Switching from infliximab to once-weekly administration of 50 mg etanercept in resistant or intolerant patients with ankylosing spondylitis: Results of a fifty-four–week study



In recent years, 2 anti–tumor necrosis factor α (anti-TNFα) drugs, infliximab and etanercept, have been approved for the treatment of ankylosing spondylitis (AS). Available data from the literature demonstrate a good efficacy of both infliximab 5 mg/kg and etanercept 25 mg twice weekly in the treatment of AS, with a 50% reduction of disease activity in ∼60% of patients (1, 2). However, in controlled trials, a variable percentage of patients, ranging between 20% and 30%, are nonresponders or achieve only a partial response and do not reach the commonly used outcome measures. Moreover, in an additional 10–20% of patients, the treatment should be interrupted due to loss of efficacy or side effects over the followup period (1–5). As observed in patients with rheumatoid arthritis (6–8), in the case of treatment failure with one agent, switching to the other agent may also be useful in patients with AS due to the different chemical structures and mechanisms of action of the 2 drugs (9, 10).

A recent French study seems to confirm the efficacy of etanercept in 11 patients with different spondylarthropathies who were unresponsive or intolerant to infliximab (11). In all studies of patients with AS, etanercept has been administered at a dosage of 25 mg twice weekly. However, in studies of patients with rheumatoid arthritis, results comparable in terms of efficacy and safety were obtained when the drug was administered at a dosage of either 50 mg once weekly or 25 mg twice weekly (12). Infliximab therapy for AS was permitted by the Italian Ministry of Health in the beginning of the year 2001, and etanercept was permitted 3 years later.

We conducted a 54-week, open-label, prospective, followup study to evaluate the efficacy and tolerability of etanercept administered at 50 mg once weekly in patients with active AS who were resistant or intolerant to previous therapy with infliximab.

Patients and Methods

A 54-week, open-label, prospective, followup study was conducted in 4 Italian secondary referral rheumatology centers (Hospital of Prato, Firenze, Reggio Emilia, and Potenza). All consecutive outpatients with AS treated with 5 mg/kg of infliximab every 8 weeks who failed to achieve or maintain a 20% clinical response according to the Assessment in Ankylosing Spondylitis criteria (ASAS20) (13) or who withdrew from therapy due to intolerance or adverse events (AEs) were recruited for the study over a 24-week period. All patients satisfied the modified New York criteria for AS (14), and all signed an informed consent form before the enrollment. All patients received infliximab at a dosage of 5 mg/kg at weeks 0, 2, 6, and then every 8 weeks. Before switching, patients received at least 6 infusions of infliximab, with infusion intervals being shortened to 6 weeks. None of the patients had complete ankylosis of the spine.

Concomitant therapies with corticosteroids, nonsteroidal antiinflammatory drugs (NSAIDs), or analgesics at stable doses were permitted, excluding any disease-modifying antirheumatic drugs (DMARDs). Because the half life of infliximab is 9–12 days and the drug is detectable in the serum 8–12 weeks after administration (15), all patients had a washout period of at least 8 weeks after the last infusion before study entry.

The primary objective of the study was to evaluate the percentage of ASAS20 responders after 54 weeks of treatment. Secondary objectives included achievement of ASAS20 at week 24; achievement of ASAS50 and ASAS70 at weeks 24 and 54; change in Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score from baseline; the number and severity of etanercept-related AEs; and the percentage of patients who were able to reduce or interrupt corticosteroid, NSAID, and analgesic intake.

Drug administration.

Because prefilled syringes containing 50 mg of etanercept for subcutaneous injections were not yet available in Italy at the study initiation, the drug was administered with 2 consecutive subcutaneous self-injections once weekly on the same day. Qualified personnel instructed all patients in proper technique and directly observed patients' first self-injections. The drug was delivered monthly to each patient in a package containing 8 prefilled syringes, and all patients were instructed to properly store the drug at 2–6°C in a place protected from light. Patients were informed to call the center for drug package substitution in the case of malfunction of home refrigerator and consequent altered storage.

