Introduction
In addition to chronic inflammation of the spine, extra-axial manifestations are common features in patients with ankylosing spondylitis (AS). Enthesitis and peripheral arthritis, predominantly of the lower limbs, occur in up to 50% of patients with AS during the course of the disease [
1‐
5]. These extra-axial manifestations of AS contribute to the burden of the disease [
6,
7].
Nonsteroidal anti-inflammatory drugs (NSAIDs) remain first-line agents for the treatment of AS and can be used for the treatment of enthesitis [
8]. Disease-modifying antirheumatic drugs (DMARDs) do not have a satisfactory effect on axial disease. Sulfasalazine has some effect on extra-axial arthritis [
9] but its benefit for treating enthesitis does not outweigh its risks [
8]. Tumor necrosis factor (TNF) antagonists, including the monoclonal antibodies adalimumab and infliximab and the TNF-receptor construct etanercept, are highly effective agents for the treatment of patients who have active AS despite NSAID treatment [
9‐
18].
In a 24-week, randomized, double-blind, placebo-controlled study of patients with active AS, 152 adalimumab-treated patients experienced a significant reduction in enthesitis compared with 81 placebo-treated patients but no significant improvement in peripheral arthritis [
15]. Similarly, other randomized controlled trials (RCTs) of TNF antagonists have not consistently demonstrated significant improvements in both enthesitis and peripheral arthritis for TNF-antagonist-treated patients compared with placebo-treated patients [
10,
12,
14,
16,
17]. Only in one RCT of infliximab in AS were significant improvements in both tender joint count (TJC) and swollen joint count (SJC) observed [
10]. We evaluated the effects of adalimumab on enthesitis and peripheral arthritis in a large cohort of 1,250 patients with active AS who were enrolled in the open-label RHAPSODY (Review of Safety and Effectiveness with Adalimumab in Patients with Active Ankylosing Spondylitis) study [
1].
Materials and methods
The patient sample and methods of the RHAPSODY study, an open-label, multicenter study conducted at 211 centers in 15 European countries, were previously described in detail [
1]. Independent ethics committees at all participating centers approved the study, and all participating patients gave written informed consent.
Patients
Patients who were at least 18 years old and who had AS according to the modified 1984 New York Criteria for AS [
19] and active disease defined by a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score of at least 4 [
20] were eligible for this study if at least one NSAID or (if stipulated by national guidelines) at least two NSAIDs had failed to control their disease. Patients were permitted to continue current AS therapy with NSAIDs, DMARDs, analgesics, and glucocorticoids (≤ 10 mg/d of prednisone equivalent) during the study. Intra-articular injections or infiltrations of extra-axial joints and tendons were not permitted within 28 days and injections of sacroiliac joints were not permitted within 14 days before screening or during the study. Previous anti-TNF therapy was permitted provided that etanercept had been discontinued at least 3 weeks or infliximab had been discontinued at least 2 months before the first adalimumab injection.
Study design
At baseline, clinicians documented the presence or absence of enthesitis and peripheral arthritis. Enthesitis was defined as at least 1 inflamed enthesis in the Maastricht Ankylosing Spondylitis Enthesitis Score (MASES) (0 to 13) [
21] or of the plantar fascia of the foot. Peripheral arthritis was defined as an SJC of at least 1 (0 to 44, excluding hip joints).
Patients self-administered 40 mg of adalimumab (Abbott Laboratories, Abbott Park, IL, USA) subcutaneously every other week for 12 weeks. The treatment effect of adalimumab for enthesitis was assessed by measuring the change from baseline to week 12 in the MASES (for patients with MASES of at least 1 at baseline) and the change from baseline to week 12 in the percentage of patients with enthesitis of the plantar fascia (for patients with inflammation of the plantar fascia at baseline). The treatment effect of adalimumab for peripheral arthritis was assessed by measuring the change in TJCs (0 to 46) and SJCs from baseline to weeks 2, 6, and 12. Overall adalimumab effectiveness was measured using the Assessment of SpondyloArthritis International Society 20% and 40% responses (ASAS20 and ASAS40, respectively) [
20,
22] and using a 50% improvement in the BASDAI (BASDAI 50) [
23,
24]. In addition, absolute values at each study visit (weeks 2, 6, and 12, as applicable) and changes from baseline were calculated for BASDAI (0 to 10) [
20], Bath Ankylosing Spondylitis Functional Index (BASFI) (0 to 10) [
25], the Patient's Global Assessment of disease activity (PaGA) (based on a visual analog scale of 0 to 100 mm), and C-reactive protein (CRP) serum concentration.
