Psoriatic arthritis (PsA) is a heterogenous inflammatory disease that may involve several different domains, including peripheral joints, entheses, nails, axial skeleton, and skin. A recent increased awareness of PsA has accompanied a large increase in available therapeutic options. In addition to traditional disease-modifying antirheumatic drugs (DMARDs), new biologics and targeted small molecules have now been shown to be effective in PsA. These agents include those targeting pathways involving tumor necrosis factor (TNF), cytotoxic T-lymphocyte–associated protein 4 (CTLA-4), interleukins (IL) 12, 17, 23, janus kinase (JAK), and phosphodiesterase 4 (PDE4). These agents have demonstrated efficacy in outcome measures developed for peripheral arthritis, such as the American College of Rheumatology 20 (ACR20) response. However, an ongoing question is whether these agents are equally effective in axial disease. Based on our experience and the existing literature, we believe that some of these agents, including PDE4 and IL-23 inhibitors, are not effective for axial disease.
Moll and Wright’s original description of PsA estimated that 5% of patients with PsA had axial disease1; however, they were describing patients in whom axial arthritis was the predominant, or the only, manifestation. There are many patients for whom axial symptoms are just one of several domains of disease activity. With this in mind, and depending on the cohort studied, the estimated overall prevalence of axial disease ranges from 7% to 32% in patients with PsA.2 This is in contrast to peripheral arthritis, a domain that occurs in most patients with PsA and is the most common manifestation of PsA.2 We believe there are differences in axial and peripheral response among some of the drugs used to treat PsA; therefore it is critical to consider both the presence and magnitude of axial involvement.
An absence of axial PsA–specific clinical trials complicates navigating this treatment domain. Most considerations regarding treatment options for axial disease in PsA are extrapolated from ankylosing spondylitis (AS) trials and experience, as is the case for the TNF and JAK inhibitors. To our knowledge, only one high-quality randomized trial, MAXIMISE, looked specifically at the treatment of axial PsA, in this case with the IL-17 inhibitor secukinumab.3 This trial demonstrated efficacy of secukinumab in reducing symptoms and acute phase reactants in patients with PsA who were categorized as having active axial disease using the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI). Other than conclusions drawn from AS trials and from this single axial PsA randomized controlled trial, data on the treatment of axial PsA are drawn entirely from observational and post-hoc analyses. As there are no consensus criteria for axial PsA, the cohorts included in these data may vary. This heterogeneity showcases the diversity in patients with PsA with axial disease but complicates the generalizability of the findings to individual patients.
Another challenge in understanding axial response to medication is the lack of specific, validated outcome measures for axial PsA. The BASDAI and, more recently, the Assessment in Ankylosing Spondylitis (ASAS) and the Ankylosing Spondylitis Disease Activity Score (ASDAS), all developed specifically for AS, are often used to measure treatment response. The BASDAI incorporates patient-reported symptoms which include fatigue, peripheral joint pain/swelling, tenderness, and morning stiffness not specifically localized to the back. The ASDAS also includes a C-reactive protein measurement.
When used to assess response in PsA, however, these patient-reported outcomes may not be precise enough to separate the impact of axial disease or symptoms from that of peripheral disease. Only question 2 on the BASDAI specifically addresses axial complaints: “How would you describe the overall level of AS-related pain you have had in your neck, back, or hips?” Even this question is vulnerable to confounding from noninflammatory causes of back pain. Although these issues exist with patient-reported outcomes, objective spinal mobility measures used in evaluation of AS, including the modified Schober test, lumbar side flexion, and cervical rotation, have been demonstrated also to perform well in axial PsA.4
This was corroborated in the INSPIRE study, which showed adequate interobserver reliability in primary AS that was equally reproducible in axial PsA, with most measures, including occiput to wall, modified Schober test, cervical rotation, lateral bending, and hip mobility, performing in a “good to excellent” manner.5 Therefore, the inclusion of these objective measures in future therapeutic studies may enhance the external validation of available data.
The Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) has established therapeutic guidelines for psoriatic disease based on currently available literature and data. Similar to previous iterations of guidelines, GRAPPA continues to recommend agents with TNF inhibition or IL-17 inhibition for patients with PsA with axial disease who have failed conservative therapy with nonsteroidal anti-inflammatory drugs (NSAIDs), physical therapy, and/or glucocorticoid injections. Newly recommended in the latest iteration of the GRAPPA guidelines, based on the efficacy of these agents in AS, is the use of JAK inhibitors for axial PsA.6
Although TNF, IL-17, and JAK targeted therapies have demonstrated more likely benefit, albeit subject to the trial limitations previously discussed, the question remains whether agents targeting PDE4 and IL-23 are an effective option for axial PsA. Studies of both PDE4 and IL-23 inhibitors in AS have not demonstrated adequate benefit, which, importantly, contrasts with the previously mentioned and recommended therapies. Additionally, there are no primary randomized control trials that have directly evaluated the efficacy of IL-23 therapy in axial PsA.
Existing data about potential benefit come from post-hoc analyses of the PSUMMIT 1 and 2 trials7-10 with ustekinumab (which inhibits IL-12 and IL-23) and the DISCOVER trials11-13 with guselkumab (a pure IL-23 inhibitor). However, these analyses relied on a physician-reported diagnosis of axial disease and not on prespecified entry criteria. This lack of uniform diagnostic criteria may introduce bias into the interpretation of the results and limit external validation. All patients in these trials had a significant burden of peripheral arthritis; therefore it is hard to know whether, even in patients with physician-reported axial disease, improvement in general outcome measures were due to true amelioration of axial disease or were confounded by improvement in peripheral and skin domains. The analysis of these trials did look specifically at patient answers to BASDAI question 2 regarding level of neck, back, or hip pain. However, it remains difficult to be certain that the results are truly a reflection of axial symptoms and are not driven by patient-perceived improvement in other disease domains and an overall positive trajectory in well-being.
In our years of practice, when we turned to biologic agents, the IL-23 inhibitors and the IL-12/23 inhibitor have not been as effective in patients with PsA who have axial-predominant symptoms. The lack of efficacy of these agents in AS, in contrast to their benefit in psoriatic skin and peripheral joint disease, raises questions about the pathophysiologic role of IL-23 in axial disease, which is yet to be fully understood. For patients with a significant burden of axial pain, in concordance with the consensus from GRAPPA,6 our strategy is to start with TNF, IL-17, or JAK targeted therapies, with the choice based on patient-specific factors, including patient comorbidities, patient administration preference, and insurance coverage. We do believe it is reasonable to try IL-23–targeted therapies in patients who have mild axial symptoms when their predominant symptoms are in other domains, such as the peripheral joints or skin. In our opinion, more convincing data supporting IL-23 inhibition are required to move this into the forefront of axial-predominant PsA therapy. Clearly the investigation of axial disease in PsA lags behind that of peripheral and skin domains. Specific classification criteria for axial PsA, as are being currently developed by GRAPPA, should facilitate more focused therapeutic trials that can better inform optimal treatment of patients with this subset of disease.