Outcome measures.

At enrollment and at every followup visit, all patients were evaluated for ASAS20 response as a primary outcome measure. Secondary outcome measures were improvement of BASDAI, Bath Ankylosing Spondylitis Metrology Index, and Bath Ankylosing Spondylitis Functional Index (BASFI) scores; chest expansion; finger-to-floor distance; tender and swollen joint counts; number of painful entheses; number of digits showing dactylitis; erythrocyte sedimentation rate; and C-reactive protein level. Enthesitis was assessed by evaluating 13 entheses according to the Maastricht Ankylosing Spondylitis Entheses (MASES) score (16), and dactylitis was assessed by evaluating finger tenderness and swelling on a 0–4 Likert scale.

Each patient was clinically evaluated by 2 independent rheumatologists on the same day. Clinical data and scoring of each outcome measure were independently recorded in patients' charts, and the final score resulted from reciprocal agreement or assuming the worse score in debated cases. Corticosteroid, NSAID, and analgesic doses were recorded at every visit. Visits were scheduled at baseline, weeks 4 and 8, and every 8 weeks thereafter until week 54.

Safety assessment.

Recommended screenings for latent or active tuberculosis (purified protein derivative [PPD] skin test and chest radiograph) were performed in all patients before starting infliximab. At study entry, PPD (10 units) skin test was repeated in all patients, and results were negative in all.

At every visit, all patients were monitored for clinical and laboratory evidence of AEs defined as mild (transient and easily tolerated), moderate (patient discomfort with interruption of usual activities), and severe (incapacitating or life threatening). Patients were also asked about domiciliary reactions at the injection site, and the last 2 injection sites were evaluated during the followup visits.

Statistical analysis.

Data were analyzed at weeks 24 and 54 using the SPSS statistical package (SPSS, Chicago, IL). Wilcoxon's matched pairs signed rank test was used to measure the changes from baseline, and P values less than 0.05 were accepted as significant. Fisher's exact test was used when necessary. Data were not analyzed for multiple variable testing and P values were not adjusted.


Over the 24-week enrollment period, 25 patients were considered suitable candidates for the study. Because 2 patients did not give consent, 23 patients were recruited. The demographic and clinical characteristics of the patients are summarized in Table 1.

Table 1. Demographic and clinical characteristics of 23 patients with ankylosing spondylitis recruited for the study*
  • *

    Values are the number (percentage) unless otherwise indicated. AEs = adverse events; NSAIDs = nonsteroidal antiinflammatory drugs.

Total number23
Age at study entry, median years43
Disease duration, median years10
B27 positive20 (87)
Psoriasis4 (17)
Peripheral arthritis3 (13)
Enthesitis/dactylitis8 (35)
Number of infliximab infusions, median16
Infliximab failure at the beginning of therapy2 (9)
Acquired infliximab resistance16 (69)
Infliximab withdrawn due to AEs or intolerance5 (22)
Prednisone intake10 (43)
Median prednisone dosage, mg/day10
NSAIDs intake20 (87)

All patients received at least 6 infusions of infliximab, with a median of 16 infusions. In 18 (78%) of 23 patients, switching to etanercept was related to a lack of efficacy of infliximab. Of these, 2 (9%) patients were nonresponders and 16 (69%) experienced acquired drug resistance. In 5 (22%) of 23 patients, infliximab was suspended because of AEs (bacterial pneumonitis in 2 patients, recurrent urinary infections in 1 patient, and severe hypotension and dyspnea during the infusion in 2 patients). At study entry, 10 (43%) of 23 were receiving corticosteroids (median prednisone dosage 10 mg/day) and 20 (87%) of 23 were receiving NSAIDs.