Statistical analysis
Patients who had received at least one injection of adalimumab were included in the analyses. Observed data were used for all effectiveness analyses, and week 12 was specified as the endpoint. Patients with MASES of at least 1 and patients with SJC of at least 1 were stratified by sex (male or female), HLA-B27 positivity (positive or negative), and concomitant DMARD therapy (yes or no) to analyze changes in MASES, TJC, and SJC. The effects of adalimumab on AS disease activity and functional disability were evaluated in four subgroups: (a) patients with no enthesitis and no peripheral arthritis, (b) patients with enthesitis and no peripheral arthritis, (c) patients with peripheral arthritis and no enthesitis, and (d) patients with both enthesitis and peripheral arthritis.
Descriptive analyses were performed by calculating counts and percentages for qualitative data and by calculating medians and first and third quartiles for quantitative data. For the three patient groups without arthritis and enthesitis, with either peripheral arthritis or enthesitis, and with both, the Jonckheere-Terpstra test was applied to test for trends in baseline values and absolute changes from baseline to week 12 in BASDAI, BASFI, PaGA, and CRP. For the median absolute change from baseline to week 12 in MASES, TJC, SJC, BASDAI, BASFI, PaGA, and CRP, nonparametric 95% confidence intervals (CIs) were calculated. Correlations between improvements in AS overall (measured by changes in BASDAI and in BASFI) and improvements in extra-axial manifestations (measured by changes in MASES, plantar fascia, TJC, and SJC) at week 12 were assessed using the Spearman rank order correlation coefficient.
Discussion
Adalimumab effectiveness and safety were investigated in a large cohort of patients with AS in an open-label, uncontrolled, multinational study designed to reflect routine rheumatology practice [
1]. The patients had long-standing, active AS with a broad range of disease manifestations typically associated with AS and were considered eligible for TNF-antagonist therapy [
9]. Herein, we specifically assessed the effectiveness of adalimumab in AS patients with or without enthesitis or peripheral arthritis. The baseline frequency of current enthesitis (54.9%) in this population and that reported in a previous adalimumab RCT (74%) [
15] are both markedly greater than the frequency reported in recent national observational studies (15% to 30%) [
1‐
3], whereas the rate of current peripheral arthritis found in this study was in accordance with rates reported in the literature [
2].
In this study, adalimumab therapy was associated with substantial improvement in enthesitis from baseline to week 12, as indicated by changes in MASES, and in inflammation across all examined entheses, including the plantar fascia. Our results are consistent with results of a previous RCT of adalimumab in 315 patients with AS [
15] and with an RCT of infliximab [
14], although comparison with the latter trial is limited by the use of disparate enthesitis scores.
In the present study, peripheral arthritis in the 281 patients who had at least 1 swollen joint at baseline improved rapidly and substantially during adalimumab therapy, with half of the patients reporting no swollen joints at week 12 (week-12 median TJC = 1 and SJC = 0). Because of the skewed distribution of values, median data are presented here. In a previous adalimumab RCT in which mean data were published [
15], the mean change at week 12 in TJC was -0.8 and the mean change in SJC was -0.4. By comparison, in our study, the mean changes from baseline to week 12 were -4.9 for TJC and -2.9 for SJC, and the median changes were lower than these mean changes (TJC, -3 and SJC, -1). Significant decreases in TJC and SJC from baseline to week 12 have been reported for infliximab therapy for AS [
10].
Women with AS had a greater number of inflamed joints and inflamed entheses at baseline than men. Although the presence of coexisting fibromyalgia in some of the female patients cannot be excluded, the pattern of enthesitis localization as measured by MASES and examination of the plantar fascia is typical for enthesitis in AS (Figure
2) [
21,
26]. Moreover, the finding that female patients improved as much as male patients after 12 weeks of adalimumab therapy suggests that coexisting fibromyalgia did not contribute substantially to the greater enthesitis count at baseline in women. The frequency of and improvement in enthesitis and peripheral arthritis appeared to be unaffected by HLA-B27 status and concomitant DMARD therapy.
