Article Text

Download PDFPDF

Concise report
Rheumatoid arthritis secondary non-responders to TNF can attain an efficacious and safe response by switching to certolizumab pegol: a phase IV, randomised, multicentre, double-blind, 12-week study, followed by a 12-week open-label phase
  1. Michael H Schiff1,
  2. Johannes von Kempis2,
  3. Ronald Goldblum3,
  4. John R Tesser4,
  5. Ruediger B Mueller2
  1. 1University of Colorado, Denver, Colorado, USA
  2. 2Division of Rheumatology, Immunology and Rehabilitation, Kantonsspital St. Gallen, St. Gallen, Switzerland
  3. 3Counter Pressure Casting Inc. (CPC), Carlsbad, California, USA
  4. 4University of Arizona Health Sciences Center, Phoenix, Arizona, USA
  1. Correspondence to Michael H Schiff, Rheumatology Department, University of Colorado, 5400 South Monaco, Englewood, Colorado CO 80111, USA; michael.schiff{at}


Objective To study the efficacy and safety of certolizumab pegol (CZP) in patients with active rheumatoid arthritis (RA) who had discontinued an initially effective TNF inhibitor (TNF-IR).

Methods A randomised 12-week double-blind trial with CZP (n=27) or placebo (n=10) followed by an open-label 12 week extension period with CZP.

Results Baseline characteristics ofthe 2 groups were similar. ACR20 response (primary end point) at week 12 was achieved in 61.5%, and none of CZP and placebo-treated patients, respectively. Weeks 12–24 showed a maximum effect for CZP at 12 weeks, and that placebo patients switched blindly to CZP attained similar results seen with CZP in weeks 0–12. Since this result was highly significant, study inclusion was terminated after entry of 33.6% of the originally planned 102 patients. Adverse events occurred in 16/27 (59.3%) CZP subjects and 4/10 (40%) placebo subjects. There were no serious adverse events, neoplasms, opportunistic, or serious infections.

Conclusions This first, prospective, blinded trial of CZP in secondary TNF-IR shows that the ACR20 response rate observed with CZP was higher than that reported in most previous studies of TNF-IR. Additionally, CZP demonstrated good safety and tolerability. This study supports the use of CZP in RA patients who are secondary non-responders to anti-TNF therapies.

  • Rheumatoid Arthritis
  • Anti-TNF
  • Treatment
  • DMARDs (biologic)
  • DAS28

Statistics from


Multiple studies have investigated the switch from one anti-tumour necrosis factor (TNF)-α therapy to another, but these were primarily open-label uncontrolled studies. In the open-label ReACT study,1 patients who had received prior therapy with etanercept and infliximab tended to have lower ACR20 response rates than patients who had previously received only one TNF inhibitor or patients who had not previously received biologic therapy. Multiple small open-label studies have shown that rheumatoid arthritis (RA) patients with an inadequate response to infliximab may be successfully switched to etanercept2–4 and, conversely, that patients with an inadequate response to etanercept may be successfully switched to infliximab.5

RA patients with an inadequate response to a TNF inhibitor may be classified as primary non-responders if they never responded or failed to respond within the first 3 months of therapy or as secondary non-reponders if they initially responded, but then lost the response after 3 months. In the ReACT study,1 ACR20 and ACR50 response rates trended higher in patients who were secondary non-responders or intolerant to the previous TNF inhibitor than in primary non-responders.6

CDAI (clinical disease activity index)7 can be employed to assess disease activity. A good correlation of CDAI and DAS28 was demonstrated by Arya et al.8 CDAI has been used in one randomised clinical trial9 and was chosen as a secondary clinical endpoint in this trial.

The objective of this double-blind prospective, placebo-controlled study was to examine the effect of 12 weeks treatment with certolizumab pegol (CZP) or placebo on measures of disease activity in patients with RA on stable concomitant conventional disease modifying anti-rheumatic drugs (DMARD) therapy, and who had discontinued a TNF inhibitor other than CZP for secondary lack of efficacy or lack of tolerability. Efficacy and safety were evaluated during an open-label treatment period of 12 weeks with CZP following the initial randomised controlled period. The rapidity and magnitude of clinical response to CZP and safety of this TNF inhibitor has been demonstrated in several studies,10 ,11 and was evaluated in this study.


