criteria (ACR20, ACR50, and ACR70), the Psoriatic Arthritis Response Criteria (PsARC) (a composite index that calls for improvement of at least 2 of the following elements: tender or swollen joint count by ≥30%, physician's or patient's assessment of global improvement by at least 1 point on a 5-point Likert scale, one of which must be joint assessment, and no worsening of any element), and a composite psoriasis score, the Psoriasis Area and Severity Index (PASI) (a composite skin score of 4 sites [head, upper extremities, trunk, and lower extremities] that assesses the extent of erythema, induration, and desquamation along with the extent of lesions per site) (57). These measures have performed reliably in clinical trials and are considered to possess adequate discriminant capability when comparing active treatment with placebo (58, 59). Efforts are under way to develop PsA-specific instruments, including assessment tools for enthesitis, axial disease, and dactylitis, to more accurately evaluate the functional and disease outcomes in PsA.
An evidence-based approach to customized management.
The efforts headed by the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) to provide a comprehensive and phenotype-specific (peripheral arthritis, spondylitis, skin and nail disease, enthesopathy, dactylitis) approach to PsA management, based on the best available evidence, have recently been published (60-68). While a complete discussion of the findings of the GRAPPA is beyond the scope of this review, their work sought to provide evidence-based recommendations obtained from the published literature, with grading according to the strength of the evidence along with expert consensus opinion for those areas in which scientific evidence was lacking. A summary of their treatment recommendations and ratings of the quality of evidence for these guidelines are outlined in Figure 2. There are a number of guidelines that still rely on expert consensus because of the lack of published information and/or accepted outcome measures specific to PsA (particularly dactylitis and PsA axial disease involvement). However, these recommendations provide the groundwork for formalizing a customized management approach to PsA. In addition, they provide a challenge for future PsA intervention trials to include specific and validated outcomes for enthesitis, dactylitis, and axial disease in PsA, which, at the current time, are lacking.
Group for Research and Assessment of Psoriasis and Psoriatic Arthritis treatment guidelines for psoriatic arthritis (PsA), including quality indices for each recommendation. Individual clinical manifestations in PsA warrant tailored guidelines for specific manifestations. Recommendation quality indices (levels, grades) are denoted in brackets next to each therapeutic agent, and correspond to the guidelines issued by the Agency for Health Care Policy Research as outlined. NSAID = nonsteroidal antiinflammatory drug; IA = intraarticular; DMARDs = disease-modifying antirheumatic drugs; MTX = methotrexate; CsA = cyclosporin A; SSZ = sulfasalazine; Lef = leflunomide; anti-TNF = anti–tumor necrosis factor; PUVA = psoralen plus ultraviolet light A; UVB = ultraviolet light B; PT = physical therapy. Adapted, with permission, from ref.60.
Conventional drugs.
Nonsteroidal antiinflammatory drugs (NSAIDs), both selective and nonselective, are often used as first-line therapy for the arthritic component of PsA, typically in patients with mild-to-moderate symptoms and without evidence of progressive joint damage. NSAIDs have the ability to reduce inflammation and improve pain and joint mobility, although gastrointestinal side effects have been a limiting factor in some PsA patients (69). Oral and intralesional corticosteroids have been associated with rebound worsening of psoriasis upon withdrawal of the drug (70), and are therefore to be used with caution.
DMARDs, including MTX, sulfasalazine, leflunomide, and cyclosporine, are used in PsA patients whose disease remains unresponsive to NSAID therapy or who exhibit progressive disease. MTX is often used as the primary DMARD in PsA because of its proven efficacy in treating both the skin and joint involvement in the disease (71, 72); however, the toxic effects of MTX, including an apparent yet unexplained increased proclivity for hepatotoxicity in psoriasis patients versus RA patients (73), have elicited concerns regarding its long-term use. A discord still exists among the dermatology and rheumatology communities regarding published guidelines for MTX monitoring (53); the American College of Dermatology recommends pretreatment liver biopsy and repeat biopsies after a 1.5 gram accumulated dose has been achieved in psoriasis patients (74), whereas the ACR guidelines (developed for RA) recommend pretreatment liver biopsies only for patients with persistently elevated liver enzymes, a history of chronic hepatitis B or C infection, or significant alcohol intake, with repeat biopsies performed as indicated by the results of liver enzyme tests (75).