The burden of disease at baseline, as measured by BASDAI, BASFI, and PaGA, was lower for patients with no enthesitis and no peripheral arthritis than for patients with either one or both of these manifestations (
P < 0.005). These results are in accordance with previous literature reports [
7,
27,
28]. Of note, two studies showed that a greater BASDAI in patients with peripheral arthritis was not associated with the joint-specific BASDAI items [
27,
29]. Although disease activity and physical impairment differed notably at baseline, the magnitude of improvement with adalimumab therapy in patients with enthesitis or peripheral arthritis or both was similar to that observed for patients without these extra-axial manifestations. At week 12, the median changes in BASDAI, BASFI, PaGA, and CRP were similar between all of the various subgroups, and no important differences in the degree of improvement were detected. However, slightly greater ASAS40 and BASDAI 50 response rates were observed for patients with no enthesitis and no arthritis than for those with enthesitis or arthritis or both. The results of this study of patients with AS categorized by enthesitis and arthritis are consistent with previous predictor analyses, which found that enthesitis and arthritis did not influence the effect of TNF-antagonist therapy [
1].
We did not detect any correlation between improvement in BASDAI and BASFI with improvement in enthesitis or peripheral arthritis. In comparison with joint counts and enthesitis counts, which capture a single domain of AS that may entirely resolve, BASDAI and BASFI capture many disease domains of AS, each of which may respond differently to therapy.
Acknowledgements
The authors thank all investigators and research nurses who contributed to patient recruitment and data collection in the RHAPSODY study. We also acknowledge the contributions of the staff at the following sites that have enrolled at least five patients with AS: Austria: Omid Zamani (Wien), Manfred Herold (Innsbruck); Belgium: Jean-Pol Dufour (Charleroi), Martin Maertens (Oostende), Filip van den Bosch (Belsele), Paul van Wanghe (Hasselt), Johan Vanhoof (Genk), Eric Veys (Gent); Denmark: Karin Grau (Kolding), Palle Holck (Silkeborg); Finland: Pentti Jaervinen (Hyvinkää), Timo Yli-Kerttula (Turku); France: Claude Benhamou (Orleans), Thierry Billey (Cahors), Maxime Breban (Boulogne), Pascal Claudepierre (Creteil), Bernard Combe (Montpellier), Xavier Deprez (Valenciennes), Liana Euller-Ziegler (Nice), Patrice Fardellone (Amiens), Pierre Fauquert (Berck), René-Marc Flipo (Lille), Jean Godde (Marseille), Philippe Goupille (Tours), J-Jean-Luc Grauer (Aix-en-Provence), Pascal Hilliquin (Corbeil Essonnes), Christian Jorgensen (Montpellier), André Kahan (Paris), Xavier Mariette (Le Kremlin-Bicêtre), Yves Maugars (Nantes), Pierre Miossec (Lyon), Alain Saraux (Brest), Jean-Louis Siame (Levin), Jean Sibilia (Strasbourg), Jaques Tebib (Lyon), Charles Zarnitsky (Montvillier); Germany: Stefan Bornstein (Dresden), Jürgen Braun (Herne), Harald Burkhardt (Frankfurt), Edmund Edelmann (Bad Aibling), A Engel (Stuttgart), Georg Gauler (Osnabrück), Gert Hein (Jena), Maria Höhle (Hamburg), Andreas Kapelle (Hoyerswerda), Herbert Kellner (München), Jörn Kekow (Vogelsang), Thilo Klopsch (Neubrandenburg), Ina Kötter (Tübingen), Andreas Krause (Berlin), Klaus Krüger (München), Brigitte Krummel-Lorenz (Frankfurt), Johannes Lohmann (Bad Bentheim), Lothar Meier (Hofheim), Ulf Müller-Ladner (Bad Nauheim), Gunther