EFFICACY AND SAFETY OF ADALIMUMAB IN ANKYLOSING SPONDYLITIS: DATA FROM REAL LIFE

Rationale. Data from controlled trials showed that adalimumab, a humanized anti-TNF monoclonal antibody, iseffective and safe in the treatment of ankylosing spondylitis (AS). Objectives. The present study aimed to observe the effi cacy and safety of adalimumab in AS in a real life clinical setting. Methods. The study observed cross-sectionaly and retrospectively the effi cacy and safety of adalimumab in all the patients admitted to the Rheumatology Department of “Sfânta Maria” Clinical Hospital between January 2008 and June 2013 who were classifi ed as having AS according to the modifi ed New York criteria. The diagnosis and follow-up of uveitic cases were done in the Ophthalmology Department of the Emergency University Hospital. Results. Within the study time-frame, 79 AS patients met the inclusion criteria: 71 (89.9%) had adalimumab for at least 24 months; 8 (10.1%) switched from adalimumab to another biological, as follows: 3 (3.8%) because of serious adverse events, 3 (3.8%) were primary non-responders and 2 (2.5%) were secondary non-responders. The clinical response was fast: after 3 months of treatment, 59 (83.1%) patients had BASDAI < 4 and 55 (77.5%) patients had BASFI < 4. Regarding safety, the serious adverse effects recorded were: infectious arthritis, pulmonary tuberculosis, pulmonary sarcoidosis. There were no cases of cancer or demyelinating disease during the study frame. Conclusions.Therapy with adalimumab in AS produces a prompt and lasting effect. The effi cacy (remission) and safety (adverse events) of adalimumab can be monitored in the real-life clinical setting using BASDAI, BASFI, and routine clinical evaluations. Clinicians may need to expect a slightly higher rate of serious adverse events and rate of treatment discontinuation than those reported by controlled trials.


INTRODUCTION
Ankylosing spondylitis (AS) is a chronic infl ammatory disease which can involve the axial skeleton, the large peripheral joint, the entheses and several extra-articular sites (e.g. uveitis, bowel disease, skin etc.). (1) Typically AS patients are in their third de-cade of lifeand present infl ammatory chronic back pain, sacroiliitis on imaging and human leukocyte antigen (HLA) B27. Without treatment, patients with AS can develop severe anatomical deformity (kyphosis by syndesmophyte formation), various degrees of functional impairment (limitation of lumbar Abbreviations ASAS -Assessment of SpondyloArthritis international Society AS -Ankylosing spondylitis BASDAI -Bath Ankylosing Spondylitis Disease Activity Index BASFI -Bath Ankylosing Spondylitis Functional Index CRP -C reactive protein ESR -erythrocyte sedimentation rate f -female HLA -human leukocyte antigen m -male NSAID -non-steroidal anti-infl ammatory drugs TNFα -tumor necrosis factor α inhibitors fl exion) and psychological distress, subsequently with poor quality-of-life outcomes and higher social costs.(2,3)These can be prevented or improved using appropriate therapy in the setting of a comprehensive and multidisciplinary management strategy. Clinically, there are four therapeutic measures which proved effective in AS: for the axial form, physical therapy, non-steroidal anti-infl ammatory drugs (NSAIDs) and tumor necrosis factor α (TNFα) inhibitors; for the peripheral form sulfasalazine in addition to the afore mentioned drugs. While physical therapy is mandatory for good functional outcomes, irrespective of medical treatments regimes, (4) NSAIDs are effective for chronic pain and are shown to slow radiographic progression, (5) bearing in mind theirmajor cardiovascular and gastrointestinal side effects. The effectiveness of TNFα blockers proved the major role played by this cytokine in the process of chronic infl ammation. All the approved anti-TNFα agents (adalimumab, etanercept, infl iximab, golimumab) seem similar regarding effi cacy and safety in AS patients. (6) Due to costs and the lack of long term safety data, anti-TNFα agents are given to selected AS patients, which should be monitored thoroughly. (7) For this purpose, in the clinical setting physicians may use two simple disease indices, BASDAI (Bath Ankylosing Spondylitis Disease Activity Index) and BASFI (Bath Ankylosing Spondylitis Functional Index).
Adalimumab, a humanized anti-TNF monoclonal antibody, has proved to be very effective in AS compared to placebo, (8-10) in short term and in long term, (11,12) regardless of radiographic progression, (13,14) peripheral and non-articular involvement, (15)(16)(17) race, (18) and age. (19) The most frequent adverse effect reported by investigators was the higher risk of serious infections, including the risk of reactivating tuberculosis. (20) The present study aimed to observe the effi cacy and safety of adalimumab in AS in a real life setting.

Study design
The study was designed to observe in a crosssectional and retrospective manner the effi cacy and safety of adalimumab in all the AS patients admitted to Internal Medicine and Rheumatology Department of our hospital between January 2008 and June 2013, using medical records kept for each visit to the rheumatologist.

Patients and data
All the AS patients included in the study met the following criteria: age above 18 years; Caucasian race; classifi cation of AS according to the modifi ed New York criteria; (21) no overlapping chronic infl ammatory disease (e.g. psoriatic arthritis, reactive arthritis); anti-TNFα naïve before initiation of adalimumab; active disease, defi ned as BASDAI > 4 and/ or BASFI > 4, which needed biological therapy according to the physician's opinion and to the Romanian guide to AS treatment, (22) which is in accordance with the ASAS/EULAR recommendations. (23) The informed consent was presumed from the written informed consent each patient gave for being admitted to a university clinic (clinical examination, blood tests, imaging studies).The data were collected anonymously and it included for each record: demographics (age, gender), infl ammation (C-reactive protein -CRP; erythrocyte sedimentation rate -ESR), disease measures (disease duration; peripheral and non-articular involvement; HLA B27 status; BASDAI; (24) (BASFI), (25) treatment variables (duration, BASDAI and BASFI treatment response, side effects, switches to other drugs). The study was approved by the local ethics committee.