Sulfasalazine has been proven effective for treating peripheral arthritis in PsA but has no proven effect on axial disease or skin lesions (76). Leflunomide was demonstrated to have statistically significant efficacy for both skin and joint involvement, compared with placebo, in a recent study of 190 PsA patients (59% of leflunomide-treated patients versus 30% of placebo-treated patients achieved a PsARC response) (77). Treatment with leflunomide was generally well tolerated, with the most common side effects being diarrhea and increased transaminase levels (77). Cyclosporine, while effective for both skin and joint manifestations, is less commonly used because of its toxic effects, most notably hypertension and nephrotoxicity (78, 79).
Other DMARDs, including hydroxychloroquine, have been used in PsA, although use of hydroxychloroquine has been reported to exacerbate skin disease (80, 81). In a case–control study by Gladman and colleagues in which 32 PsA patients were treated with chloroquine, 75% of the chloroquine-treated patients had a >30% reduction in active joint inflammation over a 6-month period, compared with 58% of control subjects. Six of the control subjects experienced a flare of their skin disease, and 6 of the chloroquine-treated patients had a psoriasis flare but no exfoliative dermatitis (82). Gold salts have also been reported to improve arthritis in PsA patients, but their use is limited due to toxicity. A discrepancy between formulations and their respective efficacies in PsA has been observed; for example, in a study of 82 patients with PsA, only intramuscular gold compounds, and not the oral formulations, resulted in a statistically significant improvement in outcomes compared with the placebo group (83). Of note, none of these conventional DMARDs are approved by the US Food and Drug Administration (FDA) for the treatment of PsA.
Targeted strategies for manipulating the mechanisms of disease
Biologic response modifiers.
Currently available TNFα antagonists.
As a central mediator in a number of inflammatory arthropathies, including RA and PsA, TNFα emerged as an attractive target for biologic therapies in the 1990s. The 3 currently available TNFα inhibitors, etanercept (a soluble TNF receptor [p75]-IgG1 fusion protein), infliximab (a chimeric monoclonal antibody specific for soluble and membrane-bound TNFα), and adalimumab (a fully human anti-TNF monoclonal antibody), have since gained US FDA approval for the treatment of RA and PsA, along with a number of other indications. Etanercept (84, 85) and infliximab (86-88) were the first agents to show impressive results in reducing the signs and symptoms of established RA, inhibiting the progression of structural damage (86, 89) and substantially improving the functional status and quality of life of patients (90, 91). Following the proven successes in RA, these agents were then studied in PsA, with demonstrated dramatic improvements in all facets of the disease, including the skin, joints, and entheses (92-95). Table 3 provides a summary of the key clinical investigations of anti-TNF therapy in PsA.
| Study reference | Agent | No. of patients | Study type/duration | Outcome measure | % of patients meeting response criteria in active treatment/placebo group | P, active vs. placebo | Comments |
|---|---|---|---|---|---|---|---|
| 92 | Etanercept 25 mg twice weekly | 60 | RDBPC/12 weeks | PsARC ACR20 | 87/23 73/13 | All <0.0001 | HAQ significantly improved with etanercept; no significant difference (in skin and joints) in MTX vs. non-MTX users |
| 96 | Etanercept 25 mg twice weekly | 205 | RDBPC/12 weeks | ACR20 PsARC PASI50 PASI75 | 59/15 72/31 47/18 23/3 | <0.001 <0.0001 ≤0.001 ≤0.001 | Inhibition of radiographic progression (TSS) at 48 weeks in etanercept group versus worsening in placebo group; no difference between groups in PsA-specific radiographic changes (osteolysis, periostitis) |
| 93 | Infliximab 5 mg/kg on weeks 0, 2, 6, then every 8 weeks | 104 | RDBPC/16 weeks; BSC/through 50 weeks† | ACR20 (week 16) ACR50 (week 16) ACR70 (week 16) PsARC (week 16) ACR20 (week 50) ACR50 (week 50) ACR70 (week 50) PsARC (week 50) | 65/10 46/0 29/0 75/21 68/69 42/53 34/39 76/74 | <0.0001 <0.001 <0.001 <0.001 <0.0001 <0.001 <0.001 <0.001 | All patients crossed over to active drug at week 17; no progression of joint damage observed in either the original infliximab-treated or the placebo-treated patients at week 50; enthesitis and dactylitis significantly improved in infliximab group |
