Abstract
Dactylitis, a hallmark clinical feature of psoriatic arthritis (PsA) and other spondyloarthropathies, may also be a severity marker for PsA and psoriasis. Traditionally, clinicians have used nonsteroidal antiinflammatory drugs and local corticosteroid injections to treat dactylitis, although conventional disease-modifying antirheumatic drugs are also used. We performed a systematic literature review to determine the most efficacious current treatment options for dactylitis in PsA. Effect sizes were greatest for the biologic agents ustekinumab, certolizumab, and infliximab, suggesting that therapy with one of these agents should be initiated in patients with dactylitis. However, the limited data highlight the need for randomized, placebo-controlled trials, with dactylitis as a primary outcome, to determine a valid, reliable, and responsive clinical outcome measure for PsA patients with dactylitis.
Dactylitis, or “sausage digit,” is considered a hallmark clinical feature of psoriatic arthritis (PsA) and other spondyloarthropathies1. The diagnosis of dactylitis, however, is challenging for clinicians not familiar with it, and is frequently misdiagnosed in cases of mild dactylitis, in obese patients, and in those with severe overlying psoriatic skin disease. Other causes of dactylitis include trauma, fracture, gout, sepsis, sarcoidosis, and tuberculosis. In cases of diagnostic uncertainty, magnetic resonance imaging (MRI) and ultrasound may help to discriminate dactylitic digits from normal ones2.
The treatment of dactylitis has largely remained empirical. Nonsteroidal antiinflammatory drugs (NSAID) are usually employed initially, but many rheumatologists rapidly progress to injected corticosteroids, which are supported by clinical evidence of response in patients with mild to moderate PsA. In resistant cases, disease-modifying antirheumatic drugs (DMARD) are used with or without biologic agents. Comparative effectiveness of these treatment strategies has not been systematically studied.
The focus of our review is to identify and evaluate the effects of therapeutic interventions used to treat dactylitis in patients with PsA. The review is part of a treatment update initiated by the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA).
Search strategy
Ovid Medline was searched from 1966 to the present, using the search terms “dactylitis” (as key search word) and “psoriatic arthritis” (treatment only). Only randomized, double-blind placebo-controlled (RCT) or open-label trials of PsA in which dactylitis was assessed as a separate outcome measure were identified. Thus, 74 articles (English-language reports only) were identified, and 29 articles were selected for full review3-13,14-24,25,26,27,28,29,30,31.
Two reviewers (SR and ST) independently extracted the data regarding study design, sample size, duration, population, agents, outcome measures, outcome data, p value, and effect size using a standardized data extraction form. A third reviewer (PH) resolved differences if needed. A fourth reviewer (WBM) examined safety data pertaining to the use of DMARD and biologic agents.
RESULTS
Of the 29 trials that assessed dactylitis as an outcome measure, a total of 6589 adult patients with PsA were studied for 12 weeks to 60 months (Table 1)3-13,14-24,25,26,27,28,29,30,31. Several studies (n = 18) were RCT with crossover design at 12–24 weeks, with or without open-label extensions. One study was prospective from a historical cohort (10 years), another was a prospective followup case-control study, and the remaining 9 were open-label studies.
Summary of clinical therapeutic studies with dactylitis as a primary or secondary outcome.
Of the 29 studies, 22 were multicenter with participation from North America, Europe, the Middle East, South Africa, Latin America, and Asia-Pacific countries. The remaining studies were single-center: 1 each from Canada, Italy, and Germany; and 2 each from Greece and the United Kingdom.
Therapeutic interventions were heterogeneous and included DMARD (methotrexate, hydroxychloroquine, leflunomide, cyclosporin A, and sulfasalazine), biologics (certolizumab, ustekinumab, golimumab, adalimumab, etanercept, infliximab, and anakinra), and the oral phosphodiesterase 4 inhibitor apremilast.
Dactylitis outcome measures were also heterogeneous and included the number of dactylitic digits (maximum 20 digits; either tender and/or nontender on 0–3 scale), percentage of patients with dactylitis, Leeds Dactylitis Index and its simplified version (LDI, LDI basic), and MRI dactylitis scores. Some studies used a simple count of dactylitic digits (based on clinician opinion), while others graded the severity 0–3 or 1–4, and all 20 digits were counted. No studies of local steroid injections or NSAID were identified.