Statistics
Normally distributed data were reported as means with standard deviations, while non-normally distributed data were reported as medians with. Differences were evaluated using non-parametric tests: binomial and χ 2 tests (or Fisher's exact test where appropriate) for nominal data; Mann-Whitney U and Kruskal Wallis tests for scale data. Correlation was established computing Spearman's coeffi cients. All tests were two-sided, were considered signifi cant if p ≤ 0.05 and were done using SPSS Statistics v.17.0.1 for Windows (SPSS Inc., Chicago, U.S.A., 2008).

General characteristics
A total of 79 AS patients met the inclusion criteria. Of these, 71 (89.9%) patients had adalimumab treatment for at least 24 months within the study time-frame ( Table 1). The other 8 patients switched from adalimumab to another biological within the study period, as follows: 3 (3.8%) because of serious adverse events (Table 3), 3 (3.8%) were primary non-responders and 2 (2.5%) were secondary nonresponders (Table 4).

Adalimumab effi cacy
The 71 AS patients who had adalimumab for at least 24 months within the study frame had a favorable disease course, with a reduction in signs and symptoms which prompted the fall of the activity and functional indices, BASDAI and BASFI ( Table 2). The clinical response was fast: by 3 months, 59 (83.1%) patients had BASDAI < 4 and 55 (77.5%) patients had BASFI < 4 ( Fig. 1 and 2).

Adalimumab safety
Of the 79 AS patients, 3 had serious adverse effects, which lead to the interruption of immunosup-pressive therapy and ultimately to the replacement of adalimumab with other anti-TNFα agents (Table 3). Of note, there were no cases of cancer or demyelinating disease during the study frame.

Persistency of adalimumab treatment
In the 24 months study frame, a total of 71 (89.9%) patients retained their adalimumab treatment, while the other 8 (10.1%) patients switched either to etanercept or infl iximab because of side effects of lack of response to adalimumab (Table 4).

Main fi ndings
In terms of effi cacy, the study observed a prompt, lasting clinical effect of adalimumab in AS patients, as pointed out by the constantly decreasing values of BASDAI and BASFI. In terms of safety, we observed a 3.8% rate of serious adverse effects as re- ported by physicians (infectious arthritis, pulmonary sarcoidosis, and pulmonary tuberculosis), where serious adverse effects meant those conditions attributed to adalimumab which needed hospitalization and specifi c multidisciplinary medical management. Almost 90% of the selected patients still had adalimumab after 24 months from its initiation, with a primary non-responder rate of 3.8% and a secondary non-responder rate of 2.5%.

Comparing with other studies
In a recent meta-analysis, Wang et al. reviewed the safety and effi cacy of adalimumab in AS patients from randomized, placebo controlled and doubleblind trials. (10)In one of their meta-analitic model of 787 AS patients treated with adalimumab, the BAS-DAI at 12 and 24 weeks was signifi cantly lower in the adalimumab group compared to the placebo group.Our results also showed that by 3 months most of the responding patients had a BASDAI < 4. As for safety of adalimumab, the cited authorsfound a 55.3% rate of any adverse events during adalimumab treatment (17.7% in our study), a 1.9% rate of serious adverse events (3.8% in our study), and a1 0.9% rate of injection site reactions (10.1% in our study). The lower rate of any adverse events in our study compared to the above-mentioned meta-analysis can be explained by the fact that physicians in a real life clinical setting do not always record the mild complains of the patients, such as headaches, nausea etc. The fact that we recorded a higher rate of serious adverse events compared to the above-mentioned meta-analysis may be due to several situations: differences in the defi nition of "serious" adverse events; social-economic and education status of the patients; a selection bias manifest in the controlled studieswhich tends to exclude patients with severe structural damage and/or important comorbidities and which underestimates the true risk of serious adverse effects in the population. The same explanation can account for the differences in the rates of discontinuation of adalimumab treatment, 2% in the meta-analysis compared to 10% in our study.

Study limitations
The study had several limitations which may raise some diffi culties in interpreting the results. First of all, the retrospective design had to deal with missing data and follow-up shortcomings. The lack of strong end-points such as the ASAS criteria for treatment response would have been more appropriate to judge the effi cacy of the drug, but these criteria are use more frequently in clinical trails than in real life settings. The lack of placebo and/orother biological drug arms did not allow comparisons with the respective comparators in order to more accu-rately size the place of adalimumab in AS therapy. Finally, the small sample size impeded any regression models for determining signifi cant predictors for clinical response.

Conclusions
Therapy with adalimumab in AS patients in which NSAIDs and/or sulfasalazine are insuffi cient is an effi cacious and safe therapeutic option, which produces a prompt and lasting clinical effect, in accordance with the new treat-to-target principles in AS. In this sense, the effi cacy (remission) and safety (adverse events) of adalimumab treatment can be monitored in the real-life clinical setting using the two simple activity and functional indices, BASDAI and BASFI, and regular clinical and biological examinations. The thorough monitoring of AS patients treated with adalimumab ensures the early detection of side effects and non-responsivity. Clinicians may need to expect a slightly higher rate of serious adverse events and a slightly higher rateof treatment discontinuation than those reported by controlled trials.