| 94 | Infliximab 5 mg/kg on weeks 0, 2, 6, then every 8 weeks | 200 | RDBPC/14 weeks | ACR20 ACR50 ACR70 PsARC PASI50 PASI75 PASI90 | 58/11 36/3 15/1 77/27 82/9 64/2 41/0 | All <0.001 | Similar results noted up to week 24; significant PASI response observed in both ACR20 responders and ACR20 nonresponders |
| 95 | Adalimumab 40 mg every other week | 313 | RDBPC/24 weeks | ACR20 ACR50 ACR70 PsARC PASI50 PASI75 PASI90 PASI100 | 57/15 39/6 23/1 60/23 75/12 59/1 42/0 29/0 | All <0.001 | Inhibition of radiographic progression (mTSS) in adalimumab group at week 24 and maintained at week 48; significant improvements in quality of life and disability measures in adalimumab group |
| 103 | Onercept 50 mg and 100 mg subcutaneously 3 times/week | 126 | DBPC/12 weeks | PsARC ACR20 | 86/45 67/31 | <0.001 0.001 | Results listed are those observed in the 100 mg arm; placebo responses were higher than in etanercept PsA trials |
- * The initial outcome measure listed for each study was the primary outcome assessed. RDBPC = randomized, double-blind, placebo-controlled; PsARC = Psoriatic Arthritis Response Criteria; ACR20 = American College of Rheumatology 20% response criteria; HAQ = Health Assessment Questionnaire; MTX = methotrexate; PASI50 = Psoriasis 50% Area and Severity Index; TSS = total Sharp score; BSC = blinded single-crossover; mTSS = modified TSS.
- † Values through week 50 are patients who were crossed over from placebo to infliximab/patients remaining on infliximab.
Etanercept.
Etanercept gained US FDA approval for the treatment of PsA in 2002, following the completion of a published phase II trial (92) and publication of data from a phase III trial (96). In the phase II trial, 47% of patients were on stable doses of MTX and had longstanding disease (a mean 20-year history of psoriasis and 15-year history of inflammatory arthritis). At the end of the 3-month placebo-controlled phase, the etanercept group showed significant improvement in all outcome measures compared with the placebo group, with the primary end point (the PsARC response) achieved by 87% of the patients receiving etanercept versus 23% of the placebo group. Of note, there was no significant difference in the response (joints or skin) in patients receiving background MTX versus those who were not taking background MTX, a finding that has also been noted in the majority of PsA trials with other anti-TNF agents.
In the large phase III multicenter trial that followed, significant improvements in skin lesions, quality of life, and function were noted in the etanercept group (Table 3). In addition, inhibition of disease progression, as measured radiographically, was demonstrated. In the 48-week open-label radiographic followup period, there was no worsening of the total Sharp score for radiographic joint damage in the active treatment group; moreover, the original placebo group, once switched over to etanercept, showed no further disease progression. As before, background use of MTX (42% of patients in this trial) did not affect the outcomes. Interestingly, there appeared to be a lag in skin response when compared with that in the joints; the full extent of joint improvement occurred sooner than the full skin response, and improvement in the skin was still occurring well into the open-label aspect of the phase II trial (92, 96, 97).
Infliximab.
Infliximab, subsequent to its successes in improving skin and joint disease in open-label studies of PsA (98, 99), was studied in 2 major placebo-controlled trials using a similar study design. In these trials, the initial Infliximab Multinational Psoriatic Arthritis Controlled Trial (IMPACT) (93) and the followup trial IMPACT II (94), background MTX was allowed, but not required, so that ∼50% of patients who were enrolled in both studies received concomitant MTX. The IMPACT study allowed continuation of DMARDs other than MTX (total of 64% receiving a background DMARD, 46% of whom were taking MTX), while IMPACT II limited DMARD use to MTX.
In the IMPACT study, enthesitis and dactylitis measures were formally assessed for the first time; these improved significantly among infliximab users. No difference in radiographic outcomes was noted between the treatment groups over 1 year, although the short duration of the placebo treatment (14 weeks) was considered to be a likely contributor to this lack of difference. Of note, the calculated annual radiographic progression rate was reduced in both treatment arms, suggesting that infliximab treatment, even after a 14-week delay in initiating treatment, inhibits radiographic progression in PsA (100).