Because of the large variability in study designs and outcome measures, and poor availability of primary data, a metaanalysis could not be performed. Significant improvement (p < 0.05) in dactylitis compared to placebo was observed with the use of certolizumab in the RAPID-PsA trial4 and with ustekinumab in the PSUMMIT1 trial6; in Phase II studies, with the use of golimumab in the GO-REVEAL trials8,9; with infliximab in the IMPACT121 and IMPACT222 trials; with a combination of infliximab plus methotrexate compared to methotrexate alone in an open-label study11; and with the use of adalimumab in prior treatment failures in other open-label studies7,15.
In contrast, no significant benefit was demonstrated in RCT of apremilast (PALACE)3, or adalimumab (ADEPT)19, or an open-label adalimumab trial23; with the use of leflunomide (TOPAS)28, or with the use of sulfasalazine in a multicenter study of US veterans31.
Although an etanercept study (PRESTA) demonstrated improvement in dactylitis scores16, a placebo-controlled trial with dactylitis as an endpoint is required. Elsewhere, the role of anakinra is still uncertain14.
Measures of treatment effect
In the few cases where primary data were available, we calculated the effect sizes for the various therapeutic interventions used in the 29 included trials (Table 1). The best available data from RCT suggested that infliximab (effect size 0.41, IMPACT1)21, certolizumab (effect size 0.50, RAPID-PsA)4, and ustekinumab (effect size 0.29, pooled PSUMMIT1 and PSUMMIT2 data)5,6 were likely to be efficacious, while effect sizes for leflunomide and sulfasalazine were 0.33 and 0.2, respectively, despite no significant difference between the treatment and placebo arms in these 2 studies28,31.
Toxicity/safety aspects related to dactylitis
A review of toxicity/safety data revealed no evidence of adverse events related to dactylitis itself. Adverse events were those typically seen in trials of either psoriasis or PsA (i.e., liver toxicity, gastrointestinal manifestations, exacerbation of psoriasis, and incidence of malignancies and autoimmune diseases).
DISCUSSION
Conclusions and limitations
This brief review reveals the dearth of evidence for treating dactylitis in patients with PsA, with highly variable study designs, dactylitis assessments, and patient populations. The most commonly used therapies, NSAID and local corticosteroid injections, have not been formally assessed. DMARD alone may be mildly effective, but the trials have not been adequately powered. Apremilast demonstrated no significant benefit.
Of the biologic drugs tested, only ustekinumab, certolizumab, and infliximab seemed promising, with golimumab as another potential candidate. Etanercept requires more dedicated study to ascertain its efficacy, and adalimumab may be ineffective. The roles of anakinra and newer small molecules and biologic therapies are uncertain.
Because dactylitis may represent a composite of pathological features, it could be argued that an assessment of tenderness and swelling in the component parts (proximal interphalangeal joint and distal interphalangeal joint) is sufficient. However, important and potentially clinically-relevant information could be lost in this simplistic definition. Imaging studies have indicated that dactylitis is a complex, multicompartment disorder, with features including tenosynovitis, enthesitis, osteitis, synovitis, capsulitis, and soft-tissue swelling32. The clinical utility of inflammatory markers and imaging studies to distinguish hot versus cold dactylitis also deserves careful scrutiny. A clear understanding of onset, duration, persistence, anatomical location (hands vs feet), and morphology will be required. Further, the training of dermatologists and other clinicians to recognize and assess dactylitis will be important not only for future clinical trials, but also to hasten referral to a rheumatologist in community practice.
Given the importance and frequency of dactylitis in PsA, future studies should include both robust and quantifiable clinical indices (e.g., the LDI), as well as imaging modalities (e.g., MRI and ultrasound), the latter of which are particularly promising as valid and sensitive measures to assess dactylitis. Importantly, future investigations using dactylitis as a primary outcome measure will determine the most appropriate treatment for this painful and damaging condition.
Acknowledgment
The authors acknowledge Laura Coates for conducting a review of the English-language dactylitis studies, Hans-Georg Zmierczak and Amir Haddad for examining dactylitis studies published in languages other than English, and Claudia Schainberg for reviewing the Juvenile PsA literature. The members of the GRAPPA Dactylitis Working Group also deserve recognition for helpful comments and suggestions pertaining to the manuscript.
APPENDIX 1
List of study collaborators. GRAPPA Dactylitis Working Group (in addition to authors listed): Ade Adebajo, UK; Sueli Carneiro, Brazil; Amir Haddad, Canada; Gurjit Kaeley, USA; Shelley Kafka, USA; Anna Moverley, UK; Oleg Nadashkevich, Ukraine; Andrew Parkinson (patient representative), UK; Ruben Queiro, Spain; Claudia Schainberg, Brazil; Ulku Ucar, Turkey; Rafael Valle-Oñate, Colombia; Hans-Georg Zmierczak, Belgium.