All patients completed the study, with no serious AEs. Minor AEs, such as injection site reaction, rash, dizziness, and transient headache, occurred in 5 (22%) of 23 patients. At week 24, ASAS20 was achieved by 18 (78%) of 23 patients, ASAS50 by 12 (52%) of 23, and ASAS70 by 9 (39%) of 23. At week 54, ASAS20 was recorded in 17 (74%) of 23 patients, ASAS50 in 14 (61%) of 23, and ASAS70 in 9 (39%) of 23.

The mean ± SD BASDAI score changed from 6.9 ± 1.3 at baseline to 3.1 ± 1.5 at week 24 (P < 0.001) and significantly changed to 2.9 ± 1.7 at week 54 (P < 0.001). Separate analysis of the single items of the BASDAI score demonstrated a significant improvement of all variables. As shown in Table 2, all outcome measures changed significantly with respect to the time of enrollment.

Table 2. Changes at weeks 24 and 54 in all outcome measures in 23 patients with ankylosing spondylitis treated with etanercept 50 mg once weekly*
OutcomeEnrollment24-week visitP54-week visitP
  • *

    Values are the mean ± SD unless otherwise indicated. PGA = patient global disease assessment; VAS = horizontal visual analog scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASAS = Assessment in Ankylosing Spondylitis; BASFI = Bath Ankylosing Spondylitis Functional Index; BASMI = Bath Ankylosing Spondylitis Metrology Index; TJC = tender joint count; SJC = swollen joint count; ESR = erythrocyte sedimentation rate (Westergren); CRP = C-reactive protein; NSAIDs = nonsteroidal antiinflammatory drugs.

  • Statistical differences calculated with respect to values recorded at enrollment.

PGA (10-cm VAS)6.3 ± 2.12.8 ± 2.2< 0.0012.9 ± 1.1< 0.001
Back pain (10-cm VAS)6.9 ± 1.52.3 ± 2.3< 0.0012.1 ± 1.1< 0.001
Inflammation (10-cm VAS)6.2 ± 1.62.7 ± 1.1< 0.0012.8 ± 0.7< 0.001
BASDAI6.9 ± 1.33.1 ± 1.5< 0.0012.9 ± 1.7< 0.001
ASAS 20%, no. (%) 18 (78) 17 (74) 
ASAS 50%, no. (%) 12 (52) 14 (61) 
ASAS 70%, no. (%) 9 (39) 9 (39) 
BASFI5.4 ± 1.62.9 ± 1.2< 0.0012.3 ± 1.4< 0.001
BASMI4.4 ± 2.13.1 ± 1.8< 0.0022.5 ± 1.6< 0.001
Modified Schober's test, cm2.8 ± 0.563.7 ± 0.87< 0.0024.1 ± 0.54< 0.001
Fingertip to floor distance, cm12.2 ± 4.66.2 ± 3.1< 0.0015.8 ± 2.9< 0.001
Chest expansion, cm3.4 ± 0.213.93 ± 0.12< 0.0023.95 ± 0.14< 0.002
TJC, no. (%)5 (22)1 (4)< 0.0011 (4)< 0.001
SJC, no. (%)3 (13)0 (0)< 0.0010 (0)< 0.001
Enthesitis/dactylitis, no. (%)8 (35)1 (4)< 0.0011 (4)< 0.001
ESR, mm/hour31.4 ± 12.312 ± 5.4< 0.00110 ± 6.3< 0.001
CRP, mg/dl2.8 ± 1.60.43 ± 0.15< 0.0010.38 ± 0.26< 0.001
Prednisone dosage, mg/day11.2 ± 4.12.3 ± 3.6< 0.0012.1 ± 2.2< 0.001
Prednisone intake, no. (%)10 (43)2 (9)< 0.0011 (4)< 0.001
NSAIDs/analgesics intake, no. (%)20 (87)2 (9)< 0.0011 (4)< 0.001

As defined by ASAS (13), a partial remission was observed in the remaining 5 (22%) of 23 patients who had a longer disease duration (median 14 years) and who did not reach 20% improvement in BASFI score and inflammation at week 54. Withdrawal of corticosteroids was possible in 9 (90%) of 10 patients, and a 50% reduction of initial dose was recorded in 1 (10%) of 10 patients. NSAIDs and analgesics interruption was recorded in 19 (95%) of 20 patients.