In IMPACT II, an analysis of skin response was carried out in ACR20 responders as compared with ACR20 nonresponders, with results showing that the median improvement in the PASI was 87% in ACR20 responders versus 74% in nonresponders, supporting the notion that skin response can be achieved with infliximab in the absence of significant joint response. Significantly less radiographic progression was noted at 24 weeks in the infliximab-treated group compared with the placebo group. As with the other anti-TNFα agents, there was no difference between the treatment groups when PsA-specific radiographic features, including gross osteolysis and pencil-in-cup deformities, were examined, and this was considered to possibly be due to the chronicity and fixed nature of such lesions (101). Once again, treatment with background MTX did not have an impact on efficacy in either of the IMPACT studies.
Adalimumab.
Adalimumab, the most recently approved anti-TNF agent for PsA, was initially observed as an effective treatment for PsA in an open trial (n = 15) (102). In the Adalimumab Effectiveness in PsA Trial (ADEPT) of 313 patients with PsA, 57% of the patients treated with adalimumab versus 15% of the placebo group achieved the primary end point (the ACR20 response). Significant improvements in skin disease along with significant inhibition of radiographic progression of disease and improvement in quality of life and functional indices were also observed in the adalimumab-treated patients during the initial 24-week ADEPT study, with continued response through 48 weeks (95) (Table 3). The safety profile of adalimumab in patients with PsA appears to be consistent with the safety profile observed in clinical trials of adalimumab in RA.
Other TNFα antagonists.
Onercept is a recombinant, fully human type I TNF receptor (p55). A phase II placebo-controlled trial of 126 patients with PsA was recently completed (103), involving treatment with 2 subcutaneous doses (50 mg and 100 mg) of onercept administered 3 times per week; the 100 mg dose of onercept was more effective than the 50 mg dose, with both doses generally well tolerated. Side effects were similar between the patients receiving onercept and the placebo group. Of note, injection site reactions were observed more frequently in the onercept group. In a phase III psoriasis trial of onercept, 2 patients developed sepsis, one of whom subsequently died (104). Two phase III clinical trials of onercept have been discontinued (105). A number of other anti-TNFα agents are under investigation in PsA, including a human monoclonal antibody for subcutaneous administration (golimumab [CNTO 148]), a PEGylated Fab fragment of a monoclonal antibody (certolizumab pegol [CDP870]), and oral compounds possessing TNFα-inhibitory action (106).
TNFα antagonists in PsA.
In general, the reported PsA trials revealed no new significant adverse events or safety issues with anti-TNF therapy other than those reported in RA trials. Regarding skin-specific issues, no increases in the frequency of Koebner phenomenon or cellulitis were observed in PsA patients.
Of note, several case series have been published describing what appears to be a paradoxical adverse reaction following treatment with the 3 anti-TNF therapies, in that all 3 therapies induced psoriasiform skin lesions (typically palmoplantar pustulosis) in previously unaffected individuals (the majority of whom had RA) without a personal or family history of psoriasis (107-112). While the underlying pathophysiologic mechanism for these observations is still unknown, hypotheses include infectious (bacterial) and autoimmune (activation of autoreactive T cells or altered immunity induced by anti-TNF activity in predisposed individuals) etiologies. Because arthritis in some of these patients continued to improve despite the cutaneous eruptions, these observations again highlight the need for more complete definition of the mechanisms of disease in the skin and joint.
Although there have been no trials directly comparing TNF antagonists in PsA, switching from one anti-TNF agent to another due to inefficacy, loss of efficacy, or side effects may be a reasonable option, as described in a small number of patients with PsA and those with SpA (113). Similar experiences with switching anti-TNF therapies in the management of RA have been reported (114, 115).
Modulators of T lymphocyte function.
Psoriasis has been described as a T lymphocyte–driven disease (116); therefore, utilizing BRMs that interfere with T cell activation and/or trafficking of T cells is a logical approach for targeted therapies in psoriasis and PsA (Figures 1 and 3). T cell activation is a 2-step process. The first signal involves the recognition of an MHC–antigen complex presented on an antigen-presenting cell (APC) by the complementary T cell receptor. The second signal consists of a variety of receptor/counterreceptor couplings, which can produce stimulatory responses (costimulatory pathways). Examples of such couplings include leukocyte function–associated antigen 3 (LFA-3) and CD2, LFA-1 and ICAM-1, and CD80/CD86 and CD28/cytotoxic T lymphocyte–associated 4 (CTLA-4). Many of these pathways have been targeted by specific therapeutic agents that are currently being investigated in clinical trials for use in psoriasis and PsA (Table 4).