Neeck (Rostock), Hans-Hartmut Peter (Freiburg), Constanze Richter (Stuttgart), Matthias Richter (Rostock), Karin Rockwitz (Goslar), Ekkehard Röther (Villingen-Schwenningen), Andrea Rubbert-Roth (Köln), Martin Rudwaleit (Berlin), A Seifert (Berlin), Wolfgang Spieler (Zerbst), Rrainer Sprekeler (Zeven), Hans-Peter Tony (Würzburg), Harald Tremel (Hamburg), Ulrich von Hinüber (Hildesheim), Jürgen Wollenhaupt (Hamburg), Silke Zinke (Berlin); Greece: Spyros Aslanidis (Patras), Kyriaki Boki (Athens), Dimitrios Boumpas (Heraklion), Michail Daniilidis (Thessaloniki), Lazaros Sakkas (Larisa), Petros Sfikakis (Athens); Italy: Fabrizio Cantini (Prato), Gianfranco Ferraccioli (Roma), Alessandro Mathieu (Monserrato), Marco Matucci Cerinic (Firenze), Ignazio Olivieri (Potenza), Carlo Salvarani (Reggio Emilia), Giovanni Triolo (Palermo), Guido Valesini (Roma); The Netherlands: Eduard Nicolaas Griep (Leeuwarden), Joanna Rosalia Maria Griep-Wentink (Den Helder), Steffen Zanen (Zwolle); Norway: Erik Rødevand (Trondheim); Spain: Alberto Alonso (Barakaldo), Melchor Alvarez Vega (Madrid), Maria Brito (Madrid), L Carreno (Madrid), Maria del Pilar Fernandez Dapica (Madrid), Jose Luis Fernandez-Sueiro (Elche), Eduardo Loza Cortina (Pamplona), Carlos Rodriguez Lozano (Las Palmas de Gran Canaria), José A. Roman Ivorra (Valencia), Agusti Sellas (Barcelona), Juan Carlos Torre (Oviedo), Juan Tovar Beltran (Elche); Sweden: Awat Jalal (Öerebro), Åke Thörner (Eskilstuna); Switzerland: Beat Michel (Zürich), R Theiler (Zürich), Peter Villiger (Bern); UK: Mohammed Akil (Sheffield), Ernest Choy (London), Robert Cooper (Manchester), Christopher Edwards (Southampton), Paul Emery (Leeds), Karl Gaffney (Norwich), Emmanuel George (Merseyside), David Grennan (Wigan), Richard Hull (Portsmouth), Paresh Jobanputra (Birmingham), Lesley Kay (Newcastle), Andrew Keat (Harrow), Bruce Kirkham (London), Robert Moots (Liverpool), Andrew Ostor (Cambridge), Millicent Stone (Bath), Paul Wordsworth (Oxford). We thank Ioanna Mantika for study management (Abbott Laboratories, UK), Christa Zaiti-Runkel for data management, Angelika Freitag and Anja Bruhn for statistical programming (Abbott GmbH & Co. KG), and Dana L Randall (Arbor Communications, Inc., Ann Arbor, MI, USA) and Michael A Nissen (Abbott Laboratories) for writing and editing support in the development and revision of this manuscript. This support was funded by Abbott. Abbott Laboratories funded the research reported in this manuscript and the manuscript's preparation. The RHAPSODY Study Group included experts from academic institutions in Europe and the US and members of Abbott Laboratories, who were responsible for the design of the original clinical trial. Collection and analyses of clinical data were performed by Abbott Laboratories.
Competing interests
MR and PC were RHAPSODY study investigators. MR has received consulting fees, speaking fees, and honoraria from Abbott, MSD (München, Germany), Schering-Plough Corporation (Kenilworth, NJ, USA), Pfizer Inc (New York, NY, USA), and Wyeth (Madison, NJ, USA). MK and HK are employees of Abbott GmbH & Co KG (Ludwigshafen, Germany), an affiliate of Abbott Laboratories, and hold shares of Abbott stock. SK is a contractor of Abbott GmbH & Co KG. RW was an employee of Abbott Laboratories at the time this study and these analyses were completed and holds shares of Abbott stock.
Authors' contributions
All authors contributed to manuscript development and reviewed and approved the content of the submitted manuscript.