AS is a chronic inflammatory rheumatic condition that occurs prevalently in patients who are HLA–B27 positive and is characterized by entheseal and synovial involvement, with progressive damage and ankylosis of the spine in the majority of patients (17). NSAIDs are usually effective in improving the symptoms and are recommended as the initial therapy (18). Different from rheumatoid arthritis, none of the traditional DMARDs have been shown to reduce the symptoms and slow the progression of axial disease in patients with AS (19).

Experimental models and studies on humans have provided a consistent body of evidence on the key role exerted by the proinflammatory cytokine TNFα in the pathogenesis of AS (20–22). Confirming these observations, recent open and controlled clinical studies have demonstrated the efficacy of anti-TNFα agents infliximab and etanercept in the treatment of AS, with a clinical response rate in terms of BASDAI score and ASAS20 criteria ranging from 57% to 71% for both drugs (1–5, 23, 24). However, long-term extension studies of controlled trials of patients treated with infliximab demonstrate that ∼20% of the originally enrolled patients withdraw from therapy due to intolerance or inefficacy (25).

Because infliximab and etanercept have a different chemical structure and a different mechanism of action in TNFα inhibition (9, 10), there is a rationale for switching from infliximab to etanercept and vice versa in clinical practice. In the case of treatment failure with one anti-TNFα agent, switching to another agent and restoring clinical response has been repeatedly reported in patients with rheumatoid arthritis (6–8, 26, 27). This therapeutic option has not yet been precisely addressed in patients with AS.

In a recent preliminary report, 9 (69%) of 13 patients with AS, psoriatic spondylitis, and undifferentiated spondylarthropathy who were switched from infliximab to etanercept achieved a satisfactory clinical response in terms of 50% BASDAI improvement (11). The drug was administered over a median followup period of 10 months and none of the patients withdrew from the study because of side effects or serious AEs.

In our clinical series, we observed a 54-week ASAS20, ASAS50, and ASAS70 response rate of 74%, 61%, and 39%, respectively, which was similar to that recorded by Delaunay et al (11). These percentages are higher than those reported in 2 controlled clinical trials of etanercept in patients with AS (2, 5), where ASAS20, ASAS50, and ASAS70 response rates did not exceed 66%, 49%, and 24%, respectively. The higher response rate recorded in our clinical series may be due to the different selection of patients. Indeed, 16 of 23 patients included in this study had initially responded to infliximab with acquired drug resistance due to an escape phenomenon despite dose adjustments.

In keeping with previous studies (5, 28), long-term administration of etanercept was characterized by a good safety profile and tolerability. Minor side effects were observed in 5 (22%) of 23 patients, with no serious AEs.

In all studies of patients with AS, etanercept has been administered at a dosage of 25 mg twice weekly. However, in a large clinical series of patients with rheumatoid arthritis, etanercept given at a dosage of 50 mg once weekly demonstrated a comparable efficacy and safety with respect to 25 mg administered twice weekly. Therefore, in this study we decided to use a dosage of 50 mg once weekly. The results seem to confirm that this dose is equally effective and safe in patients with AS. Due to the current unavailability in Italy of syringes prefilled with 50 mg of etanercept, the drug was given once weekly with 2 consecutive 25-mg injections on the same day. However, single 50-mg injections once weekly will have some advantages in terms of patient compliance, drug storage, reduction of injection number, and reduction of injection site reactions.

In conclusion, our data indicate that switching to etanercept may represent a good therapeutic option for patients with AS who are resistant or intolerant to infliximab. Switching to etanercept after infliximab treatment escape restores the clinical response in the majority of patients. Moreover, etanercept, administered at a dosage of 50 mg once weekly, is effective and well tolerated in patients with AS.