Costimulatory blockade strategies. T cell activation requires 2 signals. The first signal (signal 1) consists of recognition of a major histocompatibility complex (MHC)–antigen complex presented on an antigen-presenting cell (APC) by the complementary T cell receptor (TCR). A number of costimulatory pathways serve as second signals (signal 2) through which full T cell activation occurs. Interfering with these second signal costimulatory pathways serves as the mechanism of action for a number of biologic agents used in psoriasis and psoriatic arthritis. Alefacept is a dimeric human fusion protein that binds to CD2 on T cells and blocks its interaction with leukocyte function–associated antigen 3 (LFA-3) on APCs, in addition to binding to the FcγIII region of natural killer lymphocytes, resulting in apoptosis of T cells expressing CD2. Efalizumab is a recombinant humanized monoclonal IgG1 antibody to the CD11a subunit of the LFA-1 on T cells that prevents adhesion of LFA-1 to intercellular adhesion molecule 1 (ICAM-1) on APCs. Abatacept is a recombinant CTLA-4Ig fusion protein that selectively modulates costimulation via interruption of the CD28:CD80/86 pathway.
| Agent | Target | Status in psoriasis | Status in psoriatic arthritis |
|---|---|---|---|
| Infliximab | TNFα | FDA-approved | FDA-approved |
| Etanercept | TNFα | FDA-approved | FDA-approved |
| Adalimumab | TNFα | Phase II and III trials completed | FDA-approved |
| Alefacept | CD2 | FDA-approved | Phase II trial completed |
| Efalizumab | CD11a (LFA-1) | FDA-approved | Phase II trial completed |
| Abatacept | CD80/CD86 | Phase II trial completed | No trials completed |
- * TNFα = tumor necrosis factor α; FDA = US Food and Drug Administration; LFA-1 = leukocyte function–associated antigen 1.
Alefacept.
Alefacept, approved in the US in 2003 for the treatment of psoriasis and currently being investigated in phase II clinical trials for its use in PsA (117), is a fully human LFA-3/IgG1 fusion protein that binds to CD2 receptor on T cells to block the interaction between LFA-3 on APCs and CD2 on T cells. This blockade of the LFA-3/CD2 costimulatory pathway results in the inhibition of T cell activation and depletion of T cells via interactions through the Fc portion of the molecule (118, 119). Alefacept is administered as a weekly intramuscular injection, alternating 12 weeks on and 12 weeks off to allow for recovery of CD4 cells, the levels of which are monitored during treatment (120, 121).
Following a small, successful open-label trial of alefacept in patients with PsA (122), a randomized, placebo-controlled phase II study involving 185 patients with PsA who were taking MTX was performed, with results showing significant improvements in both skin and joint outcomes in patients receiving alefacept 15 mg intramuscularly in combination with MTX versus those receiving placebo plus MTX (117). At week 24, 54% of the patients in the alefacept plus MTX group achieved an ACR20 response, compared with 23% of patients in the placebo plus MTX group (P < 0.001). Improvements in the ACR20 response continued after the active treatment phase. No serious infections were reported and no malignancies were diagnosed in the alefacept plus MTX arm. The predictable reductions in CD4+ T cell counts in patients treated with the alefacept plus MTX combination were consistent with those reported in the phase III clinical trials of alefacept monotherapy in patients with psoriasis, with similarities in the overall safety profile observed between the psoriasis and PsA trials (117, 120, 121).
Efalizumab.
Efalizumab, approved in the US as a weekly subcutaneous injection for treatment of moderate-to-severe plaque psoriasis (123), is a humanized monoclonal IgG antibody that binds to the CD11a subunit of LFA-1 (an adhesion molecule expressed on T cells), thereby inhibiting LFA-1 binding to ICAM-1 on the APC and resulting in inhibition of T cell activation. It also interferes with migration of cells from the circulation to sites of inflammation (106). A phase II study in 117 PsA patients failed to show significant improvements in the ACR20 response at 12 weeks; 28% of the patients receiving the active drug achieved an ACR20 response versus 19% in the placebo arm (P = 0.2717) (124). Given the short duration of this trial, it is unknown if more robust improvements with efalizumab would have been evident with a longer trial (125).