Patients and study design

The study was a 12-week prospective, phase 4, double-blind, randomised, multicentre, placebo-controlled clinical trial. Patients were randomised 2:1 to receive s.c. (subcutaneous) liquid CZP (400 mg at weeks 0, 2 and 4, followed by 200 every 2 weeks) or placebo, plus stable dosing with conventional DMARDs. At week 12, all patients entered a 12-week blinded open-label phase with concomitant DMARDs and CZP (400 mg at weeks 0, 2 and 4, followed by 200 every 2 weeks). The blind was maintained. Therefore, a loading dose of CZP 400 mg at week 12, 14 and 16 was given to all patients. The study was conducted in accordance with good clinical practice and the Declaration of Helsinki, and was approved by an institutional review committee at each participating centre. All participating patients provided written informed consent (NCT01147341).


Eligible patients were diagnosed with RA, as defined by the ACR 1987 criteria12 of >6 months’ duration and a functional Class 1–3.13 Active disease was defined as ≥6 tender and swollen joints (of 28-joint count), an elevated C-reactive protein, or a CDAI ≥12 at screening.

Patients had previous secondary inadequate response or were intolerant to a TNF antagonist other than CZP that was discontinued at least 28 days prior to the baseline visit. In detail, 34 patients had loss of efficacy, and 3 patients were intolerant (migraine (n=1), injection site reaction (n=1), infusion reaction and injection site reaction to another anti-TNF agent (n=1)).

No B cell depleting agent was permitted within 6 months before enrolment.

Patients must have taken oral methotrexate or other non-biologic DMARD continuously for ≥3 months before the first study dose. Oral corticosteroids (≤10 mg/day prednisone equivalent) were permitted with stable dosing within 1 month of baseline and throughout the study period.


The primary efficacy endpoint was the proportion (%) of subjects achieving an ACR20 response in the CZP treatment group compared to the placebo group at week 12. Secondary efficacy endpoints included: the proportion of patients achieving a CDAI response, ACR50 responders, ACR70 responders, low disease activity score (DAS28 (C reactive protein; CRP) of ≤3.2 or CDAI <10), and the mean change from baseline in HAQ-DI. The same variables were analysed from week 12 onwards for the open label extension. Safety was analysed for the two groups during the entire study.

Statistical analyses

For sample size calculations, a placebo ACR20 response rate of 20% and a CZP response rate of 50% were estimated. To obtain 80% power of detecting a difference at a 5% significance level, a total sample size of 102 evaluable subjects (30 placebo-treated subjects and 60 CZP-treated, after a drop-out rate of 12%) was calculated. These analyses were performed using the Cochran-Mantel-Haenszel χ2 test.


Patient characteristics and disposition

Patient demographics and baseline characteristics were similar in the two treatment groups and indicated high baseline disease activity and long-standing disease (table 1). All patients completed the 12-week study, and one patient initially treated with placebo dropped out of the 12-week open label extension. After randomisation of 37 (36.3%) of the 102 originally planned patients, enrolment into the study was stopped by the sponsor-investigator, since an interim analysis demonstrated that the endpoint hypothesis was met.

Table 1

Demographical data

Clinical efficacy

An ACR20 response was observed in 61.5% of CZP-treated patients (figure 1A; χ2 test: 10.44,1, p<0.005), whereas no placebo patients achieved an ACR20 response. DAS28, CDAI low disease activity (CDAI<10) and CDAI decrease >13.9 (as an additional post-hoc analysis) were significantly (all p<0.0005, figure 1B, C) higher in CZP treated patients after 12 weeks as compared to patients initially treated with placebo. In parallel, the ACR 50 and 70 responses were markedly higher after 12 weeks in CZP-treated patients (figure 1A). During the 12 weeks of open label extension, the group of patients who switched from placebo to CZP treatment demonstrated significant improvement in ACR20, ACR 50 and 70 and CDAI responses and increased improvement in DAS28 scores. CZP patients initially randomised to active drug reached peak effectiveness by week 12.