Abatacept.
Abatacept, recently approved in the US for treatment of RA, is a recombinant CTLA-4Ig fusion protein that selectively modulates costimulation via interruption of the CD28:CD80/86 pathway (126). It is administered as a monthly intravenous infusion. The efficacy of abatacept has been demonstrated in a phase II psoriasis trial (127), although it has yet to be studied formally in PsA.
Other potential targets.
Do we need more targeted therapies?
Although the introduction of anti-TNFα therapies has provided a much-needed option in the armamentarium of PsA therapies, approximately one-third of patients with moderate-to-severe PsA have insufficient responses to these therapies (14). The successes and failures of the anti-TNFα agents and costimulatory blockers in the management of psoriasis and PsA have sparked considerable interest in exploring other avenues for targeting the diverse immune responses considered to be at play.
More cytokines, both pro- and antiinflammatory, to consider.
Examples of other potential therapeutic targets in PsA include a variety of antagonistic approaches (neutralizing monoclonal antibody, soluble forms of receptor antagonists, and cytokine:Fc mutant fusion proteins) to interfere with the actions of so-called proinflammatory cytokines, including IL-1, IL-6, IL-8, IL-12, IL-15, and IL-18. Anakinra, the IL-1 receptor antagonist approved for RA, has been investigated for use in PsA, although significant efficacy has not been observed (128). Plans are under way to test IL-1 Trap, a weekly subcutaneous IL-1 antagonist, in PsA (97). A monoclonal antibody to the IL-6 receptor (tocilizumab) is currently being investigated in a number of phase III clinical trials in RA and would be a logical therapy to test in PsA.
As a key cytokine in Th1 immune responses, IL-12 is also a potential target in psoriasis and PsA; a number of anti–IL-12 therapies are under development for the treatment of psoriasis. IL-15, a pleiotropic cytokine with proinflammatory effects in a number of chronic inflammatory diseases, has been demonstrated to have increased levels in the skin and synovium of psoriasis patients (129, 130), and a number of agents designed to suppress IL-15 are in development (anti–IL-15 [AMG714, HuMax-IL15], IL-15:Fc mutant [CRB-15], soluble IL-15 receptor α) (131). Arthroscopic evaluations of synovial membrane from PsA patients, obtained prior to and following a 3-month course of MTX, demonstrated that expression of IL-15 messenger RNA and protein was modified but not abrogated by MTX, providing the groundwork to consider a combination of IL-15 antagonist with MTX as a plausible therapeutic strategy for PsA (131). A pilot trial of anti–IL-15 has shown efficacy in PsA (132).
On the opposite end of the spectrum, cytokines that possess antiinflammatory effects, such as IL-10 and IL-11, could be attractive agents for the treatment of PsA. IL-10 promotes Th2-biased cytokine secretion and inhibits IFNγ (133). It appears that recombinant IL-10 is successful in attenuating skin disease but not joint inflammation, as indicated in a controlled study of PsA patients (134). Recombinant human IL-11, which has demonstrated antiinflammatory activity, has been tested in a small number of patients with psoriasis, resulting in some improvement in skin scores (135).
Additional strategies of T cell manipulation.
HuOKT3γ1 (ala-ala), a non–FcR-binding monoclonal antibody to the T cell receptor complex component CD3, anergizes type 1 lymphocytes and serves as a targeted therapeutic option for type 1 lymphocyte–mediated diseases such as PsA. In a pilot study of 7 patients with active PsA whose condition failed to respond to at least one remittive agent, improvements in skin and joint outcomes were reported (136). A humanized anti-CD80 monoclonal antibody (IDEC-114, galiximab) has been studied in a phase I/II trial of 35 patients with plaque psoriasis; improvements in skin scores and good tolerability were observed (137).
Targeting the vasculature and mediators of bone dysregulation.
The findings of elevated angiogenic factors in PsA suggest that reducing synovial vascularity may be an alternative mechanistic target in this disease. Anti-TNF therapy has been shown to produce significant reductions in the circulating concentrations of VEGF in RA (138, 139) and PsA (140). Blockade of neovascularization utilizing a number of angiogenesis inhibitor strategies (VEGF antibody, small-molecule inhibitors of tyrosine kinases) has been effective in patients with solid tumors (141, 142), and such strategies continue to be investigated. Angiogenesis inhibitor strategies could be considered for the treatment of PsA to target the highly vascular PsA synovium, likely in combination with other therapies such as anti-TNF, particularly in those patients with a suboptimal response to anti-TNF; i.e., target the inflammatory and vascular components, possibly without increasing infection risk.
Pioglitazone, a member of the thiazolidinedione family of insulin-sensitizing drugs developed for the treatment of type II diabetes mellitus, is an agonist for the peroxisome proliferator–activated receptor γ (PPARγ). Beyond its role in adipocyte differentiation, PPARγ has been considered a potential target for the treatment of inflammatory rheumatic conditions, given observations that it leads to marked inhibition of angiogenesis and down-regulation of proinflammatory cytokines (53). In an uncontrolled study of 10 patients with PsA treated with PPARγ for 12 weeks, 60% of patients were considered PsARC responders (the primary end point) and 50% reached an ACR20 response. Three patients were withdrawn from the study due to treatment inefficacy and adverse events, with edema in the lower extremities and weight gain as the most common adverse events (143).
Strategies targeting abnormalities in bone remodeling characteristic of PsA could also be considered. As described by Ritchlin and colleagues, a bidirectional model for the extensive erosive changes observed in patients with PsA may exist, whereby circulating OCPs migrate to the inflamed synovium and are induced to become osteoclasts by the NF-κB ligand (RANKL) that is expressed by synoviocytes; in addition, OCPs also traverse the subchondral bone endothelium, where they are exposed to TNFα-induced RANKL on osteoblasts and stromal cells. The result is osteoclastogenesis at the synovium and in subchondral bone (44). An inhibitor of RANKL, denosumab, which is currently being investigated in postmenopausal osteoporosis and RA, could be considered a viable target in PsA among those patients with erosive disease that is not completely controlled with their current DMARD.
MMPs, regulated by IL-1 and TNFα, and tissue inhibitors of MMPs (TIMPs) may also be attractive targets for PsA therapy. MMP-3 (an activator of other MMPs [144]), along with MMP-1, has been demonstrated in synovial tissue and skin samples from PsA patients (145). In addition, patients with polyarticular PsA were observed to have increased serum levels of MMP-1 and TIMP-1 compared with controls (146). Once again, since these strategies utilize targets “down-stream” from currently available targets (TNFα, IL-1, IL-6), they could serve as vital adjuncts to the PsA regimen, with the potential to have a favorable toxicity profile in combination with DMARD and/or biologic therapies.
As novel agents are discovered and entered into the quest to join the ranks of approved PsA therapies, they will have a large burden of proof to achieve recognition as a truly successful treatment for PsA. In developing future therapeutic strategies for PsA, one must consider the need for efficacy in both skin and joint responses, an acceptable safety and tolerability profile, the ability to inhibit disease progression as is currently measured radiographically, and the capacity to improve quality of life and functional status, all within the context of achieving favorable cost-effective results.
Conclusions
With its uniquely diverse pathophysiologic and clinical features and the ability to progress into one of the most destructive arthritides known, PsA remains a challenging disease deserving of the attention it has been receiving in recent years. A surge of critically important observations regarding the immunopathogenetic mechanisms of PsA, including the finding that the disease is paradoxically both similar to, and distinct from, other inflammatory arthropathies, has heightened the awareness that novel biologic agents used in RA cannot simply be empirically applied to PsA. Rather, these agents and others soon to come have the burden of showing efficacy in a number of immunologic, clinical, and functional outcomes specific to PsA. Measures of these outcomes are sorely needed to standardize the expected response in this disease and to determine the natural history of PsA.
Although there are clear indications that the same pathogenetic processes occurring in the skin are likely occurring in the joints and vice versa, the clinical experience with therapies that preferentially improve the condition in either the skin or the joints indicates that more factors are at play. As in RA, but perhaps even more so in PsA, there exists a clear unmet need for targeted therapies for affected patients, who are inflicted with 2 incurable diseases for which true disease-modifying agents are needed. Although the early data seem to support the potential of biologic agents to achieve significant benefits, particularly with anti-TNF therapies, we are still in the infancy of the biologic era in PsA. Further studies are needed to confirm the long-term safety and efficacy of currently available therapeutic strategies and those on the horizon.