Figure 1

Disease activity. Patients were depicted in black if treated with certolizumab pegol (CZP) throughout the randomised study and open label extension. Patients were shown in grey if initially treated with placebo (week 0–12) and switched to CZP from week 12 to 24. The percentages of patients achieving an ACR20, 50 and 70 response and clinical disease activity index (CDAI) response >13.9 points and CDAI defined low disease activity at week 12 (randomised study, (A) and at week 24 (open label extension, (B) were demonstrated for either patient group. (C) The average DAS28 level was depicted from week 0 to week 24. (D) The percentage of patients achieving a decrease in HAQ-DI ≥0.3 at week 12 (left panel) and at week 24 (right panel) was depicted for CZP (black) or initially placebo (grey) treated patients.

Patient reported outcomes

During the randomised study, the percentage of patients with a decrease in HAQ-DI of ≥0.3 was 66.7% (n=18) of the CZP patients as compared to 20% (n=2) of the placebo-treated patients (χ2 test: 3.955,1, p=0.046, figure 1D), respectively. During the open label extension, HAQ-DI improvement was also observed in the placebo patients switched to CZP during the open label extension.


Treatment-emergent adverse events (TEAE) occurred in 59.3% and 40.0% of patients in the CZP and placebo groups, respectively (table 2). The TEAEs in both treatment groups were mild (43.8% CZP; 75.0% placebo) or moderate (56.3% CZP; 25.0% placebo), as no severe events occurred.

Table 2

Adverse events: weeks 0 through 24

The most frequent TEAEs were upper respiratory infections (17.6% CZP, 0% placebo), cough (7.4% CZP, 10% placebo), or headache (7.4% CZP, 0% placebo). One serious adverse event of GI-bleeding (considered unrelated to CZP) occurred in the open-label phase. No opportunistic infections, tuberculosis or serious infections occurred.


This double-blind prospective controlled 12-week study demonstrated the efficacy of CZP as compared to placebo in treating the signs and symptoms of active RA with secondary inadequate response to TNF inhibitors. The highly significant ACR20 response rate observed with CZP (61.5%) at week 12 was similar to the response rate found in the REALISTIC study for TNF-IR.14

Early termination

After 37 patients were enrolled, an interim analysis (not preplanned) was done, and showed that the primary endpoint had been reached with a very high level of significance. Therefore, study enrolment was terminated early, as it was considered unethical to include all originally planned 102 patients and expose a third of them to placebo.

No additional effect after week 12

From week 0 to 12, the DAS28 level decreased in 18 of the 27 patients treated with CZP by an average of 1.6 (SD±1.12). The patients not responding after 12 weeks of CZP treatment were not likely to develop an additional DAS28 response despite the additional 12 weeks of open label extension treatment including three additional loading doses at weeks 12, 14 and 16. In the patients who did show a response to CZP at week 12, the average DAS improvement was 0.1 from week 12 to 24 (data not shown). As a consequence, as discussed by others,15 in patients with incomplete secondary response to methotrexate (MTX), the response to CZP peaks at 12 weeks.

The CZP safety profile was similar to that observed in previous CZP studies.

CDAI and HAQ-DI responses

The effect of CZP was analysed by the CDAI and HAQ-DI as secondary endpoints, and comparable results were found. Thus, patient-reported outcomes and CDAI are important, and are also useful for demonstrating clinical efficacy in clinical studies.


This study demonstrated a highly effective and clinically meaningful response to CZP in RA patients with secondary inadequate response or intolerance to TNF inhibitors. This is the first study to prospectively demonstrate that this response peaks at 12 weeks of therapy. This study supports the use of CZP in RA patients who are secondary non-responders to TNF inhibitors or intolerant to them, demonstrating a meaningful response and a good safety profile for this subset of RA patients.



  • Handling editor Tore K Kvien

  • Contributors MHS: Setup and conduction of study, data interpretation, manuscript preparation. JvK and RBM: Data interpretation, manuscript preparation. RG: Setup and conduction of study, data interpretation. JRT: conduction of study, data interpretation.

  • Funding UCB grant support.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The study was conducted in accordance with good clinical practice and the Declaration of Helsinki, and was approved by an institutional review committee at each participating centre. All participating patients provided written informed consent (NCT01147341).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.